BIOLOGICAL CHEbiISTRY1. INTRODUCTIONBy D. J. Manners( H e r b Watt University, Edinburgh 1)PHOTOSYNTHESIS is probably the most important and fundamentalbiochemical process in Nature. The mechanism of carbon dioxide fixationand conversion into carbohydrate by plant tissues has been established,largely by Calvin and his co-workers. More recent studies have been con-cerned with the conversion of light energy into chemical energy and formthe subject of the first Report. The end-products of photosynthesis includestarch and cellulose, which function respectively as the major reserve carbo-hydrate and cell-wall constituent of higher plants. In recent years, ourunderstanding of the biosynthesis of these polymers has had to be com-pletely revised following the demonstration by Leloir and his colleagues thatglucans are synthesised from nucleoside &phosphate sugars.The presentReport surveys these researches, and related structural and enzymic studieson glycogen and other a- and p-glucans.The proteins constitute a large class of naturally occurring macro-molecules which play a vital and unique part in metabolism, and show awide range of biological activities. For this reason, the recent AnnualReports have included general surveys on the chemical structure of a largenumber of proteins. This year the emphasis has been placed on certainproteolytic enzymes which are of particular interest. Firstly, their speci-ficities provide a means for the selective hydrolysis of a limited number ofpeptide linkages in proteins, so that identification’of the resultant mixture ofpeptides is a valuable method for determination of amino-acid sequences.Secondly, many proteolytic enzymes have been highly purified, and provide auseful system for studies on the mechanism of enzyme action.The h a 1Report describes the numerous advances made in studies on sulphur-con-taining amino-acids. These are of special importance in protein biochemistryboth as a means of providing structural units for the cross-linking of peptidechains, and also as components of those regions of peptide chains which arebiologically active. In addition, some of these amino-acids play an essentialpart in intermediary metabolism.D. (3. Smyth, Ann. Reporb, 1963,60,468; 1964, 61, 507; 1965, 62, 4882. RECENT ADVANCES IN PHOTOSYNTHESISBy C.P. Whittingham(Botany Department, Imperial College of Science & Technology, London S. W.7)Introduction.-The process of photosynthesis, which may be defined by theoverall equation6C02 + 6H20 + 6(CH20) + 60,has since the time of Blackman been separated into two reaction processes :firstly, a thermochemicsl process involving the fixation of carbon dioxideand reduction of the product to a sugar; and secondly a photochemicalprocess in which light energy, absorbed by chlorophyll, is utilised to formfrom water the reductants used in process 1. Oxygen is liberated as a con-sequence of this reaction.These two processes have been further analysed into individual partialreaction steps, the first largely as a result of studies with whole cells, eitheralgal or in leaves of higher plants, and the second largely by studies usingsubcellular particles isolated from the cell.Although it is now generallyagreed that both processes occur within the chloroplast in vivo, neverthelessin the isolation of chloroplasts from the cell, the bounding membranes aregenerally lost and substances leak out from the particles. Thus certainactivities are lost from the isolated particles, but the loss of a permeabilitybarrier makes it possible to add back key intermediates which may notpenetrate into whole cells. In this way partial reactions can be investigatedby the traditional biochemical methods.and hiscollaborators using 1*CO, a8 a tracer during photosynthesis in whole cells ledto the formulation of a photosynthetic reduction cycle in which carbondioxide was incorporated in a reaction catalysed by carboxydismutase intophosphoglyceric acid and the phosphoglyceric acid subsequently reduced tophosphoglyceraldehyde (Figure 1).In longer times, most of the carbonfixed was ultimately incorporated through hexose phosphates into carbo-hydrates, such as sucrose and starch, and into lipid components.Under conditions different from those used by Calvin, other photosyn-thetic products were obtained. For example, glycollic acid has been ob-served as a product of photosynthesis in ChZoreZZa and other green plants bymany workers.3 Maximal production is observed at lower partial pressuresof carbon dioxide and higher light intensities. Glycollic acid accumulatesin green leaves in the presence of a-hydroxysulphonates (Zelitch 3, and inChZoreEZa in the presence of isonicotinyl hydrazide (INH).4 It has beengenerally considered that glycollate is further metabolised through glyoxylateThe Thennochemical Process.-The investigations of CalvinIF. F.Blackman, Ann. Bot., 1905, 19, 281.2 M. Calvin, Angew. Chern. Internat. Edn., 1962, 1, 65.3 N. E. Tolbert and I;. P. Zill, J . Biol. Chem., 1956, 222, 895; I. Zelitch, J . Biol.Chern., 1358, 233, 1299; 0. Warburg and G. Krippahl, 5‘ Naturforsch., 1960, 15, 197;C. P. Whittingham, R. G. Hiller, and M. Birmingham, Photosynthetic Mechawmsof Green Plants ”, National Academy Science Publication 1146, 1963, p.675.4 G. Pritchard, C. P. Whittingham, and W. Griffin, J. Exp. Bot., 1962, 13, 176?580 BIOLOaIOAL OHEMISTRPand glycine to serine. The a-hydroxysulphonate is thought to inhibitglycollic oxidase and INH to inhibit the formation of serine from glycine.Glycollic oxidase, glyoxylate reductase, and a phosphatase active on phos-phoglycollate have all been observed in green tissue. The enzymes concernedin the synthesis of serine from glycine have still not been isolated from greenplant tissue. In animal tissues, Richert and co-workers have shown thatINR inhibits the formation of serine from glycine in avian livers. Prom asuggestion of Bassham and co-workers and later by Griffith and Byerrurn,’the origin of glycollic acid has been generally believed to be from the primaryphotosynthetic cycle. Bradbeer and Racker found that crystalline trans-ketolase in the presence of ferricyanide catalysed the formation of glycollatefrom fructose 6-phosphate. Others have proposed that ribulose diphosphateis cleaved to give rise to glycollate and triose phosphate.There seems littleevidence to support the view that glycollate arises from a carboxylationother than that in the primary Calvin cycle. The observations of Zelitchwhich show that in tobacco leaf discs the specific activity of glycollate isgreater than that of phosphoglycerate is not, in our opinion, conclusive.Rabson and co-workers lo have shown that [2-14C]glycollate supplied towheat leaves in air is metabolised to serine and glycine. Jimenez and co-workers n have shown that such feeding also gave rise to sucrose in which theglucose component was labelled almost equally in the 1-, 2-, 5-, and 6-carbonatoms.[3-14C]Serine gave sucrose in which the glucose was labelled in theC-1 and C-6. Wang and Waygood12 obtained similar data with wheatseedlings when [lJ4C]glycine, [2-14C]glycine and [ 1, 2-14C]glycine and[l, 2-14C] glycollate were fed. They showed further that when glycollateand glyoxylate were supplied together with radioactive glycine or serine,there was no effect on the radioactivity of the sugar formed, but additionof glycine or serine, together with [14C]glyoxylate did diminish the radio-activity (see Figure 2). Mifiin l3 has undertaken similar experiments withpea leaves and obtained essentially similar results.Furthermore he showedthat the metabolism of [14C]glycollate and [2-14C]glycine to sucrose wasinhibited either by addition of INH or by prior infiltration of the leaves withnonradioactive serine. There is therefore considerable evidence that inleaves, sucrose can be formed from exogenously supplied glycollate andthat the reaction sequence is via glycine and serine.The production of glycollate, a smaller molecule, may permit the transferof carbon from the intermediates of the photosynthetic cycle within thechloroplast to enzyme systems located in the cytoplasm. Alternatively,6 D. A. Richert, R. Amberg, and AT. Wilson, f. B w l . Chem., 1962, 237, 99.6 J. A. Bassham, A. A. Benson, L. D. Kay, A. 2. Harris, A.T. Wilson, and M.7 T. GrSth and R. U. Byerrurn, J . Biol. Chem., 1959, 234, 762.8 J. W. Bradbeer and E. Racker, Ped. Proc., 1961, 20, 88.9 I. Zelicth, J. Biol. Chem., 1965, 240, 1869.Calvin, J. Amer. Chem. SOC., 1954, 76, 1760.10 R. Rabson, N. E. Tolbert, and P. C. Kearney, Arch. Biochem. Biophys., 1962,11E. Jimenez, R. L. Baldwin, N. E. Tolbert, and W. A. Wood, Arch. Biochem.12 D. Wang and E. R. Waygood, Plant Physiol., 1962, 37, 826.13 B. Miflin, “Carbon metabolism in Chlorella”, Ph.D. Thesis, Unher&y of London,98, 1954.Biophys., 1962, 98, 172.1965WHITTINGHAM : BECENT ADVANCES IN PHOTOSYNTHESIS 581Glycollata GI yox y I at e G I y ci neCH20Hi$WOH Po _---- pkl Glyceratc5H-W 4--.-c- CH,OHHexOse $H.OHCH2WFIGURE 2 The conversion of glycollate into hexose.Tolbert l4 has suggested that glycollic acid may act as part of a permeasesystem at the chloroplast membrane.He has proposed that phosphoglycol-late can permeate the membrane and pass from the chloroplast to the cyto-plasm. Here it may be oxidised to glyoxylate which instead of undergoingfurther metabolism in the cytoplasm might re-enter the chloroplast and againbe reduced and phosphorylated. If an NADP-linked glyoxylate reductasewere present in the chloroplast and an NAD-linked reductase in the cyto-plasm as suggested by Zelitch,15 this mechanism would operate as a transfermechanism allowing the oxidation of NADPH to NADPf in the chloroplast,and the concomitant reduction of NAD+ to NADH in the cytoplasm. Buttand Peel 1 6 have shown that a-hydroxysulphonate inhibits the light-activateduptake of glucose by Chbrelh.These authors suggest that a glycollate/gly-oxylate cycle could be utilised to re-oxidise NADPH permitting a continuouscyclic electron-transport and consequent phosphorylation in the light. Thiswould be consistent with the general view that glucose uptake is dependenton a phosphorylation reaction.Partial pressures of oxygen greater than that in air have been shown toinhibit photosynthesis and this inhibition is most marked a t lower carbondioxide concentrations. Turner and co-workers 1 7 suggested that this in-hibition resulted from an inhibition by oxygen of the enzyme glyceraldehydephosphate dehydrogenase. However, in Chlorelb much of the glycollateformed is excreted from the cell; this represents an irreversible loss ofcarbon from the photosynthetic cycle intermediates.High partial-pressuresof oxygen activate the formation of glycollate and the resulting increasedloss of cycle intermediates lowers the rate of carbon fixation.lsl4 N. E. Tolbert, “Photosynthetic Mechanisms of Green Plants,” National Acad-l6 I. Zelitch, J. Biol. Chem., 1955, 216, 553.l7 J. S. Turner, J. F. Turner, K. D. Shortman, and J. E. King, AwrtraE. J . BioZ. Sci.,l* J. Coombs and C. P. Whittingham, PTOC. Roy. SOC., 1966, By 164, 511.emy Science Publication, 1145, 1963, p. 648.V. S. Butt and M. Peel, Bwchem. J., 1963,88, 31P.1958,11, 336582 BIOLOGICAL CHEMISTRYIn his early work, Hill l9 found that chloroplasts after isolation from theplant showed no reaction with carbon dioxide in the light.Arnon and hisco-workers 2o were the first to show that these particles could carry out thecomplete process of photosynthesis utilising carbon dioxide. Their experi-ments mere made under conditions of low partial-pressure of oxygen with theaddition of a cell extract to replace substances lost from the chloroplastduring isolation. Subsequent work 21 showed that certain preparations ofchloroplasts could fix carbon dioxide at significant rates even in the presenceof atmospheric oxygen. Walker and Hill 22 showed that their preparationalso produced oxygen in shicheiometric amount relative to carbon dioxidetaken up. Most recently, Jenson and Bassham 23 made preparations whichshowed rates of carbon dioxide fixation approaching that of the intact leaf.It has now been shown that the ability to fix carbon dioxide is greatest whenthe bounding membrane of the chloroplast remains intact afGer isolation.If this membrane is broken the ability to fix carbon dioxide can only berestored by the addition of tt cell extract.Both Gibbs and Walker showedwith their preparations an induction period similar to that exhibited byintact cells but that this could be eliminated by the addition of one of thecomponents of the photosynthetic reduction cycle e.g., ribo~e-5-phosphate.~lWalker and Hill 22 showed that phosphoglyceric acid not only could act as ahydrogen acceptor with their preparation but could also replace ribose-5-phosphate as carbon acceptor for carboxylation.Jensen and Bassham 23did not report a similar induction period with their preparations showinghigh rates of fixation and hence it would appear that less material had leakedfrom their chloroplasts during preparation than in those of any other workers.Even so there are striking differences between the range of substancesbecoming radioactive after feeding isolated spinach chloroplasts with 1*CO,as compared with the products in whole spinach leaves. The intermediatesof the carbon reduction cycle account for most of the fixation in the chloro-plasts together with a significant fixation into glycollic acid. There is anotable lack of sucrose in almost all the preparations so far reported.Apparently some enzyme required for sucrose formation has been inactivatedin the preparation of the chloroplasts or it must be concluded either that thisenzymic activity is not associated with the chloroplast particle in vivo, orthat under the conditions of assay other reactions successfully compete forthe precursors of sucrose.Analysis of the Photochemical System into Partial Reactions.-It is nowgenerally believed that two successive photochemical steps separated by athermoohemical reaction are involved in the production of the photochemicalreductant.The first evidence for this came from observations on the rela-tionship between the quantum yield of photosynthesis and the wavelengthof exciting light. In green plants illuminated with wavelengths longer than19 R.Hill, Adv. Enzymol., 1951, 12, 1.8 0 M. €5. Allen, D. I. h o n , J. B. Capindale, F. R. Whatley, and 1;. J. Durham,2 1 M . Gibbs and E. S. Bamberger, PZant Physiol., 1962, 34, lxii; D. A. Walker,22 D. A. Walker and R. Hill, Biochem. Biophys. Acta, (in the press).28 R. G. Jenson, and J. A. Bassham, Proc. Nat. Acud. Sci. U.S.A., 1966, 56, 1095.J . Arnes.. Chem. SOC., 1966, '47, 4149.Plant Physwl., 1965, 40, 1157WHITTINGHAM: RECENT ADVANCES IN PHOTOSYNTHESXS 583680 mp, the observed photosynthetic activity was appreciably less than wasto be expected from the absorption a t this wavelength, whereas at shorterwavelengths the quantum efficiency was found to be approximately constantthroughout the visible region. 24 The decrease in photosynthetic efficiencyoccurred in that region where chlorophyll-a was the only pigment absorbinglight.In red algae where absorption in the red is due to phycocyanin andchlorophyll-a (rather than chlorophyll-b and chlorophyll-a), the decrease inefficiency occurs at wavelengths longer than 650 n ~ p again in a region whereone pigment alone is absorbing light. In all cases it has been shown that thedecrease in efficiency observed with far-red illumination is overcome when thecells are simultaneously illuminated with a second light of shorter wave-length. From such data Emerson 25 and his collaborators concluded thatphotosynthesis involved two photochemical reaction steps sensitised by twoseparate pigments, both of which must be excited for efficient photosynthesis.The inefficiency, which is observed only at the longer wavelengths, arises inthose regions where a single pigment is absorbing radiation and heme onlyone of the two reaction steps is activated.Again studies with fluorescence have shown that a t wavelengths lowerthan 650 mp, whatever the absorbing pigment the fluorescence emitted by theplant is characteristic only of chlorophyll-a.26 It thus appears that energy istransferred from all other pigments to chlorophyll-a.Brown and French 27showed that the red absorption band in green plant cells in vivo was complexand could be analysed only in terms of the presence of three constituentforms of chlorophyll-a with absorption maxima at 670, 682, and 690 mp.Upon extraction of the chlorophyll only a single absorption peak was ob-served and it was suggested that different forms must occur in vivo due eitherto different states of aggregation of the pigment molecules or to complexformation with other cellular constituents such as proteins or lipids.Thetwo forms of chlorophyll-a absorbing at longer wavelengths, must be theforms which by themselves are ineffective in photosynthesis and also have alower efficiency for fluorescence. After excitation a t shorter wavelengthsenergy transfer is considered to take place only to the form of chlorophyll-awith an absorption maxima at 670 mp and all pigments which do so areconsidered to belong to a single pigment system, system 11. System Iconsists of the forms of chlorophyll-a with absorption maxima at 680 and690 mp.It is not clear whether energy transfer can take place betweensystem I1 and system I, but it has not so far been proven to take place.Each pigment system is considered to photosensitise only one of the twophotochemical reaction steps.Certain intermediates, which change their absorption spectrum as aconsequence of oxidation and reduction, e.g., two forms of cytochrome,cytochrome-b, and -f, are thought to participate in a thermochemical reaction24 R. Emerson, R. V. Chalmers, and C. N. Cederstrand, Proc. Nut. Acad. Sci. U.S.A.,2s R. Emerson, Ann. Rev. Plant Physiol., 1958, 9, 1.26 L. N. M. Duysens, “Transfer of excitation energy in photosynthesis,” Ph.D.1957, 43, 133.Thesis, University of Utrecht, The Netherlands, 1952.J.8. Brown, and C. S. French, Phnt Phy&oZ., 1959, 34, 305584 BIOLOGICAL CHEMISTRYbetween the two Iight-absorbing steps. This was first suggested by Duysens,S*who showed that in a red alga, Porphyridium, light quanta absorbed mainlyby pigment system I1 resulted in a reduction of both cytochromes whereaslight absorbed largely by system I (excitation at longer wavelengths than690 mp) resulted in their oxidation. Pigment system I is also considered tocontain a component referred to as P700, whose chemical identity has not yetbeen characterised. It shows reversible absorption changes in a mannersimilar to the cytochromes, undergoing reversible oxidation and reductionduring photosynthesis. It has a potential near to that of cytochrome-f and itis frequently considered that this substance is the primary reamgent for photo-reaction I.Photoreaction I1 must then produce reduced cytochrome-b withan oxidation reduction (E,,’) near 0, and it must do so utilising water as theelectron donor (Eo’ = + O ~ V ) . It has been suggested that the primaryreductant is probably plastoquinone and that cytochrome-b, and plasto-cyanin subsequently react together. A mechanism first proposed by Hill andBendall 29 in 1960 suggested the working hypothesis for most of these in-vestigations and is shown in Figure 3. The reactions dependant on photo-system I1 which are concerned in the production of oxygen fromstill largely unknown.water are0PQcyt fP 700+04Light +06LightFIGURE 3 The Hill Bendall scheme of photosynthesis.The oxidation reduction potentials(standard, p H 7 ) of some p08Sibk intemnediates is shown on the right.cyt-cytochmme ; Fd-femedoxin ; P N - p y v i d i n e nzccleotide (NA D P ) ; PQ-ptaSto-quinone ; P700-possible intemnediate whose chemical identity is not yet k m .28 L. N. M. Duysens, Ann. Rev. Plant PhySiol., 1956, 7, 26.29 R. Hill and F. Bendall, Nature, 1960, 186, 136WHITTINCHAM : RECENT ADVANCES IN PHOTOSYNTHESIS 585Photoreaction step I may be isolated from step I1 by the use of actiniclight confined to the far-red region of the spectrum. Alternatively, certainspecific poisons which interfere with the electron-transfer reactions e.g.,p = chlorophenyl dimethylurea may be added or physical treatments such asheating to 55"c which modify the biophysical structure.Detailed studies of the change in the absorption spectra during photo-synthesis for a range of green plants have indicated the existence of additionalcomponents whose oxido-reduction state undergoes changes during the reac-tion.Certain of these changes, e.g., the decreased absorption at 478 mp andthe increase at 515 mp, have not been clearly assigned to any chemicallyidentified intermediate. Witt and his collaborators 3* considered that thesechanges were related to those which occur at 648 mp and believe that all threeindicate the participation of chlorophyll-b as an intermediate in the reactionsequence.Attempts have been made to physically separate Werent types of particlefrom chloroplast preparations in the hope of obtaining some fractionsrelatively richer in one photosystem than the other.Boardman and Ander-son 31 separated by differential centrifugation, a fragmented spinach chloro-plast after treatment with digitonin. They prepared fractions with variablechlorophyll a / b ratios. The heavier fraction had an a/b ratio near 2 and wasthought by them to contain largely the pigment system 11, whereas a lightfraction with an a/b ratio approaching 6, was thought to largely consist ofphotosystem I. Subsequently, a number of workers have made similarpreparations using a range of detergents and sonication to effect variousdegrees of fragmentation, and it now seems more likely that there is nosimple relationship between particle size and type of photochemical activity.Phosphoryhtion in Relation to the Photochemical Reactions.-For manyyears it has been known that chloroplasts isolated from leaves of higherplants catalyse a photochemical oxido-reduction reaction between variousdyestuffs and water (Hill reaction 32).Later it was shown that after additionof ADP, magnesium, and phosphate in substrate amounts, ATP was formedin the light. The phosphorylation could take place under two conditions,either accompanied by a coupled oxido-reduction such that the phosphoryla-tion was stoicheiometrically related to the transfer of electrons (noncyclicphosphorylation) or in a process in which there was no net electron-transfer(cyclic phosphorylation). It was postulated that in both types of phosphory-lation the process must be related to the release of free energy in a processof electron transport.The oxido-reduction sequence would commence withthe formation consequent upon absorption of radiation of a chlorophyll statewhich could act as an electron donor, If ultimately at the end of the sequencean electron was returned to the chlorophyll radical, returning it to the groundstate, the electron flow would be cyclic and the only result of the reactionwould be the formation of a phosphate bond. The cyclic process was shownto be catalysed by a number of substances such as phenazine methosulphateThe evidence for this is not unequivocal.30 H:T. Witt, B. Rumberg, P. Schmidt-Mende, V. Sigel, B. Skern, J. Vater, and81 N.K. Boardman and J. M. Anderson, Natwre, 1964, 203, 166.8a R. Hill and R. Scarisbrick, Proc. Roy. Soc., 1940, €3, 129, 238.J. Wickard, Angew. Chem., 1965, 77, 821586 BIOLOGICAL CHEMISTRYand menadione, which were capable of acting as oxido-reduction carriers. Inthe second type of phosphorylation (noncyclic) the electrons were obtainedfrom water and transferred to an added reagent e.g., ferricyanide or quinone.Oxygen was produced and the added substance reduced to ferrocyanide andhydroquinone respectively.A class of substances have now been isolated from photosyntheticorganisms which are thought to act as the primary reductants for thephotochemical process in vivo. These substances were first isolated from thenonphoOosynthetic bacterium CZostridium pmteuriunum, but later fromvarious photosynthetic tissues including algae, bacteria, and the leaves ofhigher plants.33 This class of substances called ferredoxins, are neither haemnor flavin proteins, but contain both iron and flavin.They also containinorganic sulphur equimolar with iron. The sulphur is liberated as hydrogensulphide upon acidification, and its removal is accompanied by the loss ofiron. The ferredosin obtained from bacteria, has a molecular weight of about6000 and that from leaves of higher plants, 13,000. The ferredoxin ofspinach leaves has two iron atoms per inole, that from Chromatiurn three, andthat from Clostridiurn seven. Perredoxins are capable of oxido-reduction andhave a potential of between 400450 mv at gH 7.5, i.e., close to the potentialof hydrogen gas. Spinach ferredoxin transfers one electron per mole whereasthe bacterial ferredoxin transfers two electrons per mole.In the oxidised form the ferredoxins show characteristic absorptionbands, although the maximum occurs at different wavelengths accordingto the source of the ferredoxin.Those obtained from algae or chloroplasts ofhigher plants have absorption maxime near 463 and 420, 325 and 274 mp,whereas that from bacteria, whether photosynthetic or not, has a broad peaka t 385 mp, a shoulder at 300 mp, and a smaller peak a t 280 mp. It is con-sidered that ferredoxin is probably the primary electron acceptor for thewhole photochemical reaction sequence and the e.s.r. signal at g = 1.94 isconsidered characteristic of protein- bound reduced non-haem iron. 34 Thereduced ferredoxin is then re-oxidised by NADP to form reduced NADPHin a reaction catalysed by a flavoprotein enzyme, ferredoxin/NADP reduc-tase. The coenzyme then reduces phosphoglyceric acid to phosphoglyceral-dehyde in the carbon cycle.With preparations of “ broken ” chloroplasts un-able to utilise carbon dioxide an intermediate will react with reagents such asferricyanide, quinone, or dyestuffs. The reductase shows high specificity forNADP as distinct from NAD, thus accounting for the specificity of the photo-chemical system of the chloroplast.It has been shown by the use of monochromatic light of wavelengthlonger than 700 mp that excitation of photosystem I alone can catalyse thocyclic type of photophosphorylation.Phosphorylation can also occur in thethermal reaction believed to occur between photosystem I and photosystem11.The mechanism by which phosphorylation is related to electron flow isfar from clear both in chloroplasts and in mitochondria. One view has been53 D. I. Arnon, Experientia, 1966, 22, 273.84 J. F. Gibson, D. 0. Hall, J. H. M. Thornley, andF. R. Whatley, Proc. Nat. Acad.Sci. U.S.A., 1966, 56, 987WHITTINCHAM : RECENT ADVANCES IN PHOTOSYNTHESIS 587to suggest that an intermediate (I) forms a complex (I-R) with a reducedradical (R). This complex then reacts with a further intermediate B andwhen an electron is transferred between the two radicals an I - B complex isformed with a high-energy bond.The complex I cv B is postulated to bestable in the organelle but to react with inorganic phosphate and ADP togive ATP with the formation of free I. Direct evidence for the existence ofan intermediate I with the ability to form complexes is lacking, but thehypothesis has proved useful in the interpretation of kinetic data.An alternative hypothesis emphasises the significance of the structuralfeatures of the bounding membranes of the chloroplast as three-dimensionalstructures with an " inner " and " outer " surface. The flow of electronsthrough a sequence of carriers within the membrane is considered to beobligatorily coupled to a movement of hydrogen ions from " outside " to" inside " the structure.35 Protons can be obtained only from the outsideand following along the electron-transport chain are released only to theinner space.As a consequence, the electrochemical activity of hydrogen ionsin the internal space will rise above that in the surrounding medium. Oncethis proton gradient has been established it represents a form of potentialenergy which can be considered as the high-energy intermediate (I - B)and which, by reaction with ADP and inorganic phosphate, can give rise toATP.Neumann and Jagendorf 36 showed that chloroplasts in light catalysed anelectron-transport process which results in hydrogen uptake from the externalmedium. This was measured by monitoring the pH of an unbuffered chloro-plast suspension. The light-induced pH rise was inhibited by uncouplers tothe same extent that they inhibited phosphorylation.Neumann and Jagen-dorf 36 calculated that the pH inside the chloroplast in the absence of anybuffering would be lowered to a value of 2.5. Upon illumination, chloroplastsshow an increased scattering of light and this is probably related to thelowering of internal pH. The kinetics of both processes are nearly the same,and the pH activation curves show a similar optimum. However, a change inexternal pH in the dark produces a smaller increase in light-scattering thandoes light and probably changes in refractive index also result from illumina-tion. As a consequence of the process of proton accumulation and of dis-charge of the protons in phosphorylation, multiple layers of electron-densesheets and proton-dense sheets may be formed in the chloroplasts uponillumination.This will result in changes in the refractive index of theseareas as compared with dark conditions and this change in birefringence ispresumably a significant factor in the light-scattering response. A chloro-plast suspension also shows a decrease of apparent viscosity, a change in thesize distribution of particles, and a change in optical density upon illumina-tion.The decrease in internal pH cannot wholly explain the observed change inchloroplast volume. The decrease in volume is only measurable when thechloroplasts are suspended in a weak anion such as phosphate whereas in asalt solution such as sodium chloride, a pronounced light-dependent swellings6 P.Mitchell, Nature, 1961, 191, 144.s6 J. Neumann and A. T. Jagendorf. Arch. Biochem. Biophys., 1964, 107, 109588 BIOLOGICAL CHEMISTRYis observed. Modifications of the changes induced in chloroplasts by lightdue to ionic interference have been investigated by Packer and Siegen-thaler.37 Again various organic anions, such as phosphate and arsenate,greatly stimulated increase in the light scattering observed upon illumi-nating chloroplasts. It is suggested that loss of an anion from within thechloroplast resulks from the displacement of the equilibria of undissociatedacid across the membrane following the acidification of the interior of thechloroplast.The '' chemi-osmotic " hypothesis requires that the chloroplast mem-branes should be relatively impermeable to ions.Dilley and Vernon 38showed that when chloroplasts were suspended in tris buffer and illuminated,a release of potassium ions but an uptake of sodium ions took place. Dilleyand Vernon assumed that the ions were lost from the chloroplast to com-pensate for the increase in hydrogen ions. However, the kinetics of potassiumion movement bear little relationship to the kinetics of the pH changeobserved in the medium. Moreover the amount of ion lost is too small by afactor of 10 to account osmotically for the observed volume change referredto in the foregoing discussion. Hence, it is unlikely that the movement ofpotassium can be explained simply in terms of charge equilibration.Catalysis of the synthesis of ATP by chloroplasts in light is dependenton a light-driven oxido-reduction reaction.Jagendorf and Smith 89 foundthat if chloroplasts were treated with dilute solutions of EDTA the prepara-tion, upon re-suspension in water, could catalyse oxido-reduction reactionsbut not phosphorylation. If the EDTA extract were added together withmagnesium ions, phosphorylation was restored. The extract was consideredto contain some essential intermediates (" coupling " factors) required to linkelectron flow with phosphorylation. Again Vambutas and Racker 4Oshowed that chloroplasts after treatment with trypsin had a lowered capacityfor phosphorylation. Nevertheless, the light-induced pH rise and the light-induced change in optical properties was unchanged. This again suggested aseparation between the process of formation of some high-energy intermediateor high-energy state and the utilisation of this state for the condensation ofphosphate.It may be noted that at the present time, similar attempts toseparate phosphorylation and electron flow in mitochondria have so far notbeen successful. The reason for this is still not clear.It follows that generation within the chloroplast of electromotive forceby any means other than light should result in a state potentially capable ofgenerating ATP. In agreement with this Hind and Jagendorf 41 showed thatif chloroplasts are equilibrated at an acid pH in the dark with ADP andinorganic phosphate and the pH was then rapidly raised, phosphorylationresulted. Moreover the maximum synthesis of ATP is stoicheiometricallyrelated to the number of protons passing through the membrane during thepH equilibration.Jagendorf and Uribe 42 confirmed that the change in pH57 L. Packer and P. A. Siegenthaler, Plant PhyslsioZ., 1965, 40, 785.88 R. Dilley and L. Vernon, Arch. Biochem. Biophys., 1965, 111, 365.8s A. T. Jagendorf and M. Smith, Plant Physiol., 1962, 37, 135.4 o B . K. Vambutas and E. Racker, J. Bid. Chem., 1965, 240, 2660.4 1 G. Hind and A. T. Jagendorf. J . Biol. Chem., 1964, 240, 3195.4% A. T. Jagendorf and E. Uribe, Proc. Nat. A d . Sci. U.S.A., 1966, 55, 170WHITTINGHAM: RECENT ADVANCES IN PHOTOSYNTHESIS 589was more important than the absolute value of the initial or k a l pH. Theproduction of ATP can be so large as to make it extremely unlikely that thephosphorylation could be related to the presence of some chemical inter-mediate whose state is changed by the change in pH.This acid-base darkphosphorylation is specifically inhibited by serum prepared from the couplingfactor mentioned previously. The dark phosphorylation is also uncoupled bysuch proton conducting reagents as nitrophenols or carbonyl cyanide m-chlorophenylhydrazone.When spinach chloroplasts are rapidly changed from an acidic to a basicsuspension medium, Mayne and Clayton 43 observed that they emit for a brieftime chlorophyll fluorescence. It is considered that chlorophyll has beenexcited to the singlet state at the expense of some high-energy state.B. C. Mayne and B. K. Clayton, Proc. Nat.Acad. Sci. U.S.A., 1966, 55, 4943. THE STRUCTURE AND METABOLISM OF CfLUCANS*By D. J. Manners(Heriot - Watt University, Edinburgh I )ALTHOUGH various aspects of the biochemistry of glucans have beendescribed in previous R~ports,l-~ the substantial progress which has beenmade during the last few years now merits a more comprehensive review.Particular emphasis will be given to the use of enzymes in the structuralanalysis of glucans, and to the r81e of nucleoside diphosphate sugars inbiosynthesis. The earlier Papers dealing with sugar nucleotides have beenreviewed elsewhere so that the present account will be largely restricted towork published since 1962.The Molecular Struck6 of Glucans.-The main structural features ofmost naturally occurring polymers of D-glucose were established during theperiod 1935-1960 by the application of methylation, periodate oxidation,and partial acid hydrolysis techniques.In the case of starch-type poly-saccharides, enzymic degradation methods were also invaluable. In recentyears, there have been significant improvements in the methylation method,following the development of improved conditions for etherification,5 andthe application of g.1.c. for the separation of methylated sugars.6 The valueof periodate oxidation analyses has been increased by the development ofthe Smith degradatlion method (involving the sequence of reactions-periodate oxidation, borohydride reduction, and hydrolysis of hemiacetallinkages with dilute acid), which provides an alternative method of end-groupanalysis and a means of detecting and locating the position of periodate-resistant glucose residues.Full experimental details of these and relatedchemical methods of analysis are now available.8in studying the fine structure of clam glycogen. The average chain-lengthof six samples determined by periodate oxidation was 12-13 glucoseA combination of these chemical methods was used by Bahl and SmithD. J. Bell, Ann. Reports, 1947, 44,, 217.D. H. Hutson and D. J. Manners, Ann. Reports, 1964, 61, 429.E. F. Neufeld and W. Z. Hassid, Adv. Carbohydrate Chena., 1963, 18, 309; L. F.Proceedings, Plenary Sessions Sixth Internat. Congress, Biochemistry,” I.U.B.,K. Wallenfels, G. Bechtler, R. Kuhn, H. Trischmann, and H. Egge, Angew.Chem.F. Smith and R. Montgomery, “ Chemistry of Plant Gums and Mucilages,”“Methods in Carbohydrate Chemistry,” vol. 5, ed. R. L. Whistler, Academic0. P. Bahl and F. Smith, J. Org. Chem., 1966, 31, 1479.a D. J. Manners, Ann. Reports, 1954, 50, 288.Leloir,1964, 33, 15.Internat. Edn., 1963, 2, 515; S. Hakomori, J. Biochent. Japan, 1964, 55, 205.Reinhold, New York, 1959, p. 377.Press, New York and London, 1965.6 G. 0. Aspinall, J. Chem. SOC., 1963, 1676.* The following abbreviations are used: D P = degree of polymerisation (Le., numberof monosaccharide residues per molecule); ADPG, GDPG, IDPG, TDPG, UDPG - adenosine, guanosine, inosine, thymidine and uridine diphosphate n-glucose respectively;ADP, GDP etc. = the diphosphates of the above nucleoaides; AMP = adenosine B’-phos-phate (adenylic acid)MANNERS : STRUCTURE AND METABOLISM O F QLUCAN s 591residues.The result for one specimen from Anodonta grandis was confirmedby methylation analysis of both the glycogen and the derived polyalcohol.Partial acid hydrolysis methods continue to be widely used, although incertain experiments, it may be difficult to assess the significance of the prc-sence of trace amounts (0.3% or less) of oligosaccharides. Acid reversion fromglucose is known to give rise to small amounts of various oligosaccharides,l*but the acid-catalysed transfer of glucosyl residues a t the disaccharide levelmay be a more serious source of artefacts.ll The two processes are differ-entiated by the fact that acid reversion yields both a- and #Clinked disac-charides whereas the transfer reactions proceed with retention of the anomericconfiguration.For example, when dilute solutions of maltose (0.4%) areheated mith 0.1N-hydrochloric acid, significant amounts of isomaltose andnigerose are formed, whilst isomaltose on similar treatment gives smallquantities of maltose and isomaltotriose. l1 The presence of minute quan-tities of nigerose and isomaltotriose in partial acid hydrolysates of glycogen l2is not, therefore, structurally significant unless it can be independentlyconfirmed by an alternative analytical method. It should be noted thatneither periodate oxidation nor enzymic degradation studies provideevidence of 1,3-linked glucose residues in glycogen. The presence of 1,3-glucosidic linkages in the amylopectin component of starch is similarly con-sidered to be ~ n l i k e 1 y .l ~ ~ ~ ~The majorpart of a recent colloquium on " The contribution of enzymes to the struc-tural analysis of glycogen and starch " was devoted to studies involvingp~llulanase.~~ This enzyme is produced extracellularly by various strainsof Aerobacter aerogenes and hydrolyses pullulan, a polymer of 1,6-linkeda-maltotriosc units, quantitatively to maltotriose. In addition, it hydrolysesthe outermost inter-chain linkages in both amylopectin and glycogen, and thect-l,6-glucosidic linkages in many branched oligosaccharide a-limit dextrins.15Pullulanase will not hydrolyse the cc-1,6-2inkage attaching a single glucoseresidue to a chain of a-1,4-linked glucose residues. The smallest substrate isthe tetrasaccharide 62-cc-maltosyl-maltose.In view of the fact that pullu-lanase has a wider specificity than any other " debranching enzyme " (Tablel), and can be readily prepared, it provides the most valuable enzyme so fardiscovered for the analysis of starch-type polysaccharides.The first pullulanase preparations were acetone precipitates of cell-freefiltrates of A . aerogenes and had a low specific activity.15 Considerablepurification has been achieved by gel filtration on Sephadex G-200, and thishas revealed the existence of two forms of the enzyme with molecular weightsEnzymic Studies on Starch and Glycogen.-(a) Pullulanase.lo A. Thompson, K. Anno, M. L. Wolfrom, and M.Inatome, J. Amer. Chem. Soc.,1954,76,1309; S . Peat, W. J. Whelm, T. E. Edwards, and 0. Owen, J. Chem. SOC., 1958,586.l1 D. J. Manners, G. A. Mercer, and J. J. M. ROW^, J . Chem. Soc., 1965, 2150.la M. L. Wolfrom and A. Thompson, J. Amm. Chem. Soc., 1957, 79, 4212.lS D. J. Manners and G. A. Mercer, J . Chem. SOC., 1963, 4317; 0. P. Bahl and F.l* W. J. Whelan, Biochem. J., 1966, 100, 1P.ISM. Abdullah, B. J. Catley, E. Y. C. Lee, J. Robyt, K. Wallenfels, and W. J.Smith, J. Org. Chem., 1966, 31, 2916.Whelm, Cereal Chem., 1966, 45, 111592 BIOLOGICAL CHEMISTRYTABLE 1 Specificity of debranching enzymesGlycogenphosphory-Amylo- lase limit a-Limitpectin Glycogen dextrint dextrin Pullulani- +-* - - - - - - R-Enzyme + Amylo-1,6-glucosidase~ -+ + Isoamylase + Limit dextrinase -Pullulanase + + + + +- -I+ -* Glycogens of normal chain length (10-14 glucose residues) are not attacked by R-enzyme.The relativelyuncommon glycogens with chain lengths of about 18 glucose residues are slowly hydrolysed (ref. 35).7 The phosphorylase limit dextrin, with side-chains of four glucose residues, is attacked by pullulanaseand by isottmylase, to give maltotetraose. It is converted into a modified limit dextrin which has side-chaimof single glucose residues by the transferase. These side-chains are no longer susceptible to pullulanase, butare hydrolysed by amylo-l,6-glucosidase, to give glucose.$ All preparations show oligo-1,Pjl,4-glucantransferase activity.of about 150,000 and 50,000.16 On storage a t either 0" or 30", the heavierform was converted into the lighter form.A characteristic property of unpurified preparations of pullulanase isrelative thermostability,l7 and 30% of the activity will withstand heating at100" and pH 7 for 10 min.On heating at other pH-values, there is some lossof activity which is recovered on storage a t room temperature.Although pullulanase preparations are readily obtainable, many of theseare contaminated to a varying extent with traces of or-amylase. This factdoes not appear to have been reported in the literature. For some purposes,e.g., the structural analysis of branched oligosaccharides, this impurity isnot important. In other experiments, e.g., in examining the possible actionof pullulanase on amylose as a means of detecting branch points, the con-taminant could invalidate the observations.It is therefore essential that allpreparations of pullulanase (or any other debranching enzyme) be rigorouslytested for or-amylase using a sensitive method.Pullulanase is released from the cells of A. aerogenes during the logarithmicphase of growth on either maltose, maltotriose or pullulan. However, whenthe organism is grown in continuous culture on a mixture of 0.4% maltoseand 0.4% glucose, the enzyme is bound to the cells.18 It can be released byshaking the cells with various detergents, and can be purified by adsorptionon DEAE-cellulose and fiactionation with ammonium sulphate. The puri-fied enzyme, which has recently been crystallised,lg has a molecular weightof 145,000.Pullulanase, like the other debranching enzymes, is unable to hydrolyseall the inter-chain linkages in glycogen. About one-half of these linkagesare situated on the periphery of the molecule, and their hydrolysis releasesthe A-chains.These A-chains are side chains linked to the molecule onlyby the reducing group, in contrast to B-chains (main chains), which are notonly linked by the reducing group but also have other chains attached to16 B. M. Frantz, E. Y. C. Lee, and W. J. Whelan, Biochem. J., 1966, 100, 7P.1 7 M. Abdullah, B. J. Catley, W. F. J. Cuthbertson, and W. J. Whelan, Biochem. J.,1aK. Wallenfels, H. Bender, and J. R. Rached, Biochem. Biophys. Res. Cmm.,19 K. Wallenfels and J. R. Rached, Biochem. Z., 1966,344, 624.1966,100, 8P.1966, 22, 254MANNERS: STRUCTURE AND METABOLISM O F GLUCANS 593them.A proportion of the inter-chain linkages which are exposed by thisaction may be slowly hydrolysed, but the enzyme, presumably for stericreasons, is unable to penetrate into the interior of the molecule. This isshown by the limited increase in #?-amylolysis limit (from 48 to 56%) whichfollows pullulanase action on glycogen.l5 With amylopectin, where thedegree of branching in the interior of the molecule is only one-half that ofglycogen, pullulanase is able to penetrate to a considerable extent, as shownby the relatively large increase in #?-amylolysis limit (from 52 to 92%)following incubation with pullulanase. If access to the interior of the p l y -saccharide is facilitated either by pre-treatment with a-amylase, or byallowing #?-amylase to act concurrently, all the inner inter-chain linkages arehydrolysed. l5The average chain-length of glycogen and amylopectin may be measuredon a micro scale by the combined action of p-a.mylase and pullulanase.20 Thissimultaneous action results in the hydrolysis of all the inter-chain linkages,and the complete degradation of the individual chains.Those chains con-taining an even number of glucose residues give 100% conversion into maltose.Since /3-amylase will hydrolyse maltotriose into maltose and glucose, chainscontaining an odd number of glucose residues will give maltose and onemolecular proportion of glucose. I n a random structure there are equalnumbers of the two types of chains, so that one molecule of glucose arisesfrom every two chains.Determination of glucose, by glucose oxidase, thusgives the average chain-length. The method was tested against severalglycogens and amylopectins whose average chain-length had been deter-mined by periodate oxidation, and the enzymic results were in good agree-ment. In view of the small amounts (ca. 1 mg.) of glycogen required forenzymic assay, the method will be extremely valuable for the analysis ofglycogen from biopsy samples from patients suspected of suffering fromglycogen storage disease.The stepwise action of #?-amylase and pullulanase has been used to con-firm the multiply branched nature of glycogen and amylopectim21 Acharacteristic feature of the Meyer-type structure is the presence of approxi-mately equal numbers of A- and B-chains.The former can be selectivelydetermined by converting the polysaccharide into the 8-limit dextrin andthen treating with pullulanase. A-Chains, which have been shortened totwo or three glucose residues by #?-amylolysis, are released as maltose ormaltotriose, and can be determined by quantitative paper chromatography.With ten glycogens, the yield of maltose and maltotriose was equal to, orapproached, that calculated for molecules containing equal numbers of A- andB-chains. Potato amylopectin and waxy sorghum starch also contained asimilarly high proportion of A-chains ; glycogens do not therefore differsignificantly from amylopectins in degree of multiple branching.Since the initial attack of pullulanase involves the hydrolysis of thosea-1,6-glucosidic linkages which attach A-chains to the molecule, it provides ameans of examining the average size and distribution in length of the A-chains. This type of analysis has been applied to the phosphorylase limitE.Y. C. Lee and W. J. Whelm, Arch. Biochem. Biophya., 1966,116, 162.a1 G. N. Bathgate and D. J. Manners, Biochem., J . 1966, 101, 3C594 BIOLOGICAL CHEMISTRYdextrin (4-dextrin) of glycogen,22 and to polg saccharides synthesised fromglycogens or 4-dextrins by the action of UDPG a-glucan transglucosylaseand UDPG 22 or of muscle phosphorylase 23 and glucose l-phosphate.Pullulanase has also been used to study the action of liver branching enzymeon the outer chains of a modified glycogen and amylopectin (see p.603).24When 4-dextrin is incubated with pullulanase, maltotetraose (84%) is themajor product, with minor quantities (2-5y0) of maltose, maltotriose, malto-pentaose, and maltoheptaose. This result confirms the Walker-Whelanstructure 25 for 4-dextrin in which the A-chains (and the outer parts ofthe B-chains) contain four glucose residues.Incubation of potato and wheat amylose with pullulanase caused a,significant increase in ,&amylolysis limit and a decrease in limiting viscositynumber.26 These results are similar to those obt,ained previously 27 whenisoamylase was shown to exert a " debranching " action on potato and oatamylose, and provide additional evidence for the view that certain samplesof amylose are slightly branched and contain a small proportion of a-1,6-glucosidic inter-chain linkages which prevent complete p-amylolysis.Thebranching could arise from the limited action of Q-enzyme on linear chains ofa-l,4-linked glucose residues.27Other examples of the use of pullulanase include studies on (a) thecharacterisation of the oligosaccharides produced by the action of UDPG a-glucan transglucosylase and UDPG on a-limit dextrins,22 (b) the identificationof a small proportion (6.6%) of 1,G-linked a-maltotetraose units withinthe pullulan and (c) the improved preparation of 63-a-glucosylmaltotetraose and a-glucosylcyclomaltohexaose from amyl~pectin.~~ How-ever, the recent demonstration 3* that pullulanase action may be reversible,and may cause the partial conversion of high concentrations of maltose into atetrasaccharide (and of maltotriose into a hexasaccharide) means thatcaution is necessary in assessing the structural significance of minor amountsof higher oligosaccharides.(b) Other debranching enzymes.The use of R-enzyme (an amylopectin-debranching enzyme originally isolated from broad beans) in structuralstudies is now limited.31 This enzyme hydrolyses the outermost inter-chainlinkages in amylopectin and amylopectin p-de~trin,~Z but has no action onmost samples of glycogen.33 The first preparations of R-enzyme alsohydrolysed a-l,6-glucosidic linkages in a-limit dextrins, but this activity isnow known to be due to a separate limit dextrinase.The R-enzyme and limit82 D. H. Brown, B. Illingworth, and R. Kornfeld, Biochemistry, 1966, 4, 486.9s D. H. Brown, B. I. Brown, and C. F. Cori, Arch. Biochem. Biophys., 1966,116,24 W. Verhue and H. G. Hers, Biochem. J., 1966, 99,222.25 G. J. Walker and W. J. Whelan, Biochem. J., 1960, 76, 264.25 W. Banks and C. T. Greenwood, Arch. Bwchern. Biophys., 1966,117, 674.27 0. Kjolberg and D. J. Manners, Biochem. J., 1963, 86, 258. ** B. J. Catley, J. F. Robyt, and W. J. Whelan, Biochem. J., 1966, 100, 6P.39 J. R. Stark, Biochern. J . , 1967, 102, 27P.80 M. Abdullah and D. French, Nature, 1966, 210, 200.$1 H. G. Hers and W. Verhue, Bwchern. J . , 1966, 100, 3P.88 P. N. Hobson, W. J. Whelan, and S. Peat, J . Chem. SOC., 1951, 1461.83 S.Peat, W. J. Whelan, P. N. Hobson, and G. J. Thomas, J. Chem. SOC., 1954,479.4440MANNERS: STRUCTURE AND METABOLISM O F GLUCANS 595dextrinase activities of malted barley have been separated by columnchromatography 34 and by continuous electroph~resis.~~Studies with R-enzyme included the hydrolysis of amylopectin p-dextrinto give a high yield of maltose and maltotriose in accord with a multiplybranched structure,36 and the release of maltotetraose from amylopectin 4-dextrin,z5 thus establishing for the first time the presence of side-chains offour glucose residues.The hydrolysis of polysaccharides by yeast isoamylase has also yieldednew structural information, although the enzyme is not as stable or as readilyprepared as pullulanase.37 In addition to the studies on amylose alreadycited,27 it has been used to characterise the inter-chain linkages in numerousalgal and protozoal starches, and to examine glycogens from cases of glyco-gen storage disease.38(c) 18- and a-Amylase.The use of the amylases in the structural analysisof glycogens and starch has been reviewed re~ently.3~9 40 In using /I-amylo-lysis for the measurement of exterior chain-lengths, it now seems probablethat this property is more correctly given by (n + 2) rather than (n + 2.5)where n, is the number of glucose residues removed by p-arnyla~e.~~With certain concentrations of enzyme and substrate, there is a linearrelationship between the extent of a-amylolysis of glycogen (expressed asapparent percentage conversion into maltose) and the degree of branching.41This observation provides an alternative method for determination of averagechain-length which is also applicable to the analysis of 1-2 mg. samplesof glycogen from cases of glycogen storage di~ease.4~The isolation of oligosaccharides of DP 9-13 containing more than onea- 1,6-glucosidic linkage from a-amylolytic digests of starch or glycogenrepresents further evidence of multiple branching.40 The detailed structureof these oligosaccharides may be related to the fine structure of the interiorof the polysaccharide.Studies on Other a-G1ugans.-The dextrans are a group of bacterial glucanscontaining chains of a-1,6-linked D-glUCOSe residues, with varying degrees ofbranching, and with a- 1,4- and/or a- 1,3-g lucosidic inter-chain linkages.The dextran synthesised by Leuconostoc mesenteroides NRRL B- 1375(Betacoccus arabinosaceous, Birmingham strain) has been extensively studiedby E.J. Bourne and co-~orkers,~~ and shown, hy enzymic degradationanalysis, to contain a substantial number of side-chains consisting of singleglucose residues attached by a 1,3-linkage to a main chain of a-1,6-linkedresidues.Recent studies 44 have shown that the dextran synthesised by L. w e n -34 I. C. MacWilliam and G. Harris, Arch. Biochem. Biophys., 1959, 84, 442.35 D. J. Manners and K. L. Sparra, J . I m t . Brewing, 1966, 72, 360.36 S. Peat, W. J. Whelan, and G. J. Thomas, J . Chem. SOC., 1956, 3025.s7 Z. H. Gunja, D. J. Manners, and K. Maung, Biochem.J., 1961, 81, 392.38 0. Kjolberg, Ph.D. Thesis, University of Edinburgh, 1962.40 D. French, Biochem. J., 1966, 100, 2P.41 D. J. Manners and A. Wright, J. Chem. Soc., 1962, 1597.43 E. J. Bourne, D. H. Hutson, and H. Weigel, Biochem. J., 1963, 86, 556; D. H.O4 D. Abbott, and H. Weigel, J . Chern. SOC. (C), 1966, 816.D. J. Manners, Biochem. J., 1966, 100, 2P.0. Kjolberg and D. J. Manners, J . Chem. SOC., 1962, 4596.Hutson and H. Weigel, ibid., 1963, 88, 588596 BIOLOQIOAL OHEMISTRYteroides NRRL B-1415 has a branched structure, with about 14% of ~ - 1 ~ 4 -glucosidic inter-chain linkages. The dextran of L. mesenteroides NRRLB-1416 also has a branched structure, with an average repeating unit of sixglucose residues, and contains both a-1,3- and a-lY4-inter-chain linkages.Since catalytic oxidation and partial acid hydrolysis of the dextran fromstrain B- 1415 gave 4-O-(a-D-glUCOpyranOSyl~O~C acid)-D-glucose, andglucamylase liberated D-glucose from the dextran, it was concluded 45 thatmost, if not all, of the side chains consisted of single a-1,4-linked glucoseresidues. This conclusion has been supported 46 by the isolation of a homo-logous series of branched oligosaccharides based on isomaltose, but containinga single a-lY4-linked residue, from the enzymic degradation of this dextran.The catalytic oxidation method was also applied to the dextran from strainB-1375, and the isolation 45 of 3-O-(a-D-glUCOp~anOSyl~O~C acid)-D-glucose confirms the earlier suggestion43 of side chains of single glucoseresidues. The isolation of l-O-a-isomaltosylglycerol on Smith degradation ofthis glucan indicates that many of the side-chains are attached to twoadjacent a- lY6-linked glucose residues.46The Smith degradation method provides a useful means of determiningthe sequence of linkages in a glucan.Isolichenin is a linear polymer con-taining a-1,3- and a-1,4-glucosidic linkages in the relative proportion ofampproximately 3 : Z.47 On degradation by the Smith procedure, the majorproducts were a-glucosylerythritol (38%) and nigerosylerythritol (43%)with possibly 5% of nigerotriosylerythrito1.4* It follows that isolicheninconsists mainly of sequences of either single or pairs of a-1,3-linked D-glucoseresidues which are flanked on each side by a-lY4-linked residues.Since theyield of higher oligosaccharides was so low, sequences of three or moreadjacent a-lY3-linked residues are unlikely to occur.Structure of P-Glucm.-Several p-glucans have been studied recently,and in some instances, the new results have led to a revision of the generallyaccepted structures.Recent studiesshowed that both the soluble 50 and insoluble 51 forms of laminarin consist ofbranched chains of /3-1,3-linked D-glucose residues; some of the chains areterminated at the reducing end by a, mannitol residue which is monosub-stituted,52 and not disubstituted as previously ~uggested.5~ The branchpoints are /%glucose residues linked through C-1, C-3, and C-6 and the essen-tial difference between the two forms of laminarin is in the degree of branch-ing.60 For example, six samples of insoluble laminarin had average chainlengths (CL) of 15-19 glucose residues and degrees of polymerisation (DP)of 16-21, and were therefore almost linear.By contrast, four samples ofsoluble laminarin had CL values of 7-10 and DP values of 26-31 equivalentThe biochemistry of laminarin has been reviewed.4g45 D. Abbott, E. J. Bourne, and H. Weigel, J . Chem. SOC. ( C ) , 1966, 827.46 D. Abbott and E. Weigel, J . Chem. SOC. (C), 1966, 821.47 S. Peat, W. J. Whelan, J. R. Turvey, and K. Morgan, J. Chem. SOC., 1961, 623.48 M. Fleming and D. J. Manners, Biochem. J., 1966, 100, 24P.49 A. T. Bull, and C. G. C. Chestera, Adv. Enzymology, 1966, 28, 326.60 M. Fleming and D.J. Manners, Biochem. J., 1965, 94, 17P.61 W. D. b a n , Sir Edmund Hirst, and D. J. Manners, J. Chem. SOC., 1965, 885.63 W. D. Annan, Sir Edmund Hirst, and D. J. Manners, J . Chem. SOC., 1965, 220.53 I. J. Goldstein, F. Smith, and A. M. Unrau, Chem. and Ind., 1959, 124MANNERS: STRUCTURE AND METABOLISM OF GLUCANS 697to the presence, on the average, of 2-3 branch points per molecule. Theessentially linear molecules can compact closely to form insoluble aggre-gates; this is not possible with the branched molecules. It is thereforeprobable that in different algae, a range of laminarin polysaccharides issynthesised which differ in degree of branching, and hence, in solubility.None of the above samples of laminarin contained significant amounts ofmannose.52 (Cf.ref. 54.)Although laminarin occurs only in brown seaweeds, /%l,S-glucans arenow known to be widely distributed in Nature. These polysaccharides donot contain mannitol, but consist of chains of @-l,S-linked glucose residues,with varying degrees of branching involving @-lY6-inter-chain linkages.Callose,55 which occurs in small amounts in vascular tissues and repro-ductive structures in angiosperms, pa~hyman,~~ from the fungus Poria cowsWolf and paramylon, synthesised by the flagellates Euglena gTaciZis 57 andPeranemu trichophorum 58 are most probably linear molecules. Chryso-laminarin s9 from a mixture of diatoms (Chrysophyceae), the glucan fromPhaeodactylum tricornutum,60 leucosin from Ochromonas malbmensis ti andparamylon from Astasia ocellata 61 have it low degree of branching (see Table2).The structure of the cell-wall glucan from Sacchronyw cerevisiae hasbeen re-examined.62 An earlier methylation analysis as indicated a highlybranched structure in which chains of @-1,3-linked glucose residues were inter-linked by ,& 1 ,%-glucosidic linkages.However, a partial acid hydrolysisstudy a4 suggested a linear molecule containing certain sequences of B-13- and@-1,6-linked glucose residues. Re-investigation of yeast glucan 62 by methyl-ation, periodate oxidation, and partial acid hydrolysis has confirmed thebranched nature of the molecule, and identified the minor linkages as#L1,6-glucosidic linkages. Degradation of the glucan with a bacterial lami-narinase gave glucose, laminaribiose, and laminaritriose together with10% of a limit dextrin which consisted largely of b-1,6-linked glucose residues.The original glucan must therefore consist of main chains of p-1,6-linkedglucose residues to which are attached side chains of /l-1,3-linked glucoseresidues.A generally similar structure has been proposed independently byMisaki and Smith 65 on the basis of periodate oxidation and methylationanalyses.Many other species of fungi synthesise p-glucans containing both 1,3-and 1,g-linkages. These include polysaccharides from the mycelium ofti* F. Smith and A. M. Unrau, Chem. and Ind., 1959, 636.G. 0. Aspinall and G. Kessler, Chem. and Ind., 1957, 1296.K 6 S. A. Warsi and W. J. Whelm, Chem. and Ind., 1957, 1673.67 A.E. Clarke and B. A. Stone, Biochim. Bwphys. Acka, 1960, 44, 161.A. R. Archibald, W. L. Cunningham, D. J. Manners, J. R. Stark, and J. F.A. Beattie, E. L. Hirst, and E. Percival, Biochem. J., 1961, 79, 531.6o C. W. Ford and E. Percival, J . Chem. SOC., 1965, 7035.61 D. J. Manners, J. F. Ryley, and J. R. Stark, Biochem. J., 1966,101, 323.D. J. Manners and J. C. Patterson, Bhchem. J., 1966, 98, 19C.6a D. J. Bell, and D. H. Northcote, J . Chem. SOC., 1950, 1944.S . Peat, W. J. Whelan, and T. E. Edwards, J . Chem. Soc., 1958, 3862.A. Misaki and F. Smith, ‘‘ Abstracts, h e r . Chem. SOC., 144th Meeting,” 1963,Ryley, Biochem. J., 1963, 88, 444.14cProperty[a], (in water)[ a ] ~ (in NaOH)Hydrolysis to laminarisaccharides byRhizopus /3-glucosidase preparationIpfrared spectrum absorption peak(cm.-l)Reduotion of periodate (mol. prop.)Average chain lengthDegree of polymerisationTABLE 2 Properties of /?-1,3-glucalzsLaminarin K1- 9" + 9"+89019240.30chryso-laminarin 68-6" -+89012210.30ParamylonfromEuglenagracilis 6'+28"-+8900.02 -160ParamylonfromAstasiaocellata+ 17"-+8904350MANNERS: STRUCTURE AND METABOLISM O F GLUCANS 599Microsporum quinckeanum,66 and extracellular poly saccharides produced byClaviwps purpurea,67 Pullularia pullulunq6a and an unidentified member ofthe Pungi im~erfecti.6~ The distribution of #3-1,3-glucans is discussed indetail elsewhere.49Linear glucans containing both 8- 1,3- and 8- 1,4-linkages occur in Icelandmoss as lichenin,709 7 1 in the unicellular alga Nonodus s~bterraneus,~~ and incereal endosperms,719 T3 The presence of the two types of linkage wasoriginally established by methylation 70, 727 7 3 and partial acid-hydrolysisstudies 7 1 and their relative proportion may be assessed by periodateoxidation.The presence of 70 -+ 2% of 1,4-linkages and 30 & 2% of 1,S-linkages inboth lichenin, and in the glucans extracted from oats and barley has beenreported.74 However, lichenin differs from the cereal glucans in the sequenceof the two types of linkage.Application of the Smith degradation method tolichenin 75 gave only 8-glucosylerythritol, whereas the cereal glucans 7s, 76s 77gave this disaccharide and significant amounts of higher laminarisaccharide-erythritols. It follows that in lichenin, the 1,3-linked glucose residues arealways adjacent to 1,4-linked glucose residues, whereas in oat and barleyglucan there is a more random structure with sequences of two, three, ormore adjacent 1,3-linked glucose residues.The laminarisaccharide-eryth-ritols were rigorously identified by chemical means, and are not artefactsarising from incomplete periodate oxidation of the glucans. The yields of thehigher oligosaccharides from both cereal glucans show that the relativeproportion of pairs of adjacent 1,3-linked residues is much greater than thatof three such adjacent linkages, which in turn is greater than that of fouradjacent 1,3-linkages. The latter probably represent less than 1% of thetotal 1,3-linkages.The glucan from Monodus subterraneus also gave glucosylerythritol but nohigher saccharides when degraded by the Smith method, showing a similarityto l i ~ h e n i n ; ~ ~ it differed in having a lower proportion of 1,3-linkage~.7~An alternative method of structural analysis involves the characterisationof the products of enzymic hydrolysis.This requires a homogeneous enzymepreparation, an ideal which is difficult to obtain, and has been realised onlyin one or two recent studies. The method also assumes that a laminarinasc (endo-p- 1,3-glucanase) and a cellodextrinase (endo-p- 1,4-glucanase) arecompletely specific for the two types of linkage. However, Perlin andm H. Alfes, C. T. Bishop, and F. Blank, Canad.J . Chem., 1963, 41, 2621.6 7 A. S. Perlin and W. A. Taber, Canud. J . Chem., 1963, 41, 2278.68 H. 0. Bouveng, H. Kiessling, B. Lindberg, and J. McKay, Acta Chem. Scund., 1963,69 J. Johnson, S. Kirkwood, A. Misaki, T. E. Nelson, J. V. Scaletti, and F. Smith,’* N. B. Chanda, E. L. Hirst, and D. J. Manners, J . Chem. Soc., 1957, 1951.S. Peat, W. J. Whelan, and J. G. Roberts, J . Chem. SOC., 1957, 3916.A. Beattie and E. Percival, Proc. Roy. SOC. Edinburgh, 1962, B, 88, 171.73 0. Igarashi and Y. Sakurai, Agric. and BioZ. Chem. (Japan), 1965,29, 678.7 4 A. E. Clarke and B. A. Stone, Biochem. J . , 1966, 99, 582.7 5 M. Fleming and D. J. Manners, Biochem. J., 1966, 100, 4P.76 I. J. Goldstein, G. W. Hay, B. A. Lewis, and F. Smith, ref. 8, p. 367.7 7 0.Igarashi and Y. Sakurai, Agric. and BWZ. Chem. (Japun), 1966, 30, 642.17, 1351.Claem. and Ind., 1963, 820.C. W. Ford and E. Percival, J . Chem. Soc., 1965, 3014600 BIOLOGICAL CHEMISTRYSuzuki have shown that the laminarinase of Rhizopus arrhizus is specificfor a /%glucosidic linkage attached to a 3-O-substituted glucose residue i.e.,(4 (b)- G - G - G -the ability of the enzyme to hydrolyse linkage (b) is dependent on the natureof linkage (a). From the identity of the products isolated when lichenin isdegraded by this “ laminarinase ”, (a) must be a B-1,3-linkage9 but (b) can bseither a /3-1,3- or a p-1,4-linkage.7gs 8o It is therefore clear that with poly-saccharides containing mixed linkages, the nature of the linkage hydro-lysed by a p-glucanase cannot be unambiguously identified, on the basis ofspecificity studies carried out using simple substrates containing only onetype of linkage.It should also be emphasised that these considerations mayapply to many carbohydrases. For example, a p-1,3-glucanase from Bacilluscirculans shows a specificity requirement similar to that of the mould“ laminarinase.” 81Culture filtrates of Aspergillus niger contain a complex mixture ofp-glucosidases which have been separated by column chromatography oncalcium phosphate and Dowex-1 .a2 At least three different /3-1,3-glucanasesand two different #3- 1,4-glucanases were present. A purified endo-p- 1,4-glucanase hydrolysed cellodextrin, lichenin, cereal glucans, and two mannans,but did not hydrolyse laminarin or xylan.With barley glucan, the productsincluded 12% of cellobiose, 45% or 4-0-/3-~-~aminaribiosyl-~-glucose and16% of a related tetra~accharide.7~ It seems probable that this enzymecan specifically hydrolyse p-glucosidic linkages attached to 4-O-substitutedglucose residues. It differs from certain bacterial endo-#Lglucanases whichcan hydrolyse barley glucan as and lichenin 84 but not laminarin or cello-dextrin.Extracts of germinated barley also contain a complex mixture of8-glucosidases. 85 By a combination of dialysis, ammonium sulphate fiac-tiona.tion and chromatography on phosphorylated cellulose, two distinctendo-B-glucanases, a laminarinase, and a methoxycarbonyl cellulase wereisolated. 86 The endo-p-glucanases partially hydrolysed barley glucan to givetri- and tetra-saccharides which contained both 1,3- and 1,4glucosidiclinkages.87 With one of the endo-/?-glucanases, the products were 53% of3-O-~-D-ce~~obiosy~-D-g~ucose, 27% of 3-O-~-~-ce~~otriosy~-~-g~ucose and 8%of higher oligosaccharides which contained mainly 1,4-linked glucose residuesand, at the reducing end, a 1,3-linked residue.88 This enzyme preparationhad no action on laminarin, rnethoxycarbonylcellulose, laminarisaccharides(DP 2-6), cellosaccharides (DP 2-43), or 4-~-~-D-~amharibiosyl-D-g~ucose79 A.S. Perlin and S. Suzuki, Canad. J . Chem., 1962, 40, 50.8 8 W. L. Cunningham and D. J. Manners, Biochem. J., 1964, 90, 696.81 H. Tanaka and H. J. Phaff, J . Bucteriol., 1965, 89, 1570.8 2 A.E. Clarke and B. A. Stone, Biochem. J . , 1965, 96, 793, 802.83 E. A. Moscatelli, E. A. Ham, and E. L. Rickes, J . Biol. Chem., 1961, 236, 2858.84 E. T. Reese and A. S . Perlin, Biochem. Biophys. Res. Comm., 1963, 12, 194.86 W. W. Luchsinger, E. F. Hou, and G. L. Schneberger, Proc. West Virginia Acad.88 W, W, Luchsinger and A. W. Richards, Arch. Biochem. Biophys., 1964, 106, 65.88 W. W. Luchsinger, S. C. Chen, and A. W. Richards, Arch. Biochem. Biophys.,Sci., 1962, 34, 51.S. C. Chen and W. W. Luchsinger, Arch. Biochem. Biophys., 1964,106, 71.1965, 112, 524MANNERS : STRUCTURE AND METABOLISM OF GLUCANS 601and -cellobiose. From the identity of the major products, the Wage whichwas hydrolysed by this enzyme was probably situated between a 1,3- and a1,4-linkage, and from indirect evidence, the authors concluded that it w a ~ a1,4-linkage.89 They therefore suggested that the glucan was largely com-posed of two or three adjacent 1,4-linked glucose residues separated by single1,3-linked glucose residues.This structure is different from that obtainedby the Smith degradation meth0d.7~~ 77 It should be noted however thatthe chemical method is much more sensitive than the enzymic method forthe detection of sequences of adjacent 1,3-linked glucose residues. Theseresidues, which are resistant to periodate oxidation, are concentrated in theSmith procedure, whereas in the enzymic studies they are dispersed into smallamounts of either higher oligosaccharides of DP > 5 which are dacult tocharacterise, or are released as glucose and laminaribiose, depending upon thespecificity of the p-glucanase.The Metabolism of G1ucans.-The outstanding researches initiated byL.F. Leloir and continued by him and by W. 2. Hassid and their respectiveco-workers have clearly established that many glucans (see Table 3) aresynthesised by the successive enzymic transfer of glucose residues fromnucleoside diphosphate glucose compounds to a suitable acceptor moIecule.4In many instances, the enzyme was closely associated with, or adsorbed on to,the polysaccharide so that the assay system consisted of the glucosyl donorTABLE 3 Enzymic transfer of gluwsyl residues to glucunsGlucosyl Nature ofacceptor newly formedSource of Glucoa yl and glucosidicenzyme dbnor product linkageLiver, muscle, yeast UDPG Glycogen a-1,4Bacteria ADPG Glycogen a-1,4Higher pIants ADPG Starch a-1,4Higher plants GDPG Cellulose !-I94Higher plants UDPG Callose 8-193Flagellates UDPG Param yIon f l 4 3orUDPGRhizobiwna japon;cum UDPG /3-1,2-Glucan f3-1,2and a particulate polysaccharide-enzyme preparation.Attempts to solu-bilise the enzyme frequently caused inactivation. The first experimentswere concerned with the synthesis of glycogen from UDPG by a rat-liverpreparation, and showed the transfer of an a-glucosyl residue and theformation of UDP :-where [GI, or [GJn+l represents glycogen of DP n or (n + 1). In manyexperiments, UDP mas estimated using pyruvate kinase; later, UDPGlabelled with [la CJglucose became available, and the UDPG a-glucan trans-glucosylase could then be assayed from the rate of incorporation of into8Q W.W. Luchsinger, S. C. Chen, end A. W. Richards, Arch. Biochem. Biophys.,1965,112, 531.90 L. F. Leloir and C. E. Cardini, J . Amer. Chem. SOC., 1957, 79, 6340.UDPG + [Gln + UDP + [G]n+602 BIOLOGICAL OHEMISTRYthe polysaccharide. Since all these and subsequent experiments were per-formed on the micromole scale, new methods for the characterisation of theproducts had to be devised. With starch and glycogen, the availability ofhighly specific @- and a-amylases enable the products to be degraded tomaltose and related sugars, which can be characterised chromatographically.With B-glucans, the solubility properties were formerly used to distinguishbetween cellulose and laminarin-type polymers, but this was unsatisfactory,and micro-scale partial acid and enzyme hydrolysis methods have been used.It should be noted that in current studies, the polysaccharide is synthesised indigests of a total volume of less than 1 ml., and is not isolated by conventionalmethods. This contrasts with the classical studies on the synthesis ofamylose, amylopectin and glycogen by various phosphorylase preparationswhere gram quantities of polysaccharides were isolated and characterised bymethylation or periodate oxidation procedures.1~ 91This is synthesised fromUDPG by the concurrent action of UDPG or-glucan transglucosylase andbranching enzyme ; during catabolism the a- 1,4-glucosidic linkages aredegraded either by phosphorylase to give glucose l-phosphate, or by ana-glucosidase (glucamylase or y-amylase) to glucose, and the inter-chainlinkages are hydrolysed by amylo- 1,6-glucosidase.On fractional centrifugation of rat-liver homogenates, the UDPG or-glucantransglucosylase sedimented together with the particulate glycogen, and thespecific activity could be increased up to 300-fold by repeated mashing of theglycogen pellet.92 The activity was optimum at pH 8.4 and was increasedbetween 4- and 15-fold by the addition of physiological concentrations(lo-%) of glucose 6-phosphate.This effect is now known 93 to be due to theexistence of two different forms of the enzyme, one of which (D-form)required glucose 6-phosphate for activity, and the other (I-form) was in-dependent of this cofactor.The conversion of the D-form into the I-formrequires a heat labile subcellular fraction and Mg2f. Liver UDPG or-glucantransglucosylase could not degrade glycogen in the presence of UDP. Thedistribution of the enzyme in liver-cell fractions was examined by differentialcentrifugation and electron micr0scopy.~4 The results suggested that theenzyme was directly bound to the glycogen and was not associated withstructural elements of the liver cell. The rat-liver enzyme has recently beensolubilised and purified 1500-fold by using reversible thermal inactivationof the enzyme to remove it from the particulate glycogen.95 The level ofenzymic activity and the relative effect of added glucose 6-phosphate isunder hormonal rcgulation.96 For example, the injection of insulin into ratscaused a marked rise in activity, and a rapid deposition of glycogen.The branching enzyme from rat liver has been purified 35-fold and freedfrom or-amylase, which is an undesirable contaminant of most liver enzymeMetabolism of Glycogen.-Liver glycogen.O1 B.N. Stepanenko, A. S. Kainova,,,and N. N. Petrova, “Proceedings, ThirdD2 L. F. Leloir and S. H. Goldemberg, J . Bid. Chem., 1960, 235, 919.O3 S. Hizukuri and J. Lamer, Biochemistry, 1964, 3, 1783.O4 D. J. L. Luck, J . Biophys. Biochem. Cytology, 1961, 10, 195.O 5 D. F. Steiner, L. Younger, and J. King, Biochemistry, 1965, 4, 740.Q* D. F. Steiner, V. Rauda, and R. H. Williams, J . BWE. Chem., 1961, 236, 299.Internat.Congress Biochemistry, Brussels, 1955, 50MANNERS : STRUCTURE AND METABOLISM OF GLUCANS 603preparations. The enzyme can introduce branch points into amylose,amylopectin, and amylopectin ,!I-dextrin, the last two-substrates beingconverted into glycogen-type polysac~harides.~~ The mode of action masstudied using as substrates two polysaccharides in which the end-groupswere labelled by incubation with [14CJglucose 1 -phosphate and phosphory-lase.24 After treatment with branching enzyme, the products were analysedby periodate oxidation, by degradation with phosphorylase and amylo-l,6-glucosidase, and with pullulanase. The results showed that the branchingenzyme catalysed the transfer of a chain of at least six glucose residues from a1,4- to a 1,6-position.The enzyme had no action on a linear maltosaccharideof DP 16.Under normal conditions, mammalian liver contains appropriate con-centrations of active UDPG-pyrophosphorylase, UDPG a-glucan trans-glucosylase and branching enzyme t o convert glucose 1 -phosphate into high-molecular-weight glycogen containing 7-10% of a- 1,6-inter-chain linkages.However, abnormal glycogens may be produced under certain circumstances.In cases of glycogen storage disease Type IV, the deposited polysaccharideresembles amylopectin rather than glycogen and contains only 4-6% ofbranch points.98 A partial deficiency in branching enzyme is thus implied(see also p. 608). When chicks were fed a diet containing 16% of D-galactose,toxicity symptoms soon occurred, and the isolated liver glycogen contained asmall but significant amount (0.2%) of gala~tose.~~ When rat liver masperfused with [ 1 - 14C]-~-galactosamine, the glycogen became radioactive andcontained D-glucosamine glycosidically linked to glucose.loo Under thesetwo conditions, it is possible that other sugar nucleotides, e.g.UDP-glucosa-mine, may replace UDPG to a very limited extent. It is also of interest torecall lol the isolation of glycogen which contained a small but significantproportion of D-fructose residues, from the liver of pregnant does; a bio-chemical explanation of this observation is not available.Present knowledge of the properties of liver phosphorylase is due largelyto Sutherland and co-workers lo2 who studied the enzyme from dog liver, but,a kinetic study of the purified rabbit-liver enzyme has also been reported.103Liver phosphorylase occurs in both active and inactive forms.In rats about60% of the enzyme in normal ‘‘ resting ” liver is present in the active f0m.lo4The interconversion of the two forms, which represent a phosphorylated andnonphosphorylated protein respectively (with a serine residue a t the site ofphosphorylation), is regulated by various factors including adrenaline andglucagon, and the concentration of UDPG which is a competitive inhibitor.The inactive form of liver phosphorylase differs in several respects from thes 7 C. R. Krisman, Biochim.. Biophys. Acta, 1962, 65, 307.98 B. Illingworth and G. T. Cori, J . Biol. Chem., 1952, 199, 653.O Q J.H. Nordin and R. G. Hamen, J . Biol. Chem., 1963, 258, 489.loo F. Maley, J. F. McGarrahan, and R. DelGiacco, Biochem. Biophys. Res. Comm..1966, 23, 85.lol S. Peat, P. J. P. Roberts, and W. J. Whslan, Biochem. J., 1952, 51, xvii.loa E. W. Sutherland and W. D. Wosilait, J . Biol. Chem., 1956, 218, 459; W. D.Wosilait and E. W. Sutherland, ibid., p. 469; T. W. Rall, E. W. Sutherland, and W. D.Wosilait, ibid., p. 483; W. D. Wosilait, ibid., 1958, 233, 597.lo3 V. T. Maddaiah and N. B. Madsen, J . Biol. Chem., 1966, 241, 3873.loo V. T. Maddsiah and N. B. Madsen, Biochirn. Biophys. Acta, 1966, 121, 261.604 BIOLOGICAL CHEMISTRYcorresponding muscle enzyme (e.g., the liver enzyme is activated by sulphate),and in the amino-acid composition of the phosphopeptides derived from thecatalytic sites.105Evidence for the presence of a nonphosphorolytic pathway for liverglycogen breakdown has been accumulating.106 Torres and Olavarria lo7have shown that extracts of rat liver contain a mixture of enzymes, includingan a-amylase, and that the supernatant solution obtained on differentialcentrifugation a t 105,OOOg contained an enzyme system which releasedglucose directly from maltosaccharides and from glycogen.The glucnmylasesystem has been separated into an acid glucosidase (active over pH 3-6)and a neutral glucosidase (active over pH 4-7-5). The acid glucosidase islocalised mainly in the lysosomes and may be concerned with intracellulardigestion and autolysis. The enzyme is clearly important in normal meta-bolism.In human liver and other tissues, a deficiency of this enzyme isresponsible for the accumulation of glycogen in cases of generalised glycogenstorage disease (Type 11), even though the level of activity of phosphorylaseand amylo- 1,6-glucosidase is normal. l o 8Muscle gZycogen. In muscle tissues, glycogen is also synthesised by WDPGa-glucan transglucosylase and a branching enzyme. The former enzyme,which has been extensively studied by Leloip9 Brown,l10 and Larner,llland their co-workers, occurs in two distinct forms, one of which (r-form)does not require glucose 6-phosphate for activity, and the other (D-form) isglucose 6-phosphate dependent. The I-form can be converted into the D-form by a phosphorylation reaction requiring ATP and Mg2+.l11 Thereverse reaction is catalysed by a phosphatase.Alternatively, the Iinto D conversion may be brought about by a protein factor and calciumions.l12 Various hormones (including adrenaline 113) and other factors(including cyclic adenylic acid 11* and potassium ions 115) provide regulatorymechanisms for the control of the activity of the muscle enzyme. Theenzymes from rat,ll6 rabbit,llO, 117 and dog 118 skeletal muscle and fromtoadfish and frog muscle 119 have been highly purified. There is a markedspecies variation in kinetic properties and in sensitivity to glucose 6-phos-phate. 119105 M. M. Appleman, E. G. Krebs, and E. H. Fischer, Biochemistry, 1966, 5, 2101.108 W. J. Rutter and R. W. Brosemer, J.Biol. Chem., 1961, 236, 1247.107 H. N. Torres and J. M. Olavarria, Rcta Physiobgica Latinoamericana, 1961, 11,95; H. N. Torres and J. M. Olavarria, J . Biol. Chem., 1962, 237, 1746; 1964, 239, 2427.106 H. G. Hers, Biochem. J . , 1963, 86, 11.108 I,. F. Leloir, J. M. Olavarria, S. H. Goldemberg, and H. Carminatti, Arch.110 R. Kornfeld and D. H. Brown, J . Biol. Chem., 1962, 237, 1772.111 D. L. Friedman and J. Larner, Biochemistry, 1963, 2, 669; 1965, 4, 2261.112 E. Belocopitow, M. M. Appleman, and IS. N. Torres, J. B i d . Chem., 1965, 240,113 E. Belocopitow, Arch. Biochem. Biophys., 1961, 93, 467.114 M. M. Appleman, L. Birnbaumer, and H, N. Torres, Arch. Biochem. Bwphys.,115 H. N. Torres, L. Birnbaumer, M. D. C. 0. Fernandez, E. Bernard, and E.Belo-116 M. Rosell-Perez, C. Villar-Palasi, and J. Larner, Biochemistry, 1962, 1, 763.117 M. Rosell-Perez and J. Larner, Biochemistry, 1964, 3, 75.118 M. Rosell-Perez and J. Larner, Biochemistry, 1964, 3, 81, 773.119 M. Rosell-Perez and J. Larner, Biochemistry, 1962, 1, 769.Biochem. Biophys., 1959, 81, 508.3473.1966, 116, 39.copitow, Arch. Biochem. Biophys., 1966, 116, 59MANNERS : STRUCTURE AND METABOLISM OF GLUCANS 605;The enzyme from rat muscle 120 shows a high degree of specificity for theglucosyl donor; ADPG was 50% as effective as UDPG, but CDPG, IDPG,and ADP-maltose were inactive. Maltose and maltotriose were acceptors ofvery low efficiency but maltotetraose and higher oligosaccharides weresatisfactory, and glycogen (irrespective of source) was the best acceptor.With the rabbit-muscle enzyme, and oligosaccharide or-limit dextrins,transfer occurs exclusively to the main With glycogen, enzymeaction is also unsymmetrical and results in the elongation only of the mainchains (B-chains).This Unsymmetrical addition of 1 ,.l-linked glucoseresidues must be related to the subsequent mode of branching.Muscle phosphorylase, particularly that from rabbit skeletal muscle,has been extensively studied by E. H. Fischer and his co-workers. Theenzyme exists in two forms, one of which (phosphorylase a) is active in theabsence of adenylic acid, whilst phosphorylase b (the predominant form inresting muscle) is inactive unless adenylic acid is present. The two forms areinterconverted as follows :phosphorylase bkinase2 phosphorylase b + 4 ATP ______+ phosphorylase a + 4 ADPphosphorylasephosphat asephosphorylase a + 4 H,O ______+ 2 phosphorylase b + 4 HOPThe kinase 121 and phosphatase 122 have been extensively purified and thevarious factors (e.g., cyclic adenylic acid, Ca2+, adrenaline) controlling theiractivities examined. The phosphorylases also contain stoicheiometricamount8 of pyridoxal 5-phosphate, which is essential for enzymic activity,but functions in a different manner from that in other pyridoxal phosphate-catalysed reactions .Iz3Muscle phosphorylase from several species including rabbit heart,124human skeletal,125 cat,lB6 and lobster 1 2 7 has been highly purified.Theamino-acid composition of human and rabbit muscle phosphorylase appeart o be essentially identical,128 and the sequence of amino-acids at the activesites is now However, immunological and kinetic studies showthat the phosphorylases of skeletal and smooth muscles from the sameanimal are not identical.lS0120 S.H. Goldemberg, Biochim. Biophys. Acfa, 1962, 56, 357.lal E. G. Krebs, D. J. Graves, and E. H. Fischer, J. Bid. Chem., 1959, 234, 2867;J. B. Posner, R. Stein, and E. G. Krebs, ibid., 1965,240, 982; E. G. Krebs, D. S. Love,G. E. Bratvold, K. A. Trayser, W. L. Meyer, and E. H. Fischer, Biochemistry, 1964, 3,1022; W. L. Meyer, E. H. Fischer, and E. G. Krebs, ibid., p. 1033.122 D. J. Graves, E. H. Fischer, and E. G. Krebs, J. BioZ. Chem., 1960, 236, 805.123 J. L. Hedrick and E.H. Fischer, Biochemistry, 1965, 4, 1337; S. Shaltiel, J. L.Hedrick, and E. 11. Fischer, ibid., 1966, 5, 2108, 2117.A. A. Yunis, E. H. Fischer, and E. G. Krebs, J. BioZ. Chem., 1962, 237, 2809.lZ5 A. A. Yunis, E. H. Fischer, and E. 0. Krebs, J . Biol. Chem., 1960, 235, 3163;A. A. Yunis and E. G. Krebs, ibid., 1962, 237, 34.12( A. B. Kent, E. G. Krebs, and E. H. Fischer, J . Biol. Chem., 1958,232, 549.12' R. W. Cowgill, J. BioK Chem., 1959, 234, 3146, 3154.M. M. Appleman, A. A. Yunis, E. G. Krebs, and E. H. Fischer, J . Biol. Chem., 1963,238, 1358.laO E. H. Fischer, D. J. Graves, E. R. S. Crittenden, and E. G. Krebs, J . BWZ. Chem.,1959,234,1698; R. C. Hughes, A. A. Yunis, E. G. Krebs, and E. H. Fischer, z%id., 1962,237, 40; C. Nolan, W. B.Novoa, E. G. Krebs, and E. H. Fischer, Biochemistry, 1964,3, 542.130 E. Bueding, N. Kent, and J. Fisher, J . Bid. Chem., 1964, 239, 2099606 BIOLOffICAL CHEMISTRYIt is not possible to deal with several relevant topics, includmg details ofthe interconversion of muscle phosphorylase a and by and the regulation ofglycolpis and glycogenolysis in skeletal muscle, including the effects ofadrenaline and glucagon. These topics were considered in detail a t a recentsymposium. 31It is now generally accepted that the action of muscle phosphorylase onglycogen results in a partial degradation of the exterior chains to give alimit dextrin (9-dextrin) with outer “ stubs ” of four glucose re~idues.~5The subsequent degradation of 4-dextrin involves two reactions whichappear to be catalysed by the same protein (amylo-1,6-glucosidase). In thefirst reaction (transferase) ,132 three glucose residues are transferred from theA-chain to the B-chain producing a dextrin with the assymetric structureoriginally proposed 133 for the 6-destrin.In the second reaction (hydrolase)the lY6-linkage attaching the single glucose residue is hydrolysed. The latteractivity can also be tested using branched oligosaccharides with single-unitside chains as substrates.134 The transferase activity dif€ers from that ofother trans-cc-glucosylases in having no reaction with maltose or maltotrioseand in being unable to transfer single glucose residues.135 With [l*C]malto-triose and glycogen, the major product was [14C]maltohexaose. Attempts toseparate the transferase and hydrolyase activities by LZ wide variety ofmethods have not been successful.130The assay of amylo-1,6-glucosidase is not simple.The original methodwas based on the release of glucose from the 4-dextrin. This reaction is, to aslight extent, reversible 137 and by using [14C]glucose, it was possible toobtain a significant incorporation into the p01ysaccharide.l~~ However, theoptimum pH for the liberation of glucose is 5-6 whereas that for incor-poration is about 8.139 Alternative substrates are oligosaccharides withaingle glucose residues as side-chains or a-glucosyl Schardinger dextrins,140both of which yield glucose on hydrolysis.In certain cases of glycogen storage disease (Type 111 or limit dextri-nosis), “ glycogen ” with a 4-dextrin structure is deposited and a deficiencyin ‘I amylo-1,6-glucosidase ” is indicated.Hers 141 has suggested the exist-ence of various sub-types of limit dextrinosis since different tissue extractsahow different enzymic deficiencies depending upon the method of assay foramylo-1,6-glucosidase, and whether the defect is confined to muscle or livertissue or is generalised.There is now substantial evidence for nonphosphorolytic degradation ofglycogen in muscle tissues. This arose originally from studies on glycogen1 8 1 “ Control of Glycogen Metabolism,” CIBA Symposium, eds. W. J. Whelan andM. P. Cameron, Churchill, London, 1964.182 M. Abdullah, P. M. Taylor, and W. J. TNhelan, ref. 131, p. 123.138 G. T.Cori and J. Larner, J . Biot. Chem., 1951,188, 17.134 B. Illingworth and D. H. Brown, Proc. Nut. Acad. Sci. U.S.A., 1962, 48, 1619.135 D. H. Brown and B. Illingworth, Proc. Nut. Acad. Sci. U.S.A., 1962, 48, 1783.136 D. H. Brown and B. Illingworth, ref. 131, p. 139; D. H. Brown and B. I. Brown,1 3 7 J. Larner and L. H. Schlisefeld, Biochim. Biophya. Acta, 1956, 20, 53.138 H. G. Hers, Rev. In&. Hepatol., 1959, 9, 35.188 H. G. Hem, W. Verhue, and M. Mathien, ref. 131, p. 158.140 P. M. Taylor and W. J. Whelan, Arch. Biochem. Bzophys., 1966, 113, 500.1 4 1 H . G. Hers, ref. 131, p. 164.Biochem. J., 1966, 100, 8PMANNERS: STRUCTURE AND METABOLISM OF GLUCANS 607storage disease Type V where muscle phosphorylase is absent, and yet theglycogen content of the muscle, although greater than normal, is not exces-sive.142 The evidence for an alternative pathway also includes the demon-stration 131 that mammalian muscle tissues contain a-glucoaidms whiohhydrolyse both maltose and glycogen directly to glucose.It now seemsprobable that in normal resting muscle, glycogenolysis is minimal, and thatthe phosphorylase-amylo- 1,6-glucosidase system is used only during pro-longed muscular exercise.Other mmmlian tissues. The metabolism of glycogen in tissues otherthan liver and muscle has been studied. The UDPG a-glucan transglucosylaseof rabbit 143 and sheep 144 brain has been purified; the level of activity,although lower than in the liver, is sufficient to account for glycogen synthesisin viuo.Several glycogen-metabolising enzymes in leucocytes or erythrocyteshave been assayed (e.g., phosphorylase,l45 amylo- 1,6-gluco~idase,14~ UDPGa-glucan transglucosylase 147). In cases of glycogen storage disease, erythro-cyte enzymes are also affected, so that the biochemical analysis of bloodsamples rather than liver biopsy tissue provides a more convenient method fordiagnosis.E. L. Rosenfeld and her colleagues148 have shown that many tissuescontain a glucose-producing amylase (y-amylase). The highest activity is inthe spleen, but brain, lung, and heart tissue have a higher activity than liverand kidney. The enzyme, which has an optimum pH of 443, is absent fromblood; the relation of this enzyme to the liver and muscle a-glucosidasee isnot yet known.With glycogen, hydrolysis is incomplete (about 50%) and a,y-limit dextrin can be isolated. This presumably has outer " stubs " of onlyone or two glucose residues. Whether this dextrin would be a substrate forpurified amylo- 1,6-glucosidase has not been established.Since the publication of a comprehensiveReview 149 a new type of glycogenosis characterised by a deficiency of musclephosphofructokinase has been reported.150 Three siblings from e, singlefamily were affected.There is now substantial evidence to show that the glucose 6-phosphatase,inorganic pyrophospha t ase, and p pop hosp hat e-g lucose p hosp hotransferaseactivities of liver microsomes are due to a single enzyme.151 AdditionalIra J. Larner and C. Villar-Palasi, Proc. Nut.Acad. Sci. U.S.A., 1959, 45, 1234; R.Schmid, P. W. Robbins, and R. R. Traut, ibid., p. 1236; W. F. H. M. Mommaerts, B.Illingworth, C. M. Pearson, R. J. Guillory, and K. Seraydarian, {bid., p. 791.B. M. Breckenridge and E. J. Crawford, J . BioZ. Chem., 1960, 235, 3054.lP4 D. K. Basu and B. K. Bachhawat, Biochim. Biqphys. Acta, 1961, 50, 123.145 M. Cornblath, E. Y. Levin, E. Marquetti, and E. Y. House, Fed. Proc., 1960, 19,68; W. C. Hulsmann, T. L. Oei, and S. van Creveld, Lancet, 1961, 581; H. E. Williamsand J. B. Field, MetaboZisnt, 1963, 12, 464.146 F. Huijing, CZinica Chirn. Acta, 1964, 9, 269.147 W. L. Miller and C. Vander Wende, Biochim. Bwphys. Acta, 1963, 77, 494.Several Russian papers are reviewed by E. L. Rosenfeld in ref. 131, p.176.lrO H. G. Hers, Adv. Metabolic Disorders, 1964, 1, 1; see also R. Schrmd, ref. 131,p. 305 and subsequent Papers by D. J. Manners, B. Illingworth, D. H. Brown, H. Q.Hers, J. Larner, and E. Bueding.150 S. Tarui, G. Okuno, Y. Ikura, T. Tanaka, M. Suda, and M. Nishikawa, Bkochern.Biophy8. Rea. C'omm., 1965, 19, 517.lalM. R. Stetten, J . BWZ. Chern., 1964, 239, 3576; C. J. FisherandM. R. Stetten,Bwchh. Bwphys. Acta, 1966,121, 102; W. J. &ion and R. C. Nordlie, J . Bwl. Chem.,1964,239,2762; R. C. Nordlie and D. G. Lygre, ibid., 1966, 241, 3136.GZywgen storage diseases608 BIOLOGICAL CHEMISTRYevidence is provided by measurement of the pyrophosphatase activity ofliver homogenates from patients with Type I glycogenosis (characterisedby a lack of glucose 6-phosphatase); a markedly diminished activity wasobsemed.l52The Type II diseases (generalised glycogenosis ; Pompe’s disease) inwhich there is a deficiency of acid a-glucosidase require further study.Twopatient6 have been described with mild muscular hypotonia, and whosemuscles contained normal amounts of glycogen and glycogen-metabolisingenzymes except that a-glucosidase activity at pH 4.5 was absent.153 How-ever, the leucocytes contained a normal level of this activity.154 The clinicalsymptoms were so different from the usual cases of Type I1 disease that thetwo cases may represent the first examples of an abortive form of muscularglycogenosis. Liver homogenates from another case of Type I1 disease werealso deficient in acid a-glucosidase, but contained an a-glucosidase, active atpH 7.1, which had no action on glycogen, but hydrolysed maltose and couldalso transfer single glucose residues to or from maltose or maltotriose.155 Itis clear that the biochemical role of the various a-glucosidases in both normaland glycogenosis tissues requires further examination.New information on the Type IV disease (amylopectinosis) is nowa~ailable.l5~ Liver biopsy tissues from the fourth known w e contained305% of an amylopectin-type polysaccharide, and a liver homogenate and theleucocytes did not show branching enzyme activity. The fact that thedeposited polysaccharide still contained 6% of branch points raises theintriguing question as to how these are produced if branching enzyme cannotbe detected under the usual conditions of assay.Finally, attention is drawn to the cases of “ glycogen deficiency disease ”in which Spencer-Peet and co-workers showed the absence of UDPG a-glucanfransglucosylase in the livers of children from a single family.l57Glycogen metabolism in other species.Several aspects have been studiedincluding glycogen synthesis from UDPG by the fat bodies of the Ameri-can cockroach (Periphnetu umericunu L.),168 the silk moth (Hyalophorac e c r o p i ~ ) , ~ ~ ~ and the locust (Schistocerca cunceZlata),160 and by extracts ofbaker’s yeast, where the enzyme was purified 500-f0ld.~~l Branching en-eymes from yeast,l62 Escherichia coli 163 and Arthrobacter globqormis 164have been partially purified and shown to be similar to the branching enzymeof animal tissues.lb8 B.Illingworth and C. F. Cori, Biochern. Bhphys. Res. Comm., 1965, 19, 10.163 H. Zellweger, B. I. Brown, W. F. McCormick, and J.-B. Tu, Ann. Paediat., 1965,164 B. I. Brown and H. Zellweger, Biochem. J., 1966, 101, 16c.lS6 B. I. Brown and D. H. Brown, Biochim. Biophys. Acta, 1965, 110, 124.lS6 B. I. Brown and D. H. Brown, Proc. Nut. Acad. Sci. U.S.A., 1966, 56, 725.J. Spencer-Peet, C. M. Lewis, and K. M. Stewart, ref. 131, p. 377.lS8 A. Vardanis, Bwchim. Biophys. Acta, 1963, 73, 565.lsa T. A. Murphy and G. R. Wyatt, J . Bwl. Chem., 1965, 240, 1500.160 J. C. Trivelloni, Arch. Biochem. Biophys., 1960, 89, 149.161 I. D. Algranati and E. Cabib, J . BioE. Chem., 1962, 237, 1007.162 2.H. Gunja, D. J. Manners, and K. Maung, Biochem. J., 1960,76,441.1‘s N. Sigal, J. Cattaneo, J. P. CkELmbost, and A. Favard, Biochem. Biophya. Rea.164 L. P. T. M. Zevenhuizen, Biochim. Biophys. Acta, 1964, 81, 608.205, 413.G. M. Lewis, K. M. Stewart, and J. Spencer-Peet, Biochem. J., 1962, 84, 116P;Comm., 1965, 20, 616MANNERS : STRUCTURE AND METABOLISM OF GLUCANS 609Although Agrobacterium tumej'aciens synthesises glycogen from UDPG,lsSin several other species of bacteria, ADPG is the glucosyl donor. The pro-perties of the purified ADPG a-glucan transglucosylases from an Arthrobaderspecies 166 and from Escherichia coZi B 167 and the corresponding ADPGpyrophosphorylases have been described. 16* The latter enzymes, whichcatalyse the formation of ADPG from ATP and glucose l-phosphate, arehighly specific, and are activated by various glycolytic intermediates which,together with the concentrations of ATP and AMP, may regulate glycogensynthesis at the ADPG level.Metabolism of Starch.-Since various aspects of the metabolism of starchhave been reviewed else~here,l6~ the present discussion will be confined torecent studies on biosynthesis. As with glycogen, the synthesis of starchinvolves nucleoside diphosphate glucose intermediates.Leloir and his co-workers 170 rirst showed that bean-starch granules were associated with aninsoluble enzyme which transferred glucose from UDPG into both the amyloseand amylopectin components. Later, 171 ADPG was shown to be a much moreeffective donor than UDPG; deoxy-ADPG was also an effective d0nor.17~Although plant tissues do not contain appreciable amounts of ADPG, theycontain ADPG pyrophosphorylase.173s 1 7 4The bean-starch preparation transferred glucose from ADPG or UDPGand attached it, by an a-1,4-linkage, to starch, or to added maltosaccharides,but not to other oligosaccharides. Similar results were obtained 172 withstarch granule preparations from potatoes, wrinkled peas, and variousvarieties of maize.In potatoes, the activity was entirely confined to thestarch granules and was increased by the use of sucrose-citrate media duringisolation, which minimised inactivafion.l75 More recently, a soluble form ofthe enzyme has been obtained from potato t~bers,17~~ 177 and also, fromtobacco 178 and spinach leaves.l78 With the spinach enzyme, deoxy-ADPGwas also an alternative donor, whilst amylose and amylopectin were moreefficient acceptors than starch granules or glycogen. In contrast to theanimal enzymes, glucose 6-phosphate was not an activator. SpinachADPG-pyrophosphorylase is strongly activated by 3-phosphoglyceric acid41, 561.le5 N.B. Madsen, Bwchim. Bwphys. Acta, 1961, 50, 194; Canad. J . Biochem., 1963,166 E. Gmenberg and J. Preiss, J . Biol. Chem., 1965, 240, 2341.lS7 J. Preiss and E. Greenberg, Biochemistry, 1965, 2, 2328.lB8 L. Shen and J. Preiss, J . BwZ. Chem., 1965, 240, 2334; Arch. Biochem. Biophya.,1966, 116, 375; J. Preiss, L. Shen, E. Greenberg, and N. Gentner, Biochemistry, 1966,5, 1833.16u W. J. Whelan, Starke, 1963, 15, 247; N.P. Badenhuizen and J. H. Pazur in'' Starch: Chemistry and Te~hnology,'~ eds. R. L. Whistler and E. F. Paschall, AcademicRess, New York, 1965, pp. 65 and 133; H. R. Chandorkar and N. P. Badenhuizen,Starke, 1966, 18, 91.170 L. F. Leloir, M. A. R. de Fekete, and C. E. Cardini, J . BioE. Chem., 1961,236,636.171 E. Recondo and L. F. Leloir, Biochem. Biophys. Res. Comm., 1961,6, 85.178 R. B. Frydmctn, ATch. Bwchem. Biophys., 1963, 102, 242.175 J. Espada, J . Biol. Chem., 1962, 237, 3577.17p H. P. Ghosh and J. F'reiss, J. BioZ. Chem., 1966, 241, 4491.176 P. K. Pottinger and I. T. Oliver, Biochim. Biophys. Acta, 1962, 58, 303.176R. B. Frydman and C. E. Cardini, Bwchem. Bhphys. Res. Cmm., 1964,1'' R. B. Frydman and C. E. Cardini, Arch.Biochem. Biophys., 1966, 116, 9.178 H. P. Ghosh and J. h i s s , Biochem&ry, 1965, 4, 1354.17,407610 BIOLOGICAL CHEMISTRYand the formation of the latter during carbon dioxide fixation may represent aregulatory mechanism of starch ~ynthesis.~?~One of the unsolved problems of starch synthesis is the mechanismwhereby both linear and branched components are formed, and then organ-ised to give a granule. The granules from waxy maize contain amylopectinrather than a two-component starch, and show little or no UDPG cc-glucantransglucosylase activity 170 and only very limited ADPG a-glucan trans-glucosylase activity (about 10-20% of that shown by granules from normalMost of this activity is associated with a limited number ofgranules from the embryo and maternal tissue of the seed.180 The endo-sperm, which is the major side of starch synthesis and storage, was inactivewith ADPG.It is therefore possible that in maize, there are two separatepathways for the synthesis of the starch components; a nucleotide pathwayproducing amylose, and a phosphorylase-Q-enzyme system yielding amylo-pectin. In support of this view, the embryo of normal maize seeds has beenshown l*l to contain an ADPG a-glucan transglucosylase which differs inkinetic and other properties (but not in specificity) from the ADPG a-glucantransglucosylase present in the endosperm. In waxy mutants of maize, theactivity of the latter is selectively reduced.The synthesis of starch in ripening rice grains has been extensivelystudied by T.Akazawa and his co-~orkers.l8~-~8~ These grains containADPG-pyrophosphorylase, UDPG-pyrophosphorylase, an enzyme catalysingthe reaction :Sucrose + ADP (UDP) + ADPG (UDPG) + fructoseparticulate ADPG (or UDPG) a-glucan transglucosylases,l82 and smallamounts of ADPG and other nucleotides. The combined system can synthe-sise starch from either glucose 1-phosphate or sucrose via ADPG, or lessefficiently, UDPG. There are differences between the maize and rice en-zymes, since with the latter, glucose from ADPG was mainly incorporatedinto amylopectin whereas that from UDPG was transferred equally into thetwo components. It is possible that the conversion of sucrose into starchproceeds mainly by a reversal of UDPG-sucrose transglucosylase rather thanof the ADPG-sucrose transglu~osylase.~~3 Although the pathways for thesucrose-starch conversion in normal and glutinous varieties of rice aresimilar, the latter produces amylopectin rather than a two-componentstarch.Glutinous varieties produce a soluble ADPG a-glucan transgluco-sylase with a similar specificity to the particulate enzyme present in normalrice, but this physical difference in the enzymes does not explain the differentend-products. la4Sweet corn (Zea m y s ) synthesises both a granular two-component17* 0. E. Nelson and H. W. Rhea, Biochem. Biophys. Res. Comm., 1962, 9, 297.186 0. E. Nelson and C. Y. Tsai, Science, 1964, 145, 1194.1 8 1 T. Akatsuka and 0. E. Nelson, J . BioZ. Chem., 1966, 241, 2280.la8 T.Murata, T. Minamikawa, T. Akazawa, and T. Sugiyama, Arch. Biochem.Biophya., 1964,106,371 ; T. Murata, T. Sugiyama, and T. Akazawa, ibid., 1964,107,92.18aT. Murata, T. Sugiysma, T. Minamikawa, and T. Akazawa, Arch. Biochem.Bkphy.~., 1966, 118, 34.1 1 4 T. Murata and T. Akazawa, Arch. Biochem. Biophys., 1966, 114, 76MANNERS : STRUCTURE AND METABOLISM OF ~ L U C A N S 611starch and phytoglycogen. The endosperm contains several enzymes in-cluding UDPG-sucrose transglucosylase, ADPG-sucrose transglucosylase,ADPG and UDPG-pyrophosphorylase, and particulate ADPG and UDPGa-glucan transglucosylases .IS5 These enzymes will transfer glucose fromsucrose into the granular starch, the process being more effective in thepresence of ADP rather than UDP. The endosperm also contains a solubleADPG a-glucan transglucosylase which can transfer glucose residues fromADPG to phytoglycogen, but not from UDPG.186 This plant enzyme isunaffected by glucose 6-phosphate, calcium ions and cyclic adenylic acid;amylopectin, glycogen and maltosaccharides were good acceptors whereasamylose and starch were inactive.Sweet corn also contains two branchingenzymes l87 (one of which resembles Q-enzyme, and the other 188 is able tointroduce branch points into amylopectin), and two separate debranchingenzyrnes.l89 One of these (R-enzyme) acts only on amylopectin, whilst thesecond enzyme can hydrolyse the inter-chain linkages in both amylopectinand phytoglycogen. The location of all these enzymes within the varioussweet-corn cells, and the factors controlling their relative activities are notyet known.Many facets of starch synthesis merit continued investigation, includingthe mechanism for the concomitant production of the two components, thechange in the size and other properties of the granules, the increase in arnylosecontent, and DP of the amylose during growth, and the genetic factorsgoverning the formation of starches with low and high amylose contents.Metabolism of p-Gluciins.-AIthough the biosynthesis of cellulose hasbeen widely investigated, many details of the process are not yet k n 0 ~ n .l ~ ~Studies on the extracellular formation of cellulose by bacteria, especiallyby Acetobacter xylinurn and A . acetigenum, and also by Sarcim ~entriculi,~Slhave shown that UDPG was a glucosyl donor.Attempts to fmd a similarsystem in plants were initially unsuccessful. However, a cell-free preparationfrom mung bean seedlings and other plant tissues incorporated [14C]glucosefrom GDP-[14C]glucose into cellulose,lg2 and by chemical means, the forma-tion of new ~-1,4-glucosidic linkages was demonstrated. The enzyme pre-paration was inactive with UDPG, TDPG, ADPG, and CDPG and differedsignificantly from an A . xylinum preparation which could use both UDPGand TDPG as glucosyl donors. The same plant preparation also incorporatedmannose from GDP-mannose into a related glucomannan. 193 AlthoughGDPG is not widely distributed in plants, these tissues contain GDPG-pyrophosphorylase which catalyses the formation of GDPG and GTP anda-glucose 1 -phosphate.lg4lSb M.A. R. de Fekete and C. E. Cardini, Arch. Biochem. Biophys., 1964, 104, 173.lS6 R. B. Frydman and C. E. Cardini, Biochim. Biophys. Actu, 1965, 96, 294.ls8 D. J. Manners and J. J. M. Rowe, CJtem. and Id., 1964, 1834.IssD. J. Manners and K. L. Rowe, Arch. Biochem. Bwphys., 1967, lB, 585.loo J. A. Gascoigne, Chem. and Ind., 1963, 1580; see also H. K. Porter, Ann. Rev.lol E. Canale-Parola and R. S . Wolfe, Biochim. Biophys. Acta, 1964, 82, 403.lg2 G. A. Barber, A. D. Elbein, and W. Z. Hassid, J. Biol. Chem., 1964, 239, 4056.lg8 A. D. Elbein and W. Z. Hassid, Biochem. Biophys. Res. Comrn., 1966, 23, 311.lo4 G. A. Barber and W. Z. Hassid, Biochim. Biophya. Acta, 1964, 86, 397.N. Lavintman, Arch.Biochem. Bwphys., 1966,116, 1.Plant Physiol., 1962, 13, 319612 BIOLOCIICAL CHEMISTRYThe above observations would appear to establish GDPG as the glucosyldonor for cellulose synthesis. However, other workers 195 using differentplant tissues (e.g., Lupinus albus) and different concentrations of nucleotidesugar, Mg2+ and cysteine, have shown that UDPG may be an effectivedonor. With a cell-free oat coleoptile system, incorporation of glucose fromUDPG was greater than from GDPG, although the end-product fromUDPG, on enzymic hydrolysis, gave products other than cellobiose.~96With GDPG, cellobiose was the only product. More recently? by modifyingthe experimental conditions, glucose has been incorporated from UDPG intocellulose by a mung bean enzyme.197The studies on nucleoside diphosphate glucoses do not provide data on themode of formation of cellulose microfibrils within the plant cell-wall.Inbacteria, there is evidence of a glucose-lipid precursor produced within thecell which is converted extracellularly into cellulose.l90 A similar precursor ispresent in pea seedlings and oat coleoptiles, which may be formed in thecytoplasm, transported into the cell wall and the glucose moiety incorporatedinto the tip of a micr~fibril.~~* The physical factors involved in the layingdown of microfibrils are discussed e1se~here.l~~ It has recently been sug-gested 200 that there are two distinct stages in cellulose biosynthesis; the fistis very slow, involves nucleoside diphosphate glucose and yields only a smallamount of cellulose in the primary wall.In the second stage, there is a,rapid and substantial formation of cellulose in the secondary wall, whichmay be catalysed by a different enzyme system operating by a templatemechanism, A two-stage system of this type would be in accord with therelatively low incorporation of glucose from sugar nucleotides into cellulose(usually (20%) which is observed even under the most favourable experi-mental conditions.UDPG is the glucosyl donor for the biosynthesis of other @-glucans.During attempts to synthesise cellulose, a preparation from mung-beanseedlings catalysed the formation of an insoluble radioactive polysaccharidefrom UDP- [14C]glucose .201 It resembled cellulose in being insoluble inwater and dilute acid, but differed in being soluble in hot dilute alkali.Partial acid-hydrolysis showed it to be the P-l,S-glucan, callose.An en-zyme with a similar specScity has been isolated from the flagellate Euglenagracilis where it catalyses the synthesis of paramylon.202 Thisand the related Astasiu ocellata 204 also contain laminaribiose phosphorylasewhich catalyses the reaction :or-glucose l-phosphate + glucose + laminaribiose + inorganic phosphateIt should be noted that this reaction, and those leading to the formation of195 D. 0. Brummond and A. P. Gibbons, Biochem. Z., 1965,342,309.196 L. Ordin and M. A. Hall, Phnt Physiol., 1966, in the press.1 9 7 R. W. Bailey and W. Z. Hassid, Phytochemktry, 1967,6,293.198 J. R. Colvm, Ca&.J. Biochem., 1961,39, 1921.199 J. R. Col?,.Cad. J . Bot., 1965, 43, 339.200 M. Marx-Figmi, Nature, 1966, 210, 754, 756.201 D. S. Feingold, E. F. Neufeld, and W. 2. Hassid, J . BioZ. Chem., 1958, 833, 783.80s L. R. Marechal and S . H. Goldemberg, J. Biol. Chem., 1964, 239, 3163.103 S. H. Goldemberg, L. R. Marechctl, amd B. C. DeSouza, J . Biol. Chem., 1966,241,194 D. J. Manners and D. C. Taylor, Biochem. J., 1965, 94, 17P.45MANNERS : STRUCTURE AND METABOLISM OF GLUCANS 613/3-glucans, including a /3- 1,2-1inked glucan from Bhizobium japonicum, 205proceed with inversion of the configuration of the transferred glucoaylresidue.The enzymic degradation of cellulose, laminarin, and other /l-glucans hasbeen reviewed in detail elsewhere.49~ 2*6206 It.A. Dedonder and W. Z. Hassid, Biochim. Bwphys. Acta, 1964, 90, 239.“ Advances in Enzymic Hydrolysis of Cellulose and Related Materials,” ed. E. T.Reese, Pergamon, Oxford, 19634. THE ENDOPEPTIDASES OF VERTEBRATES *By A. P. Ryle(Department of Biochemistry, University of Edinburgh)Inhduction.-The interests which have prompted the numerous investigct-tions of proteolytic enzymes have been diverse, overlapping, and almost aanumerous as the proteinases studied. The original interest was, of course,in the function of enzymes as part of the digestive systems of organisms andthe first investigations quite quickly led to the realisation that the gastro-intestinal tracts of mammals receive secretions containing several dflerentproteinases and peptidases and that yet others were produced by otheranimals, by plants, and by micro-organisms.These interests, in the morephysiological aspects of proteinases, are outside the scope of this Report.Studies of structure and mode of action. As techniques of protein purifica-tion improved, many proteolytic enzymes were obtained in a high state ofpurity, often as crystalline preparations, although many of the crude extractsand several of the crystalline preparations contain more than one enzymicallyactive component. The purified materials then served as the starting pointfor a second type of investigation into the structure and mode of action ofthe proteinases. These enzymes have proved to be good subjects for studiesinto the mode of action of enzymes.Not only do they catalyse reactions(the hydrolysis of amide and ester bonds) whose nonenzymic counterpartsare well understood, but also it turns out that many of the enzymes them-selves are simple, being monomeric and of low molecular-weight (25,000-50,000). The structural studies have involved the specialised techniques ofanalytical protein chemistry (see the Reviews by Smyth l). The studies ofmodes of action have involved investigation of the kinetics of reactionscatalysed by the unmodified enzymes, with a variety of substrates and in-hibitors. These have often been supplemented by studies involving the useof relatively specific chemical modifications of the enzyme to produce inactivederivatiyes, or even better, derivatives with modified activity whose catalyticproperties can in turn be investigated.With such studies, two important considerations must be borne in mind.1 D. G.Smyth, Ann. Reports, 1963,40, 468; 1964,41, 607; 1965, 42, 488.~- ~ ~. ____ _ _ - * Abbreviations and symbols. The following will be used without further explanationTPCM (tosyl phenylalanyl chloromethane) : L-( 1 -tosylamido-2-phenyl)ethyl chloromethylTLCM (tosyl lysyl chloromethano) : L-( 1 -tosylamido-5-amino)pentyl chloromethylDFP : di-isopropylphosphorofluoridateDIP : di-isopropylphosphorylATEE : N-rtcetyl-L-tyrosyl ethyl esterBAEE : a-N-benzoyl-L-arginine ethyl esterAmino-acid residues: symbols of IUPAC-IUB tentative rules, Biochem. J., 1967,102,23.Km : an experimentally observed Michaelis constant, the concentration of substrateK,, K, : dissociation constants of enzyme-inhibitor and enzyme-substrate complexhat: V,-/Eo where E , = total enzyme concentration.ketoneketonewhich gives half the maximal velocity, V-.respec t i d y RYLE : ENDOPEPTIDASES O F VERTEBRATES 61 5The stoicheiometric modification of one mole of an amino-acid residue of aparticular type per mole of enzyme does not necessarily mean that only onespecific residue has undergone reaction.Ribonuclease can be alkylated byiodoacetate with the loss of one mole of histidine in the overall amino-acidcomposition, but in fact two different residues react in such a, way that thereaction of each prevents the reaction of the other in the same molecule.2In such a case it should be possible to separate the two derivatives andexamine their separate properties.The second consideration, which has often been discussed (see, forexample, ref.3), is that loss of activity on modification of a specific amino-acid residue does not necessarily imply that that residue lies in the activecentre, even if protection against modification is provided by substrates orcompetitive inhibitors. A change of conformation on binding substrate,inhibitor, or modifying reagent can equally explain the results.Proteinases cts tools for structural analysis. The structural studies in turnhave prompted a third interest, a search for proteinases with new, butsharply defined specificities, because enzymes with such properties are in-valuable in investigations into the amino-acid sequences of proteins.Trspsinis probably the most useful of the proteinases currently available, becauseof its fairly high specificit'y directed towards bonds involving the carbonylgroup of arginine and lysine residues. Protein chemists would be delightedto find more enzymes with specificities as sharp as, but different from, thatof trypsin. Hopes that a newly described endopeptidase would have thesedesirable properties have often been dashed, but thermolysin, a proteinasefrom BuciZlus thermoproteolyticu, shows a high degree of specificity for bondsinvolving the amino-group of amino-acids with bulky hydrophobic side-chains.* A proteinase in snake (Crotalus) venom,5 and an elastase fromPseudomonas 13 show similar specificities. Since these enzymes break thepeptide chain at the amino-side of residues at whose carboxyl side chymo-trypsin attacks, they could prove useful for the correct ordering in a sequenceof the peptides obtained by chymotryptic attack.In the meanwhile, chemical modifications have been used effectively tonarrow, or to widen, the specificity of trypsin.The €-amino-groups of thelysine residues of a protein can be blocked ' 9 13 with an easily removed reagentso that trypsin will then attack only at the arginine residues. After removalof the blocking agent, the lysine residues are once more susceptible to attack.The specificity of the enzyme can be, in effect, widened by conversion ofcysteine residues (naturally occurring, or produced by the reduction of cystine)into X-aminoethylcysteine residues which are susceptible to tryptic attack.9, 10a A.M. Crestfield, W. H. Stein, and S . Moore, J. BioE. Chem., 1963, 238, 2413.a G. H. Dixon and H. Schachter, Canad. J . Biochem., 1964,42, 695.H. Matsubara, R. Sasaki, A. Singer, and T. H. Jukes, Arch. Biochem. Biophys.,G. Pfleiderer and A. Krauss, Biochem. Z., 1965, 342, 85.K. Morihara, H. Tsuzuki, and T. Oka, Proc. Syntp. Enz. Chem., 17th Tokushima,' T. C. Merigan, W. J. Dreyer, and A. Berger, Biochim. Biophys. Aeta, 1962,62,122.R. F. Goldberger and C. B. Anfinsen, Biochemistry, 1962, 1, 401.H. Lindley, Nature, 1956,178, 647.1966,115, 324.1965, p. 293. Quoted in ref. 4.lo M. A. Raftery and R. D. Cole, J . Biol. Chem., 1966, 241, 3457616 BIOLOGICAL CHEMISTRYThis technique has been successfully applied to the determination of theamino-acid sequence of trypsinogen.l1 The successful conversion of serineresidues into S-aminoethylcysteine residues can be achieved by O-tosylationfollowed by displacement of the tosyl group by mercaptoethylamine.12 Ifthe reaction will proceed with large peptides sufficiently quantitatively, itshould prove a very useful weapon in the protein chemist’s armoury. Asour understanding of the specificity of other proteinases grows theirusefulness in analytical studies may also be improved by such chemicalstretching.Before leaving this topic it should be noted that trypsin is not entirelyspecific for arginyl and lysyl bonds (among those formed by natural amino-acids).The often reported fission of peptide bonds of other amino-acidsmay be due, not to a chymotrypsin contaminant, but to trypsin itself.Chromatography, treatment with acid, or with the specific chymotrypsininhibitor TPCM,139 l4 chromatography in urea solution,15 or treatment withanother specific chymotrypsin inhibitor, diphenylcarbamyl chloride,lS allfailed to abolish the apparent chymotryptic activity of trypsin. Conversely,chymotrypsin, treated with a specific trypsin inhibitor, TLCM, still showsactivity against trypsin substrates.ld It has been suggested 1 4 8 l7 thattrypsin used for structural studies should routinely be treated with a specificchymotrypsin inhibitor.The groups of proteinases. Bender and KBzdy l8 in a recent admirableReview of the mechanism of action of proteolytic enzymes, considered themdivided into groups : the serine proteinases (including chymotrypsin, trypsin,elastase, and thrombin) which are all inactivated by DFP with phosphoryla-tion of one serine residue (the “ active serine ”), the cysteine proteinases(papain, ficin, bromelain), metal-containing peptidases (aminopeptidases,carboxypeptidases, and dipeptidases) and proteinases active at low pH-values(pepsin and rennin).To these one might add the cathepsins (intracellularenzymes) which may not really form a homogeneous group, and a variety ofproteinases and peptidases of microbial origin, which may eventually befound to belong to one of the other groups.It should be noted in passing that the terms proteinase (= protease= proteolytic enz-yme) and peptidase are not clearly defined, and overlap;they now refer to what is believed to be the prime biological function of theenzyme under consideration.Many of the proteinases of the serine grouphydrolyse esters faster than peptides and are sometimes referred to asesterases. This usage seems unnecessarily to ignore their biological originand undoubted biological function.l1 K. A. Walsh, D. L. KaufTmctn, K. S. V. Sampath Kumar, and H. Neurath,I* C. Zioudrou, M. Wilchek, and A. Patchornik, Biochemistry, 1965, 4, 1811.lS 5. Maroux, M. Rovery, and P. Desnuelle, Biochim. Bwphy8. Acta, 1966, 122,l* E. B. Ong and 0. Schiilmann, 2. physwl. Chem., 1966,344,13.lo R. D.Cole and J. M. Kinkade, J. Biol. Chem., 1961, 236, 2443.l7 B. F. Erlanger and F. Edel, Biochemistry, 1964, 3, 346.l8 M. L. Bender and F. J. Kbzdy, Ann. Rev. Biochem., 1965, 34, 49.Proc. Nat. Acad. Sci. U.S.A., 1964, 51, 301.147.K. Takahashi, J. Biol. Chem., 1965, 240, 4117KYLE: ENDOPEPTIDASES O F VERTEBRATES 617The Serine Proteinases.-Trypsifi and chymotrypsin, These enzymeswere the subject of earlier Reviews.lS-21The pancreatic juice of cattle contains the zymogen trypsinogen andabout equal parts 22 of chymotrypsinogen A (iso-electric point ca. pH 9.3)and chymotrypsinogen B20 (I.E.P. ca. pH 5.2) whose products of activationhave similar specificities. Most detailed studies have been made with chymo-trypsinogen A and with trypsinogen, and with crystalline preparations of thederived enzymes.The changes which occur on activation of the zymogens have been under-stood for some time.20, 2 1 Trypsinogen undergoes autocatalytic activationby hydrolysis of a lysyl-isoleucine bond to liberate a hexapeptide (Val.-Asp,.Lys) and the enzyme, which consists of it single polypeptide chain withsix disulphide bridges. The activation is accelerated by, and the trypsinprotected from autolysis by, the presence of calcium ions.Chymotrypsinogen A is activated initially by the tryptic hydrolysis ofan arginyl-isoleucine bond to form chymotrypsin &. (The chymotrypsinswere originally identified merely by Greek-letter prefixes, but since similarenzymes may be formed from chymotrypsinogen B, it is better to use theGreek letters as suffixes to the Roman letter to identify the zymogen.)Chymotrypsin A, may then suffer chymotryptic hydrolysis of a leucyl-serinebond to liberate the dipeptide seryl-arginine and chymotrypsin 4.Inchymotrypsins A, and &, the short chain of amino-acids, the A-chain,derived from the N-terminal end of the zymogen remains attached to therest of the molecule by a disulphide bond. Chymotrypsin & can then becleaved again by chymotrypsin at two points in the long chain to liberatea dipeptide (Thr.Asn) and chymotrypsin 4 which contains three peptidechains linked by disulphide bridges. In most investigations of chymotrypsin,crystalline preparations of chymotrypsin & have been used. Two otherchymotrypsins (AB and 4) have been studied less and may bemerely differentcrystalline forms of A,.22 The molecular weights 21 of all the enzymes andzymogens are about 24,000-25,000.Preparations of chymotrypsins and trypsins and their homogeneity.Mostof the kinetic and chemical studies reported have been performed withcrystalline chymotrypsin A, of commercial origin and the kinetic parametersdetermined in different laboratories have generally been in reasonable agree-ment. Niemann 22 stressed the need for further comparative studies ofdifferent preparations and his concern has proved well justified. A recentreport 23 has shown the presence, in all of the commercial crystalline prepara-tions examined, of one or two types of contaminant which were detected andestimated by temperature- jump relaxation experiments.In these experi-ments the temperature of a system at equilibrium is very rapidly raised andthe establishment of the new equilibrium is observed. The class I contamin-l@ C. Niemann, Science, 1964, 143, 1287.2o M. Rovery, Bult. SOC. Chim. Biol., 1964, 46, 1757.21 P. Desnuelle, in “The Enzymes,” ed. P. D. Boyer, H. Lardy, and K. Myrbiick,22 P. J. Keller, E. Cohen, and H. Neurath, J. Biol. Chem., 1958, 233, 344.O 3 A. Yapel, M. Han, R. Lumry, A. Rosenberg, and D. F. Shiao, J . Amer. Chem. Soc.,1960, Academic Press, New York and London, 2nd edn., vol. 4, pp. 93, 119.1966,88, 2573618 BIOLOGICAL CHEMISTRYants, probably products of autolysis, interfere in such studies of the protoniodissociations.Enzymes may be assayed not only by measurement of therate of a catalysed reaction but aIso by " all-or-none " z4 assays in which a,reagent is used (e.g., trans-cinnamoylimidazole for chymotrypsin) whichreacts rapidly with the enzyme to form a stable intermediate and a stoicheio-metric quantity of another product (imidazole). The number of moles ofreagent used in the rapid reaction is equal to the number of moles of activecentre present. The class I1 contaminants, of unknown nature, do not inter-fere with, and so cannot be detected by, such all-or-none assays but areinhibitory in rate assays. Steady-state tests for this class of contaminantare, however, not completely reliable because of its slow dissociation fromthe enzyme. Chymotrypsin A, can fortunately be simply and reliably freedfrom both types of contaminant by passage through Sephadex G-25 in dilutehydrochloric acid at pH 3.The authors of this disturbing report state thatquantitative rate data must be treated as unreliable until verified withpurified protein, and this reservation should be borne in mind in the sectionswhich follow. Contamination of the types described is not likely to haveaffected the results of studies on the primary structure.A new preparation of trypsin and chymotrypsin from beef pancreas 26gives chymotrypsin of similar purity to tJhe crystalline enzyme in betteryield and in a shorter time. It remains to be seen whether this material isfree of the contaminants reported above.23 The presence of a t least twoactive components in three-times crystallised chymotrypsin has been re-ported.2s The components differed in their specific activity both in rateand in all-or-none assays and in their stability in the absence of calcium ions.Recently, chymotrypsins and chymotrypsinogens have been isolated fromchicken,27* 2g spiny dogfish,29 and f i n - ~ h a l e .~ ~Isolation and homogeneity of trypsin. Most investigators have used pre-parations of several-times crystallised trypsin, but there are reports of morethan one active component in such preparation^.^^ Since chymotrypsinmay be contaminated with material that can only be satisfactorily defectedby relaxation technique^,^^ it would be useful to subject trypsin preparationsto similar tests. More homogeneous trypsin might be obtained by activationof trypsinogen under controlled conditions (cf., pepsin 32); the products ofactivation in the presence and the absence of calcium ions are probablyidentical and the yield of trypsin is higher, and of inactive material lower, ifcalcium is present.33The isolation, physical properties, and amino-acid analysis of turkeytrypsin have been described.27~ 28 The enzyme is not unlike bovine trypsin.24 G. R. Schonbaum, B. Zerner, and 31. L. Bender, J . Bwl. Chem., 1961,236,2930.2s A. S. Tsiperovich and M. V. Kolodzeyskaya, Biokhimiya, 1966, 31, 564.26 B. F. Erlanger, A. G. Cooper, and A. J. Bendich, Biochemistry, 1964, 3, 1880.28 C. A. Ryan, J. J. Clary, and Y. Tomimatsu, Arch. Biochem. Biophys., 1965,110,as J.W. Prahl md H. Neurath, Biochemistry, 1966,5, 2131.36 Y. Matsuoka and A. Koide, Arch. Bwchem. Bwphys., 1966,114,422.3l A. Iachan, G. B. Domont, L. V. Disitzer, and J. C. Perrone, Nature, 1964, $308,s2 T. G. Rajagopalan, S . Moore, and W. H. Stein, J . Biol. Chem., 1966. 241, 4940.33 L. P. Chao and I. E. Liener, Biochim. Biuphys. Acta, 1965, 98, 508.C . A. Ryan, Arch. Biochern. Biophys., 1965, 110, 169.175.43; P. 0. Ganrot, Acta Chem. Scad., 1966, 20, 175RYLE : ENDOPEPTIDASES O F VERTEBRATES 619Primary structure of chymotrypsinogen, trypsinogen, and the derivedenzymes. Hartley 34 suggested a complete sequence of 246 amino-acidresidues for chymotrypsinogen A; this sequence has been modified by furtherwork from the same laboratory 35 and independent studies of the peptidechains of the enzymes 36 have provided confirmation of the suggested struc-ture.An almost complete structure determined independently by $omand his colleagues 37 differs from that of Hartley in the sequence at two pairsof residues and in the allocation of six side-chain amide groups. Thearrangement of the disulphide bonds has also been determined; 38 a skvlebond links the A-chain (by the N-terminal half-cystine) to the B-chain andanother links the B- and C-chains, the B-chain includes one intra-chaindisulphide and the C-chain two.The primary structure of trypsinogen has also been studied in twolaboratories. Walsh and Neurath 39 published an almost complete sequenceof 212 amino-acid resides with which the complete sequence of Sorm andhis colleagues 40 is in good agreement, The arrangement of the disulphidebridges has also been determined in two laborat~ries.~lThe N-terminal sequences of the trypsinogens of other species differ :Pig : Phe.Pro.Thr.Asp l.Lys-Sheep : 43 Phe.Pro .Val.Asp,.Lys-and Val. Asp ,.Lys-.It is not known whether different individual sheep produce different trypsino-gens or whether some, or all, produce both. A sequenceof 13 residues con-taining the active serine of porcine trypsin is identical with that of beef exceptfor the transposition of a glycine residue, but the sequence from the 6th tothe 16th residue at the carboxyl side of the serine is considerably ~lifferent.4~Chymotrypsin €3, chymotrypsin C, and pancreuto-peptiduae E (elustase).Chymotrypsinogen B has been purified and amino-acid analyses have beenreported 4 5 9 46 which show a composition somewhat different from that ofchymotrypsinogen A, although the molecular weight (25,000) is about the34 B.S. Hartley, Nature, 1964, 201, 1284.35 B. S. Hartley and D. L. Kaufban, Biochem. J . , 1966,101, 229.36 S. C. Glsuser and H. Wagner, Biochem. Bwphys. Res. Comna., 1965, 21, 494;S. Maroux and M. Rovery, Biochim. Biophys. Acta, 1966, 113, 126.87 B. Meloun, V. Kostka, K. VanBEek, I. Kluh, and F. Sorrn, Coll. Czech. Ohm.Comm., 1966,31,312; L. Moravek, I. Kluh, J. M. Junge, B. Meloun, and F. Som, ibid.,p. 1604.s8 J. R. Brown and B. S . Hartley, Biochem. J . , 1966, 101, 214; Z. Prusfk, B. Keil,and F. gorm, Coll. Czech.Chem. Comm., 1966,31, 2565.s9 K. A. Walsh and H. Neurath, Proc. Nat. A d . Sci. U.S.A., 1964, 52, 884.40 0. Mikeg, V. TomQFiek, V. Holeygovskf, and F. gorrn, Biochim. Biophys. Acta,1966,117,281 ; 0. MikeFi, V. Holeygovsk9, V. TomGek, and F. Bonn, Biochem. Biophys.Res. Comm., 1966, 24, 346.41D. L. KaufTmsn, J . Mol. Bbl., 1965, 12, 929; V. HoleySovskf, V. Tomsliek,0. Mike;, A. S. Dsnilova, and F. Sorm, Coll. Czech. Chem. Cmm., 1965, 30, 3936.42 M. Charles, M. Rovery, A. Guidoni, and P. Desnuelle, Biochim. Biophys. Acta,1963, 69, 116.4 3 S. Bricteux-Gregoire, R. Schyns, and M. Florkin, Biochim. Biophys. Acta, 1966,12'7, 277.IP J. Travis and I. E. Liener, J . Biol. Chem., 1965, 240, 1967.0. Guy, D. Cratecos, M. Rovery, and P. Desnuelle, Biochim.Biophya. Acta, 1966,115, 404.48 L. B. Smillie, A. G. Enenkel, and C. M. Kay, J. Biol. Chem., 1966, 241, 2097620 BIOLOGICAL CHEMISTRYmme. The N-terminal sequence differs a t only one place in the first 17residues, serine-14 being replaced by alanine. Activation depends ontryptic hydrolysis of the arginyl-isoleucine (15-16) bond, as in chymotryp-sinogen A, to form chymotrypsin B,, but the 0-terminal dipeptide (rtlanyl-arginine) is not removed from the short ( A ) chain by autolysis. The bondsin the region of residue 150 of chymotrypsinogen B, which are hydrolysed bychymotrypsins A, or B, to form activatable neochymotrypsinogens B, havebeen identified.47 The B- and C-chains of a chymotrypsin B have beenseparated and analysed.48 The disulphide bonds of chymotrypsinogen Bare arranged similarly to those of chymotrypsinogen A although somedifferences in the amino-acid sequence are apparent.49The status of chymotrypsin C , which has been isolated from porcinepancreas and has a broader specificity than bovine chymotrypsin & (ref. 50)is less clear. The zymogen has been isolated chromatographically and isquite similar in amino-acid composition, N-terminal sequence, and activationbehaviour to a zymogen obtained as a fraction of bovine procarboxypepti-d a ~ e . ~ ~ Chymotrypsin C and its precursor may be identical to a crystallineesteroproteolytic enzyme S2 and its zymogen 53 isolated by a differentmethod from porcine pancreas. The zymogen was only separated withdifficulty from an apparent anionic porcine trypsinogen which has not other-wise been reported.Elastase has been purified from a commercial pancreatic extract bychromatographic methods, and has a broad specificity (on the chains ofoztidised insulin) different from that of trypsin or ~hymotrypsin.~~ It issensitive to inhibition by DFP, and a phosphorylated serine residue is thenfound in the same sequence (Asp.Ser.Gly) 55 as with the other serine pro-teinases.The proteolytic and elastolytic activities ran in parallel duringthe chromatographic fractionation, and a similar preparation has been suc-cessfully used for the identification of unique amino-acid sequences aroundthe cystine residues.56 These sequences are similar to those of trypsin andchymotrypsin.Despite this evidence of homogeneity, the protease activityof a similar preparation was strongly inhibited by soybean trypsin inhibitor,while the elastase activity was unaffected and strong salt solutions had thereverse effect.5' It was suggested that the enzyme possesses two activesites responsible for the different kinds of activity. However, the possibilitythat the effect of the inhibitors is not on the enzyme, but on the proteinsubstrates should not be neglected. A proteolytic component, lacking4 7 0. Guy, M. Rovery, and P. Desnuelle, Biochim. Biophys. Acta, 1966, 124, 402.48 C. 0. Parkes and L. B. Smillie, Biochirn. Biophys. Acta, 1966, 113, 629.49 L. B. Smillie and B. S. Hartley, J . MoZ. Biol., 1966, 12, 933.5 0 J. E. Folk and P.W. Cole, J. BioZ. Chem., 1966, 240, 193.51 J. R. Brown, R. N. Greenshields, M. Yamasaki, and H. Neurath, Biochemistry,1963, 2, 867; J. R. Brown, M. Yamasaki, and H. Neurath, ibid., p. 877.5 2 E. C. Gjessing and J. C. Hartnett, J . Biol. Chem., 1962, 237, 2201.53 B. McConnell and E. C. Gjessing, J . Biol. Chem., 1966, 241, 573.64 M. A. Naughton and I?. Sanger, Biochem. J., 1961, 78, 156.5 6 M. A. Naughton, F. Sanger, €3. S. Hartley, and D. C. Shaw, Bwchem. J., 1960,56 B. S. Hartley, J. R. Brown, D. L. Kadfman, and L. B. Smillie, Nature, 1966,m R. L. Walford and B. Kickhbfen, Arch. Bwchem. Bwphys., 1962, 98, 191.77, 149.207, 1157RYLE : ENDOPEPTIDASES OF VERTEBRATES 621elastase activity has also been separated from crystalline elastase; 58 itsrelationship to the other enzymes of the pancreas has not been clarified.The similarities of primary structure of the serine proteinmes.Structuralsimilarities within this group of enzymes were foreseen by Sorm and Keil 59and have already been discussed by several authors.20, 393 569 60 They includethe following. (a) Identity, or near-identity of the sequence of amino-acidsround the active serine residue (see Table). In another serine proteinase,TABLE Amino-acid sequences round the active serine residues ofSome enzymesChymotrypsin A (beef)56 -Cys.Met . Gly . Asp.Ser . Gly .Gly .Pro .Leu.Val .C ys-Chymotrypsin B (beef)66 -Cys.Met.Gly.Asp.Ser.( Gly,,Pro,Leu).Val.Cys-Trypsin (beef)ssm *O -Cys . Gln . G1 y . Asp. Ser . G1 y .G1 y .Pro .Val .Val.Cys-Trypsin (pig) 44 -C y s . Gln . Gl y . G1 y . Asp. Ser. Gly .Pro .Val.Val . C ya -Elastase (pig)56 -Cys.Gln. Gly. Asp.Ser. (Gly,,Pro) .Leu.His.Cys-Subtilopeptidaseb -Asn.GIy.Thr.Ser.Met.Ala.Ser.Pro.His-subtilopeptidase, (outside the scope of this Report) despite many othersimilarities to the mammalian enzymes, the active serine is not preceded byan acidic amino-acid residue. (b) Trypsin and chymotrypsins A and Rcontain a pair of histidine residues (Nos. 40 and 57 in the chymotrypsinogenA sequence) quite widely separated in the linear polypeptide chain, butheld close together by an intra-chain disulphide bond:I IHis.Phe.Cys . . . . . . . . . His.CysIn elastase the same arrangement is found with replacement of phenylalanineby threonine.e2 The imidazole rings of histidine-57 in chymotrypsin and ofthe homologous residue in trypsin are alkylated a t N-3 upon inhibition ofthe enzymes with the substrate analogues TPCM and TLCM 66-68respectively. (c) The new N-terminus liberated in the essential activationstep by tryptic hydrolysis ia isoleucyl-valyl-glycyl- in trypsin and chymo-trypsin. (d) If a few deletions in one sequence or the other are permittedabout 40% of the amino-acid residues of trypsinogen and chymotrypsinogenin homologous positions are identical.39 ( e ) Four of the five disulphidebridges of chymotrypsinogen are exactly homologous with four of the sixbridges in trypsinogen. (f) Six of the nine proline residues of the twozymogens are found in homologous positions.These imino-acids placerestrictions on the possible conformation of the polypeptide chain, as do the68 V. Ling and R. A. Anwar, Bwchem. Biophys. Res. Comm., 1966, 24, 593.ss F. gorm and B. Keil, A&. Protein. Chem., 1962, 17, 167.6o F. gorm, V. Holeygovskf, 0. MikeB, and V. Tomaiiek, Coll. Czech. Chem. Comm.,1965,30, 2103.61 R. A. Oosterbaan and J. A. Cohen, in “Structure and Activity of Enzymes”,ed. T. W. Goodwin, J. I. Harris, and B. S. Hartley, Academic Press, New York, 1964,p. 87.62 L. B. Smillie and B. S. Hartley, Biochem. J., 1966, 101, 232.63 E. B. Ong, E. Shaw, and G. Schollman, J . BWZ. Chem., 1965,240, 694.G. Schollman, Biochem. Z., 1965, 343, 103.65 B. Meloun, D. PosfiiilovB, Biochem. Biophys. Acta, 1964, 92, 152.V. TomBEiek, E.S. Severin, and F. sorm, Biochem. Biophys. Res. Comm., 1966,6 7 0. Scholhnan, 2. Naturforsch,. 1966, 21b, 194.68 P. H. Petra, W. Cohen, and E. N. Shaw, Biochem. Bwphys. Res. Comm., 1965,20, 545.21, 612622 BIOLOGICAL CHEMISTRYdisulphide bonds. It has been suggested 6o that the similarity of structureof trypsinogen and chymotrypsinogen will be found to extend to the second-ary and tertiary levels.Secondary and tertiary structure of chymotrypsinogen A and chyrnotrypsins.Detailed results are not yet available. X-Ray diffraction studies of isomor-phous crystals containing heavy-atom replacements have identified anarrangement in which pairs of molecules in the crystal are related by a 180"rotation; 69 this may be connected with the dimerisation of chymotrypsinin concentrated sol~tion.7~, 71 By an analysis of the structure of chymo-trypsinogen at 4 A res~lution,~~ the absolute configuration of the moleculehas been identified and some parts of the peptide chain, but not its wholecourse, could be traced.Extensive regions of a-helix were not apparent.Studies of the optical rotatory dispersion and circular dichroism 73 indicatedlittle change in the context of a-helix on activation of chymotrypsinogen,but Fasman, Foster, and Beychok,74 using the same methods over a greaterrange of wavelength, concluded that about 13% (i.e., 32) of the residues ofchymotrypsinogen A are in a-helical regions and that on activation these areextended to include 10-15 more residues.The location of the active sites within the crystal has been plotted bycomparison of the diffraction of crystals of different specifically inhibitedchymotryp~ins~7~ and Kraut and co-workers 73 stressed that X-ray diffrac-tion studies would be greatly helped by the availability of isomorphousderivatives having heavy atoms covalently bound at well-defined sites.Advantage has been taken of the specific enzymic activity to prepare suchderivatives.76 Kraut,?' reviewing the X-ray diffraction studies of theseproteins, reports that the difference Patterson series of native chymotrypsinand the p-mercuribenzenesulphonyl derivative revealed no significant differ-ences, except those due to the substitution by the reagent itself.The impli-cation that changes of conformation do iiot occur on sulphonylation isdifficult to reconcile with physicochemical data which indicate a change inconformation upon acylation or phosphorylation of the enzyme at theactive site.78-826 9 D.M. Blow, H. G. Rossmann, and B. A. Jeffery, J . MoE. BioE., 1964, 8, 65.7 0 G. W. Schwert and M. A. Eisenberg, J . Biol. Chern., 1949,179, 665.71 F. 5. KBzdy and M. L. Bender, Biochemistry, 1965, 4, 104.7 2 J. Kraut, D. F. High, and L. C. Sieker, Proc. Nut. Acud. Sci. U.S.A., 1964,51,839.73 D. N. Raval and J. A. Schellman, Biochim. Biophys. Acta, 1965, 107, 463;R. Biltonen, R. Lumry, V. Madison, and H. Parker, Proc. Nut. Acad. Sci. U.S.A., 1965,54, 1018, 1412.74 G. D. Fasman, R. J. Foster, and S . Beychok, J. 1MoE. Biol., 1966, 19, 240.75 P.B. Sigh, H. C. W. Slchner, C. L. Coulter, 3'. Kallos, H. Braxton, and D. R.Davies, Proc. Nat. Acad. Sci. U.S.A., 1964, 51, 1146; P. B. Sigler, B. A. Jeffery, B. W.Matthews, and D. M. Blow, J . MoZ. BWE., 1966, 15, 175.76 V. M. Stepanov and L. P. Matyash, Biochem. Biophys. Acta, 1966, 124, 406;D. Rizok and J. Kallos, Biochem. Biophys. Res. Comm., 1965, 18, 478.7 7 J. Kraut, Ann. Rev. Biochem., 1965, 34, 247.78 H. L. Oppenheimer, B. Labouesse, and G. P. Hess, J . BioE. Chem., 1966,241,2720.7B T. C. Bruice, Proc. Nat. Acad. Sci. U.S.A., 1961, 47, 1924.8o (a) A. Y. Moon, J. Mercouroff, and G. P. HOSS, J . BioE. Chenz., 1965, 240, 717;( b ) A. Y. Moon, J. M. Sturtevant, and G. P. Hess, ibid., p. 4204.81 H. Weiner and D. E. Koshland, jun., J .MoE. Biol., 1965, 12, 881.82 I. A. Bolotina, M. V. Volkenstein, and 0. P. Chikalovs-Luzina, Biokhirniya, 1966,51, 241RYLE : ENDOPEPTIDASES OF VERTEBRATES 623Chymotrypsin. T h acyl-enzyme intermediate. Massive evidence l8 nowsupports the hypothesis that the chymotrypsin-catalysed hydrolysis of labilesubstrates occurs in three steps : the formation of an enzyme-substrate(Michaelis) complex, conversion of the complex into an acyl-enzyme withliberation of one of the products and then hydrolysis of the acyl-enzyme toregenerate the free enzyme (Equation 1).k+, k+s k+*E + S e E S + E S ’ - - + E + P,k-1 +PIEvidence supporting the hypothesis includes the isolation of stable di-isopropyl phosphoryl- or acetyl-enzyme after stoicheiometric (1 : 1) reactionwith DFP or p-nitrophenyl acetate, observation of a “ burst ” of nitro-phenol liberation on addition of the enzyme in the second reaction, thedetection by spectrophotometric means of many other acyl-enzymes, andisolation of stable, inactive, diphenylcarbamyl-, and sulphonyl-enzymes.Degradation of the acetyl- and phosphoryl-enzymes 83 has shown thatthese groups are attached to the side chain of a specific ccactive serine”residue (No.195 in the chymotrypsinogen A sequence) and the same residuehas been identified as the site of substitution by phenylmethanesulphonyland dimethylaminonaphthalenesulphonyl groups. 84Two questions concerning the acyl-enzyme intermediate have causedconsiderable debate : whether such an intermediate exists in the hydrolysisof specific as well as of labile substrates, and whether the acyl-serine is theobligatory form of the intermediate.The experiments of Bender and hiscolleagues 85--89 answer the first question affirmatively. Two objections tothe acyl-serine intermediate as the obligatory normal form have been raised.The degradation studies may involve transfer of the acyl group t o theactive serine ” when the enzyme is denatured or degraded. Although theabsorption spectra of denatured acylacryloyl-chymotrypsins are consistentwith an ester linkage, the direction of the shift of absorption maximum ondenaturation is the opposite to that expected for transfer of the chromo-phoric acyl group to water from a less polar environment such as the activecentre is held to be.9lBender and KQzdy take the view that since the same spectrophoto-metric anomaly is seen with several enzymes, including the less-closely(683 R.A. Oosterbaan, P. Kunst, J. van Rotterdam, and J. A. Cohen, Biochim.Biophys. Acta, 1958, 27, 549, 556; 0. H. Dixon, D. L. Kauffimn, and H. Neurath, J .Bid. Chem., 1958, 233, 1373; R. A. Oosterbaan, M. van Adricliem, and J. A. Cohen,Biochim. Biophys. Acta, 1962, 68, 204.8 4 A. M. Gold, Biochemistry, 1965, 4, 897.B. Zerner, R. P. M. Bond, and M. L. Bender, J . Amer. Chem. Soc., 1964,86,3674.86 M. L. Bender, G. E. Clement, F. J. Kbzdy, and H. d‘A. Heck, J . Amer. Chem. SOC.,87 F. J. KBzdy, G. E. Clement, and M. L. Bender, J . Amer. Chem. SOC., 1964,86,3690.8 8 M. L. Bender, G.E. Clement, C. R. Gunter, and F. J. Khzdy, J . Amer. Chem. Soc.,1964,86, 3680.1964,86, 3697.M. L. Bender and F. J. KBzdy, J. Amer. Clbem. Soc., 1964,86, 3704.S. A. Bernhard, S. J. Lau, and H. Noller, BiochemGtry, 1965,4,1108.Dl J. Kallos and K. Avatis, Biochemistry, 1966, 5, 1979; R. Wildnrcuer and W. J.Crcnady, ibid., p. 2885624 BIOLOGICAL CHEMISTRYrelated subtilisin, it is more likely to be due to a common physical perturba-tion than a common chemical migration. The serine location of the acylgroup is also supported by the thermodynamics of hydrolysis of the acyl-enzyme g2 and by the stoicheiometry of the proton liberation during itsformation.93 Evidence that specific substrates too may acylate the seriaehas been obtained 94 by isolation of a large peptide believed to include theactive serine and containing N-acetyl-3-nitrotyrosine from a reaction mixtureof chymotrypsin and the corresponding ester at low pH.It has been possible to explain the absence of catalytic activity in stableacyl-enzymes e.g., DIP-chymotrypsin, while denying an essential catalyticrole to the serine residue, and accepting that the serine is substituted in thenative enzyme, by postulating steric hindrance by the bulky substituent.This argument is not valid for an inactive derivative in which the " activeserine " has been converted specifically into a (smaller) dehydroalanineresidue .95The balance of the evidence supports an acpl-serine enzyme as an inter-mediate in the hydrolysis of all substrates.Chemical modi$btion of other groups in chymotrypsin.The inhibition ofchymotrypsin by the substrate analogue TPCM63-65 has already beennoted; the reagent specifically alkylates one histidine residue (no. 57). Thegeneral technique of " af6nity labelling " to modify groups in or near theactive centre has also been applied to chymotrypsin with a Werent type ofreagent, which modifies another residue. A number of reagents related toa-bromoacetanilide, having reactive bromine atoms at varying distancesfrom an aromatic ring, have been used for " mapping " the distances be-tween different parts of the active centre.g6 A methionine residue, No. 192,lying near the active serine, is alkylated to give derivatives with loweredcatalytic activity. A bifunctional reagent, the p-nitrophenyl ester ofbromoacetyl-waminoisobutyric forms an acyl-enzyme which at pH 7mostly decomposes to give the free enzyme; a t lower pH-values alkylationof methionine-192 occurs and the acyl-enzyme bond can then be hydrolysedto give a partly active derivative.These methionine-alkylated enzymes arestill capable of reacting with DFP 96 and the low activity in standard rate-assays is largely due to increases in K,-values and but little to decreasesin Vmx. The methionine residue thus seems to be involved in binding thesubstrate, but not in the catalytic activity.A similar result is obtained with chymotrypsin in which methionine-192has been specifically oxidised to the sulphoxide by photo-oxidation98 orwith periodate 99 or hydrogen peroxide.98, 100 However, when both methion-h e residues are oxidized with hydrogen peroxide in urea solution 100 or in9% P.W. Inward and W. P. Jencks, J . BWZ. Chem., 1965, 240, 1986.93 J. Keizer and S. A. Bernhard, Biochemistry, 1966, 5, 4127.9 4 V. Shalitin and J. R. Brown, Biochem. Biophys. Res. Cornm., 1966,84, 817.95 H. Weiner, W. N. White, D. G. Home, and D. E. Koshland, jun., J . Amer. Chem.9s H. J. Schramm and W. B. Lawson, 2. physiol. Chem., 1963, 332, 97.97 W. B. Lawson and H. J. Schramm, Biochemistry, 1965, 4, 377.98 H. Schachter and G. H. Dixon, J . BioE. Chem., 1964, B9, 813.99 3. R. Knowles, Biochem. J., 1965, 95, 180.SOC., 1966, 88, 3851.100 H. Weiner, C. W. Batt, and D. E. Koshland, jun., J . Bhl. Chem., 1966,841,2687RYLE : ENDOPEPTIDASES O F VERTEBRATES 625the presence of substrate a t low pH-values,lol 70% of the activity in all-or-none assays can be recovered on removal of the urea, but the specikity ofthe enzyme is markedly altered.It has no activity with N-acetyltyrosineamide, while the activity with N-acetyltyrosine ethyl ester is only slightlyreduced, like that of the mono-sulphoxide enzyme.Studies of the specificity of chymotrypsin 102-106 have led to the recog-nition of three parts of the active site which normally bind the ,!?-awl side-chain, the acylamido-group, and the sensitive bond respectively. ‘‘ Wrong-way” binding may occur, for example, when the acylamido-group is aro-matic and the ,!?-substituent small; the acylamido-group is then bound inthe aryl-binding site.,It is not clear, with the bromo-derivatives mentionedabove, to which site the bromo-group will be bound. With a-bromo-N-phenylethyla~etamide,~~ which alkylates an unidentified methionine, thephenyl group presumably enters the aryl site, and the a-bromoacetyl group(which is the same distance from the phenyl group as is the acetyl group in anormal substrate ATEE) may enter and react in the acylamido-site, but it ispossible that instead it enters the catalytic site. The same difficulty ariseswith N-a-bromoacetyl-a-aminoisobutyric acid nitrophenyl ester. Does thebromoacetyl group enter the aryl or the acylamido-site? It thus becomesimport,ant to discover which methionine residue is alkylated by cc-bromo-N-phenylethylacetamide and also by iodoacetyl phenylalanine esters 107 forwhich the ambiguity of the orientation of the inhibitor on the enzyme surfacepresumably does not arise.It is also clearly important to measure theactivity with amide as well as ester substrates. From the data presentlyavailable, it seems likely that methionine-192 is accessible in both theacylamido- and aryl sites. Knowles ss has pointed out that it is probablysignificant that trypsin, which has no affinity for aryl side-chains hasglutamine, not methionine, at the position three places from the activeserine .Earlier chemical probings for groups affecting the activity have beenreviewed. lo8, lo9 More recently, inhibition by diphenylcarbamyl chlor-ide and by cyanate ll1 apparently at the active serine has been reported.Diphenyldiazomethane 112 yields a partly active product presumably byesterification of a carboxyl group for whose role independent evidenceIo1 H.Schachter, K. A. Halliday, and G. H. Dixon, J. Biol. Chern., 1963, 238,loa G. E. Hein and C. Niemann, Proc. Nut. Acad. Sci. U.S.A., 1961, 47, 1341.lo3 G. E. Hein and C. Niemann, J . Amer. Chem. SOC., 1962, 84, 4487.lo4 G. E. Hein and C. Niemann, J . Amer. Chem. SOC., 1962, 84, 4495.l o 5 S. G. Cohen, L. H. Klee, and S. Y. Weinatein, J . Amer. Chem. SOC., 1966, 88,5302; S. G. Cohen, Z . Neuwirth, and S. Y. Weinstein, iEid., p. 5306; S. G. Cohen, R. M.Schultz, and S. Y. Weinstein, ibid., p. 5315.lo6 J. B. Jones, C. Niemann, and G. E. Hein, Biochemistry, 1965, 4, 1735.lo7 0.Gundlach and F. Turba, Biochem. Z . , 1962, 335, 573.lo* D. E. Koshland, jun., D. H. Strumeyer, and W. J. Ray, jun., Brookhawn Nationallo9 G. H. Dixon and H. Schachter, Canad. J. Biochem., 1964, 42, 695.B. F. Erlanger, A. G. Cooper, and W. Cohen, Biochemistry, 1966,5, 190.ll1 D. C. Shaw, W. H. Stein, and S. Moore, J . Biol. Chem., 1964, 239, PC 671.l l a A . A. Aboderin and J. S. Fruton, Proc. Nut. ACME. Sci. U.S.A., 1966, 56,PC 3134.Symposium in Biology, 1962, 15, 101.1252626 BIOLOGICAL CHEMISTRYexists.93 Some of the tryptophan residues are reactive 113 and some of thereactive tyrosine residues have been identified.ll4The mechanism of action of chymotrypsin. Bender and KBzdy 18, 89criticise several earlier suggestions and suggest a mechanism in which theessential features are symmetry of formation and hydrolysis of an acyl-serine intermediate in conjugate acid-base-catalysed reactions, in which theacid and base are a pair of imidazole groups of histidine residues.A weakpart of the argument lies in the assumption of a pair of histidine residues forwhich the strongest evidence is the structural similarity of the pair linkedby a disulphide bond in trypsinogen and chymotrypsinogen A and now 62in chymotrypsinogen B and elastase. In this mechanism, the role of a groupof pK cu. 9 required as an acid in acylation but not in deacylation wassomewhat di.fGcult. It now appears that this group, probably the a-amino-group of the N-terminal isoleucine residue of the B-chain 78 is involved,through a change of conformation of the protein, in the binding step andnot the acylation step.789 115, l 1 6 Only the conformation with the amino-group protonated is able to bind the substrate. A conformation change onbinding substrate has often been observed 78-82 and can explain the differenttitration curves of chymotrypsin and DIP-chymotrypsin,80b although thisdifference was not observed by other workers.ll7 The suggestion thatchymotrypsin, but not trypsin, has an acidic group of pK cu.8.7 with acatalytic role 1l8 now needs reconsideration.Studies with a wide range of sub-strates 102-105, 119-121 and inhibitors 106 of flexible or of rigid 121 conforma-tion in which the sensitive bond is placed at varying distances from thearyl side-chain, the acylamido-group, or a group serving to replace eitherof these are in accord with a model lo4, 120 in which three regions of theenzyme surface normally bind specifically the p-aryl, the acylamido-, andthe reactive group of the substrate.Some special substrates may bind inunproductive modes and so show lorn reactivity but firm binding.lls Thespecificity of the binding and the catalytic activity are not entirely inde-pendent,122-12* the interaction presumably reflecting the change of con-formation of the protein on binding substrate. It has been proposed18The specijcity of chymotrypsin.113 J. H. Swinehart and G. P. Hess, Biochim. Biophys. Acta. 1965, 104, 205; T. F.114 S. K- Dube, 0. A. Roholt, and D. Pressman, J . BioE.Chem., 1966, 241,116 M. L. Bender, M. J. Gibian, and D. J. Whelan, Proc. Nut. Acad.Sci. U.S.A., 1966,llS A. Himoe and G. P. Hess, Biochern. Biophys. Res. Comm., 1965, 23, 234.117 M. A. Mkhi and F. Behr, Biochim. Bwphy8. Acta, 1964, 89, 309.118 B. F. Erlanger, W. Cohen, S. M. Vratsanos, M. Castleman, and A. G. Cooper,ll9 J. R. Rapp, C. Niemann, and G. E. Rein, Biochemistry, 1966, 5, 4100.120 S. G. Cohen, J. Crossley, E. Khedouri, R. Zand, and L. H. IZlee, J . Amer. Chem.121 M. C. Silver, J . Amer. Chern. SOL, 1966, 88, 4247.122 J. R. Knowles, J . Theor. BWE., 1965, 9, 213.lZ3 J. B. Jones, T. Kunitake, C. Niemann, and G. E. Hein, J . Amer. Chem. SOC., 1965,lZ4 D. W. Ingles, J. R. Knowles, and J. A. Tondinson, Biochem. Biophys. Res.Spande, N.M. Green, and B. Witkop, Biochemistry, 1966, 5 , 1926.4665.56, 833.Nature, 1965, 205, 868.Soc., 1963, 85, 1685.87, 1777.Comm., 1966, 23, 619RYLE : ENDOPEPTIDASES OF VERTEBRATES 627that the ratio kJKm is a useful measure of the specifkity of an enzymicreaction, valid even for multi-stage reactions.A beginning has been made on the exploration of the secondary specificity,that is to say the effect of amino-acid residues adjacent to that providingthe carbonyl group of the sensitive bond, by consideration of the bonds splitby chymotrypsin in proteins of known amino-acid sequence.f25have discussedthe extensive evidence which suggests that the mechanism of action oftrypsin is very like that of chymotrypsin. The most striking features arethe similarity of the rate-constants for deacylation of the enzymes carryingthe same acyl group,l26 the similarity of the pH-dependence of the kineticparameters, and similarities in the kinetics of hydrolysis of acetylglycineethyl ester,l27 a substrate laclring the specific @-subs tituent for either enzyme.One strong piece of evidence against a similar mode of action is evidencethat the existence of an acyl-enzyme for trypsin is unlikely.Barman andGutfreund,l28 using a rapid-flow apparatus and a quenching technique, wereunable to detect any " burst " of ethanol liberation in the reaction of trypsinwith BAEE, and concluded that the detectable intermediate 129 was not anacyl-enzyme. On the other hand a t low pH-values a " burst " of nitrophenolis spectroscopically detectable with a-N-benzyloxycarbonyl-L-lysine p-nit,ro-phenyl ester 130 as substrate.The rate constants for the hydrolysis of thisand the corresponding methyl and benzyl esters are nearly identical 131despite their widely different reactivities towards nucleophiles, and a labile,enzymically inactive intermediate has been detected in the reaction of trypsinwith BAEE or with ben~oy1arginine.l~~ All these results strongly supportthe acyl-enzyme hypothesis. It would be valuable to test the reaction oftrypsin with the nitrophenyl ester above by the rapid-flow and quenchingmethod.In experiments with amides and esters of benzoyl- and tosyl-arginine,further evidence for the acyl-enzyme intermediate has been 0btai11ed.l~~Use was made of the fact that (as with chymotrypsin) the apparent acylationstep is rate-limiting for amide substrates, and the apparent deacylation stepfor ester substrates, to obtain the individual rate constants.Trypsin, unlike chymotrypsin,134 is activated by excess of substrate.135The activation observed with ester substrates was not detected with amidesubstrates 136 but since the solubility of the latter limits the concentrationsThe rnechunism of action of trypsin.Bender and Khzdyla5 G. L. Neil, C. Niemann, and G. E. Hein, Nature, 1966, 210, 903.126 M. L. Bender, J. V. Killheffer, jun., and F. J. Kezdy, J . Amer. Chena. SOL, 1964,lZ7 T. Inagami and E. Mitsude, J . Biol. Chem., 1964,239, 1388.lZ8 T. E. Barman and H. Gutfreund, Proc. Nat. Acad. Sci.U.S.A., 1965, 53, 1243.12B S. A. Bernhard and H. Gutfreund, Proc. Nat. Acad. Sci. U.S.A., 1965, 53, 1238.IsoM. L. Bender, G. J. KBzdy, and J. Feder, J . Amer. Chem. SOC., 1965, 87, 4953.131 M. I;. Bender and F. J. KBzdy, J . Amer. Chem. SOC., 1965, 87, 4954.132 M. L. Bender, F. J. KBzdy, and J. Feder, J . Amer. Chem. SOC., 1965, 87, 4955.133 J. J. Bechet, J . Chim. phys., 1964, 61, 584; 1965, 62, 1095.la4 D. W. Ingles and J. R. Knowles, Bwchem. J., 1966, 99, 275.lS5 C. G. Trowbridge, A. Krehbiel, and M. Laakowski, jun., Biochmbtry, 1963, 2,843; J. J. Bechet and J. Yon, Bwchim. Bwphys. Acta, 1964,89, 117; N. J. Baines, J. B.Baird, and D. T. Elmore, Biochem. J., 1964, 90, 470; S. M. Howard and J. W. Mehl,Biochim. Bwphys. Acta, 1965,105, 594.86, 5330.lS6 J.Chevallier and J. Yon, Biochim. Biophys. Acta, 1966, 122, 116628 BIOLOGICAL CHEMISTRYwhich can be used, it is not possible to draw the conclusion that activationby excess of substrate affects the deacylation, but not the acylation step.Symmetry of these steps is thus not excluded.As with chymotrypsin, there is evidence,13’ in this case from the kineticsof hydrolysis of BAEE a t low pH-values, for a basic group of pK ca. 4,presumably a carboxylate group, active either in the formation of theMichaelis complex, or in the acylation step. Analogy with chymotrypsinfor which no pK of 4 is seen in acylation 138 favours the former role as doconsiderations of symmetry.A proposal 139 for a mechanism for trypsin is the same as one proposedearlier 79 for chymotrypsin and involves an acyl-serine, not as obligatoryintermediate but in a cul-de-sac in equilibrium with an acylimidazole.Bender 89 rejected this proposal on the ground that it did not explain thekinetic effect of deuterium oxide as solvent, although both the other authorsclaim 79, 139 this as evidence for their scheme.Thrombin.This enzyme, reviewed by Laki and Gladner,140 is formed inthe blood from prothrombin by a process which is incompletely understood,and then possesses fibrinogen-clotting, proteinase, and esterase activities.New methods for preparing thrombin 141, 142 and prothrombin 143 have beendescribed. The molecular weight of thrombin is about 34,000, which isabout one-half that of prothrombin.144 The enzyme is sensitive to inhibitionby DFP and has the sequence -Gly.Asp.Ser.Gly- a t the active ~entre.14~Chromatographic analysis of the products of activation 146 and N-terminalanalyses 1 4 1 3 142 suggest that activation involves removal of peptides fromthe N-terminus of the precursor.The specificity of the enzyme is similar to that of trypsin,141 and it evenactivates trypsinogen l 4 7 but the yield of trypsin is only 70% of that obtainedby tryptic activation, probably because of further degradation. Thrombin,like trypsin, is not completely inactive with “ specific ” substrates forchymotrypsin ; it hydrolyses phenylalanine methyl ester and is inhibited by2-nitro-4-carboxyphenyl-NN-diphenyl carbamate.17 9 148 Some of the dis-crepancies in the literature concerning the kinetics of the clotting offibrinogen by thrombin are attributed to changes in the fibrinogen substrateon ageing.149Plasmin.This enzyme is produced from plasminogen in the blood wherethe characteristic action is the lysis of clots. It is sensitive to DFP; the13’ J. A. Stewart and J. E. Dobson, Biochemistry, 1965, 4, 1086.13* J. A. Stewart, H. S. Lee, and J. E. Dobson, J . Amer. Chem. Xoc., 1963,85, 1537.139 M. Lazdunski, Bull. SOC. Chim. biol., 1965, 47, 301.14O K. Laki and J. A. Gladner, PhysioZ. Rev., 1964, 44, 127.144 G. Y. Shinowara, Biochim. Biophys. Acta, 1966, 113, 359.143 H. C. Moore, S. E. Lux, 0. P. Malhotra, S. Bakerman, and J. R. Carter, Bbchk.1l4 C. R. Harmison, R. H. Landabaru, and W. H. Seegers, J .BWZ. Chem., 1961,145 J . A. Gladner and K. Laki, J . Amer. Chem. SOC., 1958,80,1263; G. J. S. Rao and146 D. L. Aronson and D. Mhnach6, Biochemistry, 1966,5, 2635.1 4 7 A. Engel, B. Alexander, and L. Pechet, Biochemistry, 1966,5, 1543.148 E. R. Cole, J. L. Koppel, and J. H. Olwin, Canad. J . Biwhem., 1966, 44, 1051.149 E. A. Gryaznukhina and V. A. Belitser, Biokhimiya, 1965, 30,696.S. Magnusson, Arkiv Kemi., 1965, 24, 349.Biophys. Acta, 1965, 111, 174.236, 1693.N. Chandrasekhar, Ann. Biochem. Exp. Med., Calcutta, 1961, 21, 233RYLE : ENDOPEPTIDASES O F VERTEBRATES 629pB-dependence of activity is similar to that of trypsin,150 and plasmhogencan be activated by thrombin.147 The reciprocal activation of the pre-cursors of thrombin and plasmin by the enzymes may be of significance inthe control of the blood-clotting process.The heterogeneity of human andbovine plasminogens detectable on electrophoresis in starch gel is at leastpartly attributable to changes occurring during the isolation of the zymogens,but there appear to be two natural components in human serum.151 Theactivation of bovine plasmhogen probably proceeds through intermediatestages 162 which may be analogous to n- and 6-chymotrypsins in the forms-tion of a-chymotrypsin.Plasminogen is activated by an activator now identified 153 as an equi-molar complex of streptokinase and plasmin. Activator-activity is alsoformed from streptokinase plus plasminogen but this reaction is time-dependent and inhibited by &-aminohexanoate, a known inhibitor of plas-minogen activation, so that the only true activator must be the plasmin-streptokinase complex.The molecular details of the complex formationand of the activation reaction remain to be investigated.Cathepsins. Studies on these intracellular proteinases (reviewed byFruton 154) have been hampered by difficulties in obtaining pure material.They are widely distributed in animal tissues, and in the living cell meprobably found in the ~ ~ s o s o M ~ s . ~ ~ ~ Cathepsins A, B, and C were originallypurified by precipitation procedures and characterised by their pH optima,their catalytic activity with peptide substrates, and the sensitivity ofcathepsins B and C to activation by cysteine.l56 Cathepsin A has not beenidentified in more recent chromatographic fractionations and its status isunclear.Cuthepins D, E, and B.Cathepsin D has been isolated from bovineand rabbit 15gp 159 spleen. It hydrolyses proteins optimally at pH 3-4, buthas no activity with benzyloxycarbonylglutamyltyrosine, the characteristicsubstrate for cathepsin A, although its specificity is otherwise similar to thatof pepsin. The bovine enzyme comprises at least ten components, some ofwhich were obtained in apparently homogeneous state.157Cathepsin E was obtained more readily from bone marrow 159 than fromspleen. It has an optimum pH m. 2.5 with protein substrates, is not affectedby DFP or cysteine and does not hydrolyse peptide substrates for cathepsinsA, B, or C. Its specificity of attack on the B-chain of oxidised insulin israther narrower than that of cathepsin D.l60 Another cathepsin from pig150 E.Ronwin, Canad. J . Biochem., 1962,40, 57.151 J. Y . S. Chan and E. T. Mertz, Canad. J. Biochem., 1966, 44,469, 475.lSa J. Y. S. Chan and E. T. Mertz, Canad. J, Bwchem., 1966,44, 487.15s C.-M. Ling, L. Summaris, and K. C. Robbins, J . Bwl. Chem., 1965, 240, 4213;lS4 J. S. Fruton, in “ The Enzymes,” ed. P. D. Boyer, H. Lardy, and K. MyrbLick,ls5 J. M. W. Bouma and M. Gruber, Biochim. Biophp. Acta, 1966,113, 360.166 H. H. Tallan, M. E. Jones, and J. S. Fruton, J . BhZ. Chem., 1952,194, 793.lS7 E. M. Press, R. R. Porter, and J. Cebra, Biochem. J., 1960, 14, 501.15* T. Webb and C. Lapresle, Nature, 1960, 188, 66.159 C. Lapresle and T.Webb, Biochem. J . , 1962,84,455.16* H. Rangel and C. Lapresle, Biochim. Biophys. Acta, 1966, 128, 372.B. C. W. Hummel, F. F. Buck, and E. C. de Renzo, i&id., 1966, 241, 3474.Academic Press, New York, 1960, 2nd edn., vol. 4, p. 233630 BIOLOGICAL CHEMISTRYkidney is reported to have a slightly different specificity with the samesubstrate .I61Cathepsin B has a specificity similar to that of trypsin-its typical sub-strate is benzoylarginineamide. If the lysine residue of the substratephthaloylglycyl-lysine methyl ester is replaced by 4-oxalysine or 4-thialysinethere are only minor changes in K , with cathepsin B but the values forVmax for the lysine, oxalysine, and thialysine derivatives are in the ratio1 : 2.7 : 0.19; the responses of papain and trypsin to the same changes weredifferent.162 The dramatic effect of a small change in the substrate far fromthe sensitive bond on Vmax must be the result of differences in conformationof the enzyme substrate complexes.Cathepsin C (dipeptidyl transferme).It now appears that the early pre-parations of cathepsin C contained a t least two enzymes, a proteinase nowfurther purified,163 and a peptidase devoid of proteinase activity also furtherpurified 16* and obtained in a homogeneous form.lG5 This second enzymehas a molecular weight 164, 165 of 21,000 and hydrolyses, or polymerises bytransamidation reactions, dipeptide esters or amides with liberation ofalcohol or ammonia. The enzyme requires chloride ions166 as well assulphydryl compounds for complete activation.Fruton and his colleagues 165have suggested that this enzyme should be named dipeptidyl transferaseand it really has no place in a review of endopeptidases.While cathepsins B, D, and E may be thought to function in the intra-cellular degradation of proteins, either of foreign ones, or on the death ofthe cell, the unusual specificity of dipeptidyl transferase makes it quiteunsuitable for this purpose or for a general role in protein anabolism.Acid Rotebases.-This group of enzymes, which are all extracellularand of gastric origin, includes pepsins, rennin, and gastricsin. The statusof the so-called gastric cathepsin 167 is obscure; some of its manifestations,such as activation of proteolytic activity by cysteine 168 may be artefacts(effect on the substrate rather than the enzyme) and others, such as thedemonstration of physically separable enzymic acti~ities,l~~ may be explainedby the presence of one of the better-characterised enzymes to be discussedbelow.Pepsin. Preparation and homogeneity.The known inhomogeneity ofIslT. P. Levchuk, M. I. Levyant, and V. N. Orekhovich, Biokhimiya, 1965,1132 G. I. Tesser, R. J. F. Nivard, and M. Gruber, Biochint. Biophys. Acta, 1964, 89,163 S. C. Dhar and S. M. Bose, Leather Science (Madras), 1964, 11, 309 (Chem. A h . ,164 R. J. Planta and M. Gruber, Biochim. Biophys. Acta, 1964, 89, 503.165 R. i\l. Metrione, A. 0. Neves, and J. S. Fruton, Biochemistry, 1966, 5,1e6 J. K. McDonald, T. J. Reilly, B. B. Zeitman, and S.Ellis, Biochem. Biophys. Rea.113' E. Freudenberg, Enzymologia, 1940, 8, 385; S. Buchs, ibid., 1953, 16, 193.168 S. Buchs, 2. physiol. Chm., 1954, 296, 129.169 R. Merten, C. Schramm, W. Gra.ssman, and K. Hannig, 2. physwl. Chem., 1952,289, 173; V. N. Orekhovich, L. A. Lokahina, V. A. Mantev, and 0. V. Troitskaya,Doklady Akad. Nauk S.S.S.R., 1956, 110, 1041; W. H. Taylor, Biochem. J., 1959, 71,373; V. 31. Stepanov, E. D. Levin, and V. N. Orekhovich, Doklady Akad. Nauk S.S.S.R.,1961,136, 1238.80,986.303.1965, 62, 2 9 7 0 ~ ) .1597.Cmm., 1966, 24, 771RYLE : ENDOPEPTIDASES O F VERTEBRATES 631most preparations of pepsin 170, 1 7 l has inhibited serious study for some timebut the last year or two have witnessed a sudden flowering of interest.Theenzyme almost universally used is that from the pig and has been reviewedearlier. l 7Neutral extracts of mucosa contain pepsinogen and the three minorpepsinogens, B, C , and D,173--176 which can be separated chromatographicallybut of which only pepsinogen C separates from pepsinogen in the presenceof a high proportion of the latter. The single peak of zymogen obtained bychromatography of crude commercial pepsinogen 177 may thus containpepsinogens B and D which would distort the amino-acid analysis rep0rted.17~Pepsinogen D may be identical with dephosphorylated pepsinogen 176 andif so, would not contribute any error to the amino-acid analysis of pepsinogen.Crystalline pepsinogen appears to be homogeneous and has been used 171for the preparation of pepsin more homogeneous than any other, on theevidence of its single N- and C-terminal residues (isoleucine and alaninerespectively).The amino-acid compositions of crystalline pepsinogen andof pepsin obtained from it are similar to those previously reported l7l, 17'9 17*and give molecular weights (39,000 and 34,000) in good agreement withphysical data. 1 7 7Primary structure of pepsin and pepsinogen. The sole C-terminal amino-acid residue of both proteins is alanine, and the sole N-terminal residue isleucine in the zymogen and isoleucine in the enzyme 1 7 1 in agreement withearlier rep0rts.l7~, 18* The apparent similarity of the C-terminal sequencesof pepsin and pepsinogen, together with the differing N-terminal residueshas been taken to show that pepsin is the C-terminal portion of the pepsin-ogen chain.If the similarity a t the C-terminus is now restricted to the onlyamino-acid (alanine) released by carboxypeptidase 1 7 1 from the purestpreparations, the force of the argument placing pepsin as the C-terminalportion of pepsinogen is weakened.The N-terminal amino-acid sequence of pepsin has been identified asIle.Gly.Asp.Glu-lgl and that of pepsinogen,l82 tentatively, asLeu.Va1.Leu.Glu.Pro.Ala. Glu.Phe.Ser .Leu.Lys.Asp.Gly.Lys.Val. (Asp,Pro) .Leu.-Pepsin and pepsinogen both contain one gram-atom of phosphorus per mole,1713 A. P. Ryle and R. It. Portor, Biochem. J., 1959, 73, 75.171 T. G. Rajagopolan, S. Moore, and W. H. Stein, J . Biol. Chem., 1966, 241, 4940.1 7 * ( a ) F.A. Bovey and S. S. Yanari, in " The Enzymes," ed. I?. D. Boyer, H. Lardy,and I<. Myrbiick, Academic Press, New York, 1960, 2nd edn., vol. 4, p. 63; (b) R. M.Herriott, J . Gen. Physiol., 1962, 45, suppl., p. 57.173 A. P. Ryle, Biochem. J., 1960, 75, 145.174 A. P. Ryle, Biochem. J., 1965, 96, 6.175 A. P. Ryle, and M. P. Hamilton, Biochem. J . , 1966, 101, 176.176 D. Lee and A. P. Ryle, Bhchem. J . , 1967 (in the press).177 R. Arnon and G. E. Perhnn, J . BioZ. Chem., 1963,238, 653.178 H. van Vunakis and R. M. Herriott, Biochim. Biophys. Acta, 1957,23,600; 0.0.17* R. C. Williams and T. G. Rajagopalm, J. Bwl. Chem., 1966,241, 4951.180 K. Heirwegh and P. Edman, Biochim. BiOphys. Acta, 1957, 24,219.lE2 L. A. Lokshina and V.N. Orekhovich, DokWy Akad. Nauk S.S.S.R., 1960,133,Blumenfeld and G. E. Perlmann, J . Ben. Physiol., 1959,42,553.P. Edman, Proc. Roy. Azcstral. Ciwm. Inst., 1957, 24, 434.472632 BIOLOGICAL CHEMISTRYprobably as a diester.ls3 The site of binding of the phosphate has been identi-fied 184 as the serine residue in the sequence -Glu.A.la.Thr.Ser.Glu.Glu.Leu-.No other site was identified, and it is possible that the other bond of thepresumed diester was hydrolysed under the conditions used for degradationof the protein. No data were reported for recovery of phosphate so that itis also possible that a basic phosphopeptide or a pair of neutral peptideslinked by a phosphodiester bridge was retained on the cation-exchangecolumn used for isolation of the acidic phosphopeptide.Pepsin is a very acidic protein, containing only two arginine, and onehistidine and lysine residue per mole.l7l Two of these basic residues occurin a nonapeptide sequence: 185-Asp.Arg. Ala. Asn. Asn. Lys. Val. Gly. Leu.-One of the four methionine residues precedes valine, and two more precedeaspartic acid (or asparagine).lS6 Lastly, sequences accounting for all sixhalf-cystine residues have been identified 187 in peptides from enzymicdigests of pepsin :-G1 y .Cys. Ser . Gly .Cys . Glu-I-Asp. (Cys, Ser,,Thr ) . Gly-I-1le. (C ys , Ser ) . Ser-I I-Gln.Asp.His. (Cys, Ser, Asp) .Ah. (Cys,Ser). Ser .Leu-The last sequence includes the single histidine residue which had earlier beenidentified lS8 by the phenylthiohydantoin method in a rather differentsequence as the fifth residue in the polypeptide chain of pepsin.If thislocation is correct, one of the disulphide bonds of pepsin closes a small loopnear the N-terminus and another makes a still smaller loop elsewhere. Muchof the chain must therefore be unrestricted by covalent cross-links, and eventhat provided by the phogphate is not essential for enzymic activity.183In agreement with this conclusion, reduction of the disulphide bonds ofpepsinogen in 8M-urea causes little unfolding beyond that produced by ureaa1one.ls9 Pepsinogen is one of the group of proteins whose biological activitycan be restored by re-oxidation after such treatment.lsgActivation ofpepsinogen. The activation of pepsinogen is very slow a tpH-values above 5 and is optimal a t pH ca.2. At pH 4.6, the reaction ispurely autocatalytic, but at lower pH-values it becomes more complicatedlS3 G. E. Perlmann, J . Gen. Physiol., 1958, 41, 441.lS4 V. M. Stepanov, E. A. Vakhitova, C. A. Egorov, and S. M. Avaeva, Biochim.le6 Yu S. Kutznetsov, G. G. Kovaleva, and V. M. Stepanov, Biochim. Bbphy8. Acta,!:6 V. M. Stepanov, S. P. Katrukha, and V. I. Ostoslavskaya, Khim. prirod. Soedi-B. Keil, L. MorAvek, and F. f3orm, 1967, Coll. Czech. Chem. Comm. (in the press).lSs M. B. Williamson and J. M. Passmann, J . Bwl. Chem., 1956, 222, 151.R. F. Steiner, V. Frattali, and H. Edelhooh, J. Biol. Chem., 1965, 240, 128.Biophys. Acta, 1965, 110, 632.1966,118,219.nenaz., Akad. Nauk. Uz. S.S.R., 1966, 8, 138RYLE : ENDOPEPTIDASES O F VERTEBRATES 633although an autocatalytic component can be demonstrated lS0, lgl evendown to pH 0.One of the products of activation is a basic peptide ofmolecular weight 192 cu. 3000. At pH-values above 5.0, this peptide isbound to the enzyme and inhibits &-clotting activity. At lower pH-values, the complex dissociates and the inhibitor is itself digested lS3 (optim-ally at pH cu. 3.5). The rate of activation is a linear function of ionicstrength,l94 suggesting that electrostatic bonds stabilise the zymogen and thepepsin-inhibitor complex.Pepsinogen, activated at pH 3, rather than pH 2, gives a product thathas hydrolase activity but lacks the usual transpeptidase activity of pepsin.lg5This product, which may be identical with one of the chromatographic frac-tions of pepsinogen activated under similar conditions, should well repayfurther investigation.A survey 196 of the bonds which are splitby pepsin in studies on the primary structures of proteins shows a con-siderable preference for tyrosine, phenylalanine, or leucine as the amino-acid providing either the carboxyl or the amino-group for the sensitivebond. This requirement is reflected in the nature of the synthetic peptideswhich are good sub~trates,~72a, l979 198 but many other bonds are hydrolysedat slower rates.Pepsin also has some esterase activity with a substrate(acetylphenylalanyl-~-phenyl-lactate)lgg with a bulky hydrophobic side-chain at both sides of the sensitive bond.A hydrophobic binding site for substrates is also suggested by the in-creasing effectiveness of aliphatic alcohols as competitive inhibitors withincreasing chain-length.2wPepsin catalyses trans-peptidation reactions, but unlike those catalysed by the enzymes of the" serine " group and by papain, it is the amino-group and not the carboxylgroup which is transferred. 201 This alone indicates a pathway different fromthat of the other enzymes, and suggests an amino-enzyme intermediate.The specificity shown in t'ranspeptidation is similar to that in hydrolysis.202The enzyme also catalyses the exchange of 1 8 0 between water and thecarboxyl group of the virtual substrate benzyloxycarbonyl-L-phenylalanine,but not of the D-isomer or of the free L-amino a~id,~O3~ suggesting that theThe speci$city of pepsin.ATEE was not hydrolysed.The pathway of the pepsin-catalysed reaction.loo R.M. Herriott, J . Gen. Physiol., 1938, 21, 501.lgl R. M. Herriott, J . Gen. Physiol., 1938-39, 22, 65.192 H. van Vunakis and R. M. Herriott, Biochirn. Bwphys. Acta, 1956, 22, 537.lg3 R. M. Herriott, J . Gen. Physwl., 1941. 24, 325.lg4 P. T. Varandani and M. Schlamowitz, Biochim. Biophys. Acta, 1963, 77, 496.lS5 H. Neumann and N. Sharon, Biochim. Biophys. Actu, 1960, 41, 370.lg6 J. Tang, Nature, 1963, 199, 1094.K. Inouye, I. M. Voynick, G. R. Delpierre, and J. S. Fruton, Biochemistry, 1966,lo* L. E. Baker, J . Biol. Chem., 1951, 193, 809.Igs L. A. Lokshina, V. N. Orekhovich, and V. A. Sklyankina, Nature, 1964, 204, 580.2oo J.Tang, J . Bwl. Chem., 1965, 240, 3810.201 H. Newnann, Y. Levin, A. Berger, and E. Katchalski, Biochem. J., 1959,73,33;J. S. Fruton, S. Fujii, and M. H. Knappenberger, Proc. Nut. Acad. Sci. U.S.A., 1961,202 N. I. Ma'tsev, L. M. Ginodman, V. N. Orekhovich, T. A. Valueva, and L. N.Akirnova, Biokhimiya, 1966, 31, 983.5, 2473.47, 759634 BIOLOOICYAL CHEMISTRYcarboxyl group must also be activated by the enzyme. In the transpeptida-tion reaction 20sb between benzyloxycarbonyl-L-phenylalanyl-L-tyrosine andacctyl-L-phenylalanine, the new oxygen of the carboxyl group of the bemyl-oxycarbonylphenylalanine product is derived from water, and not from thecarboxyl group of the acetylphenylalanine acceptor as would be the case ifdonor and acceptor reacted directly together in a four-centre exchangereection on the enzyme surface.The available evidence thus supports a pathway which involves bindingof both the amino- and the carboxyl group to the enzyme.Groups essential for peptic activity.Modification of the amino-groups ofpepsin causes no loss of activity, but iodination, acetylation, or diazotisationof the tyrosine residues or esterification of the carboxyl groups a11 lead toprogressive loss of activity (see Review l 7 9 . Iodination or diazotisationof up to 6 tyrosine residues of pepsinogen can occur without loss of potentialactivity.204 Almost complete alkylation of the methionine residues of pepsinoccurred without loss of activity so that the inactivation produced byN-bromosuccinimide, which reacts with both the methionine and the trypto-phan residues, has been attributed to the reaction with the latter.205 Thetyrosine residues were not oxidised.Diphenyldiazomethane 206 also inactivates pepsin and pepsinogen butthe reaction is not confined to a single carboxyl group.The fact that thezymogen is also inactivated suggests that either inactivation is not due toreaction at the active centre, or that the zymogen itself is inactive notbecause the centre is masked but because it is not formed, and that part ofit is free to react with the inhibitor. The inactivation of pepsin by thereagent is enhanced by some substrates suggesting that inactivation occursby reaction elsewhere than at the active centre. A particularly interestingtreatment is that of pepsin with acetylimidazole;207 when 9 tyrosine residueshave been acetylated the activity against haemoglobin is reduced to 40%but that against acetyl-DL-phenylalanyl-L-di-iodotyrosine is more thandoubled.A possibility here is that acetylation prevents inhibition by theDL-diastereoisomer.All the above results are difficult to interpret because of lack of specificityof the reagent. Two others have the desired specificity. Inactivation byp-bromophenacyl bromide 208 occurs in a stoicheiometric (1 : 1) reactionwhich shows the same sort of pH-dependence as the catalytic activity andprotection by substrates. A single aspartic acid residue 209 found *08 in apeptide with the composition (Gly,,Asp,Ser,Glu) is esterified. The techniqueof '' affinity labelling " has been applied to pepsin with the use of diazoacetyl-208 (a) N.Sharon, V. Grisaro, and H. Neumann, Arch. Biochern. Biophys., 1962,97,219. ( b ) N. I. Mal'tsev, L. A. Ginodman, and V. N. Orekhovich, Doklady Alcad. Nauk.S.S.S.R., 1965, 165, 1192.204 H. Neumann and N. Sharon, Proc. I V Cong. Sci. Studies. Rehovoth. Israel, 1961.206 L. A. Lokshina and V. N. Orekhovich, Biochem. (U.S.S.R.), 1964, 29, 300.206 G. R. Delpierre and J. S. Fruton, Proc. Nut. Acad. Sci. U.S.A., 1965, 54, 1161.207G. E. Perlmann, J . Bid. Chem., 1966, 241, 153.208 B. F. Erlanger, S. M. Vratsanos, N. Wassermann, and A. G. Cooper, Biochem.2oa E. Gross and J. L. Morell, J . Bid. Chem., 1966, 241, 3638.B<ophys. Res. Comm., 1966, 23, 243RYLE : ENDOPEPTIDASES O F VERTEBRATES 635norleucine methyl ester 210 in a reaction which is rapid and stoicheiometric(1 : 1) in the presence of cupric ions.Denatured pepsin failed to react andpepsinogen reacted only slowly and unspecifically and remained fully activ-atable. The results with both these reagents support those with less specificones which suggested a catalytic role for a carboxyl group in pepsin.Kinetic studies and the possible mechanism of pepsin action. Kinetic datahave recently been obtained with N-benzyloxycarbonyl- and N-acetyl-L-phenylalanyl-L- tyrosine,z11 N-acet yl-L-phenylalanyl-L-dibromot yrosine, 21N-acetyl-~-phenylalanyl-~-di-iodotyrosine,~~~ and the four possible N-acetyldipeptides of L-tyrosine and L-phenylalanine.Z1P These substrates have thedisadvantages of inconveniently low solubilities, and a carboxyl groupcapable of ionising in the pH range of interest.A new series of substrates,acylated peptide esters incorporating a histidine residue to confer improvedsolubility, has been introduced. 197Some of the data presented appear to confht. The first-order kineticsof hydrolysis of acetyl-L-phenylalanyl-L-tyrosine 211, 215 (pH 2, 37 ") havebeen ascribed 216 to inhibition by the product acetyl-L-phenylalanine withKi = Km = cu. 2 m ~ , but Ki was not directly determined. First-orderkinetics were also found for the hydrolysis of acetyl-L-phenylalanyl-L-di-bromotyrosine 212 (pH 2,25") and ascribed to the same cause (Km = 0.093m~,KI = O - l h ~ ) .However no inhibition was observed for acetyl-L-phenyl-alanyl-L-di-iodotyrosine 213 (pH 2, 37", Km = 0*075m) and the sameworkers confimed the fist order kinetics of hydrolysis of acetyl-I;-phenyl-alanyl-L-tyrosine but found that the Ki for acetyl-L-phenylalanine was2 3 m ~ (pH 2, 37"). The values given allow one to calculate for the dissocia-tion of the pepsin-acetylphenylalanine complex AH" = + 78000 cal./mole,AS" = + 245 cal./mole/degree. These values are of the same order ofmagnitude as those for the thermal denaturation of trypsinz17 and mayimply that the enzyme suffers a temperature-dependent conformation changeor that such a change occurs on dissociation of the enzyme-inhibitor complex.Further data, with control of the ionic strength, are needed.Studies of the pH-dependence of the hydrolysis of a neutral substrate(acetyl-L-phenylalanyl-L-tyrosine methyl ester) 2 l 8 gave a bell-shaped curvefor kc,, with pK-values of 1.8 and 3.5.The pR-values were increased by0 . 3 4 . 5 units in D,O, bEt the magnitude of kMt was not affected by D20, soruling out a proton transfer in the rate-determining step.It is not possible, on the meagre data available, to suggest a detailedscheme for the mechanism of action of pepsin. One proposal 1% 218 involvesformation of a carboxyl anhydride in the enzyme which then reacts with thel l o T. G. Rajagopolan, W. H. Stein, and S. Moore, J . Biol. Chem., 1966, 241, 4295.211 M. S. Silver, J. L. Denburg, and J. J. Steffens, J. Amer. Chem. SOC., 1965, 87,ala E.Zeffrsn and E. T. Kaiser, J . Amer. Chem. SOC., 1966, 88,3129.na W. T. Jackson, M. Schlamowitz, and A. Shaw, Biochemistry, 1965, 4, 1537.W. T. Jackson, M. Schlsmowitz, and A. Shaw, Biochemistry, 1966,5,4105.*16 L. E. Baker, J . Biol. Chem., 1954, 211, 701.*16 N. M. Green, Aature, 1956, 178, 145; L. E. Baker, {bid., p. 146.*17 M. L. Anson and A. E. Mirsky, J . Ben. PhysioE., 1934, 17, 393; M. Dixon andE. C. Webb, "Enzymes," Longmans Green and Co. Ltd., London, 1964,Znd edn., p. 148.886.G. E. Clement and C. L. Snyder, J . Amer. Chem. Soc., 1966, 88, 5338.636 BIOLOGICAL CHEMISTRYsubstrate to form covalent links with both halves; these are then liberatedin hydrolytic reactions. In this connection, the detection with hydroxyl-amine of two ester-like groups in pepsin may be significant.208Conformation of pepsin and pepsinogen.Optical rotation measurementsshow that while pepsin contains very Little a - h e l i ~ , ~ ~ ~ pepsinogen has suchfolded regions, 220 stabilised in part by electrostatic interactions involvingthe basic peptides removed on activation. 220, 221 Carboxyl hydrogen bondshave also been proposed to play a part in stabilising the conformation ofpepsin 222 and pepsinogen; 223 these may be bonded to abnormally-ionisingtyrosine residues 224 as in ribon~clease.~25The conformational relationships of pepsin and pepsinogen have alsobeen studied by immunological methods. 226 Denatured pepsin is antigenic-ally less similar than pepsin to pepsinogen so that some of the structuraldeterminants of the native enzyme must be present in the zymogen.Inview of the apparent large entropy change on combination of pepsin with acompetitive inhibitor, investigation of the optical absorption and rotationof such systems should be interesting.Pepsins By C, and D and their zyrnogens. The gastric mucosa of the pigcontains pepsinogen and three other acid proteinase zymogens each account-ing for 5-10% of the total potential activity.l73, l 7 6 The zymogens andenzymes have physicochemical properties similar to those of pepsinogen andpepsin. Pepsinogens B and C were first recognised as their respectiveenzymes (parapepsins I and 11) l 7 0 which differ from pepsin in their speci-ficity. In standard assays, pepsin B does not show any activity with hzmo-globin, while pepsin C is inactive with acetyl-L-phenylalanyl-L-di-iodotyro-sine.Pepsin D hydrolyses both substrates with about the same specificactivity as pepsin. Pepsin C, like pepsin, catalyses amino-transfer trans-peptidations 227 so providing further evidence of the similarity of theenzymes but its amino-acid composition and that of its zymogen 175 showthem to be more basic proteins than the major ones. All three minor enzymeshave different N-terminal amino-acid residues from their zymogens ;173,174,176pepsin D and pepsinogen D have the same C-terminal amino-acid, andpossibly sequence,l76 as pepsin and pepsinogen. It thus seems likely thatactivation occurs by removal from the N-terminal end of the zymogens ofpeptide material which, in the case of pepsinogen C and pepsinogen Dincludes inhibiting peptides.228 The only detected chemical differencebetween pepsin D and pepsinogen D on one hand, and pepsin and pepsinogenon the other, is that the first two (like the B and C proteins) do not contain219 G.E. Perlmann, Proc. Nut. Acad. Sci. U.S.A., 1959, 45, 915.220 G. E. Perlmann, J . Mol. Biol., 1963, 6, 452.231 G. E. Perlmann, Biochem. J., 1963,89, 45P.2*Z H. Edelhoch, Biochim. Biophys. Acta, 1960, 38, 113.223 H. Edelhoch, V. Frattali, and R. F. Shiner, J . Biol. Chem., 1965, 240, 122.294 G. E. Perlmann, J . BioZ. Chem., 1964, 239,3762.225 L. K. Li, J. P. Riehm, and H. A. Scheraga, Biochemktry, 1966,5, 2043.226 H. van Vunakis, H. I. Lehrer, W. S.Allison, and L. Levine, J . Gem. PhyaiOl., 1963,46, 589; J. F. Gerstein, H. van Vunakis, and L. Levine, Biochemistry, 1963, 2, 964; R.Amon and G. E. Perlmann, J . BWZ. Chem., 1963,288,963.227 A. P. Ryle, BuU. SOC. Chim. biol., 1960, 42, 1223.228 K. K. Oduro and A. P. Ryle, unpublished work; D. Lee, Can&. J. Bhchm.,1967 (in ths press)RYLE : ENDOPEPTIDASES O F VERTEBRATES 637phosphate. It seems likely that pepsinogen D is merely dephosphorylatedpepsinogen, but whether the phosphate has been removed during the isolationof the zymogen or whether it has never been attached is not known.Humn gastric proteinases. Several pepsinogens have been detected byzone electrophoresis and immunological methods in the gastric mucosa ofman 229 and other species.230 One group of workers has crystallised anenzyme, gastricsin 231 from human gastric juice and have isolated azymogen 232 by chromatographic means from human mucosa.Gastricsinhas a higher pH-optimum than pig pepsin (haemoglobin substrate) andbehaves less acidically on electrophoresis in starch gel. Although thezymogen appears to migrate as a single band on electrophoresis, on acidifica-tion it gives rise to both pepsin and gastricsin, but in a ratio which apparentlyvaries according not to the pH of activation, but to the final pH of thesolution before it is fractionated by chromatography. The homogeneityof the enzymes obtained, and the possibility that one of them is a complexof enzyme and activation peptide should be investigated, particularly inview of the low milk-clotting activity of gastricsin.More than one peakhas also been observed on chromatography of homogeneous pig pepsinogenactivated a t pH-values above 3. 32 The immunological cross-reactions ofthe human and porcine enzymes and zymogens have been investigated; 23sthe inter-species differences are greater than those between homologousenzyme and zymogen.Another group of workers has separated three pepsinogens 23* and threeor four pepsinsz35 from neutral and acidified extracts of human mucosaand have presented tentative evidence 236 for the formation of pepsin-inhibitor complexes from the different zymogens. It is not possible to relatethe human enzymes to those from the pig because the latter are characterisedby their substrate specificity whilst the former have not been tested withany peptide substrate.Rennin. The milk-clotting proteinase from the stomach of the calf hasbeen extensively reviewed.237 Two prorennins of molecular weight 36,000and three rennins of molecular weight 31,000 have been separated bychromatography and are rather less acidic proteins than pepsin and pep-sinogen.Activation of prorennins occurs at pH-values below 5 in a reactionwhich is at least partly autocatalytic and involves removal of peptides fromthe N-terminal end (of prorennin B). The specificity of the enzyme is similarto that of pepsin, as judged by its activity with synthetic peptide substratesor the B-chain of oxidised in~ulin,2~8, 239 but it does not remove the229 I.Kushner, W. Rapp, and P. Burtin, J. Clin. Invest., 1964, 43, 1983.230 W. B. Hanley, S. H. Boyer, and M. A. Naughton, Nature, 1966, 209, 996.231 J. Tang, S. Wolf, R. Caputto, and R. E. Trucco, J . Biol. Chem., 1959, 234, 1174.232 J. Tang and K. I. Tang, J. Biol. Chem., 1963, 238, 60.233 M. Schlamowitz, A. Shaw, and W. T. Jackson, Biochemistry, 1964, 3, 636.2 3 4 M. J. Seijffers, H. L. Segal, and L. L. Miller, Amer. J. Physiol., 1963, 205, 1099.*35 M. J. Seijffers, H. L. Segal, and L. L. Miller, Amer. J. Physiol., 1963, 205, 1106.236 M. J. Seijffers, L. L. Miller, and H. L. Segal, Biochemistry, 1964, 3, 1.237 B. Foltmann, Compt. rend. Trav. Lab. Carlsberg, 1966, 35, 143.238 J. C. Fish, Nature, 1957, 180, 345.239 V. Bang-Jensen, B. Foltmann, and W.Rombauts, Compt. rend. Trav. Lab.Carlsberg, 1964, 34, 326638 BIOLOGICAL CHEMISTRYC-terminal tetrapeptide sequence from ribonu~lease.~~~ Uncertainty stillexists ti37 over the nature of the primary effect of rennin in clotting K-caseinand the possibility remains that an ester bond is split. Rennin is inactivatedby photo-oxidation 240 in the presence of Methylene Blue. Tryptophan,methionine, and histidine residues become oxidised, and since the loss ofactivity is correlated well with loss of histidine and since no loss of activitywas observed on modification of the first two types of residue with 2-hydroxy-5-nitrobenzyl bromide or hydrogen peroxide respectively it was suggestedthat the inactivation on photo-oxidation is due to oxidation of histidineresidues.Rennin is also partly inactivated, but only at pH 8, not at pH 6.5,by reaction with dimethylaminonaphthalenesulphonyl chloride with incor-poration of 1-5-2 fluorescent groups per mole; 2 4 l there is, however, noevidence that this inactivation, or that caused by photo-oxidation specificallyinvolves the active centre.Conclusion.-Although good progress has been made in the understandingof the mode of action of chymotrypsin, and to a lesser extent, trypsin, littleor no information is available for the other proteinases. In the ajerineproteinase group, comparative studies have often been useful in underliningthe significance of an observation (e.g., the common sequence round theactive serine) and such comparisons may be expected to be useful among theacid proteinases too.The biological significance, and the presumed common ancestral origin,of the serine enzymes have often been discussed 2O, 39, 5% 6O but it is stillnot clear what biological advantage is conferred by having such a mulfi-plicity of pepsins or of serine enzymes.Some advantage is presumablyconferred by a " spare wheel" effect-a mutation that renders an enzymeinactive will not prove lethal if other enzymes are still available to performits functions, but this idea is no5 very satisfying.It is significant that the activation of the zymogens discussed here, andalso of pro-carboxypeptidase 242 occurs by proteolytic removal of materialfrom the N-terminal end. Since proteins are synthesised from this end, theinhibitory region is formed first and the active part later.Thus, the ribosomeis never embarrassed by having attached to itself chains of nearly-completedenzyme with proteolytic activity. One would suppose that the need for asimilar device was as great with other enzymes, especially ribonucleases, butno zymogen for ribonuclease has been reported.240 R. D. Hi11 and R. R. Laing, 1965, Biochinz. Biophys. Acta, 99, 352.241 R. D. Hill and R. R. Laing, Nature, 1966, 210, 1160.242 K. S. V. Sampath Kumar, J. B. Clegg, and K. A. Walsh, Biochemistry, 1964, 3,17285. THE BIOCHEMISTRY OF SULPHUR-CONTAINING AMIBO-ACIDSBy Gc. A. Maw( 0 b S h @ & ~ Crop8 Research Inetitute, Rwr&ingtota, 8ussex)LIKE most facets of biochemistry, the study of sulphur metabolism hasexpanded considerably in the last few years.This has been accentuated bythe fact that sulphur participates in a relatively large number of types ofchemical linkage, practically all of which occur naturally and many of whichhave fundamental roles in intermediary metabolism. It has consequentlybeen necessary to limit severely the scope of this Report, which is confinedto developments in the biochemistry of sulphur-containing amino-acids and~ome of their derivatives.New Naturally Occurring Sulphur Amino-acids.-The range of sulphuramino-acids known to occur naturally has been reviewed by Fowden 1 andby Virtanen.2 The number of these compounds has increased appreciablyin recent years. S- (2-Carboxy-n-propyl)-~-cysteine and S-(2-carboxyethyl)-L-cysteine have been identified in children's urine,3 S-( 1,2-dicarboxyethyl)cysteine has been found in adult urine and pig kidney,4 and S-carboxy-methylcysteine in plants.X-Methylcysteine and its sulphoxide have beendetected in human urine and S-methylglutathione in bovine brain.' S-Sulphocysteine (cysteine-8-sulphonate), previously reported to be formedfrom thiosulphate and serine in Aspergillus niduhnq8 has been found in theurine of the blotched Kenya genet: and the corresponding glutathionederivative has been isolated from rat small-intestine.10A methionine derivative of paramount interest, N-formylmethionine, isinvolved as an intermediate in protein chain synthesis (see later section).Homomethionine (5-methylthionorvaline) has recently been detected incabbage.11 The first isolation of methionine sulphoxide from a naturalsource, the blowfly, has been reported.12 Lanthionine has previously beenregarded as an artifact formed from cystine residues during the hydrolysisof wool proteins.It has now been shown to occur in the amino-acid pool ofinsects and in the chick embryo.]'L. Fowden, Ann. Rev. Bwchem., 1964,33, 192.a A. I. Virtanen, Angew. Chem., 1962,1, 299; Phytochemistry, 1965,4, 207.S . Ohmori, T. Shimomura, T. Azumi, and S. Mizuhara, Biochem. Z., 1965,343, 9.T. Kuwaki and S. Mizuhara, Biochim. Biophya. Acta, 1966,115,491.ti C. Buziassy and M. Mazelis, Biochim. Biophys. Acta, 1964,86, 186. * F. Tominaga, S. Kobayashi, I. Muta, H. Takei, and M. Ichinose, J.Biochem.(Japun), 1963,54,220; F. Tominaga, K. Ob, and H. Yoshida, ibid., 1965,57, 717. ' A. Kanazawa, Y. Kakimoto, T. Nakajima, and I. Sano, Bbchim. Bbphys. Acta,1965, 111, 90.T. Nakamura and R. Sato, Natwe, 1963,198,1198.J. C . Crawhall and S. Segal, Nature, 1965, 208, 1320.lo H. C. Robinson and C. A. Passternak, Biochem. J., 1964, 93, 487.l1 M. Sugii, Y. Sukata, and T. Suzuki, Chem. and Phum. BUZZ., (Jupn), 1964, 12,l* F. Lucaa and L. Levenbrook, Biochem. J., 1966,100,473.l8 D. R. Rao, A. H. Ennor, and B. Thorpe, Biocbm. Bwphys. Res. Comm., 1960,l4 N. H. Sloane and K. G. Untch, Biochemistry, 1966,5, 2658.1115.22, 163640 BIOLOGICAL CHEMISTRYA new homo cyst eine derivative , 8- h ydroxymeth ylhomoc y s t eine , has beenisolated from the alga Chondrus 0cellatus.1~Cysteine md Cystine.-Cysteine affects the activity of many enzymes.It has been shown to protect nitrite and nitrate reductases in plants frominactivation by polyphenol oXidases,16 and to inhibit choline acetyltrans-ferase l7 and alkaline phosphatase.l* The effect of cysteine on the lastenzyme is believed to be a chelation with the zinc atom a t the active site.Cysteine also inhibits respiration and catalase synthesis in Succharmycescerevisiue,lQ and histidine uptake and potassium retention in brain slices.20The radioprotective effect of cysteine towards yeast has been found tobe optimal only when a certain amount of the amino-acid has entered thecells.It seems probable that it is the free cysteine pool and not protein-cysteine which is radioprotective.21 Studies have been made of cysteinetransport into kidney tissue with the aid of dithiothreitol (Cleland’s reagent)to maintain the amino-acid in its reduced form,22 and determination of thecysteine/cystine ratio in kidney cortex using N-ethylmaleimide has shownthat cysteine predominates over the disulphide.23 Cystine transport has beenexamined in the intestine and in Escherichia ~oEi.2~ Cysteine has beenshown t o be the most effective sulphur amino-acid in the prevention ofnutritional muscular dystrophy in chicks.26The incorporation of cys tine into protein involves a cysteine residue eitherbeing attached to other amino-acids by peptide bonds or forming a disulphidelink with a cysteine residue already in the protein chain.A study byWilliamson 27 of the two types of bonding in regenerating wound tissue in ratsshowed that approximately 7% of the incorporation occurred by disulphide-bond formation.Metabolism. A cystathionine-cleaving enzyme in Neurospora has beenfound to actively degrade cystine to pyruvate.2* Two distinct cystine-degrading systems have since been detected in cabbage leaf hom~genates.~~One of these is a particulate enzyme and, like the Neurospora enzyme, alsoforms pyruvate. It is inactive towards cystathionine, however, but can cleavecysteine-S-sulphonate at a greater rate than cystine itself.The oxidation of cysteine to cysteinesulphinic acid has been studied in ratliver preparations.30, 3l Sorbo 30 was able to show the accumulation of thel6 M.Takagi and A. Okumura, Nippon Sukan Cfakkaishi, 1964,30,837.l6 E. Pojnsr and E. C. Cocking, Biochern. J . , 1964, 91, 29P.l7 D. Morris, C. Hebb, and G. Bull, Nature, 1966, 209, 914.18 S. G. Agus, R. P. Cox, and M. J. Gritltin, Biochim. Biophys. Acta, 1966,118, 363.19 C. Bhuvaneswaran, A. Sreenivasan, and D. V. Rege, Bwchem. J., 1964,92, 504.20 K. D. Neame, Nature, 1964, 203, 1067; J . Neurochem., 1964,11, 67.21 U. Schaedel, E.-R. Loehmann, and W. Laskowski, Natzwe, 1966,211,431.22 J. C. Crawhall and S. Segal, Biochint. Biophys. Acta, 1966, 121, 215.s3 J. C. Crawhall and S. Segal, Biochem. J . , 1966, 99, 19C.24 R. P. Spencer, K. R. Brody, and H. 0. Mautner, Nature, 1965, 207, 418.26 L. Leive and B. D. Davis, J . Biol.Chem., 1965, 240,4362.26 J. N. Hathcock and M. L. Scott, Proc. SOC. Exp. BioE. Med., 1966, 121, 908.27 M. B. Williamson and 0. H. Clark, Arch. Biochem. Biophya., 1966, 114, 314.28 M. Flavin and C. Slaughter, J . Biol. Chem., 1964, 239, 2212.29 M. Tishel and M. Mazelis, Nature, 1966, 211, 745.30 B. Sorbo and I;. Ewetz, Biochem. Biophy8. Rea. Cornm., 1965,18, 359; L. Ewetz91 A. Wainer, Biochim. Biophys. Acta, 1965, 104, 405.and B. Sorbo, Bwchim. Biophya. Acta, 1966, 128, 296MAW : SULPHUB-CONTAINING AMINO-ACIDS 641sulphinic acid by addition to the system of hydroxylamine, which inhibitsfurther desulphination and decarboxylation of the product. The oxidationwas stimulated by ferrous ions and by the addition of NADPH. Cysteine hasalso been shown to be an efficient precursor of j9-cyanoalanine in both E .wliand plant extracts.32The occurrence of sulphur-containingnucleotides as minor constituents of bacterial and mammalian sRNA has beenreported.33-S5 Two such nucleotides so far identified are 4-thiouridylicacid 33 and a second containing a 2-thiopyrimidine.34 Lipsett has obtainedstrong evidence that the sulphur moiety of the former compound is derivedfrom ~ y s t e i n e , ~ ~ and has further shown that the sulphur may be introducedinto uracil residues of sRNA at the macromolecular level. In E. wli prepara-tions, trans-sulphuration from cystine to E. coli sRNA required pyridoxalphosphate, ATP and Mgz+. E. coli ribosomal RNA and yeast and liversRNA’s were inactive as sulphur acceptors.36 Similar findings have beenobtained by Hayward and Weiss with cysteine in E.coli extracts.37Glutathione.-A number of new metabolic roles for glutathione has beenestablished. Glutathione is implicated as a co-factor in the oxidation ofeIementa1 sulphur to sulphite in Thiobucillus t h i o p a r ~ , ~ ~ and is requiredspecifically in the synthesis of dimethyl selenide from selenite in mouseliver.3B Administration of oxidized glutathione to vitamin B,,-deficient ratsproduces an increased excretion of formiminoglutamic acid. 4O Pyridoxineappears to be an important regulator of glutathione metabolism, for adeficiency of the vitamin is associated with an increased content of gluta-thione in the liver and erythrocyte^.^^ In E . coZi extracts, glutathione formsa derivative with spermidine in the presence of ATP and magnesium i 0 n ~ .~ 2The oxidation of glutathione has been followed in rat liver fractions 45and in lens.44 Other glutathione-oxidizing systems which have been studiedrecently include glutathione-organic nitrate reductase in rat liver 45 andglutathione-insulin transhydrogenase, present in pancreas and liver.46 Theglutathione reductase of yeast has been purified and its properties and reac-tion mechanism investigated.*78uZphur-wntuining nucleotides.s2 P. M. Dunhill and L. Fowden, Nature, 1965,208,1206; H. G. Floss, L. Hadwiger,s3 A. Peterkofsky and M. N. Lipsett, Bwchem. Biophys. Res. Comm., 1965, 20, 780;34 J. A. Carbon, 1;. Hung, and D. S . Jones, Proc. Nat. A d .Sci. U.S.A., 1965, 53,36 T. Schleich and J. Goldstein, Science, 1965,150, 1168.s6 M. N. Lipsett and A. Peterkofsky, Proc. Nat. A d . Sci. U.S.A., 1966, 55, 1169.37 R. S. Hayward and S. B. Web, Proc. Nat. A d . Sci. U.S.A., 1966, 55, 1161.s8 I. Suzuki and M. Silva, Bwchim. Bkphys. Acta, 1966,122,22.4 O N. P. Sen and P. L. McGeer, Canad. J . Biochem., 1966, 44,286.41 J. M. Hsu, E. Buddemeyer, and B. F. Chow, Bwchem. J., 1964, 90, 60.4* C. W. Tabor, H. Tabor, and I;. de Meis, Fed. Proc., 1966,25, 709.43 B. 0. Chistophersen, Biochem. J . , 1966, 100, 95.A. Pirie, Biochem. J., 1965, 96, 244.46 P. Needleman and F. E. Hunter, Mol. Phapynacol., 1965,1, 77.P. T. Varandani and H. H. Tomizawa, Biochim. Biophys. Acta, 1966, 113,498;4’ R. F. Colman and S.Black, J. BiOZ. Chem., 1965,240,1796; V . Massey and C. H.and E. E. Conn, ibid., p. 1207.M. N. Lipsett, J . Biol. Chem., 1965, 240, 3975.979.H. E. G-anther, Biochemistry, 1966, 5, 1089.M., 1966, 118, 198; H. M. Katzen and F. Tietze, J . Biol. Chern., 1966, 241, 3561.WiUiame, ibid., p. 4470642 BIOLOQICAL CHEMISTRYGZututhione in blood. The role of glutathione in blood has been discussedby several authors.*g, 40 The peptide is present in relatively large amountsin erythrocytes and, on account of its ready oxidation by glutathione per-oxidase, is considered to act as a detoxicant of the hydrogen peroxideproduced continuously in cells, so protecting haemoglobin and other cellularconstituents from oxidative Supporting this is the finding that insubjects possessing an inherited blood glutathione deficiency the erythrocytelife span was markedly shortened and the cells were abnormally susceptibleto oxidative breakdowa61 Glutathione can exert a regulatory action on theactivity of the hexose monophosphate pathway in red cells, through thedemand for NADPH required specifically in the reduction of oxidizedglutathione by glutathione reductase.62Disulphides.-The chemistry and biochemistry of thiol-disulphideexchange reactions was reviewed in 1965 by Lumper and Zahn.63 The morerecent literature on this topic is too voluminous to be treated fully here.Naturally occurring disulphides include cysteine homocysteine disulphidepresent in the urine of cystinurics 64 and as an intermediary metabolite inNeurospora c r a s s c ~ , ~ ~ and coenzyme-A glutathione disulphide in liver andkidney.5s A specific reductase for the latter compound, distinct from gluta-thione reductase and glutathione-insulin transhydrogenase , has been identi-fied in rat liver and bovine kidney.67 Penicillamine administered to cysti-nurics is excreted partly as cysteine penicillamine disulphide, and has provedof value in reducing cystine levels in plasma and urine by converting theamino-acid into the more soluble mixed disulphide.58Cysteine and glutathione interact with s e m albumin and other proteins,including hit?m~globin,~~, 49, 69 most probably forming protein thiol mixeddisulphides. There is some evidence for the formation of a disulphide linkbetween glutathione and certain enzymes, e.g., glucose-6-phosphate dehydro-genase, aspartate aminotransferase and acid phosphatase.sO An unusualdisulphide link between a cysteine residue in Streptococcal proteinase and anas yet unidentified volatile thiol has been reported.sl Some of the disulphidesdB J.D. Harley, Arature, 1965, 206, 1054; A. Hochberg and E. Dimant, Biochim.Biophys. Acta, 1965, 104, 53.49 J. Niv, A. Hochberg, and E. Dirnant, Biochim. Biophys. Acta, 1966,127, 26.6 0 P. Hochstein and G. Cohen, Acta Biol. Med. Qer. Suppl. 3, 1964, 292; A. S. Hill,A. Haut, G. E. Cartwright, and M. M. Wintrobe, J. Clin. Invest., 1964, 43, 17; H. S.Jacob, S. H. Ingbar, and J. H. Jandl, ibid., 1965, 44, 1187.61 H. K. Prins, M. Oort, J. A. Looa, C. Ziircher, and T.Beckers, Blood, 1966,27,145.I* H. S. Jacob and J. H. Jandl, J . Biol. Chem., 1966, 241, 4243.Is L. Lumper and H. Zahn, Adv. Enzymol., 1965, 27, 199.I4 G. W. Frirnpter, J . Biol. Chem., 1961, 236, PC51.6b J. I,. Wiebers and H. R. Garner, Biochim. Biophys. Acta, 1966, 117, 403.66 S. H. Chang and D. R. Wilken, J . Biol. Chem., 1966,240, 3136; R. N. Ondarza,Biochirn. Biophys. Acta, 1965, 107, 112.67 R. N. Ondarza and J. Martinez, Biochim. Biophys. Acta, 1966, 113, 409; S. H.Chang and D. R. Wilken, J . Biol. Chem., 1966,241,4251.5 8 J. C. Crawhall and C. J. Thompson, Science, 1965,147,1459; J. E. McDonald andI?. H. Henneman, New Engl. J . Med., 1965, 273, 678.6 9 R. Frater and F. J. R. Hird, Biochem. J., 1963, 88, 100; B. E. Davidson andF.J. R. Hird, ibid., 1965, 96, 890; J. Wagner and V. Janata, Folia Haematol., 1966,83, 524.6 0 E. Bottini and G. Modiano, Biochem. Biophys. Res. Comm., 1964, 17, 260; H.Walter and J. C. Caccam, Biochem. J., 1966,100, 274.61 W. Ferdinand, W. H. Stein, and S. Moore, J . Biol. Chem., 1965, 240, 1160MAW: SULPEUR-CONTAINING AMINO-ACIDS 643listed above may possibly be formed or degraded by enzymes concerned withprotein disulphide reduction and interchange.62A number of metallofiavoproteirzs (succinic dehydrogenase, xanthineoxidase, etc.) contain small amounts of bound sulphide. Massey 63 considersthat this sulphide is linked with ferric iron and a cysteine sulphur atom toform a labile disulphide structure which may be involved in the catalyticfunction of these enzymes:Mercapturic Acids.-Aromatic hydrocarbons and their halogen and nitro-derivatives have long been known to be excreted as X-substituted derivativesof N-acetylcysteine when administered to animals.More recently, severalother polycyclic hydrocarbons have been shown to be conjugated in thisway.64Aliphatic halogen compounds, such as i~domethane,~~, 66 br~moethane,~'brornopr0pane,~8-~1 and higher homologues are also excreted in smallamounts as the corresponding mercapturic acids, as are some nitroparaf6ns.6sIn addition, alkyl halides give rise to a number of other related metabolites.Bromoethane and bromopropane are excreted partly as the mercapturic acidsulpho~ides,~~, 72 and bromopropane and its higher homologues are alsoconverted into the corresponding hydroxyalkylmercapturic acids.70~ 71, 73, '*Iodomethane gives rise to S-methylcysteine and to methylmercaptoaceticacid and N- (methylmercaptoacety1)glycine. 65 The last two compounds havebeen identified as catabolites of X-methylcysteine and S-methylglutathionein the r ~ ~ t .7 ~ ~ 76There is now adequate evidence that the cysteine moiety of mercapturicacids originates from glutathione. Liver glutathione levels fall when mermp-turic acid precursors are fed to Furthermore, within 1 hr. after theadministration of iodomethane to rats, S-methylglutathione accumulated inthe liver, accounting for 45-50% of the dose of halide givene66 ConversionC. M. Brown and J. S. Hough, Nature, 1966, 211, 201 ; F. De Lorenzo, R.F.Coldberger, E. Steers, D. Givol, and C. B. Anfinsen, J . Biol. Chem., 1966, 241, 1562.63 R. W. Miller and V. Massey, J . Biol. Chem., 1965, 240, 1453; P. E. Brumby,R. W. Miller, and V. Masssy, ibid., p. 2223.64 E. Boyland and P. Sims, Biochem. J., 1964,90,391; ibid., 1964,91,493; P. Sims,ibid., 1964, 92, 621; E. Boyland and P. Sims, ibid., 1965, 95, 788; P. L. Grover and P.Sims, ibid., 1965, 96, 521.66 E. A. Barnsley and L. Young, Biochem. J . , 1965, 95, 77.66 31. K. Johnson, Biochem. J., 1966, 98, 38.6 7 A. E. R. Thomson, E. A. Barnsley, and L. Young, Biochem. J., 1963, 86, 145.68 H. G. Bray, J. C. Caygill, 8. P. James, and P. B. Wood, Biochem. J., 1964,90,127.6Q T. H. Grenby and L. Young, Biochem. J . , 1960, 75, 28.70 E. A. Barnsley, T.H. Grenby, and L. Young, Biochem. J., 1966,100, 282.71 E. A. Barnsley, Biochem. J . , 1966, 100, 362.7 2 E. A. Barnsley, A. E. R. Thompson, and I;. Young, Biochem. J., 1964,90, 688.73 E. A. Barnsley, Biochern. J., 1964, 93, 15P.74 S. P. James and D, J. Jeffery, Biochem. J . , 1964, 93, 16P.7 5 E. A. Barnsley, Biochim. Biophys. Acta, 1964, 90, 24.C. J. Foxwell and L. Young, Biochem. J . , 1964,92, 50P.7 7 M. M. Barnes, S. P. James, and P. B. Wood, Biochem. J., 1959,71,680; T. Suga,I. Ohata, and M. Akagi, J . Bwchem. (Japan), 1966, 59, 209644 BIOLOGICAL CHEMISTRYof aromatic compounds into &substituted glutathiones also occurs inand has been obtained in rat tissue preparations.79The initial conjugation of mercapturic acid-forming compouncb withglutathione is primarily an enzymic process,79# 8o although with idomethaneand possibly other reactive alkyl halides some nonenzymic reaction withglutathione may take place.66 Four distinct enzymes have so far beenidentified which mediate in the conjugations and which appear to be fairlyspecific with respect to the type of compound undergoing conjugation:glutathione S-aryltransfera~e,~~~ 81 glutathione S-alkyltransferase,82 gluta-thione S-epo~idetransferase,~~ and a glutathione transferase catalysing thereaction of the tripeptide with certain unsaturated corn pound^.^^ All fourenzymes are present in the liver and to some extent in the kidney of a varietyof animal species.The epoxide transferase is probably involved in mercap-turic acid formation from polycyclic hydrocarbons, which are believed to bemetabolized initially to epoxides by a perhydroxylation system.83 As yetthe physiological role of these enzymes is obscure.A conjugation of gluta-thione with isovalerate in the presence of AT9 occurs in liver homogenatestogives-( l-carboxyisobuty1)-glutathione,sSand the product may be cleaved bykidney homogenates or glutathionase preparations to 8- (l-carboxyisobuty1)-cysteine (isovalthine), an amino-acid excreted by hypocholesteraemic sub-jects.86The subsequent metabolism of X-substituted glutathiones to mercapturicacids is thought to proceed through hydrolytic cleavage by glutathionase,present in liver and kidney, forming the corresponding S-substitutedcysteines, followed by an N-acetylation,87 since various S-alkylglutathiones 7gand S-alkylcysteines,67, 7% 75 when administered to animals by injection areconverted extensively into the mercapturic acids.The participation ofglutathionase has recently been confirmed in rat-kidney microsome prepara-tions.88 Apparently the hydroxyalkylmercapturic acids are metabolites onlyof alkyl halides and are not formed in vivo from X-alkylglutathiones or S-alkylcysteines. Their formation may possibly involve the conversion of thehalide into an epoxide or halogeno-alcohol prior to the reaction with gluta-fhi0ne.~1, 7*Methionhe.-Numerous studies of the incorporation of methionine intovarious proteins and tissues have been rep0rted.8~ Methionine taken up by7 8 A. J. Cohen and J.N. Smith, Biochem. J . , 1964, 90, 449, 457.?9 J. Booth, E. BoyIand, and P. Sims, Biochem. J., 1960,74, 117; P. Sima and P. L.8 0 J. Booth, E. Boyland, and P. Sims, Bhchem. J., 1961, 79, 516; S. Al-Kassab,81 P. L. Grover and P. Sims, Biochem. J . , 1964, 90, 603.83 E. Boyland and K. Williams, Biochem. J., 1965,94, 190.84 E. Boyland and L. F. Chasseaud, Biochem. J . , 1966, 98, 13P.86 T. Kuwaki, Acta Med. Okayama, 1964, 18, 333.S6 T. Kuwaki, J. Bwchem. (Japan), 1965, 57, 125.87 H. G. Bray, T. J. Franklin, and S . P. James, Biochem. J., 1959,71, 690.T. Suga, H. Kumaoka, and M. Akagi, J . Biochem. (Japan), 1966,60,133.89 J. L. Sirlin, J. Jacob, and C. J. Tandler, Biochem. J., 1963,89, 447; Z. Pokorny,J. Neuwirt, J. Borova, and K. Sule, Acta Biol.Med. Ger., Suppl. 3, 1964,300; A. Wiernyand H. Bergner, Arch. Tieremaehr., 1964,14,317; E. L. Gadsden, C. H. Edwards, A. J.Webb, and G. A. Edwards, J . Nutrition, 1966,87, 139; A. I. Nikolaev and A. I. (Taziev,Grover, ibid., 1965, 95, 156; P. Sims, ibid., 1966, 98, 215.E. Boyland, and K. Williams, &id., 1963, 87, 4.M. K. Johnson, Biochem. J., 1966, 98, 44MAW: SULPHUR-CONTAINING AMINO-ACIDS 645the liver fluke Fusciokz hepatica is not incorporated into protein, but isdegraded and the methyl-carbon and the sulphur atom are metabolized bydifferent pathways.90 Methionine transport in ChloreZEcc eruZga4-is has beenreviewed by Sh~ift.~l The amino-acid acts as an antimutagen in Schixo-sacchrmyces pmbe,Q2 and inhibits respiration in turnip slice~.~3 The lattereffect is annulled by adenine, and this may be a counteraction of the effect ofmethionine in reducing the intracellular levels of ATP and ADP throughthe formation of S-adenosylmethionine.Methionine has a specific stimu-latory effect on cephalosporin synthesis in CephaZosporium acremonium,acting as a precursor of cysteine via the cystathionine pathway.gQA methionine-requiring mutant of E. coli when starved of methionhe isable to synthesize RNA in the absence of protein synthesis,95 codrming anearlier observation. In this mutant, the normally stringent control of RNAsynthesis by amino-acids appears to be relaxed. Methionine has also beenfound to produce a preferential stimulation of RNA synthesis in certainstrains of E .coli, with little change in the growth rate.96 The excess RNAwas located mainly in the ribosomes.In chicken liver homogenates both isomers of the a-hydroxy-derivative ofmethionine give rise to methionine via oxidative and transaminative steps.Q7A study of the parallel conversion in rat tissue preparations has shown 8requirement for a flavin coenzyme, partly replaceable by NAD or NADP,and a requirement for glutamine which can be partly replaced by a~paragine.~~In yeasts, methionine can be converted into the a-hydroxy-derivative,which is then released into the medium.Qs Recent work by Goodwin andothers,lW using 14C-methyl, 35S-labelled methionine indicates that thisamino-acid provides an intact thiomethyl unit for the biosynthesis of thethiazole ring of thiamine in yeast.The data presented suggest that thecomplete carbon chain of methionine is incorporated. An enzyme systemhas been detected in rat liver and kidney preparations capable of oxidizingthe methyl group of methionine and other S-methyl compounds to CO,.The oxidation does not involve an initial transmethylation step and appearsto be the result of the peroxidatic action of catalase in the presence of aH,O,-generating system.101Numerous reports on the evolution of ethylene from plant tissues haveappeared in recent years. Mapson and others lo2 found that in a model systemVoprosy Med. Khim., 1965, 11, 66; M. ROUX, Cornpt. Rend. SOC. Biol., 1965, 159, 709;E. R. Smith, D ~ P . Abs., 1965, 26, 1123; S. A. Morenkova; Nature, 1966, 209, 917.~~ _ _ _ ~ _ _ ~~ - ~____ ~ _ _E.M. Pantelouris, Res. Vet. Sci., 1965, 6, 334.A. Shrift, Plant PhgsioE., 1966, 41, 405.a a C. H. Clarke, J . Qen. Microbiol., 1965, 39, 21.a3 D. D. Davies, J . Exp. Botany, 1964, 15, 538.84 P. G. Caltrider and H. F. Niss, Appl. Microbiol., 1966, 14, 746.e5 G. Turnock and D. G. Wild, Bwchem. J., 1965, 95, 597.O6 W. H. Matchett, Bucteriol. Proc., 1966, 93; E. Z . Ron and B. D. Davis, J. MoE.BioE., 1966, 21, 13.R. S. Gordon, Ann. New York Acad. Sci., 1965, 119, 927.B. W. Langer, Bwchem. J., 1965, 95, 683.g g 0. A. Maw and C. M. Cope, Arch. Biochem. Biophys., 1966, 117, 499.loo D. 33. Johnson, D. J. Howella, and T. W. Goodwin, Biochem. J., 1966, 98, 30.Iol E. J. Kuchinskas, Arch. Biochem. Biophys., 1965, 112, 605, 610.Io2 M.Lieberman, A. T. Kunishi, L. W. Mapson, and D. A. Wardale, Bioch.em. J.,1965, 97, 449646 BIOLOGICAL CHEMISTRYmethionine in the prcsence of ascorbate and cupric ions was degraded aero-bically to ethylene and a variety of other products, including acrolein andmethanethiol. The two-carbon ethylene unit was established as originatingfrom carbon atoms 3 and 4 of the amino-acid. Fission of the G-S bond wasconsidered to be facilitated by complex formation as follows:Methional (3-methylthiopropionaldehyde) proved to be a more effectimsource of ethylene, and may be an intermediate in the reaction. Methioninesulphoxide in this system gave little ethylene, but appreciable quantities ofmethane .It was subsequently shown that methionine stimulates ethylene produc-tion in apple tissue slices, the synthesis being inhibited by the presence ofcopper-binding reagents.loS The biosynthesis of ethylene may thereforeinvolve a copper-containing enzyme, possibly a peroxidase.In an enzymesystem obtained from cauliflower florets which degrades methionine toethylene, methional was again found to be a more active substrate. Cleavageof the aldehyde in the presence of enzymically generated hydrogen peroxideappears to take place.lO*Ethylene has been reported to be formed both enzymically and non-enzymically from a nonprotein fraction of pea seedlings.106 The nonenzymicreaction has been further investigated and the ethylene precursor identifiedas methionine.106 This system gave conversions ranging from 50-80~0,and required light and flavin mononucleotide or riboflavin.Methional alsoacted as a source of ethylene, as did ethionine, homocyst(e)ine and the a-hydroxy-derivative of methionine.N-Formylmethionine and Protein Synthesis.-One of the more significantdevelopments in the field of sulphur biochemistry has been the recognition ofN-formylmethionine as an apparently unique initiator of protein synthesis.Studies on the nature of the terminal amino-acid residues of various proteinshas revealed that in many instances the terminal residues do not exhibit arandom distribution. In the soluble and ribosomal proteins of E . coli the endgroups are accounted for largely as methionine and to a lesser extent asalanine, with serine and threonine in minor amounts.107 Certain algal bili-proteins also contain methionine as the major N-terminal reaidue.lO* TheloS M.Lieberman, A. T. Kunishi, L. W. Mapson, and D. A. Wardale, Plccrct Phyhl.,lo' L. W. Mapson and D. A. Wardale, Biochem. J., 1966,101, 6P.lo6 S. F. Yang, H. S . Ku, and H. K. Pratt, Biochem. Bwphys. Res. Comm., 1966,24,lo7 J.-I?. Waller, J . Mot. Bhl., 1963, 6, 483.lo* P. 6Carra, Biochem. J., 1965, 94, 171.1966, 41, 376.F. B. Abeles and B. Rubinstein, Biochim. Biophys. Acta, 1964, 93, 676.739NAW : SULPHUR-CONl'AINING AMINO-ACIDS 647occurrence of this nonrandom distribution of methionine and alanine intro-duced the possibility that these amino-acids might constitute starter signalsfor the initiation of protein synthesis.Marcker and Sanger 109 showed that sRNA from E.coli in a cell-freesystem reacts with methionine, and that after the initial attachment theor-amino-group of the methionine residue may become formylated. Twodistinct species of methionine-accepting sRNA's have been identified.One of these gives rise to a methionyl-sRNA (met-sRNAF) which cansubsequently undergo formylation ; the other gives a methionyl-sRNA (met-sRNAM) which is not formylated. The possible role of formylmethionyl-sRNA as an initiator of peptide chain synthesis was suggested, and this hasbeen confirmed, at least for some bacterial proteins.11oIn a cell-free system from E. coli using bacteriophage RNA as messenger,N-formylmethionine became incorporated into several proteins, includingphage coat protein and, furthermore, the amino-acid residue adjacent toformylmethionine was found to be alanine.The occurrence of alanine as anN-terminal residue in native coat protein could be accounted for by theenzymic removal of the formylmethionyl group from the nascent protein.Likewise, the appearance of methionine as an N-terminal residue in otherproteins could result from the cleavage of the formyl group alone.Subsequent work by Capecchi ll1 suggests that in some proteins theamino-acid residue adjacent to formylmethionine may be serine or 8ome otheramino-acid, rather than alanine. Enzymic removal of the formyl or formyl-methionyl moieties could thus explain the observed distribution of N-terminal residues in E.wli proteins.107Marcker and others,112 using a similar E. coli cell-free system withbacteriophage RNA as messenger, have shown that formylmet-sRNA, givesrise to at least two different polypeptides, both containing formylmethionineas the N-terminal residue, the adjacent amino-acid residues being alanine andserine, respectively. It is significant that met-sRNAF in the absence of atransformylating system gave rise to peptides with N-terminal methionineresidues, indicating that the formyl group is not essential in locating amethionine residue in the N-terminal position. Its presence appears toincrease the rate of formation of the first peptide bond. The ability to directthe location of methionine residues resides in the sRNA, since met-sRNAHunder these conditions gave peptides with methionine solely in internal posi-t ions. Met - sRNA, resembles other aminoacyl- sRNA' s in ribosomal- bindingproperties. On the other hand, met-sRNAF can bind to the site which isspecific for polypeptidyl-sRNA, and this met-sRNA species must act as aninitiator of protein synthesis by virtue of its affinity for this site.The methionine analogues, norleucine and ethionine, can substitute formethionine in the acylation of sRNA, and both can be as readily formylatedloS F.Sanger and K. A. Marcker, J . Mol. BioZ., 1964, 8, 835; K. A. Marcker, ibid.,J. 111. Adams and M. R. Capecchi, Proc. Nut. Acad. Sci. U.S.A., 1966, 55, 147;ll1 M. R. Capecchi, Proc. Nut. A d . Sci. U.S.A., 1966, 65, 1517.11* B. F. C. Clark and K.A. Marcker, Nature, 1966, 211, 378; M. S. Bretscher and1965, 14, 63.R. E. Webster, D. I;. Engelhardt, and N. D. Zinder, ibid., p. 155.K. A. Marcker, ibid., p. 380648 BIOLOGICAL CHEMISTRYafter attachment to sRNA.l13 These analogues might therefore be expectedto appear in N-terminal as well as internal positions in peptide chains.Several Papers have been published on the coding requirements forformylmethionine. The triplets AUG, GUG, and UUG promote bindingof met-sRNAp or formylrnet-sRNAF to ribosomes. AUG also promotesbinding of met-sRNAM.114S-Adenosylmethionine.-The chemistry and biochemistry of S-adenosyl-methionine have recently been extensively reviewed,115 and a simplified andimproved procedure for its production from yeast has been described.l16This compound accumulates in the vacuoles of yeast cells, particularlyduring growth in the presence of methionine, but is released into the culturemedium during sporulation.l17 S- Adenosylmethionine has been identifiedas the cofactor required with thiamine pyrophosphate for the conversion ofpyruvate to acetylcoenzyme A and formate in E.coti.ll8 The levels of thesulphonium compound in the white blood-cells of subjects with chronicmyelocytic leukamia are reported to be as much as four times normalvalues,11g although the significance of this increase is so far unknown.Further studies have been made of the varioustransmethylations requiring X-adenosylmethionine as the methyl donor. Abacterial S-adenosylmethionine-magnesium protoporphyrin methyl trans-ferase has been described,l20 and the X-adenosylmethionine-homocysteinemethyl transferase of yeast has been further purified and characterized.lalX-Methylmethionine is almost equally effective a methyl donor substrate asX-adenosylmethionine for the latter enzyme.This is in contrast to the homo-cysteine methyl transferase of jack bean for which X-adenosylmethionine hasonly one-tenth of the substrate activity of S-methylmethionine.122 X-Adenosylmethionine-histamine methyl transferase is reported to be moreabundant in male than in female rat kidneys. This parallels the sex differencein histamine and N-methylhistamine excretion among rats.123 S-Adeno-sylmethionine yields methyl groups for the synthesis of sterols in yeast.Thistransmethylation is greatly stimulated by the presence of ~arb0nate.l~~Transmethylation.113 J. Trupin, H. Dickerman, M. W. Nirenberg, and H. Weissbach, Biochem. Biophys.Res. Comm., 1966, 24, 50.114 €3. F. C. Clark and K. A. Marcker, Nature, 1965, 207, 1038; J . MoZ. BioZ., 1966,17, 394; T. Nakamoto and D. Kolakofsky, Proc. Nut. Acad. Sci. U.S.A., 1966, 55, 606;D. A. Kellogg, B. P. Doctor, J. E. Loebel, and M. W. Nirenberg, ibid., p. 912; R. E.Thach, K. F. Dewey, 5. C. Brown, and P. Doty, Science, 1966,153,416; T. A. Sundarara-jan and R. E. Thach, J . MoE. Biol., 1966, 19, 74.116 F. Schlenk, Fortschy. Chem. org. NaturstofSe, 1965, 23, 61 ; S. K. Shapiro and F.Schlenk, “ Transmethylation and Methionine Biosynthesis,” Univ. of Chicago Press,Chicago, Illinois, 1965.116 F. Schlenk, C.R. Zydek, D. J. Ehninger, and J. L. Dainko, Enzymologia, 1965,29, 283; S. K. Shapiro and D. J. Ehninger, Analyt. Biochem., 1966, 15, 323.117 G. Svihla, J. L. Dainko, and F. Schlenk, J . Bucteriol., 1964, 88, 449.118 J. Knappe, E. Bohnert, and W. Briimmer, Biochim. Biophys. Acta, 1965, 107,llS R. J. Baldessarini and P. P. Carbone, Science, 1965, 149, 644.120 K. D. Gibson, A. Neuberger, and G. H. Tait, Bwchem. J., 1963, 88, 325.lZ1 S. K. Sha,piro, D. A. Yphantis, and A. Almenas, J . Biol. Chem., 1964,239,1561;122 L. Abrahamson and S. K. Shapiro, Arch. Biochem. Biophys., 1965, 109, 376.123 S. H. Snyder and J. Axelrod, Biochim. Biophys. Actu, 1965, 111, 416.124 J. R. Turner and L. W. Parks, Biochim.Biophys. Acta, 1965, 98, 394.603.S. K. Shapiro, A. Almenas, and J. F. Thomson, ibid., 1965, 240,2512MAW : SULPHUR-CONTAINING AMINO-ACIDS 649Met h yla tion of p hosp hat idylmonome t h yle t hanolamine to p hospha tid y lcho -line by 8-adenosylmethionine has also been demonstrated in liver prepara-Methylated purines and pyrimidines are known to be constituents ofDNA and RNA, particularly transfer RNA. The N-methyl groups of thesebases originate from methionine, and S-adenosylmethionine functions as themethyl donor, the transmethylations taking place a t the polynucleotideRNA methyl transferases, for example, have been detected in awide variety of species, and in a methionine-requiring mutant of S. cerevisimmethionine deprivation resulted in the formation of submethylated RNA.lZ7Infection of E.coli with certain bacteriophages causes an increase in theactivity of DNA methyl transferase. This appears to be a phage-directedincrease in enzyme synthesis, rather than an enzyme activation.128Baoteriophage infection of E. coli also resulted in the appearance of anenzyme degrading S-adenosylmethionine to 5’-methylthioadenosine andhomoserine, which is absent in normal E. c0Zi.1~9CataboEism. An enzyme apparently confined to the pituitary gland whichsplits S-adenosylmethionine to S-adenosylhomocysteine and methanol hasbeen identified.13* This reaction could be the source of the methanol foundnormally in urine and breath. The further metabolism of S-adenosylhomo-cysteine, the demethylation product of S-adenosylmethionine, has beenstudied by Duerre and others.A nucleosidase present in E. coli and othergram-negative bacteria is able to split off the adenine moiety, yieldingS-ribosylhomocysteine. The homocysteine residue of this product is incor-porated into protein-methionine in E. coli, and this appears to proceedthrough the further cleavage of S-ribosylhomocysteine to homocysteine andribose. The cleavage enzyme was found in E. cold but not in liver or yeast.lalThe above findings lead to the following catabolic sequence for S-adeno-sylmethionine, and account satisfactorily for the regeneration of the com-pound from S-adenosylhomocysteine and homocysteine :tions.S-Ad-methionine ---+ S-Ad-homocysteine ---+ S-RibosylhomocpteineA further catabolite of S-adenosylmethionine, 5’-methylthioadenosine,is formed together with homoserine lactone by direct enzymic cleavage, andalso arises from the decarboxylation of S-adenosylmethionine and transfer ofE.F. Marshall, T. Chojnacki, and G. B. Ansell, Biochem. J., 1965, 95, 30P.128 E. Borek and P. R. Srinivaaan, Ann. Rev. Biochem., 1966, 35, 276.127 K. Kjellin-StrBby and H. 0. Boman, Proc. Nut. Acad. Sci. U.S.A., 1965,58,1346.128 R. Hausmann and M. Gold, J. Biol. Chem., 1966,241, 1985.lag M. Gefter, R. Hausmann, M. Gold, and J. Hurwitz, J . Biol. Chem., 1966, 241,130 J. Axelrod and J. Daly, Science, 1965, 150, 892.J. A. Duerre, J . Bwl. Chem., 1962, 237, 3737; J. A. Duerre and P. M. Bowden,Bwchern. Biophys. Res. Comm., 1964,16,150; J.A.,Duerre and C. H. Miller, J . Bacterwl.,1966, 91, 1210.1995650 BIOLOGICAL CHEMISTRYthe propylamino-group to putrescine. The nucleoside has been shown toundergo reconversion to 8-adenosylmethionine in Cundida utilis.132 Thispathway affords a second mechanism whereby fragments of the S-adenosyl-methionine molecule can be re-utilized and, like the conversion of S-adeno-sylhomocysteine into homocysteine, may well represent a cellular economy of8-adenosylmethionine in group transfer reactions.Csstathionine.-Cystathionine is now k l y established as an inter-mediate in trans-sulphuration reactions between cysteine and methionine inboth animals and micro-organisms. This thioether accumulates in the brainof pyridoxine-deficient animals 133 and is present in relatively large amountsin normal human and monkey brain.134 It originates solely from dietarymethionine in animals, and from cysteine in bacteria, whereas both pathwaysare operative in moulds and yeasts.l35Studies by Rowbury and Woods,13* and by Flavin and co-workers ls5~ lS7have shown that in E.coli and Salmonella typhimurium, cystathioninesynthesis from cysteine involves two main steps, namely the initial formationof O-succinylhomoserine from homoserine and O-succinylcoenzyme A, andthe subsequent replacement of the succinyl group by cysteine. There is alsoevidence that in Neurospora, O-acetylhomoserine rather than O-succinyl-homoserine may be the form in which homoserine reacts with cysteine.ls8Further purification of the cystathionine synthetase systems of rat liver,139and of Salmonella,140 have been described.Cleavage of cystathionine in liver proceeds by it pyridoxal phosphate-dependent y-elimination, giving cysteine, a-ketobutyrate, and ammonia. Thebacterial cystathionase yields homocysteine, pyruvate and ammonia by ap-elimination.Both types of cleavage enzyme have been identified inNeurospora and yeasts, accounting for the reversible trans-sulphurationsoccurring in these organisms.28, 135, 141 The distribution of the y-cleavageenzyme and of cystathionine synthetase in various tissues of a number ofanimal species has been r e ~ 0 r t e d . l ~ ~Control of Sulphur Amino-aid Biosynthesk-The general pathway ofcysteine biosynthesis in sulphate-assimilating bacteria and in yeasts has beenadequately reviewed el~ewhere.1~3 Interest has also centred on the controlmechanisms affecting cysteine synthesis and its conversion to methionine.In E .wli, Bacillus subtilis and 8. typhimurium the synthesis of ATP sulphury-lSa F. Schlenk and D. J. Ehnhger, Arch. Biochem. Biophys., 1964, 106, 95.133 D. B. Hope, J . Neurochem., 1964,11, 327.lS4 H. H. Tallrtn, 5. Moore, and W. H. Stein, J . Biol. Chem., 1958, 230, 707; H.ShimiZu, Y. Kakimoto, and I. Sano, J. Newrochem., 1966,13, 65.lS6 C. Delavier-Klutchko and M. Flaxin, J. BWZ. Chem., 1965, 240, 2537.IS6.R. J. Rowbury and D. D. Woods, J . Uen. Alicrobiol., 1964, 36, 341; R. J. Row-bury, zbzd., 1964, 37, 171; Bzochern. J., 1964, 93, 20P.lS7 M. Flavin, C.Delavier-Klutchko, and C. Slaughter, Science, 1964,143, 50; M. M.ICaplan and M. Flavin, Biochim. Biophys. Acta, 1965, 104, 390.la* S. Nagai and M. Flavin, J. Biol. Chem., 1966, 241, 3861.lSB A. Nagabhushanam and D. M. Greenberg, J. BioZ. Chem., 1965,240, 3002.14* M. M. Kaplan and M. Flavin, J. BWZ. Chern., 1966, 241, 4463.141 M. Flavin and A. Segal, J. Biol. Chem., 1964, 239, 2220.Ira S. H. Mudd, J. D. Finkelstein, F. Irreverre, and L. Laster, J. Biol. Chem., 1966,143 H. D. Peck, Bacterial. Rev., 1962, 26, 67.240,4382MAW : SULPHUR-CONTAINING AMINO-ACIDS 651lase, adenylylsulphate (APS) kinase, 3'-phosphoadenylylsulphate (PAPS)reductase and sulphite reductase is repressed by growth in the presence ofcyst(e)ine.l44, 145 Sulphite and sulphide are also repressors of sulphat'eactivation, and sulphide represses sulphite reduction, but Pasternak 145 con-siders that these compounds act only after conversion into cysteine.Therepression by cystine of sulphate activation and reduction, and sulphite re-duction is coincident a t high concentrations (0.85 mM) while low cystinelevels (0.05 mM) repress sulphate activation only. This differential repressionenables the organism to utilize intermediates, such as sulphite, and at the sametime blocks the synthesis of unwanted enzymes in the pathway. The sul-phate-transport system in S. typhimurium is inhibited by sulphite and isrepressed by ~ysteine.1~~ These last mechanisms, by imposing an immediatecontrol on the amount of sulphate entering the cell, represent the pre-liminary regulatory process in sulphate metabolism.In yeasts a number of regulatory mechanisms have been identified.147Sulphate activation to AX'S is inhibited by APS and by PAPS (product in-hibition), by sulphide (feedback inhibition), and is repressed by meth-ionine.h addition, sulphite reductase is repressed by cysteine.Rowbury and WoodslP8 have found that the enzymes mediatingmethionine biosynthesis from cysteine in E. coli, namely, homoserine O-trans-succinylase, cystathionine synthetase, cystathionase, and the homo-cysteine methylase complex, are repressed by growth in the presence ofmethionine. In addition, homoserine O-tram-succinylase is inhibited bymethionine (feedback inhibition), and cystathionase is inhibited by homo-cysteine (product inhibition).The control mechanisms of cysteine synthesis from methionine in mam-malian liver have also been studied.149Ethionine.-Metabolic efects.Earlier work on the biochemistry ofethioninc has been summarized in the comprehensive reviews of Stekol150and of Farber.151 This amino-acid has continued to attract much attentionin view of its biochemical role as a metabolic trap for ATP and as a livercarcinogen. The injection of ethionine in animals initiates a sequence ofbiochemical events in the liver which have been well described by Farber andothers,15a namely the development of an acute cellular deficiency of ATP,then a marked inhibition of RNA synthesis followed by an inhibition ofIr14 J. Droyfusa and K.J. Monty, J . Biol. Chem., 1963, 238, 3781.Ids R. J. Ellis, S. K. Humphries, and C. A. Pasternak, Biochem. J., 1964, 92, 167;C. A. Pasternak, R. J. Ellis, M. C. Jones-Mortimer, and C. E. Crichton, ibid., 1965, 96,270; J. F. Wheldrake and C. A. Pasternak, ibid., p. 276.146 J. Dreyfuss, J. Biol. Chem., 1964, 239, 2292.14' P. C. de Vito and J. Dreyfuss, J . Bacterial., 1964, 88, 1311.148 R. J. Rowbury and D. D. Woods, J . Gen. Microbiol., 1964, 35, 145; ibid., 1966,42, 155.IrlB A. Kato, M. Ogura, H. Kimura, T. Kawai, and M. Suda, J . Biochem. (Japan), 1966,69, 34; A. Kato, M. Ogura, and M. Suda, ibid., p. 40.160 J. A. Stekol, Adv. Enzyml., 1963, 25, 369.162 E. Farber, K. H. Shull, S. Villa-Trevino, B. Lombardi, and M. Thomae, Nature,1964, 203, 34; S.Villa-Trevino, K. H. Shull, and E. Farber, J . BioZ. Chem., 1966,241, 4670; K. H. Shull, J. McConomy, M. Vogt, A. Castillo, and E. Farber, &bid.,p. 5060.E. Farber, Adv. Cancer Res., 1963, 7, 383652 BIOLOGICAL CHEMISTRYprotein synthesis. This in turn is followed by an accumuIation of fat in theliver, the release of triglyceride fatty acids being impaired.153Injection of ethionine induces a marked hypoglycsmia in female rats, andit has been suggested that this may be an important factor in the genesis ofethionine fatty livers.154 Glucose administration decreases the accumulationof hepatic triglycerides, and the role of the hypoglycaemia produced byethionine may be to increase the transport of triglycerides to the liver.Biotin may also be involved in the development of ethionine fatty livers.155In female rats the intestinal transport of triglycerides containing long-chain fatty acids is reduced after an injection of ethi0nine.1~~ Inhibition ofprotein synthesis in the intestinal mucosa, leading to diminished chylomicronformation is considered to account for this effect on lipid transport.Methionine, adenine, and ATP are able to annul many of the lesionsproduced by ethionine.However, the last two compounds are selective intheir action. They fully alleviate the inhibition of methionine incorporationinto liver protein, but afford only partial protection against the inhibition oftransmethylations to lecithin precursors.157 These findings strengthen theview that ethionine exerts multiple biochemical effects in the whole organism,due to different basic mechanisms.Some effects, e.g., inhibition of proteinsynthesis, are the result of an induced deficiency of ATP in the liver, whileothers are due to an inhibitory effect of hdenosylethionine on transmethyla-tion reactions involving S-adenosylmethionine.In contrast to the effect of injected ethionine in rats,152 ethionine added t'othe diet was found to stimulate RNA synthesis,158 and when added to themedium of a methionine-requiring auxotroph of E. wli, the compoundproduced an increased synthesis of DNA.15g In 8. typhimurium adenineenhances the growth-inhibitory effect of ethionine.l60 Adenine stimulatesethionine uptake by the cells and may exert its effect by increasing the rate atwhich the inhibitor can reach its primary site of action.A number of other metabolic effects of ethionine have been reported,including the inhibition of tropolone biosynthesis in Penicillium stipitatum,181the induction in rats of premature birth,ls2 experimental porphyria,lS3 anincreased afltinity of the liver for iron,l64 and effects on biliary secretion.ls6Feeding of ethionine to animals leads to changes in the levels of variousenzymes in the liver, for example a decrease in glycogen synthetase,ls8 and153 A. Bezman-Tarcher, P.J. Nestel, J. M. Felts, and R. J. Havel, J . Lipid Res., 1966,154 B. Combes and S. Schenker, Nuture, 1966, 209, 911.166 M. Marchetti, V. Ottani, and P. Puddu, Arch. Biochem. Biophys., 1966, 115, 84.156 D. E. Hyams, S. M. Sabesin, N. J. Greenberger, and K. J. Isselbacher, Biochim.157 L. S. Gordon and E. Farber, Arch. Biochem. Biophys., 1965,112,233.158 M. K. Turner and E. Reid, Nature, 1965, 203, 1174.159 R. C. Smith and W. D. Salmon, J . Bacterwl., 1965, 89, 687.160 R. C. Smith and W. D. Sahon, J . Bacterial., 1965, 89, 1494.1 8 1 R. Bentley, J. A. Ghaphery, and J. G. KeiI, Arch. Biochem. Biqhys., 1965, 111, 80.162 B. F. Chow and C. E. Agustin, Nature, 1966, 210, 1271.163 A. Palma-Carlos, I;. Palma-Carlos, M. Gajdos-Torok, and A. Gajdos, Nature,164 T. D. Kinney, N. Kaufmann, and J. V. Klavins, N d w r e , 1966, 211, 857.166 G. Barber-Riley, Experkntia, 1966, 22, 233.166 H. G. Sie and A. Hablanian, Biochem. J., 1965, 97, 32.7, 248.Biophys. Acta, 1966, 125, 166.1966, 211, 977MAW: SULPHUR-CONTAINING AMINO-ACIDS 653ornithine transcarbamoylase,l67 and an increase in glucose-6-phosphatedehydrogenase,fe% 167 and arginase.167 Liver cystathionase is increased bya single injection of ethionine.lss Ethionine by injection also produces a,decrease in liver NADP.l‘j9Metabolism. The primary metabolite of ethionine in both animals andmicro-organisms is S-adenosylethionine,l50~ l 7 0 most probably a key inter-mediate in the various transethylation reactions involving ethionine whichhave been reported.170, 1 7 1 The major excretory product of the amino-acidin the rat is reported to be 5’-ethylthioinosine.172 In addition, a small butsignificant fraction of the ethyl group is converted into ethanol and acetate.In yeasts the sulphur of ethionine becomes incorporated to a significantextent in protein- cystine and methionine. 73 This trans-sulphurationprobably takes place via the initial formation of X-adenosylethionine, followedby de-ethylation to 8-adenosylhomocysteine, with homocysteine as a subse-quent intermediate. The a-hydroxy-derivative of ethionine has been identi-fied as a further catabolite in yeasts.QgEthionine resistance. The development of micro-organisms resistant to theeffects of ethionine has been the subject of a number of investigations.Ethionine-resistant mutants of 8. cerevisiae and N . crassa, also yeasts maderesistant by growth in the presence of elevated concentrations of theinhibitor,showed a diminished ability to take up ethionine.173~ 174 In the ethionine-resistant yeasts there was an enhanced ability to metabolize ethionine-sulphur to cystine and rnethi~nine.~’~ C. utilis adapted to very high con-centrations of ethionine released considerable amounts of methionine into thegrowth and an ethionine-resistant mutant of N . crassa has beenfound to excrete a variety of methionine derivatives.176 In the latterorganism, control of methionine synthesis appears to have been lost.Sulphur-aminoacidurias.-An earlier account of these hereditary diseaseshas been given by Crawhall.177Cystathioninuria. The nature of this defect has been discussed byF r i m ~ t e r , ~ ~ ~ who has confirmed previous hdings that it is associated with adeficiency of cystathionase in the liver. Addition of pyridoxal phosphate topreparations of cystathionase from the livers of cystathioninurics produced anincrease in the activity of the enzyme, and it was suggested that the defeotcould be due to a structural alteration of the apoenzyme, resulting in its167 P. McLean, Biochem. J., 1966, 99, 776.16* 0. Durieu-Trautmann and F. Chatagner, Bull. SOC. Chim. biol., 1966, 48, 77.16@ A. L. Greenbaum, J. B. Clark, and P. McLean, Biochem. J . , 1964, 93, 17C; T. F.Slater and B. C. Sawyer, ibid., 1966, 101, 24.170 L. W. Parks, J . Bid. Chem., 1958, 232, 169; R. C. Smith and W. D. Salmon,Arch. Biochem. Bwphys., 1965, 111, 191.171 A. D. Argondelis and D. J. Mason, Biochemistry, 1965, 4, 704. E. L. Patterson,J. H. Hash, M. Lincks, P. A. Miller, and N. Bohonos, Science, 1964, 140, 1691; S. K.Shapiro, A. Almenas, and J. F. Thomson, J . Biol. Chem., 1965, 240, 2512.172 Y. Natori and H. Tamer, Biochim. Biophp. Acta, 1965, 107, 136.173 G. A. Maw, Arch. Biochem. Biophys., 1966, 115, 291.174 W. A. Sorsoli, K. D. Spence, and L. W. Parks, J . Bwteriol., 1964, 88, 20. M. S.176 M. Musilkova and Z. Fencl, Folia Microbial., 1964, 9, 374.176 S. B. Galsworthy and R. L. Metzenberg, Biochemistry, 1965, 4, 1183.177 J. C. Crawhall, Ann. Reports, 1964, 61, 484.178 G. W. Frimpter, Science, 1965, 149, 1095.Kappy and R. L. Metzenberg, Biochim. Biophys. Acta, 1965, 107, 425654 BXOLOQICAL CHEMISTRYinability to combine normally with the coenzyme. This deficiency of cysta-thionase in cystathioninuric liver was found by Mudd and co-workers179to be accompanied by a deficiency of homoserine dehydratase, confirmingGreenberg’s observation lSo that cystathionase and homoserine dehydrataseare two activities of the same enzyme.Homocystinuria. A substantial number of cases of this condition havenow been identified by the Carson group and by Schimke and co-workers,resulting in a clearer characterization of the disease. lS1 Homocystinuricsoften show elevated levels of methionine as well as homocystine in the blood,and low or negligible levels of cystathionine in the brain.ls2 The funda-mental biochemical disturbance is a deficiency of cystathionine syntheta~e.1~3In this condition there is a marked impairment of cystine formation frommethionine, and cystine may consequently become nutritionally essential forthe maintenance of nitrogen balance.184The urinary excretion of homocystine and methionine accounts for only asmall fraction of the dietary intake of methionine, even though the majordegradative pathway for methionine is blocked a t the point of cystathioninesynthesis.185 In the Course of an examination of homocystinuric urine foradditional sulphur compounds to explain this discrepancy, small amounts ofhomolanthionine, a higher homologue of cystathionine were detected,1s6also a new derivative of homocysteine which was identified as 5-amino-4-imidazolecarboxamide-5’-5- homocysteinylriboside : 87HO OHThe metabolic origin of this compound is not as yet apparent. Possibilitiesdiscussed by Perry and others lS7 include an enzymic interaction betweenhomocysteine and aminoimidazolecarboxamide, or a hitherto unrecognizedJ. D. Finkelstein, S. H. Mudd, F. Irreverre, and L. Laster, Proc. Nut. A d . Sci.U.S.A., 1966, 55, 865.N. A. J. Carson, C. E. Dent, C. M. B. Field, and G. E. Gaull, J . Pediat., 1965,66,565; R. N. Schimke, V. A. McKuaick, T. Huang, and A. D. Pollack, J . Amer. Med.A ~ ~ o c . , 1965, 193, 711.D. P. Brenton, D. C. Cusworth, and G. E. Gad, Pediatrics, 1965,35, 50.laa S. H. Mudd, J. D. Finkelstein, F. Irreverre, and L. Laster, Science, 1964, 143,1443; ibid., 146, 785.lE4 D. P. Brenton, D. C. Cusworth, and G. E. Gaull, J. Pediat., 1966, 67, 58; D. P.Brenton, D. C. Cusworth, C. E. Dent, and E. E. Jones, Quart. J . Med., 1966, 35, 325;D. P. Brenton and D. C. Cusworth, Clin. Sci., 1966, 31, 197.L. Laster, S. H. Mudd, J. D. Finkelstein, and F. Irreverre, J . CGn. Invest., 1965,44, 1708.T. L. Perry, S. Hansen, H.-P. Bar, md L. MacDougall, Science, 1966, 152, 776.180 Y. Matsuo and D. M. Greenberg, J . Biol. Chem., 1958,230,545.la6 T. L. Perry, S. Hansen, and L. MacDougall, Science, 1966,152, 1750MAW: SULPHUR-CONTAINING AMINO-ACIDS 655degradation of S-adenosylhomocysteine , with S-inosylhomocysteine as anintermediate.HypermetiLioninaemia. A number of cases have been reported of severeinfantile liver disease, characterized by a 30- to 50-fold increase in themethionine level in the blood, and by a gross aminoaciduria.188 Among theamino-acids excreted in abnormally large amounts were methionine, methio-nine sulphoxide, homocystine , and cystathionine. a-Keto-acids, includingthe a-keto-analogue of methionine, were also present. The nature of themetabolic defect in this condition is nnknown.Cystinuria. In addition to cystine, the urine of cystinuric subjects con-tains cysteine homocysteine mixed di~ulphide.~4, Urinary cystine origin-ates solely from plasma cystine.190 The intestinal transport of cystine incystinurics has been shown to be defe~tive,l8~, 191 and defects in the renaltransport of cystine have been rep0rted,18~ although Segal and others 181found that cystine uptake by kidney cortex slices from cystinurics wasnormal. Segal lg2 has presented evidence that the condition is tt geneticerror of dibasic amino-acid transport, and that the associated cystinuria isdue to an impairment by dibasic amino-acids of the cysteine efflux from thekidney. In a later Paper Segal has classified the cystinurias into threebiochemically and genetically different disea~es.1~3lB8 T. L. Perry, D. F. Hardwick, G. H. Dixon, C. L. Dolman, and S. Hansen, Pedi-la@ T. H. Foley and D. R. London, CEin. Sci., 1965, 29, 549.l@O G. W. Frimpter, Clin. Sci., 1966, 31. 207.d T t k , 1965, 36, 236.lg1 M. Fox, S. IThier, L. Rosenberg, w. Kiser, and S. Segal, New Engl. J. Med.,1964, 210. 556.lia L. -Schwartzman, A. Blair, and S . Segal, Biochern. Bwphys. Rea. Conzm., 1966,lgs L. E. Rosenberg, S. Downing, J. L. Durant, and S. Segal, J. CZin. Invest., 1966,23,220.45, 365