1. INTRODUCTIONBy D. J. Manners(The Heriot- Wutt University, Edinburgh, 1 )DURING the past decade, investigations of enzymes have developed fromdescriptive studies of isolation, purification, and properties to the examinationof the detailed stereochemical specificity and mechanism of action. Sub-stantial progress has followed from the application of physicochemicaltechniques and, in particular, of isotopic labelling. The present Reportincludes an authoritative account of such studies with special emphasis onwork published since 1961. From the many examples quoted, those describ-ing the stereochemical changes involved in the conversion of mevalonateinto squalene, lanosterol, and cholesterol are particularly important. In thiswork, the stereospecific synthesis of mevalonate labelled with deuterium ortritium at each of the two 4-positions was a significant preliminary.The microbiological oxidation of aromatic compounds was last consideredin 1956, and it is appropriate to report on the substantial progress which hasbeen made in this field since then. Alternative routes of metabolism to thosealready established have been discovered and considerable progress made instudies on both the purification and mechanism of action of osygenases, inwhich oxygen atoms from molecular oxygen are directly incorporated intothe substrate.The regulation of the synthesis of enzymes concerned iii themetabolism of trytophan and mandelate are also considered.It is customary for the Biological Chemistry section to include at leastone survey of the major classes of naturally occurring high polymers. Sincenucleic acids are being described elsewhere (p.402) and neutral polysac-charides have been reviewed in 1964 and 1965, attention is now focused oncarbohydrate sulphates, and on proteins. Although certain aspects of thechemistry of polysaccharide sulphates have been included in previousReports, recent progress in this field has been so substantial (see for example,Tables 1 and 2, pp. 472, 481) that a more complete account is required.The final section describes the continued and significant progress whichhas been made in studies on peptides and proteins2. THE STEREOSPECIFICITY OF ENZYMIC REACTIONSBy J. W. Cornforth and G. Ryback(Milstead Laboratory of Chemical Enzymology,“Shell ” Research Ltd., Sittingbourne, Kent.)Introduction.-The stereospecificity observed in the great majority ofenzymic reactions can plausibly be attributed to two main causes: the needto accept a limited range of molecular species as substrates for conversioninto a correspondingly limited range of products; and the need for specificgeometrical relationships between the functional groups taking part in thereaction, if the activation energy is to be minimized.A subtle manifestation of stereospecificity in enzymes is the capacity todiscriminate between formally identical groups in a molecule. Characteris-tlically, an enzyme modifying an a group in a substrate Caabd will select thegroup having one particular orientation with respect to b and d.This canbe demonstrated when the product of the enzymic reaction is asymmetric[as when glycerokinase converts glycerol exclusively into (A)-a-glycero-phosphate1] or when asymmetry in substrate or product is deliberatelyinduced by isotopic labelling of one of the a groups. The second is by farthe more general possibility and isotopic labels have been increasingly usedin biochemistry to explore the stereochemistry of enzymic reactions.The subject was excellently reviewed up to 1961 by Levy, Talalay, andVennesland;2 the present Report is largely devoted to progress since thatdate.* With few exceptions, this Report deals with asymmetry revealedby isotopic labelling, since the stereochemical relation between a substrateand a product which are asymmetric when unlabelled is self-evident whenthe absolute configurations are known. In addition, some representativework on exploration of the stereochemical environment of the active sites ofenzymes is mentioned.For most substrates of type Caabd, a is hydrogen, and modern physicalmethods have greatly facilitated recognition of asymmetric treatment of amethylene group by an enzyme. First, polarimeters may now have sensi-tivities of 10-4 to degrees of rotation from the visible region to near thelimit of the quartz ultraviolet.Optical activity due to stereospecific re-placement of one hydrogen atom in a methylene group by deuterium canthus be detected in samples of a few milligrams. Secondly, enzymic reac-tions leading to the loss of one hydrogen atom from a methylene grouplabelled stereospecifically with a hydrogen isotope can be followed by massspectrometry when deuterium is the isotope, or by measuring a shift inT : 14C ratio from that of a doubly-labelled substrate when the asymmetry ofC.Bublitz and E. P. Kennedy, J . Biol. Chem., 1954, 211, 963.H. R. Levy, P. Talalay, and B. Vennesland, Progr. Stereochem., 1962, 3, 299.* The practice, used in that Review,2 of designating acidic substrates of enzymicexperiments as their anions, e.g., “ succinate,” in the text, but as their acidic forms inpictorial formulae, is continued hereCORNFORTH AND RYBACK : STEREOSPECIFICITY O F REACTIONS 429the methylene group is due to tritium. Conversely, stereospecific non-enzymic elimination reactions may be used to demonstrate enzyme-inducedasymmetry in a labelled methylene group, provided that an adjacent asym-metric centre can direct the course of elimination. Thirdly, the alterationproduced in the n.m.r.spectrum of a substance, when one or more of itshydrogen atoms is replaced by deuterium, can be used to relate the absoluteconfiguration of an asymmetrically deuteriated methylene group with thatof an adjacent centre of known absolute configuration. Fourthly, and sofar uniquely, a specimen of monodeuterioglycolate, produced by the actionof muscle lactic dehydrogenase on deuterioglyoxylate (1 ), has been shown 3by X-ray and neutron-diffraction measurements with the 6Li salt to haveE n z ., NADHIl___fI(I) L C02H - ~ n z ., NAD+t,he (&')-configuration (2). The anomalousD x"HO C02Hneutron-scathering(2)amplitude of6Li and the differing neutron-scattering amplitudes of H and D gave Bijvoetinequalities from which the absolute configuration was deduced.Propionyl-coenzyme A Carboxylase, Methylmalonyl-coenzyme A Epi-merase, Succinyl-coenzyme A Mutase.-This enzymic sequence mediatesinterconversion of propionate and succinate in mammalian tissue and in someba~teria.~ Propionyl-coenzyme A (or CoA) carboxylase, which has beenobtained crystalline from pig heart, converts propionyl-CoA, in the presenceof biotin, ATP, and bicarbonate, into methylmalonyl-CoA. Methylmalonyl-CoA is rearranged by a vitamin B ,,-coenzyme-dependent isomerase, methyl-malonyl-CoA mutase, to succinyl-CoA.The carboxylase is stereospecificand the methylmalonyl-CoA which it produces is the enantiomer (or strictlyspeaking the epimer, since coenzyme A is itself asymmetric) of the formutilized by the mutase. A third enzyme, methylmalonyl-CoA epimerase,must therefore intervene to interconvert the two molecular species, spon-taneous interconversion being quite slow. The stereochemistry of all stageshas now been investigated and the conclusions are summarized as shownThe stereochemistry of the carboxylase product was demonstratedindependently by two groups of workers. Both proofs depended on reduc-tion of the product (4) by Raney nickel to a partially racemized but stilloptically active 3-hydroxy-2-methylpropionic acid (7 ; R = H).converted this into the dextrorotatory phenylurethane (7 ; R = COeNHPh),which was synthesized from (2S,3S)-2-amino-3-methylsuccinic acid (8) ofknown absolute configuration.The other group used enzymic carboxyla-tion of 2,2-dideuteriopropionyl-CoA with KH14C03 to obtain a 14C-labelledproduct (4: H* = D). Racemization on reduction of this deuteriated(3-6).One groupC. K. Johnson, E. J. Gabe, M. R. TayIor, and I. A. Rose, J. Amer. Cheni. SOC.,Y. Kaziro and X. Ochoa, Adv. Enzymol., 1964, 26, 283 (Review).M. Sprecher, M. J. Clark, and D. €3. Sprinson, Biochem. Biophgs. Res. Comm.,J. RBtey and F. Lynen, Biochern. Z., 1965, 342, 256.1965, 87, 1802.1961, 15, 581430 B I 0 LO G I C AL C H E MI S TRYH MeH Me Degradn. H O 2 C P > i l H( 6 stages) H NH,product with nickel was slight ; the 3-hydroxy-2-methylpropionic acid wasconverted into the hydrazide (9; H* = D), portions of which were mixedwith larger amounts of synthetic (B)-, (@, and (RS)-hydrazide, preparedfrom the resolved optical isomers of 3-hydroxy-2-methylpropionic acid.The radioactivity due to the enzymically prepared materia,l remained withthe (R)- and (RS)-hydrazide on recrystallization, but was largely lost fromthe (8)-hydrazide; the material of enzymic origin was therefore the (R)-hydrazide (9).The absolute configurations of the synthetic material hadbeen assigned by st correlation through the oxazolidone (10) with (8)-alanine(ll), from which the enantiomorph of (10) was prepared by reduction to(8)-alaninol and ring-closure with phosgene.Both investigations led to thesame conclusion: that the product of enzymic carboxylation is (&)-methyl-malonyl-CoA (4). The method of deducing stereochemistry from the separa-bility or inseparability of radioactivity from synthetic enantiomorphs isadvantageous since the enzymic experiments can be on the usual scale of afew micromoles.Me H k' HZN COlHLiAIHT - HMe 'OH(13)Rktey and Lynen6 also determined the absolute configuration of thehydrogen displaced on carboxylation. (2R,3R)-2,3-Epoxybutane (12) wasreduced by lithium aluminium tritide to a butanol(l3; H* = T) oxidized byhypobromite to (5)-2-tritiopropionic acid, the coenzyme A derivative ofwhich (3; H* = T) was used for carboxylation on the enzyme.The productafter conversion (with suitable dilution) into the oxazolidone (9; H* = T)retained 80% of the tritium originally present. It follows that, in theenzymic carboxylation, the carboxyl group in the product occupies thCORNFORTH AND RYBACK : STEREOSPECIFICITY O F REACTIONS 431spatial position of the displaced hydrogen, provided that the retention oftritium is not an isotope effect operating on a non-stereospecific process.This latter possibility was dismissed by examining (R)-2-tritiopropionyl-CoA,which lost all its tritium on carboxylation.7The absolute configuration (5) of the methylmalonyl-CoA utilized by themutase follows from that of (4). The mechanism of action of the epimerasehas been established4 as an exchange of the tertiary hydrogen with protonsfrom the medium (and not, for example, a shift of the SCoA group or adecarboxylation-carboxylation) ; thus, a plausible intermediate is the planarenolate-ion (14).It is still unknown whether the initial rate of racemizationis equal to the initial rate of hydrogen exchange : in other words, whether theintermediate (14) can add a proton from the medium indifferently on eitherside, or whether its choice is limited, by attachment to the enzyme, to accept-ing a new proton with inversion of configuration and exchange, or re-attach-ing the original proton with retention and no exchange. Since the hydrogenexchange can be measured in deuterium oxide by n.m.r.,8 and the rotationalchange should be measurable in the same medium by a modern polarimeter,the experiment appears possible.The mutase which interconverts (R)-methylmalonyl-CoA and succinyl-CoA has been shown to induce an intramolecular 1 : 2-shift of the CO*S*CoAgroup accompanied by a 1 : 2-shift of hydrogen in the opposite direction.*When (R)-methylmalonyl-CoA is available and epimerase is absent, theMeHOZC \v A SCoA0- (14)succinyl-CoA contains no stably bound hydrogen from the medium; but ifpropionyl-CoA is converted into succinyl-CoA by the sequence of three en-zymes in deuterium oxide, the (22)-methylmalonyl-CoA contains deuterium( 5 ; Ht = D) and this persists in the succinyl-CoA.This experiment hasbeen done,5 the enzyme preparation consisting of an acetone powder frombeef liver mitochondria; the succinic acid derived from hydrolysis of thethiolester was dextrorotatory and hence had the (8)-configuration(6; HT = D), since (R)-rnonodeuteriosuccinic acid is laevor~tatory.~ Dis-crepancies between the optical rotatory power and the deuterium content ofthe succinic acid suggested that other enzymes had also been at work.Succinic Dehydr0genase.-Conversion of succinate to fumarate by pre-parations of this enzyme in the presence of electron acceptors is a trans-elimination of hydrogemlo However, anmobic incubation of succinatewith the enzyme in the presence (necessary for a useful rate of reaction) offumarate leads to exchange of succinate hydrogen with the aqueous medium.D.Arigoni, F. Lynen, and J. RBtey, Helv. Chim. Acta, 1966, 49, 311.P.Overath, G. M. Kellerman, F. Lynen, H. P. Fritz, and H. J. Keller, Biochem. Z.,J . W. Cornforth, G. Ryback, G. Popjak, C. D o d g e r , and G. Schroepfer,1962, 335, 500.Bwchern. Biophys. Res. Comm., 1962, 9, 371.lo T. T. Tchen and H. van Milligan, J . Amer. Chem. SOC., 1960, 82, 4115432 BIOLOGICAL CHEMISTRYWith the artificial substrate (8)-chlorosuccinate (15), Gawron et uZ.11 foundthat, in a deuterium oxide medium, one hydrogen of the substrate wasexchanged preferentially ; with succinate, two hydrogens could be exchangedon prolonged incubation. The succinate isolated after shorter incubationtimes was, however, largely monodeuteriated,12 and this species was shownby its optical rotatory dispersion to be (S)-monodeuteriosuccinate.Thisenantiomorph was also shown to be formed by hydrodechlorination of thedeuteriated (8)-chlorosuccinate, which must therefore have been (28)-chloro-(3R)-[3-D]succinate (15 ; H* = D). trans-Elimination of hydrogenfrom succinate to give fumarate necessarily removes one (R) and one ( S )hydrogen: it appears that the (8) hydrogen is the first to be labilized. (Wefollow here a suggestion of Prof. V. Prelog that an a substituent in a moleculeCaabd may be designated as (22) [or (S)] if promotion of this group to ahigher rank than the other a group confers (R) [or ( S ) ] asymmetry on theC atom.) The use of negative-ion mass spectrometry in the measurementsof deuteriosuccinic acid is noteworthy.Aspartwe Analogues.-The known trans-geometry of the addition ofammonia to fumarate a,nd ifs elimination from aspartate on the enzymeaspartase was utilized to study the stereospecificities of argininosuccinase 13and of adenylosu~cinase.~~ Both these fumarate-yielding eliminations arealso trans.From analysis of the proton coupling constants observed withthe N-acetyl anhydrides of the diastereoisomeric 3-methylaspartates, andcomparison of these with similar measurements on the O-acetyl anhydride of(2S,3R)-[3-D] malate it was concluded l5 that the methylaspartate active inthe methylaspartate ammonia-lyase reaction has the (28,3X) configurationand hence that the addition and elimination catalysed by this enzyme aretrans.This assignment supports an experiment on the stereochemistry of theglutamate mutase reaction, a process requiring vitamin B,, coenzyme ; for(ZX,3S)-3-methylaspartate is also a substrate for this enzyme. Sprecherand Sprinson 16 incubated mesaconate in deuterium oxide with a cell-freeextract of Clostridium tehnomorphum and isolated a, 4-deuterioglutamic acid(17) (18)(17).Chemical oxidation of this gave (R)-monodeuteriosuccinic acid (18).Since the glutamate (17) presumably was formed on the mutase from(ZS,3X)-[3-D]3-methylaspartate (16), the conclusion was that inversion ofl1 0. Gawron, A. J. Glaid, J. Francisco, and T. P. Fondy, Nature, 1963, 197, 1270.l2 0. Gawron, A. J. Glaid, and J. Francisco, Biochem. Biophys. Res. Comm., 1964,l3 H . D. Hoberman, E. A. Havir, 0. Rochovansky, and S. Ratner, J. Biol. Chena.,l4 R.W. Miller and J. M. Buchanan, J. Biol. Chem., 1962, 237, 491.l5 H. J. Bright, R. E. Lundin, and L. L. Ingraham, Biochemistry, 1964, 3, 1224.l6 3%. Sprecher and D. B. Sprinson, Ann. New York Acad. Sci., 1964, 112, 655.16, 156.1964, 239, 3818CORNFORTH AND RYBACK : STEREOSPECIFICITY O F REACTIONS 433configuration occurs a t the carbon atom on which a glycine residue is re-placed by hydrogen.Citrate Condensing Enzyme, Aconitase, 5-Dehydroquinate Dehydratase.-Hanson and Rose17 incubated 5-dehydroshikimate (19) and NADH in tri-tiated water with an extract of Aerobacter uerogenes, obtaining a 6-mono-tritiated quinic acid (20). From the known absolute configuration of quinicacid, the tritiated citric acid (21) obtained from this specimen by oxidationbad the (322) configuration shown.The tritium atom in (21) proved to bethat removed as water by the enzyme aconitase. This confirmed the earlierassignments,2 based on Hudson's lactone rule, that the (A)-methoxycarbonylgroup of citrate is that derived from oxaloacetate, and that the eliminationsand additions of water catalysed by aconitase are trans. Further, theaddition of water to 5-dehydroshikimate is, uniquely, cis. This was ex-plained by supposing that addition of hydroxyl ion to the ap-unsaturatedketonic system occurs as a first stage, followed by protonation of the resultingenol.Isocitric Dehydrogenase.-This universal enzyme mediates the conversionof (2&38)-isocitrate (22) into 2-oxoglutarate (23) and carbon dioxide.Nicotinamide-adenine dinucleotide phosphate (NADPS) is the coenzyme,and a proton from the medium appears in the product.The stereochemistryof this change was demonstrated independently by two groups. Englardand Listowsky ran the enzymic reaction in deuterium oxide, and oxidizedthe product (23; H* = D, Hi = H) to (8)-monodeuteriosuccinic acid(24; H* = D, Ht = H). Lienhard and Rose 19 established that Z-oxoglu-tarate in contact with the enzyme and the reduced form of the nucleotideE n ~.,H:o_I)_.f--Enz.Ht HaT HH 02CHO Hl7 K. R. Hanson and I. A. Rose, Proc. Nut. Acad. Sci. U.S.A., 1963, 50, 981.l8 S. Englard and I. Listowsky, Bwchem. Biophys. Res. C O ~ ~ L ~ . , 1963, 12, 356.Is G. E. Lienhard and I. A. Rose, Biochemi8try, 1964, 3, 185434 BIOLOGICAL CHEMISTRYexchanged with the medium only the hydrogen which is introduced ondecarboxylation of isocitrate.This enzymic exchange carried out with2-oxoglutarate randomly labelled with tritium in the 3-position theIefore gavezb 2-oxoglutarate (23; H* = H; Ht = T) labelled at the non-exchangeablehydrogen. This 2-oxoglutarate with glutamic dehydrogenase gave glu-tamate (25) converted by chemical degradion into (8)-aspartate (26) and(&)-malate (27). The tritium in these specimens was almost entirely retainedon treatment with aspartase and fumarase; since these enzymic reactions areknown to be trans-addition-eliminations, the stereochemistry of the tritiummust have been as shown. It is interesting that here, as with propionyl-CoAcarboxylase, the introduced hydrogen occupies the steric position of thedisplaced carboxyl.Isocitrate Lywe.-This enzyme reversibly changes (2R,3X)-isocitrate(28) into succinate and glyoxylate.In a detsrium oxide medium, unlabelledisocitrate and a partially purified enzyme from yeast gave succinic acidcontaining 1.4 atoms of deuterium per molecule.20 The optical rotatoryH COlHEnz., D 2 0+ , COlHC02H - H OZC *C H 2 E n z . HOIC.CH2 C02H CH 0H "OHdispersion of this specimen showed dextrorotrttions m. 1.4 times the corre-sponding lzvorotations of (R)-monodeuteriosuccinic acid. The presump-tion was that some initially formed (8)-deuteriosuccinate, (29) because of thereversibility of the reaction and the symmetry of the succinate ion, wentthrough another cycle to emerge as (88)dideuteriosuccinate.It is interest-ing that here the introduced hydrogen atom does not occupy the spatialposition of the displaced group but produces an apparent inversion ofconfiguration. This conclusion is supported by other experiments 20@ withtritiated substrates.Enzymes of Squalene Biosynthesk-The enzymes converting six mole-cules of mevalonate (31) into one molecule of squalene (35) have been iden-tified and several have been extensively purified; the usual sources are mam-malian liver and yeast. The course of the biosynthesis (Scheme 1) was dis-cussed by Bloch 21 in a 1964 Nobel lecture, and Clayton 22 has recently reviewedthe subject. The three methylene groups of mevalonate all change theirbonding during biosynthesis ; the stereochemical consequences of labellingthese groups asymmetrically with hydrogen isotopes have been ~tudied.~3Stereospecific chemical synthesis (Scheme 2) was used to prepare meva-2o M.Sprecher, R. Berger, and D. B. Sprinson, J . BioE. Chem., 1964, 239, 4268.zoaK. R. Hanson, Fed. Proc., 1965, 24, 229.2 1 K. Bloch, Science, 1965, 150, 19.22 R. B. Clayton, Quart. Rev., 1965, 19, 168, 201.23 G. Popjik, Biochem. J . , 1965, 96, 1P; J. W. Cornforth, ibid., p. 3P; J. W. Corn-forth, R. H. Cornforth, C. Donninger, G. Popjbk, C. Ryback, and G. J . Schroepfer,Proc. Roy. SOC., 1966, B , 163, 436; C. Donninger and G. Popjhk, ibid., p. 465; J. W.Cornforth, R. H. Cornforth, C. Donninger, and G. Popj&, ibid., p. 492CORNBORTH AND RYBACK : STEREOSPECIFICITY OF REACTIONS 435M e ?H Enz.Me OH H * HH02C - H02C & CH 0OH (30) S C O K H I / H*"Ht (31)RH' H H+ HEnz. ~OPPH"OPPHf"Hf(32) (3 3 ) dOPP + /J + ,* H* tiOPP OPPH+ H" 'H4 H* H*(33) (32) (32)v f "'*634)OPP(35)SCHEME 1.R = ADP-ribose; R' = ADP-ribose phosphate.lonate labelled with deuterium or tritium a t each of the two 4-hydrogenatoms. The lactone (36) and the trans-hydroxy-acid (41) were convertedinto the geometrically isomeric benzhydrylamides (37) and (42), each ofwhich was epoxidized by perbenzoic acid. The epoxides (38) and (43) werereduced by lithium borodeuteride or lithium borotritide to mevalonylbenzhydrylamides (39 and 44; Htt = D or T). The mevalonic acids (40and 45; Htt = D or T) obtained by hydrolysing these amides were accom-panied by their enantiomorphs, but since (3X)-mevalonate is not utilized fo436(36)BIOLOGICAL CHEMISTRYCHzOH( 3 7)(4 4)SCHEME 2.R = GO*NH.CHPh2; Reagents: 1, (C,HllN),C; 2, Ph,CHATH2;3, PhC0,H; 4, LiBH,tt; 5, NaOH.polyisoprenoid biosynthesis the enzymically active forms had the absoluteconfigurations shown.Each molecule of mevalonate participating in squalene biosynthesis losesa hydrogen atom from C-4, either when isopentenyl pyrophosphate (32) isisomerized to dimethylallyl pyrophosphate (33) or when coupling of C, unitsoccurs.The two deuterio-mevalonates (40 and 45; Htt = D) were usedseparately as substrates for the synthesis of farnesyl pyrophosphate (34) by apartially purified enzyme preparation from rat liver.The mass spectrum offarnesol obtained from this product by the action of alkaline phosphataseshowed the presence of three atoms of deuterium when (4R)-[4-D]rnevalonate(45; Hft = D) was the precursor, and of none when (48)-[4-D]mevalonate(40 ; Htt = D) was used. Similarly, squalene from (4R)-[4-T,2-14C]mevalo-nate (45; Htf = T) had the same 14C: T ratio as the mevalonate, whereassqualene from the (48)-mevalonate (40 ; Hit = T) was tritium-free.Asymmetric labelling of one hydrogen atom on C-5 of mevalonate waseffected by reduction of mevaldate (30) with stereospecifically (4R) tritiatedNADH in the presence of mevaldic reductase, for when the product mevalo-nate was converted enzymically via farnesyl pyrophosphate into farnesol,enzymic oxidation of this alcohol to farnesal by NADf and liver alcoholdehydrogenase caused loss of all tritium from C-1.This carbon (and there-fore C-5 of the parent mevalonate) thus had the (R)-configuration providedthat oxidation of farnesol on liver alcohol dehydrogenase follows the samestereocheinical course as that established for ethanol.24 The latter assump-24 R, U. Lemieux and J. Howard, Canad. J . Chenz., 1963, 41, 308CORNFORTH AND RYBACK : STEREOSPECIFICITY O F REACTIONS 437tion was supported25 by oxidation on the same enzyme of geraniol stereo-selectively labelled with tritium a t C-1 by Streitwieser's 26 chemical pro-cedure.Asymmetrically deuteriated mevalonate (31; Ht = D) was now pre-pared from mevaldate and (4R)-[4-D]NADH on mevaldic reductase and wasconverted into squalene (35; Ht = D).Lawulinic acid from ozonolysis(Scheme 3) of this squalene on oxidation by hypoiodite gave (R)-mono-Y v v4'''(46) (46)SCHEME 3.Reagents: I, 0,; 2, HC0,H; 3, NaOl.deuteriosuccinic acid (46; Hb = D). It followed that each association of C,units in squalene biosynthesis is accompanied by inversion at tjhe carbonatom from which the pyrophosphate ion is eliminated.Association of two molecules of farnesyl pyrophosphate (34) to onemolecule of squalene is accompanied by displacement of a hydrogen atomfrom one of the carbon atoms forming the new C-C bond and intiroduction inits place of a (48)-hydrogen 27 from NADPH. The steric position occupiedby the introduced hydrogen was determined 28 by using, as the precursor inan enzymic synthesis of squalene, mevalonate fully deuteriated at C-5.Ozonolysis of this squalene gave, from the four central carbon atoms (as),the optically active (8)-trideuteriosuccinic acid (49).The stereochemistryof the introduced hydrogen is therefore that shown in (35) as HS. Thisassignment was confirmed 29 independently by biosynthetic conversion ofsqualene, itself biosynthesized from synthetic [14Clfarnesyl pyrophosphateand (M)-[4-T]NADPH, into cholic acid via cholesterol. The T : 14C ratio25 C. Donninger and G. Ryback, Biochem. J., 1964, 91, 1lP.28 A. Streitwieser, J. R. Wolfe, and W. D. Schaeffer, Tetrahedron, 1959, 6, 338.27 G. Popjhk, G. Schroepfer, and J.W. Cornforth, Biochem. Biophys. Res. Comni.,28 J. W. Cornforth, R. H. Cornforth, C. Donninger, G. PopjQk, G. Ryback, and29 B. Semuelsson and De 1%'. S. Goodman, J . Biol. Cheni., 1963, 239, 98.1961-1962, 6, 438.G. F. Schroepfer, Biochem. Biophys. Res. Comm., 1963, 11, 129438 BIOLOGICAL CHEMISTRYD HD (49)in the cholic acid (50) was halved on conversion into methyl 3%,7cr-diacetoxy-12-oxo-5~-cholanoate (51) and reduced alinost to zero by subsequent enoliza-tion to exchange the C-11 hydrogens. This is the result expected fromcyclization of squalene tritiated as in (35; HO = T) since [lZg-T]- and[ lla-TI-cholesterols would be formed in equal amounts and the 12B-tritium,which should be undisturbed by the biosynthetic 12a-hydroxylation, wouldbe eliminated on oxidation to (51).Both (5R)-[5-D]- and (SR)-[5-T]-mevalonates (31; Kt = D or T) werefound to be converted into squalene without loss of hydrogen isotopes;23hence, the hydrogen eliminated during the enzymic conversion of farnesylpyrophosphate into squalene originated as a (58) hydrogen of mevalonate.Thus, squalene biosynthesized from (5$)-[ 5-Dlmevalonate must have, a tthe central carbon at which exchange of hydrogen has occurred, the ccn-figuration (35) where Ht= D and He = H, the other central carbon carryingone deuterium atom.When this squalene is degraded by ozonolysis, thesecentral carbons appear as the methylene groups of a 2,3-dideuteriosuccinicacid, which is either (88) (if configuration is retained at the carbon whichdoes not exchange hydrogen) or (XR) (if configuration is inverted).In theevent, the succinic acid was the optically inactive meso (SR) form (47 ; Ht = D,HO = H); thus, inversion of configuration occurs here as with the coupling ofThe availability of (4R) -[4-T ; 2-14C] and (45) -[4-T ;2-14C]meva~lonates ofhigh specific activity opened new lines of investigation in polyisoprenoidbiosynthesis. Thus, it was found 30 that rubber biosynthesized from thesemevalonates by Hevea brasiliensis latex incorporated all of the tritium fromthe (48)-mevalonate and none of the tritium from the (4R)-mevalonate.This is the exact opposite of what happens in squalene biosynthesis. Thedifference is presumably due to the &-geometry of the double bonds inrubber : trans, trans-farnesyl pyrophosphate simultaneously formed by thelatex preparations showed the ‘‘ squalene ”, not the “ rubber ,” pattern ofretention and elimination of tritium.The biosynthesis of phytoene and 8-carotene from mevalonate in carrotslices follows the ‘‘ squalene ” ~ a t t e r n , ~ l the (4S)-hydrogen being eliminated.30 B.L. Archer and D. Barnard, Biochem. J., 1965,96,1P; B. L. Archer, D. Barnard,E. G. Cockbain, J. W. Cornforth, R. H. Cornforth, and G. PopjBk, Proc. Roy. SOC.,1966, By 163, 619.3l T. W. Goodwin and R. J. H. Williams, Biochem. J., 1965, 94, 5C.c5 unitsCORNFOETH AND RYBACK : STEREOSPECIFICITY O F REACTIONS 439From the eight molecules of mevalonate incorporated in each molecule ofthe carotenoids, no (4R)-hydrogen is lost when phytoene is formed, but twoof the eight (4R)-hydrogens are absent from #?-carotene, as shown by changesin the T : 14C ratio.Presumably these two are lost when the cyclohexanerings are closed (e.g., 52-453).UST q k '/ ,... rnzr = T q ....(5 2) (5 3)The established intermediate stages 23 in the biosynthesis of cholesterolfrom six molecules of (4R)-[4-T ;2-14C]mevalonate require that, of the six(4R)-hydrogen atoms, five shall persist in lanosterol (54) and only three incholesterol (55) (assuming that elimination and not rearrangement occursfrom C-3 and C-5 of the steroid nucleus). Of the six labelled carbons,all survive in lanosterol and five in cholesterol. Thus, if the T : 14C ratiois 1 : 1 in mevalonate, it should be 5 : 6 in lanosterol and 3 : 5 in choles-terol.This was verified for lanosterol and cholesterol synthesized from(4R)-[4-T ;2-14C]mevalonate by microsomal and soluble enzymes from ratliver.32 Of the three surviving tritium atoms, one was located at C-17 andthe other two in the side-chain (presumably a t C-20 and C-24), in agreementwith theory. The 3 : 5 ratio found for cholesterol was also found for fuco-sterol (56) synthesized by Fucus spiralis from the same mevalonate pre-cursor.33 This implies that the tritium atom which occupied (3-24 has notbeen eliminated but rearranged, wost probably to C-25 as forecast earlier,3*on alkylation of the side-chain.Stearate Dehydrogenation.-The efficient conversion of stearic acid (57)into oleic acid (58) by growing cultures of a strain of Corynebacterium diph-theriae has been shown 35 to involve stereospecific loss of two hydrogen atoms(H* and Ht in formula 57).The product is not seriously contaminated byendogenous stearic or oleic acid or by products of further metabolism.(S)-9-Tritiostearic acid (60) was synthesized from (R)-9-hydroxystearic acids2 J. W. Cornforth, R. H. Cornforth, C. Donninger, G. PopjQk, Y. Shimizu, S. Ichii,E. Forchielli, and E. Caspi, J . Amer. Chem. SOC., 1965, 87, 3224.33 L. J. Goad and T. W. Goodwin, Biochem. J . , 1965, 96, 79P.34M. Castle, a. Blondin, and W. R. Nes, J . Amer. Chem. Soc., 1963, 85, 3306.35 G. F. Schroepfer and K. Bloch, J. Biol. Chem., 1965, 240, 54440 BIOLOGICAL CHEMISTRYby reducing the derivative (61) with lithium aluminium tritide (inversion ofconfiguration a t C-9) and oxidizing the resulting primary alcohol.Theenantiomer of (60) was similarly obtained from (S)-9-hydroxystearic acid .j.;l-.H f H(57)T HRH D H H D - E n z . K%H D-N:N-D - R%R, + R%~'(C. d r p h t h e r i a e ) *R' H b D \ H( i ) LiAITrHO HR L R 'HO HR = *[CH,I7*CO,H; R' = .[CHJ7*CH,.(62) derived from (61) by a Walden inversion. Loss of tritium from the(R)-enantiomer of (60), but not from (60) itself, during conversion into oleicacid established the stereochemistry of the reactioii a t C-9. Similar experi-ments based on natural (R)-10-hydroxystearic acid (63;H* = H) led to thesame conclusions regarding C-10 but, since the absolute configuration of (63)was inferred from its having the same sign of rotation as (R)-9-hydroxy-stearic acid, the results were confirmed by incubating the racemic mixture(59), prepared from oleic acid by cis-addition of deuterium, with C.diphtheriaeand comparing the proportions of di- , mono-, and non-deuteriated moleculeswith the ratios expected from the two stereochemical possibilities a t C-10.A considerable isotope effect was observed for the elimination of hydrogenfrom C-9, but not for that from (3-10, suggesting that the removal of the twohydrogen atoms is not synchronous. It is pointed out that, since the mech-anism of the reaction is not known, interpretation of the results as an enzymic&-dehydrogenation need not follow.The stereochemistry of the dehydrogenation performed by C.diphtheriaehas been used 36 to show that, when oleic acid is hydrated by a fermentingculture of a Pseudomonm species to give (R)-10-hydroxystearic acid (63), ahydrogen atom from the medium is introduced at C-9 in a configuration(H* in formula 63) consistent with a trans-addition of water t o oleic acid.Propanedio1dehydrase.-The stereospecificity of the reactions alreadydiscussed is so complete that an enzyme utilizing two enantiomeric substrateswith equal ease and differing stereospecificities comes as something of ashock. Dioldehydrase from Aerobacter aerogenes has been extensivelypurified37 and the most active preparations gave a single active band onstarch-gel electrophoresis. (a)- and (8)-Propanediols are both readily con-verted by this enzyme into propionaldehyde; no hydrogen from the medium36 G.F. Schroepfer, J. Amer. Chm. Soc., 1965, 87, 1411.3 7 H. A. Lee and R. H. Abeles, J . Biol. Chem., 1963, 238, 2367CORNFORTH AXD RYBACK: STEREOSPECIFICITY OF REACTIONS 441is incorporated during the dehydration which must therefore proceed bymigration of hydrogen (or conceivably of a methyl group). Vitamin B,,coenzyme is a necessary co-factor.Frey, Karabatsos, and Abeles 38 reduced (8)- and (8)-lactaldehydes (64)and (68) with liver alcohol dehydrogenase and deuteriated NADH. Theexpected stereospecificity of deuterium introduction was checked by con-verting the product propanediols into their 4-nitrobenzylidene acetals (67)and (71), which gave n.m.r.spectra different from each other and from thatof a third specimen, the acetal of propanediol prepared by non-enzymic(LiAlD,) reduction of (38)-lacta,ldehyde. It follows that both enzymicHO Ye CHOMe H A HO CHOR = ADP-ribose.reductions were stereospecific (otherwise all three spectra would have beenidentical) and that the stereospecificity of insertion of deuterium was thesame for both substrates (otherwise the first two acetals would have beenenantiomorphs with identical spectra). If the known stereospecificity ofliver alcohol dehydrogenase for acetaldehyde applies also to lsctaldehydes,the products from (R)- and (8)-lactaldehyde were respectively (65) and (69).Each propanediol was converted on the dehydrase into propionaldehyde.Examination of the n.m.r.spectra of the 2,4-dinitrophenylhydrazonesshowed that the (R)-propanediol gave 2-deuteriopropanal (66) and the(8)-propanediol gave 1 -deuteriopropanal (70). A large kinetic isotope effectwas noted only with the (R)-propanediol, with which apparent migrationof deuterium occurs.(8)-[ 2-DILacticmid (72), prepared from pyruvic acid and deuteriated NADH on lacticThis work has lately been confirmed and extended.3958 P. A. Frey, G. L. Karabatsos, and R. H. Abeles, Biochm. Biophgs. Res. Comm.,1965, 18, 551.(The legend for curve B in Figure 1 of this Paper refers to curve C andcice versa.)30 J. R&y, A. Umani-Ronchi, and D. Arigoni, Experientia, 1966, 22, 72.442 BIOLOGICAL CHEMISTRYdehydrogena.se, was reduced to (S)-[Z-D]propane-l,2-diol (73).A 1,l-dideuteriated (R)-propanediol (74) was also prepared by reduction of (R)-H D kMe M e D x Me (i) M e t h y l a t e A ( i i ) L i A I i - i i = HO CH2OH HO CDlOH M e C02H HO CO2H(72) (73) (74) (75)lactic acid with lithium aluminium deuteride. Enzymic dehydration ofboth these specimens and oxidation of the resulting propionaldehydes gavein each case the dextrorotatory (28)-[%D]propionic acid (75) identified bycomparison of its optical rotatory dispersion with that of an authenticspecimen.6These results exclude methyl migration and show that the hydrogenmigration occurs with inversion of configuration a t C-2. It was pointedout 389 39 that the attachment of both (R)- and (S)-propane-l,Z-diol to theenzyme can have the same stereospecificity with respect to the methylgroup, the 2-hydroxyl group, and the hydrogen atom of the l-hydroxylgroup.When the (R)- (76) and S-diols (77) are arranged thus, the orientation..HCof the migrating hydrogen (H*) can in each case be anti to that of the depart-ing hydroxyl, as the inversion of configuration should require. Stereo-specificity of binding to the enzyme a,nd a favoured steric relation of thereacting groups are thus preserved (but differently) in each case. This expla-nation assumes (with experimental support 37) that only one enzyme isconcerned.Stereochemical Exploration of Active Sites.-The specificity of an enzymefor a variety of " unnatural " substrates may serve as a means of exploringthe immediate environment of the active centre; for, in general, substratespecificity and even product stereospecificity could be controlled by ob-struction of unfavourable structures or conformations, as well as by selec-tive binding of favourable ones.For example, glutamine synthetase 4O(from sheep's brain) converts both (R)- and (#)-glutamate into glutamines,but of the " unnatural " substrates (R)- and (S)-2-methylglutamate onlythe (&')-enantiomorph (78) was amidated. 3-Aminoglutarate is also a sub-.-4O H. M. Kagan, L. R. Manning, and A. Meister, Biochemistry, 1965, 4, 1063CORNFORTH AND RYBACK: STEREOSPECIFICITY OF REACTIONS 443strate for this enzyme and is converted, stereospecifically, into (3R)-3-aminoglutaramate (79) .41 These results were explained by supposing thatthe enzyme (since it does not amidate aspartate) imposes a fully extendedconformation on the glutaric acid chain in order that the two carboxylgroups shall be bound to the enzyme.Observing this condition, it i s possibleto arrange models of (Qglutamine, (R)-glutamine, and (3R)-3-aminoglu-taramic acid so that the carboxyl groups, the aminocarbonyl groups, and theamino-groups occupy essentially the same positions in space, but this is notpossible with (3S)-3-aminoglutaramic acid. In these models it could alsobe seen how an a-methyl group might prevent binding to the enzyme sur-face in the (R)- but not in the (8)- glutamic acid series.Oxido-reductases dependent on nicotinamide nucleotides have receivedspecial attention. Prelog 42 and his collaborators have studied the ratesand products of reduction of a large number of alicyclic ketones on horseliver alcohol dehydrogenase and on an extensively purified oxido-reductasefrom Curvularia falcatu, and have developed an ingenious system for sosynthesizing the results as to obtain a picture of the space available to sub-strates a t the active centre of the enzyme.Briefly, the configuration of theproduct alcohol is taken as more representative of the transition state thanthat of the substrate ketone; therefore, the hydroxyl group, the hydrogenatom introduced from the riucleotide, and the carbon atom to which they areboth attached are all given a fixed orientation. Molecular models of allproducts from substrates reducible by the enzyme are oriented in accordancewith those requirements, and a diamond-lattice skeleton is constructedwhich will just accommodate all their carbon atoms at lattice points.Asmore substrates are studied it may be possible to extend the lattice, and thestudy of non-reducible substrates which protrude from the lattice can beused to define regions where steric inhibition from enzyme or co-enzyme issevere. The limited number of conformations available to the alicyclicsubstrates limits also the alternative arrangements which have to be takeninto account. The lattice proposed for the Curvularia enzyme is shown inthe Figure.Laitice section (heavy lines) of oxido-reductwe f r m Curvularira falcata;42= '' forbidden position.' '41 E. Khedouri and A.Meister, J. Bid. Chm., 1965, 240, 3357.4a V. Prelog, Pure Appl. Chem., 1964, 9, 119444 BIOLOGICAL CHEMISTRYA similar approach has been used 43 to assemble the data obtained fromthe rates and products of reduction of cyclohexanones, 2-decalones, and10-methyl-2-decalones on liver alcohol dehydrogenase, and quantitativeinhibition factors were calculated for particular regions : this allowed pre-diction of rate and stereochemistry of reduction for a substrate not previouslyincluded in the lattice. A tentative orientation of the coenzyme in relationto the lattice was proposed. It will be interesting to see how far this approachcan be developed.43 J. M. H. Graves, A. Clark, and H. J. Ringold, Biochemistry, 1965, 4, 26553. THE MICROBIOLOGICAL DEGRADATION OF AROMATICCOMPOUNDSBy D.W. Ribbons(Milstead Laboratory of Chemkal Enzymology,“Shell ” Research Ltd., Sittingbourne, Kent.)IT is nine years since attention was focused on the microbiological oxidationof aromatic compounds in these Reports.1 Numerous major contributionshave considerably modified the general patterns of metabolism that hadbeen established; they have extended our knowledge of the mechmisms ofoxygenase reactions and clarified the ways by which the syntheses of theenzymes catalysing these catabolic reactions are controlled.It iscleu that catechol, protocatechuate, gentisate, and homogentisate are notthe only substances that are susceptible to ring fission; and, further, thatthe oxygenative cleavage of the former two compounds does not alwaysyield a muconic acid. Simple benzenoid compounds now known to yieldaliphatic compounds by oxygenase reactions also include 3-methylcate-~ h o l , ~ - ~ 4-methylcatechol,3-5 2,3-dihydro~yphenylpropionate,~~ homopro-tocatechuate,7~ quinol, hydroxyquinol, 9 and 2,3 -dihydroxybenzoate.l OOxygenations of catechol and protocatechuate are also catalysed by catechol2,3-oxygenase 2, 1 1 3 12 and protocatechuate 4,5-oxygenase 2,133 l4 to yieldsubstituted muconic semialdehydes.Several other enzymic reactions ofthis type are known to rupture the rings of condensed polynuclear aromatichydrocarbons and also those of some heterocyclic compounds, includingtryptophan and its metabolites.The purification and crystallization of some oxygenases has been mainlyresponsible for the progress of studies on the mechanism of their action.15The use of newer techniques of mass spectrometry and electron spin reson-ance spectroscopy has also largely contributed to success in this area.The analysis of the nature of the sequential induction and also repressionof the enzymes involved in some well-established pathways of metabolismW.C. Evans, Ann. Reports, 1956, 53, 279.S. Dagley, W. C. Evans, and D. W. Ribbons, Nature, 1960, 188, 560.H. Nakagawa, H. Inoue, and Y. Takeda, J. Biockm. (Japan), 1963, 54, 65.D. W. Ribbons, Proc. Internat. Congr. Biochem., New York, 1964.S. Dagley, P. J. Chapman, D. T. Gibson, and J. M. Wood, Nature, 1964,202,775.S. Dagley, P. J. Chapman, and D. T.Gibson, Biochim. Biophys. Acta, 1963,’ H. Eta, M. Kamimoto, S. Senoh, T. Admhi, and Y. Takeda, Biochem. Biophys.P. J. Chapman and S. Dagley, Biochem. J., 1960, 75, 6P.@P. Larway and W. C. Evans, Biochem. J., 1965, 95, 52P.lo D. W. Ribbons, J. @en. Microbiol., 1966, in the press: Biochem. J., 1966, inl1 S. Dagley and D. A. Stopher, Biochem. J., 1959, 73, 16P.l2 Y. Kojima, N. Itads, and 0. Hayaishi, J . Biol. Chem., 1961, 236, 2223.Is S. Trippett, S. Dagley, and D. A. Stopher, Biochem. J., 1960, 76, 20P.14D. W. Ribbons and W. C. Evans, Biochem. J . , 1962, 83, 482.l6 0. Hayaishi, Plenary Sessions 6th Internat. Congr. Biochem., New York City,Many alternative routes of metabolism have been discovered?78, 781.Res. Comm., 1965, 18, 66.the press.1964, I.U.B., 33, 31446 BIOLOGICAL CHEMISTRYhas been facilitated by more detailed kinetic studies, by the use of mutantsthat have deletions in the catabolic pathway, and by the use of gratuitous(non - meta boliz a ble ) inducers .I6 9 17The three main themes of this Report are: (a) alternative and generalpathways of metabolism; (b) oxygenases; and (c) regulation of the metabolicpathways involved.Pathways of Metabolism.-The methods by which benzenoid compoundsare dissimilated fall into two distinct classes: (a) cleavage of a catechol be-tween carbon atoms bearing hydroxyl groups, generally yielding 3-0Xoadi-pate, e.g., as shown in Scheme 1; and (b) cleavage of the ring between ahydroxylated carbon and a non-hydroxyhted carbon atom, generallyyielding pyruvate, e.g., as for gentisate oxidation or as shown in Scheme 2.Formation of 3-oxoadipate.In 1956, it was shown by Gross, Gafford,and Tatum l8 that the route of protocatechuate dissimilation to 3-oxoadipate(6) by extracts of Neurosporca crama was distinct from that utilized byPseudomonas putida A3. 12.19 (-)-p-Carboxymuconolactone (y-carboxyme-thyl-p- carboxy-Aa-butenolide) (1 1) is accumulated by purified extractsof N . crassn incubated with cis,cis-/?-carboxymuconate (9). This lactone isnot an intermediate for bacterial pathways. Neurospora and Pseudomonaspathways of protocatechuate degradation also Mered in the distributionof carbon atoms from protocatechuate into 3-oxoadipate (6). The C-3 of3-oxoadipate was derived exclusively from C-6 of protocatechuate inNeurospora but formed randomly from C-1 and C-6 of protocatechuate inPs.putidu A3.12. The possible symmetrical intermediates, &4ihydroxy-adipate and its di-y-lactone,20 were unable to act as precursors of 3-oxoadipatein the bacterial pathway, although the dilactone was non-enzymicallyisomerized a t pH values greater than 6.0 to (A)-muconolactone (a), and oneisomer of this was metabolized to 3-oxoadipate by Moraxella Iwofii Vibrio(O/l), Nocnrdia erythropolis, and certain pseudomonads but not by Ps.putida.20 ( + )-Muconolactone was isolated as a product of /I-carboxy-muconate metabolism by heat-treated extracts of M . Iwofii 21 and meta-bolized by untreated extracts to 3-oxoadipate. These results indicated amultiplicity of routes by which 3-oxoadipate could be formed from /I-carboxy-muconate : (i) via; p-carboxymuconolactone, (ii) via a symmetrical inter-mediate, and (iii) via a muconolactone.Scheme 1shows the pathway of /?-carboxymuconate metabolism to 3-oxoadipatethrough y-carboxymuconolactone (10) and this route may be common to all18 R.Y. Stanier, G. D. Hegeman, and L. N. Omston, “ Regulation chez les micro-orgtmisms,” Coll. Int. C.N.R.S., Marseilles, 1963, 227.17 J. Mandelstam, “ Regulation chez les micro-organisms,” Coll. Int. C.N.R.S.,Marseilles, 1963, 221.18 S. R. Gross, R. D. Gafford, and E. L. Tatum, J . Bid. Chern., 1956, 219, 781;E. L. Tatum and S. R. Gross, ibid., p. 797; S. R. Gross, ibid., 1959, 233, 1146.l9 D. L. MacDonald, R.Y. Stanier, and J. L. Ingraham, J . Bid. Chern., 1954, 210,809.8O €3. B. Cain, D. W. Ribbons, and W. C. Evans, Bkchem. J., 1961, 79, 312; S. R.Elsden and J. L. Peel, Ann. Rev. Microbwl., 1958, 12, 145.81 R. B. C&, Biochem. J., 1961, 79, 298.38 L. N. Ornston and R. Y. Stanier, Nature, 1964, 204, 1279: J . Biol. Chern., 1966,in the press.Ornston and Stanier 22 have now provided a clearer pictureRIBBONS : MICROBIOLOGICAL DEGRADAI'ION 4-47h (4) ccO4Non-enzymic IABacteria IA I \(7) (9) (1 '1Fungalbacteria, including Ps. M . Iwofii and putidcc. Furthermore, they haveelegantly elucidated the reactions leading to the formation of 3-oxoadipatefrom catechol which were not, until recently, entirely clear. The metabolismof catechol through &,cis-muconate (3) and ( +)-y-carboxymethy1-Aa-butenolide (4) was established in 1951,23 and the lactonizing and delactoniz-ing enzymes were separated.24 The delactonization of ( + )-muconolactone(4) was presumed to yield the enol form of 3-oxoadipate, possibly via, anenol-lactone (5).239 24 Formal proof of the participation of a second lactonehas now been provided.22 The enol-lactone, y-carboxymethyl- AD- buteno-lide (5) was isolated 22 as a product of cis,cis-/3-carboxymuconate (9) meta-bolism, using purified extracts of Ps.putidiz A3.12. Mutants (ELH-mutants) of this strain, which no longer contain the ferminal enzyme of thereaction sequences of Scheme 1 cannot form 3-oxoadipate from catechol.Extracts of benzoate-induced ELH- mutants do not accumulate 3-oxoadi-pate enol-lactone ( 5 ) when incubated with catechol or &,cis-muconate (3) ;instead (+ )-muconolactone (4) appears, suggesting that the equilibrium ofthe muconolactone-isomerase-catalysed reaction is almost completely infavour of (+)-muconolactone.These same extracfs also catalyse the con-version of 3-oxoadipate enol-lactone into a mixture of ( + )-muconolactoneand cis,&-muconate. The inability of ELH- mutants to grow on benzoateor p-hydroxybenzoate had suggested that the routes of 3-oxoadipate forma-tion from catechol or profocatechuate were convergent a t some point before3-oxoadipate. The common intermediate is the enol-lactone of 3-oxoadipate.W. C. Evans, B. S . W. Smith, R. P. Linstead, and J.A. Elvidge, Nature, 1951,168, 772.24 W. R. Sistrom and R. Y . Stanier, J . BioE. Chem., 1954, 210, 821448 BIOLOGICAL CHEMISTRYThe y-carboxymuconolactone proved to be too unstable to isolate asa product of 8-carboxymuconate metabolism. Its structure is deduced bythe ease with which the enol-lactone is formed non-enzymically by de-carboxylation, and also by its ultraviolet spectrum, which has a maximuma t 230 mp, somewhat higher than expected.It now seems that the isolation of (+)-muconolactone as a product ofprotocatechuate metabolism by extracts of M . Iwofii 21 and its non-enzymicformation from y-carboxymuconolactone 22 (10) was due to the activityof muconolactone isomerase which would cstalyse its formation from theenol-lactone.22 This enzyme is undoubtedly present in extracts of manybacteria induced to protocatechuate since muconolactone can act as aprecursor of 3-0xoadipate;~O furthermore, the enol-lactone hydrolase isparticularly thermolabile.Ps. putida A3.12 appears to be the exceptionsince extracts from uninduced cells of this strain do not catalyse the con-version of cis,&-muconate or muconolactone t o 3-oxoadipate.l8-20 Never-theless, the randomization of C-1 and C-6 of protocatechuate into C-3 andC-4 of 3-oxoadipate has been explained by the ready isomerization of theenol-lactone to muconolactone, and equilibration of this with cis,&-inuconate,22 which is the only symmetrical aliphatic compound in Scheme 1 ;i.e., the particular extracts used for the isotope experiments 18 may havebeen able to effect this equilibration.In N .crassa the formation of 3-oxoadipate from 8-carboxymuconolactone(11) is presumed to proceed via the enol-lactone ( 5 ) but this has not beendemonstrated. An intermediate between #I-carboxymuconolactone and theenol-lactone of 3-oxoadipate may be a carboxyenol-bctone since concomitantdecarboxylation during the formation of the enol-lactone ( 5 ) from 8-car-boxymuconolactone (11) is not a necessity as it is in the bacterial conversionof y-carboxymuconolactone (10) into the enol-lactone (5) of 3-oxoadipate.3-Oxoadipate is also a product of hydroquinone metabolism by bacteria.9Extracts of a soil pseudomonad were able to cleave the nucleus of quinol andtransform the product, y-hydroxymuconic semialdehyde, to 3-oxoadipate.Pormation of pyruvute. Bacteria: and possibly other micro-organisms,25are able to form aliphatic compounds from catechol and protocatechuateby a second mode of ring fission. Thus, the enzymes catechol 2,3-osy-genase 2,119 12926 and protocatechuate 4,5-oxygenase,29 14, 2 7 open the nucleusof their substrates to form substituted muconic semialdehydes.Pyruvatewas shown to arise from these intermediates. This route of degradation isgenerally easily recognized by two characteristics : (a) the initial products ofring fission, the muconic semialdehydes, are bright yellow a t neutral andalkaline pH values, and ( b ) these products react non-enzymically withammonium ions to form pyridine carboxylic acids (Scheme 2, Table 1).Theformer property has been used as a spot test for the presence of catechol2,3-oxygenase 28 and protocatechuate 4,5-oxygenase 2 9 in whole cells.2 5 R. F. Bilton and R. B. Cain, J . Gen. Microbiol., 1966.26 S. Dagley and D. T. Gibson, J . BioE. Chern., 1964, 239, PC 1284; S. Dagley and27 S. Dagley and M. D. Patel, Biochem. J., 1957, 66, 227.28 E. S. Pankhurst, J . Appl. Bmteriol., 1965, 28, 309.* @ N. J. Palleroni and R. Y. Stanier, J. Ben. Microbiol., 1964, 35, 319.D. T. Gibson, Biochem. J., 1965, 466RIBBOXS : MICROBIOLOGICAL DEGRADATIOF 449Although the discovery of the meta-fissions is relatively recent, it is clearthat they are widespread and that the enzyme-catalysed reactions leading topyruvate from different ring-cleavage substrates are of a similar type, andmay be presented in a general pattern as seen in Scheme 2, and Table I(p.450). Dagley, Chapman, Gibson, and Wood have established thisR' * COzH R'R'C02HI,co M e (16)general metabolic route for catechol,59 26 3-methylcatechol, 4-methylcate-ch0l,5 protocatechuate,5 2,3-dihydroxyphenylpropionate,5, 6 , 3O and 3,4-dihydroxyphenylacetate.31 It is evident that all these substrates are noteatabolized by the same series of enzymes, although there is evidence thatsome of the enzymes that catdyse these analogous reaction sequences arenot specific. Catechol, 3-methylcatechol, and 4-methylcatechol may forma group that are catabolized by the same series of enzymes. The specificityof catechol 2,3-oxygenase does not, however, extend to protocatechuate,2,3-dihydroxyphenylpropionate9 or homoprotocatechuate, which are oxidizedby protocatechuate 4,5-oxygenase,14 2,3-dihydroxyphenylpropionate oxy-genase,6? 31 and 3,4-dihydroxyphenylacetate 2,3-oxygenase or 3,4-dihy-droxyphenylacetate 4,5-0xygenase,~O respectively. Similarly, the aldolasethat splits y-hydroxy-a-oxovalerate to acetaldehyde and pyruvate (as forexample in the sequence for catechol) is distinct from that which forms twomolecules of pyruvate from one molecule of y-hydroxy-y-methyl-a-oxoglu-tarate (part of the reaction sequence for protocatechuate oxidation).Dis-crimination between the ring-cleavage oxygenases may also be made by theease wit'h which some of them are dissociated from the Fez+ ions essential totheir activity.Catechol2,3-oxygenase,l5 protocatechuate 4,5-oxygenase,l*, 272,3-dihydroxyphenylpropionate oxygenase, and 3,4-dihydroxyphenylace-tate 4,5-oxygenase 30 are easily resolved from Fez+ ions by precipitationwith ammonium sulphate or by dialysis; but only prolonged storage or treat-ment with hydrogen peroxide produms in 3,4-dihydroxyphenylacetate2,3-oxygenase the requirement for added Fe2f.Although the integrating rbles of 3-oxoadipate and pyruvate are wellestablished for many reaction sequences leading from the benzene nucleus,these two metabolites are not always the earliest common intermediates ofchemically analogous pathways. For pathways involving " cis,cis-muconic30 S. Dagley, P. J. Chapman, and D.T. Gibson, Biochem. J., 1965, 97, 643.31 S. Degley and J. M. Wood, Biochim. Biophys. Acta, 1965, 99, 383; S. Dagley,J. M. Wood and P. J. Chapman, Biochem. J . , 1962, 84, 9PPTABLE 1 Intermediates formed during metabolism of substituted catechols byRing-cleavage substrateCatechol3 -Methylcatecho14-Methylcatechol2 , 3 -Dihydr oxy - /3-phenyl-propionateProtocatechuate3,4 -Dihydroxyphenylacetate3,4-Dihydroxyphenylacetateandrosta-1,3,5( 10)-triene-9,17-dione (29a)3,4-Dihyclroxy-9, ~ O - S ~ C O -RlHMeHHHSee (a)in Scheme 4R2HHMeHHHCH2C02HMeR,HHHHCO,HCH2C02HHHHydroxy -acidMoiety (14)y-Hydroxy- a-oxo-valeratey-Hydroxy- a-oxo-valerate~-HY&OXY- a-oxo-caproatey-Hydr~xy- a-oxo-valeratey-Hydroxy - y-methyl-a-oxoglutaraMhydroxy- a-oxovalerate[ pimelatecaproatey-Cmbox~thyl-y-1 y-Hydroxy- a-0x0-~-HY&oxY- a-OXO-AcidFormateAcetateFormateSuccinateFormateFormate[FormateRIBBONS : MICROBIOLOUICAL DEURADATION 451acids," the enol-lactone of 3-oxoadipate is the first common intermediate ofcatechol and protocatechuate metabolism (Scheme 1) ; for reaction sequencesof the type shown in Scheme 2, y-hydroxy-cc-oxovalerate is the fist commonintermediate arising from catechols where R2 and R3 are hydrogen atoms.The formation of y-hydroxy-cc-oxovalerate occurs independently of the sub-stituent a t R1, be it hydrogen, methyl, or p-propionyl, since the carbonatom to which it is attached is separated as a carboxylic acid from the restof the carbon atoms derived from the benzene nucleus (Scheme 2).The enzymes, first extracted from pseudomonads by Dagley et aZ.cata-lysing the conversion of catechol into formate, acetaldehyde, and pyruvateand the conversion of 2,3-dihydroxyphenylpropionate into succinate, acetal-dehyde, and pyruvate have been studied in more detail.26s31 The aldolasewhich, in both enzyme sequences, catalyses the formation of acetaldehydeand pyruvate from one of the enantiomers of y-hydroxy-cc-oxovaleraterequires Mg2+ ions for high activities; the Achromobackr enzyme (2,3-dihydroxyphenylpropionate sequence) also responds to Mh2+ ions but theseare not as effective as Mg2+ ions.3lIt is well established that catechol, the methyl-substituted catechols, andprotocatechuate may be oxygenated a t two different sites; however, the twodistinct sites of cleavage on the 3,4-dihydroxyphenylacetate molecule areboth meta to the o-dihydroxy-group.Products of both enzymic reactionshave been characterized as derivatives and the sites of cleavage determinedby the formation of characteristic pyridine carboxylic acids by non-enzymicreaction with NH,+ ions. The muconic semialdehyde, obtained from thereaction catalysed by the 2,3-oxygenase, yielded with NH,+ ions a carboxy-methylpyridinecar boxylate oxidized by permanganat e to p yridine -2,s-dicarboxylate 32 (23) whereas the 4,5-oxygenase-catalysed reaction gave in t'heOHA I- 2,3-OxygenaseB = 4,s -0xygenases2 K. Adachi, Y.Takeda, S. Senoh, and H. Eta, Bwchim. Bhphgs. Acta, 1964,93, 483452 BIOLOGICAL CHEMISTRYsame way an isomeric carboxymethylpyridinecarboxylate (25) decarboxyl-ated to pyridine-4-acetic acid (26) 3O (Scheme 3).The Pseudomoms studied by Hayaishi and co-workers1~ appears toutilize an alternative route for a-hydroxymuconic semialdehyde metabolismto that established by Dagley et aL5 (Scheme 2). The sequential formation of4- oxalocr otonate , 4 - hydroxy-2 - oxovalerate , ace t o p yr uvat e, and acetate pluspyruvate was proposed. The enzyme decarboxylating 4-oxalocrotonate t o4-hydroxy-2-oxovalerate was also present in extracts of other bacteria thatmetabolize catechol to pyruvate without carbon dioxide ev~lution,~ i.e.,according to Scheme 2.Enzymes that decarboxylate 3-oxoglutarate occurin Pseudomanas induced to oxidize homoprotocatechuate (4,5-fission) byreactions of Scheme 2 , 6 9 30 but 3-oxoglutarate is not an intermediate.30Furthm examples of the general pathway illustrated in Scheme 2 are tobe found amongst micro-organisms degrading the A ring of certain steroids.Cleavage of ring A of steroids. The microbiological transformation ofsteroids has received special attention, mainly because of commercial pros-pects. Many of these are oxygenative reactions where hydroxyl groups arestereospecifically introduced into almost any position of the nucleus (forReview see Hayano 33). Less widespread, however, are ring-cleavage oxy-genases. Sih and Wang 34 and others 35 showed that androst-4-ene-3, 17-0 nCHO MeMe I + ':--- pMeC? CO2H iCH y x k r M eI- CPH/ * CO2HR(32)('33) Pe + C3 fragmentMef- -HZN COZH A H2N CO2H(3 5 )R = 0f--0 NMeR(3')H2C H2C3 3 M.Hayano, " Oxygenases," ed. 0. Hayaishi, Academic Press, New York, 1963,3 4 C. J. Sih and K. C. Wang, J . Amer. Chem. Soc., 1963, 85, 2136.35 R. M. Dodson and R. D. Muir, J . Amer. Chem. Soc., 1961, 83, 4627.p. 181R I B B O N S : MICROBIOLOGICAL DEGRADATION 453&one (27) was converted into perhydro-7ap-methyl- 1,5-dioxo-lH-3a~-indane-4-propionic acid (32; R = a) via the seco-phenol, 3-hydroxy-9,lO-secoandrosta- 1,3,5( 10)- triene-9,l'l-&one (28) by Pseudomonas sp., Myco-bacterium smegmatis, and Nocurdia restrictus. Cleavage and loss of the Catoms of the A ring of the seco-phenol had clearly occurred.3,4-Dihy-d.roxy-9,10-secoandrosta-1,3,5( IO)-triene-9,17-dione (29; R = a) was alsooxidized to the acid (32; R = a). 4-Hydroxy-2-oxocaproic acid (as thelactone) (34) and 4(5), 9( l0)-diseco-3-hydroxyandrosta-l( 10),2-diene-5,9,17-trion-4-oic acid (30; R = a) were detected as intermediates, and thelatter (30; R = a) was converted non-enzymically with ammonia into itspyridine compound ( 3 1 ) . 3 s y 37 The reactions shown in Scheme 4 (29)-(33)are chemically analogous to those in Scheme 2. A similar series of reactionswould also account for the conversion of progesterone into l-acetylperhydro-7a~-methyl-5-oxo-1H-3aa-indane-4-propionic acid (32 ; R = b) by M . smeg-m~tis.~* The degradation of androst-4-ene-3,17-dione (27) by Ps.testo-steroni, however, yields 2-amino-cis-hex-4-enoate (35) and alani1.1e.3~ Re-actions leading to these compounds are also shown in Scheme 4.Oxidative metabolism of condensed polynuclear aromatic compowzds. Thissection logically follows the last since reactions which degrade condensedpolynuclear aromatic compounds (Scheme 5) are similar in many respects tothose which degrade catechol to pyruvate, as shown in Scheme 2, once the" polynuclear 1,2-diol" has been formed. Thus, oxygenative cleavage ofX\/-.Y k'R2 "'0HO(43)(3 7) (38)Repeat sequenc3 Afor phe5anthrGne/R'OHd C02H z(42)enzymicCHO(41) + fO2H ' 0.co Me\\A Maleylpyruvatefrom 2-naphthoI?36 C. J. Sih, S. S.Lee, Y . Y . Tsong, and K. C. Wang, J . Amer. Chem. SOC., 1965,37 C. J. Sih, K. C. Wang, D. T. Gibson, and H. W. Whitlock, jun., J . Amer. Chem.38 K. Schubert, K.-H. Bohme, and C. Horhold, 2. physiol. CJzem., 1961, 325, 260.3s D. A. Shaw, L. F. Borkenhagen, and P. Talalay, Proc. Nut. Acad. Xci. U.S.A.,87, 1385.SOC., 1965, 87, 1386.1965, 54, 837454 BIOLOGICAL CHEMISTRYthe '' catechol " ring occurs between carbon atoms 1 and 8a, to yield a sub-stituted enol-pyruvate. The next established product is a, substitutedsalicylaldehyde (41) formed by loss of a C3 fragment as pyruvate. If theformation of the aldehyde and pyruvate occur by mechanisms similar tothose in Scheme 2, hydration a t C-3 and C-4 would occur, followed by aretro-aldol fission.It is possible that a complete analogy with the reactionsdiscovered by Dagley et aL6 (Scheme 2) occurs in the degradation of poly-nuclear compounds. Ruptureof the substitutingring in Scheme 5 between car-bon atoms 4a and 8a of structure(40), loss of pyruvate from the hydrated pro-duct, and aromatization of the substituting ring would yield a salicyaldehyde.The general scheme presented for the dissimilation of polynuclear com-pounds seems to account for all the observations so far made for naphthaleneand various substituted derivatives. Table 2 includes the details of theintermediates established for the various polynuclear growth substrates.The most thorough studies to date, in accord with this scheme, have beenmade by Davies and Evans4* for naphthalene and Evans, Fernley, andGriffiths 41 for phenanthrene and anthracene.Pseudomonads that havebeen grown on napthalene as sole source of carbon release small quantities ofD-trans-lY2-dihydro-1,2-dihydroxynaphthalene 42 and salicylate 439 44 into themedium. Extracts of these cells cleave the hydroxylated ring of 1,2-dihy-droxynaphthalene with the consumption of one mole of oxygen per mole ofthe diol ; the product, cis-o-hydroxybenzalpyruvate, was isolated as thepyrylium perchlorate salt. The ease with which this intermediate formscoumarin now explains the earlier observations upon the accumulation ofcoumarin in culture fluids and during enzymic oxidation of 1,2-dihydroxy-naphthalene.45 Carbon dioxide is not evolved enzymically during the meta-bolism of l ,2-dihydroxynaphthdene, but only when conditions permitlactonization, e.g., a t low pH values.o-Hydroxybenzalpyruvate is degradedto pyruvate and salicylate by an NAD specific dehydrogenase. a-Hydroxy-muconic semialdehyde (Scheme 2), the product of catechol cleavage by the2,3-ouygenase, is formed from naphthalene, salicylate, and catechol withcells that have been aged or stored at - 15". Epoxidation of the naphthalenering is probably the initial stage of naphthalene degradation, followed by atrans-hydration of the epoxide to form the dihydro-diol; high-speed super-natant fractions of extracts obtained from pseudomonads grown on naptha-lene when supplemented with NADH, glutathione (reduced), alcohol, andexcess of alcohol dehydrogenase yield the dihydro-diol. This is metabolizedfurther, upon addition of NAD, to cis-o-hydroxybenzalpyru~afe.~~ All ofthe naphthalene-utilizing organisms so far examined appear to follow thispathway;40 thus, the tentative identification of 3-oxoadipate as a meta-bolite of naphthalene 47 and its inclusion in metabolic maps, requires furtherinvestigation.4 O 6.I . Davies and W. C. Evans, Biochem. J., 1964, 91, 251.4 1 W. C. Evans, H. N. Fernley, and E. Grifliths, Biochsrra. J., 1965, 95, 819.43 N. Walker and G. H. Wiltshire, J . Gen. Mdwobiol., 1953, 8, 273.43 R. J. Strawinski and R. W. Stone, J . Bacterial., 1943, 45, 16.44 V. Treccani, Ann. Microbiol., 1953, 5, 232.L6 H. N. Fernley and W. 0. Evans, Nature, 1958, 182, 373.4 e E .Graths and W. C. Eva-, Bwchem. J., 1965, 95, 51P.47 J. F. Murphy and R. W. Stone, Canad. J . Microbwl., 1955, 1, 579Polynuclear carbon sourceNaphthalene1 -Chloronaphthalene2-Chloronaphthalene1 -Brornonaphthalene1 -Methylnaphthalene2 -Methylnaphthalene2 -Hydroxymethylnapthalene2 -Methoxynaphthalene2-E thoxynaphthalene2-HydroxymphthalenePhenanthreneAnthraceneTABLE 2 Cmpounds isolated as metabolites of polynuclearSubstituents forScheme 3R' R' R8H H Hc1 H HH c1 HBr H HMe H HH Me HH CH,OH HH Me HH OEt HH OH HCH:CH*CH:CH HH :CH.CH :CHDihydro-diol42596059505161Products40, 42,59605950, 6464, 5667484858(a) 1-HY~~(b) Salicylate(a) 3-HY~~(b) Salicylatenaphthoatenaphthoat456 BIOLOGICAL CHENISTRYScheme 5 does not account for all the metabolites found during dissimi-lation of 2-methylnaphthalene by Pseudomonas aeruginosa 54 and otherbacteria.57 An alternative pathway exists in which the methyl group isoxidized to carboxyl.The further met'abolism of 2-naphthoate remains to beelucidated, although precursors of it apparently participate in reactionsdepicted in Scheme 5It seems that only mono-nuclear hydrocarbons are degraded to 3-oxoadipate by various species ofbacteria. Ps. aeruginosa, Mycobacterium rhodochrons, and Nocardia sp.oxidize benzene via catechol and &,cis-muconate to 3-oxoadipate.62-6*The trans-diol, cyclohexa-3,5-diene-1,2-diol, is suspected of being inter-mediate between benzene and catechol ; extracts of Aerobncter aerogeneshave been shown to catalyse dehydrogenation of this compound to cate-ch01.~~ Epoxidation of benzene may precede formation of the diol, butproof of these steps is lacking.trans-trans-Muconate has been isolated as ametabolite in Micrococcus sphaeroides 66 and Nocardia corallina,66-6B sincethe stability of the isomers of the muconic acids was elucidated;23 the routeof its formation, as in animals,6g remains to be investigated.The degradation of alkylbenzenes has been little studied, except fortoluene. The ability of Ps. aeruginosa to oxidize benzyl alcohol, benzalde-hyde, benzoate, and catechol, only when the cells had been exposed totoluene suggests that side-chain oxidation occurs before ring cleavage.iOSuch a pathway is however not ubiquitous;70-72 Claus and Walker i2 havedemonstrated that other pseudomonads yield different enzyme systems tometabolize toluene.Extensive whole-cell experiments with Achromobactersp. and Pseudomonas sp. grown on toluene, benzyl alcohol, or benzene showthat benzaldehyde and benzoate are not metabolized except after a period48 R. J. W. Byrde, D. F. Downing, and D. Woodcock, Biochem. J., 1959, 72, 344.4 9 C. Colla, C. Biaggi, and V. Treccani, Atti Acad. naz. Lincei, Rend. Classe Sci.50 L. Canonica, A. Fiecchi, and V. Treccani, Rend. Inst. Lombard0 Sci. Lettere, B,61 C . Colla, A. Fiocchi, and V. Treccani, Ann. Microbiol. Enzimol., 1959, 9, 87.5 2 M. H. Rogoff and I. Wender, J . Bacteriol., 1957, 73, 264.53 M.H. Rogoff and I. Wender, J . Bacteriol., 1957, 74, 108.5 4 M. H. Rogoff and I. Wender, J . Bacteriol., 1959, 77, 783.5 6 V. Treccani and G. Baggi, Rend. Inst. Lombard0 Sci. Lettere, B, 1962, 96, 32.5 6 JT. Treccani and A. Fiecchi, Atti IX Congr. Naz. Microbiol. Palermo, 1956, 139.j7 V. Treccani and A. Fiecchi, Ann. Microbiol. Enzimol., 1958, 8, 36.58 K. Walker and K. D. Lippert, Biochenz. J., 1965, 95, 5C.j 9 X. Walker and G. H. Wiltshire, J . Gen. Microbiol., 1955, 12, 478.6o C. Arnaudi and V. Treceani, Sci. Reports Ist. Super. Sanitd, 1961, 1, 378.61 V. Treccani, Progress Indust. Microbwl., 1962, 4, 3.6 2 A. C. Van Der Linden and G. J. E. Thijsse, Adv. Enzymol., 1965, 27, 469.63 E. I<. Marr and R. W. Stone, J . Bacterwl., 1961, 81, 425.6 4 V.Treccani and B. Bianchi, Atti X Congr. Naz. Microbiol., 1959, 207.6 5 P. Iz. Ayengar, 0. Hayaishi, M. Nakajima, and T. Tomida, Biochirn. Biophys.6 6 A. Kleinzeller and Z. Fend, Chem. Abs., 1953, 47, 4290.67 T. Wieland, G. Griss, and B. Haccius, Arch. Mikrobiol., 1958, 28, 383.68 B. Haccius and 0. Helfrich, Arch. Mikrobiol., 1958, 28, 394.69 D. V. Parko and R. T. Williams, Biochem. J., 1954, 51, 339.7o 31. Kitagawa, J . Biochem., 1956, 43, 653.71 J. R. Forro and R. W. Stone, Bact. Proc., 1965, 90.7 2 D. Claus and N. Walker, J. Gen. Microbiol., 1964, 38, 107.(Table 2).-Metabolism of other aromatic hydrocarbons.$8. mat. nut., 1957, 23, 66.1957, 91, 119.Acta. 1959, 33, 111RIBBONS : MICROBIOLOGICAL DEGRADATIOK 457of induction, although these cells contain enzyme systems of low specificity.Furthermore, 3-methylcatechol was detected as a metabolite of toluene, andwas oxidized by cells taken from toluene media but not by those harvestedfrom catechol media. The pathway of toluene degradation is not yet entirelyclear, although a route through 3-methylcatechol and fissions according toreactions of Scheme 2 might be suggested, since yellow 0x0-acids, acetate,and pyruvate have also been detected as metabolite^.^^ Further supportfor such a scheme comes from the isolation of glycollic acid as a metaboliteof benzyl alcohol in toluene-grown cells.The hydroxymethyl group wouldbe RI in Scheme 2 and the unidentified phenol in culture filtrates may be2,3-dihydroxybenzyl alcohol.Extracts obtained from a toluene-grownpseudomonad degrade toluene and 3-methylcatechol to a compound withspectral properties similar to 2-hydroxy-6-oxohepta-2,4-dienoate (13 ;R1 = Me);10, 7l benzaldehyde, benzoate, and catechol, however, yielda-hydroxymuconic semialdehyde (13).Very little is know-n about the microbial oxidations of other alkylben-zenes. Nocardia sp. which oxidize decyl- dodecyl-, and octadecyl-benzenesyield phenylacetic acid ;73 ethyl and n-butylbenzenes also yield this productwhen they are used to supplement media containing n-alkanes, but they donot support growth.'4 Alliylbenzenes containing an odd number of carbonatoms yield cinnamic acid.74MetaboEism of tryptophan. Pseudomonads are known to degrade trypto-phan to carbon dioxide, water and ammonia by two main routes called (a)the aromatic pathway and (b) the quinoline pathway (Scheme 6).The mainsteps of the aromatic pathway were established 75 before 1951, but elucida-tion of the route of kynurenic acid degradation is comparatively recent.Invariably micro-organisms that utilize the quinoline pathway can oxidizeD-tryptophan as well as the L-isomer. The D-isomer is metabolized by itsown stereospecific enzymes to kynurenic acid (49) ; 76 the final stage is cata-l>-sed by D-kynurenine oxidnse, and this reaction is distinct from the transa-mination of L-kynurenine (46) to kynurenic acid (49). Behrman 77 isolateda pseudomonad that oxidized tryptophan by the aromatic pathway andutilized D-tryptophan. A tryptophan racemase was demonstrated in thisorganism, and D-kynurenine is not metabolized.A third pathway is sug-gested for Fhuobacte~ia metabolizing D-tryptophan. 78 Electron acceptors,such as phenazine methosulphate, are required in addition to an amino-groupdonor. Formation of indolepyruvic acid and transamination to L- tryptophanappear to occur.The quinoline pathway of tryptaphan metabolism was discovered whenextracts of cells supplemented with reduced coenzymes yielded L-glutamate,D- and L-alanine, acetate, and carbon dioxide, and it was found that the73 D. M. Webley, R. B. Duff, and V. C. Farmer, Nature, 1956, 178, 1467.73 J. B. Davis and R. L. Raymond, AppZ. Microbiol., 1961, 9, 383.7s R. Y. Stanier and 0. Hayaishi, Science, 1951, 114, 326.76 M.Tashiro, T. Tsukada, S. Kobayashi, and 0. Hayaishi, Biochem. Biophya.7 7 E. J. Behrman, Nature, 1962, 196. 150.Res. C'omm., 1961, 6, 155.'* J. R. Martin and K. N. Durham,- Biochem. Biophp. Res. Comm., 1964, 14,388458 BIOLOGICAL CHEMISTRY+ NH3Q U i no I i ne pathwayglutamate was derived from the carbocyclic ring.79-81 The intermediates7,8- dihydroxykynurenate ( 5 1 ) , 5- ( y - carboxy- y - oxopropyl) -4,6-dihydroxy-picolinate (53), 5- (p-formylethyl) -4,6-dihydroxypicoliate (54), and 5- (p-carboxyethyl)-4,6-dihydroxypicolinate (55) have apparently been charac-terized for Pseudomom sp . , 82y 83 and 5 - ( y - carb oxy - y - oxopropenyl) -4,6 -dihydroxypicolinate (52) and the acid (55) for an Aerococcus 84 metabolizingkynurenic acid.It seems likely that a 7,8-epoxide is intermediate betweenkynurenate and the dihydro-diol (50) as for naphthalene metabolism.Experiments using l 8 0 support this v i e ~ . ~ 3 The origin of the end productsof this reaction scheme is less clear, Kuno et aZ.82 have demonstratedstoicheiometric formation of a-oxoglutarate, ammonia, and oxaloacetatefrom 5-(~-carboxyethyl)-4,6-dihydroxypicolinate (55) ; extracts of Aero-coccus also yield glutamate, aspartate, and pyruvate. 84 The enzyme systemsfor this dissimilation are obviously still too crude to yield a definitiveanswer.For pseudomonads that utilize the aromatic pathway, the formation of7B E. J. Behrman and T. Tanaka, Biochem. Bwphys. Res. Comm., 1959, 1, 257.0. Hayaishi, H. Taniuchi, M.Tashiro, and S. Kuno, J . BioE. Chem., 1961, 238,81 K. Horibata, H. Taniuchi, M. Tashiro, S. Kuno, and 0. Hayaishi, J . BioE. Chem.,88 S. Kuno, M. Tashiro, H. Taniuchi, K. Horibata, 0. Hayaishi, S. Senoh, T.e8 H. Taniuchi and 0. Hayaishi, J . BioE. Chem., 1963, 238, 283.84 S. Dagley and P. A. Johnson, Biochim. Biophys. Acta, 1963, '78, 577.2492.1961, 236, 2991.Tokuyama, and T. Sakan, Fed. P~oc., 1961, 20, 3RIBBONS : MICROBIOLOGICAL DEGRADATION 459catechol from anthranilate appears to be the only established route. How-ever, an alternative route of anthranilate metabolism is exhibited by a soilAchromobacter sp ; 5-hydroxyanthranilate and gentisate are implicated asintermediates in this pathway.", 86 Phvobacteria, after growth on anthra-nilate, oxidize sali~ylate.~'Metabolism of phenylpropanoid structures.Although a high proportionof organic carbon is returned to the soil as phenylpropanoid structures, thereis little detailed information as to how these compounds are metabolized.Specific studies have been made on the microbial degradation of phenyl-propionate, cinnamate, and coumarin.Some species of Arthrobacter utilize coumarin as sole source of carbon;88changes in ultraviolet absorption and chromatographic characteristics sug-gested that o-coumarate and melilotate (o-hydroxyphenylpropionate) weremetabolites. o-Coumarate is reduced to rnelilotate by an NAD or NADPoxidoreductase which is present in cells only after growth on o-coumarate.2,3-Dihydroxyphenylpropionate has also been tentatively identified as aproduct of melilotate when incubated with crude extracts and NADH.89,It would be of interest to know if 2,3-dihydroxyphenylpropionate is meta-bolized by reactions shown in Scheme 2 by Arthrobacter sp.Growth of pseudomonads on cinnamate generally yields cells able tomet a b olize p henylpropionate .Melilot at e and 2,3 -dihydr ox yphenylpr o -pionate have been isolated from culture filtrates of cinnamate 91, 92 andphenylpropionate media.32 Reduction of this side-chain appears to befairly widespread, since cinnamate, .Q-hydroxyci~amate, and 3,4-dihydroxy-cinnamate yield their respective dihydro-compounds from media supportingthe growth of Lactobacillus pastorknus par. q u i n i c ~ s . ~ ~ Phenylpropionicacid has also been detected as a metabolite of cinnamate by Pseudomonas;92hydroxylation of the ring a t position 3 before reduction of the side-chain hasalso been suggested for this strain;92 in this respect, it is interesting to notethat the specscity of 2,3-dihydroxyphenylpropionate 2,3-oxygenase extendsto 2,3-dihydroxycinnamate.In addition to these reactions, substitutedphenylpropanoid compounds are decarboxylated to yield their phenylethanesand styrenes by L. pastorianum 93 and Aerobacter,g* respectively.Studies on the bacterial degradation of lignans and other lignin-relatedcompounds have also been rep0rted,~5 and in some detail for Agrobac-t e r i ~ . ~ ~ , 97 a-Conidendrin-degrading Agrobacteria oxidize several otherlignans indicating low specificity of response to the inducer.Sequential85 J. N. Ladd, Nature, 1962, 194, 1099.86 J. N. Ladd, Austral. J. BioZ. Sci., 1964, 17, 153.87 T. Higashi and Y. Sakamoto, J. Biochern. (Japan), 1960, 48, 147.C. C. Levy and G. D. Weinstein, Nature, 1964, 202, 596.C. C. Levy and (3. D. Weinstein, Biochemistry, 1964, 3, 1944.C. C. Levy, Nature, 1964, 204, 1059.C. B. Coulson and W. C. Evans, Chern. and Id., 1959, 643.8a E. R. Blakley and F. J. Simpson, Canad. J. Microbiol., 1964, 10, 175.g3 G. C. Whiting and J. G. Cam, Nature, 1959, 184, 1427.Q 4 B. J. Finkle, J. C. Lewis, J. W. Corse, and R. Lundin, J. Biol. Chem., 1962, 237,96 H. Ssrensen, J. Gen. Microbiol., 1962, 2'7, 21.V. Sundman, J . Gen. MicrobioE., 1964, 36, 185.97 V. Sundman, J. Ben. MicrobioE., 1964, 36, 171.2926460 BIOLOGICAL CHEMISTRYinduction experiments suggest that or-conidendrol is an intermediarymetabolite of a-conidendrin. Extracts of these organisms should yield avariety of interesting new enzymes.Oxygenases Involved in Aromatic Ring Metabolism.-The progress madein the elucidation of the mechanisms of oxygenase reactions is almost en-tirely due to the admirable researches of Hayaishi and his colleagues.lj9 98The demonstration that pyrocatechase catalysed the incorporation of bothatoms of a molecule of oxygen into the carbon substrate to yield labelledmuconic loo and that phenolase catalysed the incorporation ofmolecular oxygen into an aromatic substrate in the presence of reducingsubstances,lol initiated the studies on oxygenase reactions.A particularlygood account of oxygenases has been published by Hayaishi.ls During thepast two years three ring-cleavage enzymes, pyrocatechase,lo2 catechol2,3-oxygenase,lo3* lo4 and 3,4-dihydroxyphenylacetic acid 2,3-oxygensse 7, lo5have been crystallized. The problems of instability that previously pre-cluded such purifications were solved by maintaining the enzymes in areduced condition, notably with sodium borohydride under anaerobic con-ditions,lO6j 107 and by protecting them against inactivation by air in thepresence of organic solvents.Cadecho1 1,2-oxygenase (pyrocatechase) . This has been extensively purifiedand crystallized,15, loa the final preparations being homogeneous in theultracentrifuge and upon electrophoresis. Molecular weights of 95,000,108&4,OOOY1O9 and 78,0003 are reported for preparations from Pseudomonus,Micrococcus, and Brevibacterium, and for Pseudomonas pyrocatechase twoatoms of iron are bound to each protein molecule. No other cofactors havebeen found.Although this enzyme was the first of its type discovered, theactual demonstration of an Fe2+-ion requirement has been diffcult and con-fusing. Suda and co-workers have removed Fe2+-ion with o-phennn-throline from pyrocatechase obtained from Psedmonas sp. and reactivatedthe enzyme with Fe2+. They also demonstrated that an enzyme preparationMicrococcus ureae was able to exchange its protein-bound Fe2+ with eso-genous 59Fe2f only when both substrates (cstechol and oxygen) were pre-88 0.O9 0.loo 0.l01 H.2914.102 Y.1963, 85,Io3 M.1963, 11,lo4 M.lo6 H.100 0.lo’ H.Biop h ya.lo8 H.829.lo9 K.5450.Hayaishi, ‘‘ Oxygenases,” Academic Press, New York, 1962.Hayaishi, M.Katagiri, and S. Rothberg, J . Amer. Chem. SOC., 1955, 77,Hayaishi, M. Katagiri, and S. Rothberg, J . Biol. Ckem., 1957, 229, 905.S. Mason, W. L. Fowlks, and E. Peterson, J . Amer. Chem. SOC., 1955, 77,Kojima, J. Nakazawa, H. Taniuchi, and 0. Hayaishi, J . Jap. Biochem. SOC.639.Nozaki, H. Kagamiyama, and 0. Hayaishi, Biochem. Biophya. Res. Crnnm.,65.Nozaki, H. Kagamiyama, and 0. Hayaishi, Biochem. Z., 1963, 388, 582.Kita, J . Biochem. (Japan), 1966, 58, 116.Hayaishi, H. Taniuchi, and Y. Kojima, Fed. Proc., 1962, 21, 52.Taniuchi, Y.Kojima, F. Kanetsuna, H. Ochiai, and 0. Hayaishi, Biochem.Res. Comm., 1962, 8, 97.Taniuohi, Y. Kojima, A. Nakazawa, and 0. Hayaishi, Fed. Proc., 1964, 23,I Tokuyama, M. Suda, and Y. Shimomura, Proc. Internat. Symp. EnzymeM. Suda, K. Hashimoto, H. Matsuoka, and T. Kamahora, J . Biochem. (Japan),Chem., Japan, 1957, 197.1951, 38, 289RIBB ON S : MICROBIOLOGICAL DEGRADATION 461sent, and the rate of exchange was a function of the rate of enzymic reactionoccurring.l*s The exchange data indicated that this pyrocatechase was ableto exchange only one atom of iron per molecule of protein.Pyrocatechase obtained from Ps. arlviWct is red, with a broad absorptionband between 400 and 600 mp, and its e.s.r. spectrum shows a sharp linea t g = 4.2 and a signal width of 35 gauss between the peaks.1ll Dithioniteabolishes the visible and e.s.r.spectra, both of which are partially re-stored by oxygenation. These results indicate the presence of one Fe3+ ionbound to each protein molecule. Protein-bound Fez+ ion could not bedetected spectroscopically even with dithionite-reduced enzyme, althoughchemical analysis was positive. A correlation between bound Fe3+ ion,visible absorption spectra, and enzymic activity was found during varioustreatments. Addition of catechol to the enzyme (anaerobic) abolishes thee.s.r. signal but the 400-600-mp spectrum is retained and a new 700-mppeak appears. The e.s.r. signal reappears when the catechol has been trans-formed to muconate by exposure to oxygen; a t the same time the 700-mppeak disappears.It would be of interest to know if both atoms of iron inthis pyrocatechase preparation are exchangeable during reaction, or alter-natively to know if the M . ureae preparation contains one or two atoms ofiron per enzyme molecule. It is clear that preparations of this enzyme fromdifferent bacteria, especially Pa. arvilh, N. ureae, Ps. JZuorescens, mdBrevibacterium ficscum are markedly different with respect to the effect ofFe2+ ions and reducing compounds upon activity and substrate specificity.It has been tacitly assumed in reactions involving the addition of twoatoms of oxygen to the carbon substrate that both atoms of oxygen arederived from the same molecule, and further that both atoms of oxygen donot become attached to the same carbon atom.With the availability ofhighly enriched 180 and " high-molecular-weight " mass spectrometers ithas been possible to test and verify this assumption. Thus, in an atmospherecontaining only the 180-180 and 160-160 species of oxygen, the dimethylester of the cis,cis-muconate formed during pyrocatechase-catalysed oxygena-tion of catechol had molecular masses of 170 and 174 only.112 Furthermore,the fragmentation pattern showed that only one atom of l 8 0 was incor-porated into each of the carboxyl groups of ci&s-muconate.llzCatechoE 2,3-oxygenase (metapyrocatechase). Catechol 2,3-oxygenase wascrystallized from Ps. arvilh after only a 30-fold purification of extracts.103,The molecular weight of 140,000 was calculated from sedimentations of thehomogeneous protein in the ultracentrifuge.Colorimetric analysis revealedone atom of iron per protein molecule and, unlike pyrocatechase, catechol2,3-oxygenase was colourless. E.s.r. spectral data indicate that the ironis in the Fe2* state. Native catechol2,3-oxygenase does not give a signal a tg 4.2, but this appears when catechol is also added to aerobic systems.15 Itseems unlikely that a ferric state arises only when reaction occurs or whenthe substrate (or analogue) can form a complex with the metal. Unlikepyrocatechase, catechol 2,3-oxygenase is easily resolved from its iron bylfl T. Kakazawa, Y. Kojima, IE. Fujisawa, 3X. Nmaki, and 0. Hayaishi, J . Biol.112 N. Itade, Biochtm. Biophys.Res. Comnz., 1965, 20, 149.Chenz., 1965, 240, PC3224462 BIOLOGICAL CHEMISTRYhydrogen peroxide oxidation and dialysis, and competitively inhibited byFe2+-chelating agents.l53,4-Dihydroxyphenylacetate 2,3-0xygenase. Pseu&nwnus ovalis, aftergrowth on p-hydroxyphenylacetate, yields extracts from which 3,4-dihy-droxyphenylacetate 2,3-oxygenase has been crystallized. '9 32,105 The crys-tals were colourless but showed a maximum absorption a t 280 mp and ashoulder at 292 mp. Ultracentrifugal analysis revealed a homogeneouspreparation with a molecular weight of 100,000; each molecule of enzymecontained four to five atoms of iron. It seems that the iron in these pre-parations is in the Fe2+ form, like catechol2,3-oxygenase, since it is inacti-vated by hydrogen peroxide and reactivated by Fe2+ ions.Tryptophctn oxygenase (tr yptophctn pywoktse).When L-tryptophan pyr-rolase preparations from Pseudomoms extracts are inactivated by ageing,addition of ascorbate or hzmatin (ferriprotoporphyrin IX) can restoreenzymic activity.15 Tryptophan itself, after prolonged incubation, will alsoreactivate this enzyme, especially under anaerobic ~onditions.1~~ The reacti-vation process is accompanied by changes in the absorption spectrum of theenzyme, suggesting that ascorbate and tryptophan reduce the iron to theFe2+ form; in fact the activity of the enzyme is directly related to its redoxstate. Spectral and redox changes that occur in tryptophan pyrrolase duringcatalysis show reduction of the Fe3+ form to the Fe2+ form, but this is notstoicheiometric initially.Anaerobic incubation of the pyrrolase with trypto-phan results in substrate-dependent reduction of enzyme-hzmatin to enzyme-haeme ; oxygen reoxidizes this, suggesting a cyclic oxidation-reduction of thehaematin during catalysis. The hzematin analogue, protoporphyrin IXinhibits oxygenase activity which can be restored by excess of hzmatin.Difference spectra of the reaction mixture suggest the occurrence of an inter-mediate complex, probably of enzyme, tryptophan, and oxygen.15 Theactive coenzyme forms of haematin and haeme, and their participation in thecatalysis, make this enzyme exceptional among oxygenases that catalyse theaddition of two atoms of oxygen to the carbon substrate.However, specula-tion concerning the mechanism of reaction must be reserved until the r61eof copper has been elucidated. Feigelson et ~ 1 . ~ ~ 3 have recently found largeamounts of bound copper in their purest pyrrolase preparations, and theactivity of the enzyme is inhibited by specific chelators for cuprous or cupricions; other chelating agents do not inhibit.The enzyme catalysing the formation ofcatechol from anthranilate has been obtained from Ps. a e r u q i n ~ s a , l ~ ~ ~ 115Ps. $uorescens 115 -117 and Micrococcus ureae .115 This complex reactionconsumes one mole each of oxygen and NADH, with formation of one moleAnthranilate hydroxylase.113 P. Feigelson, Biochim. Biophys. Acta, 1964, 92, 187; P. Feigelson, Y. Ishimura,and 0.Rayaishi, ibid., 1965, 96, 283; idem., Biochem. Biophys. Res. Comm., 1964, 14,96; H. Maeno snd P. Feigelson, Biochem. Biophys. Res. Comm., 1965, 21, 287.I1*T. Higashi and Y. Sakamoto, J . Bwchem. (Japan), 1960, 48, 147.115 A. Ichihara, K. Adachi, K. Hosokawa, and Y. Takeda, J . Biol. Chem., 1962,28'7, 2296.116 H. Taniuchi, M. Hatanaka, S. Kuno, 0. Hayakhi, M. Nakajima, N. Kurihara,J . Biol. Chem., 1964, 239, 2204.I1'S. Kobayashi, S. Kuno, N. Itada, and 0. Hayaishi, Biochem. Biqhys., Res.Comm., 1964, 16, 556RIBBONS : MICBOBIOLOGICAL DEGRADATION 463each of carbon dioxide and ammonia per mole of anthranilate used and cate-chol formed.ll6 It was suggested that intermediary formation of an epoxide,and its subsequent hydrolysis, occurred. However, Kobayashi et aL117showed that the four atoms of oxygen in cis&-muconate are all derivedfrom molecular oxygen and not fkom water.Consequently, it seems likelythat an epoxide is not formed, but that both atoms of an oxygen moleculeadd across carbon atoms 1 and 2 of anthranilate to yield a cyclic peroxide.The unusual requirement of Fe2+ ions for maximal hydroxylase activitylends further support to this hypothesis of addition of a single molecule ofoxygen to anthranilate; an experhen$ similar to that conducted by Itada 11%with pyrocatechase should verify this point.PhenyZatanine hydroxylase. Kaufnirtn lls has studied this enzyme frommammalian sources in some detail but only recently has a similar enzymebeen extracted from Pseudomoms sp.Guroff and Ito 119, 120 have demon-strated a requirement for Fez+ ions, tetrahydropteridine and molecularoxygen for tyrosine formation with purified extracts. Preincubation withFe2+ is usually necessary for full activity. The involvement of pteridine inthis reaction parallels the mammalian enzyme; however, there are somedifferences in the specificity of response to pteridine analogues. The incor-poration of ISO, into the tyrosine formed has been shown with whole cells ofPs. aeruginosa.121Salicytate hydroxylase. During purification of a salicylate hydroxyla,sefrom cells of a Gram-positive coccus, a requirement for flavin adeninedinucleotide (FAD) was demonstrated.151122 Catalytic quantities of FADare required if a source of NADH, or NADPH, is available.The stoicheio-metry of the reaction corresponds to:FADSalicylafe + NADH, + 0, --+ Catechol + NAD + H,O + COa2,3-Dihydroxybenzoate, phenol, benzoate, and anthranilate are not meta-bolized by this system, and the introduction of the second hydroxyl isassumed to replace the carboxyl group. Inhibition of the enzyme by metalchelators can be reversed by dialysis ; metal-ion additions were withouteffect.Several other microbial aromatic oxygenases havebeen described but not in sufficient detail to discuss possible mechanisms.Those that have not been referred to previously include benzoate hydroxyl1ase,117 kynurenic oxygenase, naphthalene hydroxylase,4* protocatechuate3,4-oxygenase,123 protocatechuate 4,5-oxygenase,% l 4 gentisate oxyge-n a ~ e , l ~ ~ , 125 7,8-dihydroxykynurenate oxygena~e,~~ 5-(p-carboxyethyl)-4,6-11* S.Kaufman, ‘‘ Oxygenases,” ed. 0. Hayaishi, Academia Press, Xew York andLondon, 1962, p. 129.lZo G. Guroff and 1. Ito, J . BWE. Chem., 1965, 240, 1175.121 K. Takashima, D. Fujimoto, and N. Tamiya, J . Biochem. (Jupun), 1964, 55,12* M. Katagiri, S. Yamamoto, and 0. Hayaishi, J . Biol. Ohern., 1962, 237, PC2413.123R. Y. Stmier rand J. L. Ingraham, J . Biol. Chem., 1954, 210, 799.1 2 4 S . Sugiyams, K. Yano, H. Tanaka, K. Komagata, and K. Arims, J . Cen. AppE.126 L. Lmk, Biochim. Biophys. Acta, 1959, 34, 117.0 t h - oxygenases.C. Guroff arid T. Ito, Biochim. Biophya. Acta, 1963, 77, 159.122.Microbiot?. (Japan), 1958, 4, 223464 BIOLOGICAL CHEMISTRYdihydroxypicolinate oxygena~e,~~ and homogentisate oxygenase,126 ex-cluding other heterocyclic oxygenases.Regulation of metabolism.-Regulation of the synthesis of the enzymeswhich catalyse the catabolism of aromatic compounds has only recentlybeen studied.It is known that these enzymes are generally elaborated inresponse to specific substrates and the process was thought to be sequen-tia1,12' e.g., substrate A induced enzyme a for its metabolism to compound B,which in turn acted as inducer for the synthesis of enzyme b for its ownmetabolism to compound C, and so on. Sequential induction undoubtedlyregulate the synthesis of some enzyme sequences, but it is supplemented byco-ordinate induction, whereby a group of enzymes is synthesized in re-sponse to a single inducer.Co-ordinate induction of a group of enzymes ispresumably controlled by a single operon or regulon.12SRegulation of the aromatic pathway of tryptophan metabolism. The aro-matic pathway of tryptophan oxidation by Ps. Jluorescens proceeds throughcatechol as shown in Scheme 6. The formation of the whole sequence ofenzymes involved is initiated by exposure of cells to L-tryptophan or L-kyn~renine.~~ Palleroni and Stanier 2 9 showed that L-tryptophan itselfdoes not act as an inducer; low but measurable quantities of tryptophanoxygeiiase (tryptophan pyrrolase) and formylkynurenine formamidase arepresent in uninduced cells and these enzymes form L-kynurenine, thereal inducer, which then induces co-ordinately and sequentially the forma-tion of all the enzymes of this pathway.The sequential nature of the induc-tion of enzymes by kynurenine and anthranilate was demonstrated by atechnique that might be exploited further. Cells were exposed to L-trypto-phan for a period sufficient to induce only the enzymes for its own catabolismto ant,hranilate. Ultraviolet irradiation of the suspension to prevent furtherprotein synthesis and incubation with more tryptophan resulted in an almostquaiititative conversion into anthranilate. Anthranilate serves as inducerfor enzymes catalysing reactions succeeding but not those preceding itselfin the sequence; other sequential steps are probably involved.Theenzymes employed for DL-mandelate oxidation by Ps. putida were alsoshown to be induced co-ordinately and sequentially.l6.12'3 129 The first fiveenzymes 130-132 (the " mandelate " group), mandelic racemase (El), L-mandelic dehydrogenase (E2), benzoylformic carboxylase (EJ, and benz-aldehyde dehydrogenases (E4J and (ESb), appear to be controlled by asingle operon 128 and are co-ordinately induced by either DL-mandelate orbenzoylformate, and also by the gratuitous inducer phenoxyacetic acid(Scheme 7). Benzoate does induce formation of these enzymes, but onlythose lower in the metabolic sequence, and is the first sequential inducer inRegulution of the synthesis of enzymes of the mandelate pathway.12u P. J. Chapman and S. Dagley, J . Gen. Microbiol., 1962, 28, 251.12' R. Y. Stanier, J. Bacteriol., 1947, 54, 339.lP8 F. Jacob and J.Monod, J . Mol. Biol., 1961, 3, 318; W. K. Maas and E. McFall,Ann. Rev. Mhrobiol., 1964, 18, 96.12* I. L. Stevenson and J. Mandelstam, Biochem. J., 1965, 96, 354.130 I. C. Guwlus, C. F. Gunsalus, and R. Y. Stanier, J . Bacterwl., 1953, 68, 538.131 R. Y . Stanier, I. C. Gunsalus, and C. F. Gunsalus, J. Bucteriol., 1953, 66, 543.132 C. F. Gunsalus, R. Y . Stanier, and I. C. Gunsalus, J . Racteriol., 1953, 66, 548RIBBONS : MICROBIOLOGICAL DEGRADATION 465this pathway. It was not determined unequivocally if a second sequentialstep occurs for the induction of the catechol oxygenase or succeedingenzymes, but this seems likely. Similarly, the protocatechuate group ofenzymes (Scheme 7 ) may not all be induced co-ordinately; one or moresequential steps may be involved.Numerous gratuitous aromatic inducers cause complete derepression(induction) of the enzymes of the mandelate group in some mutants that areconstitutive for enzymes of this group, i.e., formed in the absence of exo-genous inducer.The catechol group of enzymes is unaffected. When theinducer is also a substrate then maximal synthesis of enzymes of the catecholgroup also occurs. Vanillic acid is a gratuitous inducer for the group ofenzymes converting protocatechuate to 3-oxoadipate in 1N. crassa l8 (Scheme1); it seems likely that this group of fungal enzymes is also formed co-ordina t el y.It is clear that an induction mechanism for the synthesis of enzymes oran enzymic sequence would not control the quantity of enzymic activityrequired. Additional methods of regulation are required for this, and specificmetabolic repressions of inducible enzymes have been amply demonstratedby Mandelstam.Thus, pyruvate, a product of the reaction catalysed bytryptophanase in Escherichia coli, specifically represses the formation of thistryptophan-inducible enzyme.133Specific end-product repression is also a regulating mechanism for thesynthesis of the sequence of enzymes that catalyse mandelate and p-hydroxy-mandelate oxidation. The first five enzymes of the mandelate group(Scheme 7) are subject to end-product repression by succinate and acetate.However, several of the intermediates of mandelate (and p-hydroxymande-late) oxidation are able to repress the enzymes of this group and other enzymesof the sequence that catalyse their formation; these are the products of eachsequential step of the pathway, i.e., catechol (or protocatechuate) and ben-zoate (or p-hydroxybenzoate).Succinate and acetate are also able to repressthe formation of enzymes of the catechol (or protocatechuate) group and ben-zoate (or p-hydroxybenzoate) hydroxylase. The resultant control mechan-ism is thus called multi-sensitive end-product repression 129, 134 (Scheme 7).Benzaldehyde and p-hydroxybenzaldehyde were also tested as possiblerepressors and were found to repress almost completely the synthesis of theenzymes of the mandelate group. However, the ability of these cells to growon p-hydroxybenzaldehyde was not impaired. A third non-specific benzal-dehyde dehydrogenase is induced under these conditions and this is notcontrolled by the same operon as the mandelate group.The ability of thehenzaldehydes to repress the formation of the enzymes of which they aresubstrates is almost analogous to the ability of kynurenine to induce thesynthesis of the enzyme of which it is the product.Induction and repression of the mandelate group of enzymes is the resultof a balance of inducer and repressor, in that end-product repression may bereversed by higher concentrations of the inducer.129~134The recent renewed interest in the metabolism of and the effect of133 E. McFall and J. Mandelstam, Nature, 1963, 197, 880.lS4 J. Mandelstam and G. A. Jacoby, Biockem. J., 1965, 94, 569CHOH*COZH QPRO1GBenzoatehydroxylasep-Hydroxybenzoate (OH)(5 7) (58) hydroxylasRIBBONS : MICROBIOLO~ICAL DEGRADATION 467halogenated benzoates on the enzymes degrading benzoate should yieldfurther information about the control mechanisms of the synthesis of in-ducible enzymes.135Non-speci$c multi-enzyme systems.The earlier suggestiom 131, u6 thatp-hydroxymandelate and mandelate are oxidized by the same sequence ofenzymes has been conhrmed.129 The mandelate group of enzymes catabolizeboth substrates to yield either benzoate and p-hydroxybenzoate and thesethen induce specifically enzymes for their own degradation (Scheme 7).A similar non-specific oxidation sequence appears to operate in Ps.aerugilzosa T1 when o-, m-, or p-cresol or phenol are the substrates of meta-bolism.1° These cells do not distinguish between these four carbon sourcesas inducers nor as substrates.All induce a sequence of enzymes which w i l lmetabolize them, wia catechol (from phenol) or a methylcatechol (from thecresols), according to Scheme 2. This situation may extend to other ba~teria.~It has been suggested 41 that catechol 2,3-oxygenase specificity alsoextends to the polynuclear substrates of ring fission (Scheme 5). In viewof the homology of several of the enzymic steps that occur during the meta-bolism of polynuclear compounds and of some catechols, it is tempting tosuggest that other common enzymes mediate these reactions.The ability of a variety of substrate-inducers to initiate the synthesis of amulti-enzyme sequence that can metabolize several compounds eventuallyto common metabolites represents a great economy to the cell.As a conse-quence of such low specificities, the cell is not required to carry and replicate‘‘ extra ” genetic information as DNA for each reaction sequence. When thecell is presented with more than one of these non-specific substrate-inducers,the synthesis of a single multi-enzyme system to deal with all of them is asaving on both protein and RNA synthesis. The example, par excellence, isthe metabolism of certain bicyclic terpenoids by a non-specific multi-enzyme system.l37I n contrast, it has also been demonstrated that some bacteria? e.g.,Vibrio 0/1, are able to synthesize either catechol 172-oxygenase or catechol2,3-oxygenase depending on the nature of the primary inducer.13* Itappears that the more non-polar substrate inducers, e.g., the cresols ornaphthalene, lead to catechol2,3-oxygenase synthesis, whilst polar substrate-inducers, e.g., benzoate, yield the 1,2-0xygenase.~~ 138 However, this is notso in every case.lO? 104In the review by Evans and in these Reports attention has been devotedmainly to bacterial metabolism of aromatic compounds.It is clear thatmost groups of micro-organisms are able to catabolize these compounds.Fungal metabolism of some aromatic compounds has been reviewed re-~ent1y.l~~ Many yeasts are also able to grow on simple phenolic compounds ;I4*ls6 F. Bernheim, J . Biol. Chem., 1953, 203, 775; D. E. Hughes, Biochem. J . , 1965,96, 181; A.G. Callely and J. G. Jones, Biochem. J., 1965, 97, 11C.136 S. E. Gunter, J . Bacterial., 1953, 66, 341.la’ I. C. Gunsalus, P. J. Chapman, and J. F. Kuo, Biochem. Bwphys. Res. Comm.,1965, 18, 924.13* E. GrifEths, D. Rodriques, J. I. Davies, and W. C. Evans, Biochem. J., 1964,91, 16P.13# M. E. K. Henderson, Pure Appl. Chew., 1963, 7 , 589.140 G. Harris and R. W. Ricketts, Nature, 1962, 195, 473468 BIOLOGICAL CHEMISTRYthis is an area where there is only a sparse knowledge of the metabolicroutes involved.25$ 141 It seems likely that algae and protozoa will alsoyield enzymes that degrade benzenoid compounds; in this respect somephotosynthetic bacteria have no difficulty in photometabolizing aromaticcompound~.~4~9 l43Recent work describing the biosynthesis of aromatic cornpo~nds,~~Pthe catabolism of heterocyclic compounds such as nicotine,145 and the vita-mins pyridoxine,146 nicotinic acid,l4’ and riboflavin,l48 and structures such asflavonoids lP9 is readily available.A review on the microbiological oxidationof pesticides has also appeared.lS0141 J. S. Hough and D. A. J. Wase, J . gen. ITlicrobiol., 1965, 39, v.148 M. H. Procter and S. Scher, Biochem. J., 1960, 76, 33P.148 E. R. Leadbetter and A. Hawk, Bact. Proc., 1965, 22.146 R. L. Gherna, S. H. Richardson, and S. C. Rittenberg, J . Biol. Chem., 1965,lJ6 E. E. Snell, A. A. Smucker, E. Ringlemann, and F. Lynen, Biochem. Z., 1964,E. J. Behnnan and R. Y. Stanier, J . Biol. Chem., 1957, 228, 923.14*D. R. Harkness y d E.R. Stadtman, J . Biol. Chem., 1965, 240, 4089.lQS G. H. N. Towers, Biochemistry of Phenolic Compounds,” ed. J. B. Harborne,15* R. W. Okey and R. H. Bogan, J . Water Pollution Control Fed., 1965, 37, 692.B. A. Bohm, Chena. Rev., 1965, 65, 435.240, 3669.341, 109.Academic Press, London and New York, 19644. CARBOHYDRATE SULPHATESBy D. A. Bees(Chemktrg Department, The University, West Mains Road, Edinburgh 9)THE sulphate half-esters of carbohydrates have never before been the soletopic for a Report in this Series, and it is 16 years since the last comprehen-sive Review in any journal.1 Some individual aspects have, however, beencovered separately from time to time, and reference will be made later tothe more recent of such Reviews. The enzymic hydrolysis of carbohydratesulphates will not be discussed; this has been adequately reviewed else-where.2,General.-Carbohydrate sulphates occur naturally in animal tissues andin marine algae.True carbohydrate sulphates would appear to be rare inhigher plants; the “plant sulpholipid” which is so widespread, is not asulphate ester but a sulphonic acid. Carbohydrate sulphates rarely, if ever,play such an active part in metabolism as do many carbohydrate phosphates.Their biological functions in cell walls, membranes, intercellular regions,and other situations, would appear to be connected with their physicalproperties more usually than with their chemical reactions. The presenceof ionised sulphate groups, and often of consequent polyelectrolyte character,must greatly influence physical properties.It is a striking coincidence thatmany different polysaccharide sulphates appear to contain two differentsugar units in a strictly alternating linear structure-an arrangement that israre in other polysaccharides. There may be some relation between thecharge of the sulphate and the sizes of the sdphate group and disaccharideunit which leads to conformations and interactions that are biologicallydesirable. It is important that the detailed molecular structures be deter-mined so that their full biological significance can be understood.Prom a chemical point of view, structure determination, especially ofhigh-molecular-weight carbohydrate sulphates, presents a number of inter-esting problems and opportunities.Sulphate groups can normally beretained during methylation of the hydroxyl groups, and information abouttheir location can therefore be obtained by subsequent hydrolysis. Theinformation is, however, ambiguous, and additional evidence is required todistinguish between the site of the sulphate and of glycosidic substitution.This may be obtained by methylation of the desulphated polysaccharide,4periodate oxidation before and after de~ulphation,~ alkaline eliminati~n,~a~ g s 7E. G. V . Percival, Quart. Rev., 1949, 3, 369.K. S. Dodgson and B. Spencer, Ann. Reports, 1956, 53, 318.J. R. Turvey, Adv. Carbohydrate Chem., 1965, 20, 183. ’ (a) R. W. Jeanloz and P. J. Stoffyn, Fed. R o c . , 1958, 17, 249; R. W. Jeanloz,P.J. Stoffyn, and M. TrBmGge, ibid., 1957, 16, 201; (b) S. Hirano, P. Hoffman, and K.Meyer, J. Org. Chem., 1961, 26, 5064; (c) T. C. S. Dolan and D. A. Rees, J. Chem. Soc.,1966,& 3534.(a) J. P. McKinnell and E. Percival, J . Chem. SOC., 1962, 3141; ( b ) E. Percivaland J. K. Wold, ibid., 1963, 5459.eD. A. Rees, J . Chem. SOC., 1961, 5168.7D. A. Rees, J . Chem. Soc., 1963, 1821470 B I0 L 0 G I C 9 L CHEMISTRYinfrared spectroscopy, 7, isolation of sugar sulphates after partial fragmenta-tion,’, and measurement of the rate of hydrolysis of the sulphate ester.’? 10Aqueous acid hydrolysis of sulphate esters occurs a t about the same rate asglycosidic hydrolysis, and a complicated mixture of products is therefore tobe expected in structure analysis of polysaccharide sulphates by straight -forward partial acid hydrolysis.Mild treatment with rnethanolic hydrogenchloride (a heterogeneous reaction) effects desulphation of many poly-saccharides without substantial glycosidic cleavage.ll This method isnormally applied in conjunction with methylation or periodate oxidation(see above). Less frequently used methods are acetylative desulphation 12and reductive removal of the mixed ester which can be prepared with diazo-methane.13 A simplification of the partial hydrolysis products is achieved bythe use of rnercaptolysis ** or acetolysis,15 because desuIphation oecurs muchfaster than the rate of depolymerisation under both these conditions.There are additional problems in the structure determination of carbo-hydrate sulphates from animal tissues, owing to their covalent combinationwith protein or lipid.The investigation of the linkages between glycosamino-glycan sulphates and protein is a t a particularly interesting and excitingstage, and a full report of recent work is given later. Sulphated-carbo-hydrate-protein complexes of other types have also been isolated fromanimal tissues, but these are less well characterised at present.The biosynthesis of carbohydrate sulphates would appear to be similarto other sulphate esters, the transfer of stdphate being from adenosine 3’-phosphate 5‘-sulphatophosphate (PAPS) .16 An improved synthesis ofthe analogue, adenosine 5’sulphatophosphate, involves the reaction of the1 -ethoxyvinyl derivative of barium 2-cyanoethyl sulphate with adenylicacid, and then removal of the protecting gr0ups.l’Methods.-The fmal volume of the “Methods in CarbohydrateChemistry ” series includes some sections on carbohydrate sulphates.lsBecause of the importance in medicine of polysaccharide sulphates fromanimal tissues, their identification and analysis have been much discussed.Methods have been described for detection and determination of polysac-charide sulphates in animal tissues and extracts,lg, 20 for their separation8 A.G. Lloyd, K. S. Dodgson, R. B. Price, and F. A. Rose, Bwchim. Biophys.Acta, 1961, 46, 108; A. G. Lloyd and K. S. Dodgson, ibid., p. 116.9 S . Suzuki and J. L. Strominger, J . Bid. Chem., 1960, 235, 2768; S. Suzuki,ibid., p.3580; J. R. Turvey and D. A. Rees, Nature, 1961, 189, 831.10 D. A. Rees, Biochem. J., 1963, 88, 343.11 R. Johnstone and E. G. V. Percival, J . Chem. Soc., 1950, 1994; T. G. Kantorand M. Schubert, J . Amer. Chern. SOC., 1987, 79, 152.12 M. L. Wolfrom and R. Montgomery, J . Amer. Claena. SOC., 1950, 72, 2859; T.Dillon and P. O’Colla, Proc. Roy. Irish Acad., 1951, 54, B, 51.13 G. Coleman, M. Higgs, A. Holt, and M. B%ulvin, Chewa. and Ind., 1963, 376.l4 A. N. O’Neill, J . Amer. Chem. Soc., 1955, ‘77, 6324.15K. Morgan and A. N. O’Neill, Cunad. J . Chem., 1959, 37, 1201.16 P. W. Robbins in “ The E n ~ p e s , ” eds. P. D. Boyer, H. Lardy, and K. Myrbiick,17 G. R. Banks and D. Cohen, J . Chem. Soc., 1965, 6209.18 “ Methods in Carbohydrate Chemistry,” vol.V, ed. R. L. Whistlsr, Academia19 S. A. Barker, C. N. D. Cruickshank, and T. Webb, Carbohydrate Em., 1965,80 M. Schmidt and A. Dmochowske, Bwchm. Biophys. Acta, 1964, 83, 137; C. A.Academic Press, New York, 1962, vol. 6, p. 383.Press, New York, 1965.1, 52REES : CARBOHYDEATE SULPHATES 47 1and detection by electrophoresis,21 and for the determination of their com-ponent sugars by enzymic,L2 colorimetric,23 and automatic colorimetric 24methods, and by gas ~hromatography.~5 Problems inherent in the deter-mination of hexosamines after hydrolysis in the presence of protein havebeen studied.26 There have been further refinements in Yaphe’s convenientmethod for the colorimetric determination of 3,6-anhydrogalactose in algalpolysaccharide sulphates.2’ This sugar can also be determined by polaro-graphy.28 The sulphur in polysaccharides can be determined by neutronactivation analysis.29The histochemistry of glycosaminoglycans has been re~iewcd.~O Methodsfor distinguishing different polysaccharide sulphates in tissue sections 31include selective suppression, using electrolytes, of their interaction withAlcian Blue dye,32 and microfractionation 33 and specific enzymolysis 34 inconjunction with histochemical studies. Glycosaminoglycan sulphates canbe stained for electron microscopy by the use of a colloidal iron reagent.35PolysacchaPide Sulphates born Animd Tissues.-These usually occur ascarbohydrateprotein macromolecules. A later section covers the carbo-hydrate-protein linkages ; the structure and metabolism of the carbohydratepart is discussed here.These are the best known and character-ised polysaccharide sulphates from animal tissues.The work of the pastdecade which has led to the determination of the main structural features ofthe polysaccharide chains, has been reviewed by M ~ i r , 3 ~ and by Brimacombeand Webber in their remarkably up-to-date b00k.~7 The overall picture issummarised in Table 1. Unfortunately, the nomenclature is not a t allClycosaminoglycan sulphates.Antonopoulos, S. Gardell, J. A. Szirmai, and E. R. de Tyssonslr, ibid., p. 1; D. S. Trundleand G. V. Mann, ibid., 1965, 101, 127; K. E. Kuettner and A. Lindenbaum, ibid.,p. 223; R. G. Brown, G. M. Button, and J. T. Smith, Analyt. Biochem., 1965, 12, 195;G.Manley, Nature, 1965, 206, 1253; J. S. Mayes and R. G. Hansen, Analyt. Biochem.,1965, 10, 15.21 J. H. Brookhart, J . Chromatog., 1965, 20, 191; S. Magnusson, Arlciv Kemi, 1965,24, 211; A. M. Saunders and S. Thomsen, Nature, 1965, 205,497; L. B. Jacques, R. E.Ballieux, and C. van Arkel, Fed. PTOC., 1964, 23, 459; S. Okhuma, T. Shinohara, andC. Miyauchi, Nature, 1965, 207, 527.23 0. Luderitz, D. A. R. Simmons, 0. Westphal, and J. L. Strominger, Analyt.Biochem., 1964, 9, 263; J. M. Sempere, C. Gancedo, and S. Asensio, ibid., 1965, 12,509.23 T. A. Good and S. P. Bessman, AnaZgt. Biochem., 1964, 9, 253.24 E. A. Balazs, K. 0. Bernsten, J. Karossa, and D. A. Swann, AnaZyt. Biochem.,25 M. B. Perry, Canad. J . Biochem., 1064, 42, 451.26 E.F. Hartree, Analyt. Biochem., 1964, 7, 103.27 W. Yaphe and G. P. Arsenault, Analyt. Biochem., 1965, 13, 143.28 T. Fujiwara, K. Morihara, and C. Araki, Bull. Chem. SOC. Japan, 1964, 37, 760.29 E. L. McCandless, AnaZyt. Biochem., 1964, 7, 357.30 R. C. Curran, Internat. Rev. Cytology, 1964, 17, 149.31 T. Sugiyama, Acta Path. Japon., 1964, 14, 413.32 J. E. Scott, J. Dorling, and C. Quintarelli, Biochem. J., 1964, 91, 4P.3 3 J. A. Szirmai, Biochem.. J . , 1964, 90, 1P.34 T. J. Leppi and P. J. Stoward, J . Histochem. Cytochem., 1965, 13, 406.35 R. C. Curran and A. E. Clark, Bwchem. J., 1964, 90, 2P; R. C. Curran, A. E.36 H. Muir, Internat. Rev. Connectizqe Tissue Res., 1964, 2, 101.37 J. S. Brimacornbe and J. M. Webber, “ Mucopolysaccharides,” Elsevier,1965, 12, 547, 559; D.A. Swann and E. A. Balazs, ibid., p. 565.Clark, and D. Lovell, J. Anat., 1965, 99, 427.Amsterdam, 1964TABLE 1 i i a i n structural features of glycosaminoglgcanRepeating unitChondroitin 4-sulphateChondroitin 6 -sulphateDematan sulphateKeratan sulphateHeparinHeparan sulphate( 1-t 4) -0 - P-D- (glucopyranosyluronic acid) -( 1+ 3) -2-acetamido-2 -deoxygalactopyranosyl 4-sulphate( 1 --t 4) -0 - P-D - (glucopyranosyluronic acid) - ( 1 --+ 3) -2 -acetamid0 -2 -deoxy(galactopyranosyl 6-aulphate)(1+4)-0- a-~-(idopyranosyluronic acid)-( l+ 3)-2-acetamido-2-deoxy-galactopyranosyl 4-aulphate( 1 4 4) -2 -acetamido-2 -deoxy -0- / l - ~ - (glucopyranosyl 6 -sulphate) -( 1 --f 3)galactop yranos y 1( 1 --f 4) -0 - a-D - (glucopyranosyluronic acid 2 -8ulphate) - ( 1 + 4) - 2 -deoxy-2-amino-0-a-D-glucopyranosyl 6-sulphateSulphated polymer of glucosamine and gliicuronic acid with a structurehas features in common with heparin.It differs in degree of sulphation,that some glucosamine units are N-acetylated, and perhaps also inrespects.~ o t e . These structures are idealised. Different preparations may not always be sulphated as fully or ar, regularlyIn most eases it has yet to be definitely of sugar units near the linkage to protein in all cases SO far examined.featuresREES : CARBOHYDRATE SULPHATES 473s tandardised. Jeanloz's suggestions 38 will be followed here, and the readeris referred to his article 38 for a useful '' Dictionary " which might help inrelating this Report to other literature.The essentially 1 4 linked structure of heparin, which was shown bypartial hydrolysis of the carboxyl-reduced desulphated polysa~charide,~~ andby methylation of heparin 40 and its carboxyl-reduced N-acetylated desul-phated derivative,41 has been further confirmed by partial hydrolysis of3'-acetylated desulphated he~arin.4~ It seems that earlier 43 indicationsof 1+6 linkages might have been reversion artefacts.Synthetic 0-cc-D-glucopyranosyl-( 1+4)-2-amino-2-deoxy- cc-r>-glucose is identical with aproduct from carboxyl-reduced desulphated heparin, in confirmation ofthe a - 1 4 linkage between the glucuronic acid and glucosamine units inhe~arin.~4 The heparins from dog, beef, hog, sheep, and human tissues arebiologically, chemically, and physically similar ;45 analytical variations mightbe due to variable proportions of non-covalently bound sulphate 46 and otherimpurities. Glucosamine units in which the amino-groups are sulphated orunsubstituted, are converted into 2,5--anhydromannose by nitrous aciddeamination, with cleavage of the adjauent glycosidic link.Under carefullycontrolled conditions, only N-sulphated units react .47 This has been usedt o show that heparan sulphate from the livers of patients with the Hurlersyndrome (see p. 478) is heterogeneous with respect to N-sulphate content .48A polysaccharide of the heparin-heparsn sulphate type has been isolatedfrom whale organs. It contains a proportion of N-acetylated glucosamineunits, which would appear from the nature of the deamination products tooccur together in A crude heparinase from a Flaaobacterium hasbeen fractionated into a glucuronidase and an eliminase.The latter enzymecleaves heparin and heparan sulphate (but not desulphated heparin) in themanner that is now well known for many other uronic-acid containingpolysaccharides, with the formation of oligosaccharides with 4,5-unsaturateduronic acid end-gro~ps.~~ Comparison of the products from the eliminaseaction on heparin and heparan sulphate confirms that these polysaccharidesare to some extent structurally similar and promises to show whether thereare any differences other than in the pattern of sulphation and N-acetylation.38 R. W. Jeanloz, Arthritis and Rhewnatisnz, 1960, 3, 233.39 M.L. Wolfrom, J. R. Vercellotti, and D. Horton, J . Org. Chem., 1964, 29, 540,4 0 0. NominB, R. Rucourt, and D. Bertin, Bull. SOC. chim. France, 1961, 561.41 I. Danishefsky, H. B. Eiber, and A. H. Williams, J. Biol. Chem., 1963, 238,2895; RI. L. Wolfrom, J. R. Yercellotti, and D. Horton, J . Org. Chem., 1964, 29, 547.4 2 I. Danishefsky and H. Steiner, Biochem. Biophys. Acta. 1965, 101, 37.4 3 I. Danishefsky, H. B. Eiber, and E. Langholtz, Biocheni. Biophys. Bes. Coinin.,1960, 3, 571; J. A. Cifonelli and A. Dorfman, ibid., 1961, 4, 328.41 M. L. Wolfrom, H. El Khadem, and J. R. Vercellotti, J . Org. Chem., 1961, 29,3284.4 5 G. H. Barlow, L. J. Coen, and M. M. Mozen, Biochem. Biophys. Acta, 1964,46 J. R. Helbert and M.A. Marini, Biochim. Biophys. Acta, 1964, 83, 122.4 7 J. A. Cifonelli, Fed. Proc., 1965, 24, 354; compare D. Lagunoff and G . Warren,Arch. Biochem. B i o p h y ~ . , 1962, 99, 396; A. B. Foster, E. F. Martlew, and M. Stacey,Chem. and Id., 1953, 825.48 J. Knecht and A. Dorfman, Biochem. Biophys. Res. Comm., 1965, 21, 509.4 9 Z . Yosozawa, Biochein. Biophys. Res. Comm., 1964, 16, 336.so A. Linker and P. Hovingh, J . Biol. Chem., 1965, 240, 3724.Qand earlier Papers.83, 272474 BI 0 LO GI C AL CHEMI S TRYKeratan sulphate from bovine cornea can be separated on the basis ofethanol solubility into two sub-fractions 51 which differ, as do various samplesof corneal keratan ~ u l p h a t e , ~ ~ in the ratio of sulphate to hexosamine and insome physical propertiea.Keratan sulphates from skeletal tissues alsoshow variations.52, 53 There are, however, more profound differences be-tween the preparations from the different types of tissue, particularly in theassociated peptide and in a t least some of the carbohydrate-peptide linkages,but also in other respects.52, 53 It is still an open question whether thereare fundamental differences in the molecular structures of the polysaccharidechains. The accepted structure for keratan sulphate (Table 1) was estab-lished using a preparation from Significant amounts of galactos-amine appear to be combined in skeletal keratan sulphates, in a form that isnoti removed by fractionation or treatment with hyaluronidase and that isnot associated with a molar equivalent of glucuronic acid.This has ledto the suggestion that galactosamine is a molecular component of skeletallreratan sulphate . 52There have been few developments in the understanding of biosynthesisof the golysaccharide chains since the full account of this topic last year.=The possible biosynthetic intermediates chondrosine 1 -phosphate and uridinediphosphate chondrosine have been synthesised.55 A compound that appearsto be a sulphated cytidine monophosphate has been isolated from rat skinand might be involved in polysaccharide sulphate bios~nthesis.~6 Dermatansulphate has a less broad molecular-weight distribution than would beexpected if it were synthesised by random p~lymerisation.~~ Glucosaminecan be utilised locally for glycosaminoglycan synthesis a t the sites of con-nective-tissue formation in rats.5s Recent work on the enzymic degradationof glycosaminoglycans suggests that many of the degradative enzymes inanimal cells might be located in lysosomes.5* These cytoplasmic bodiesappear to contain many other hydrolytic enzymes that are presumablyinvolved in intracellular digestion.GOSome information about polysaccharide sulphate conformations has beenobtained by physical methods.An X-ray study of oriented films of a sub-fraction of chondroitin 4-sulphate containing only one sulphate for everythree disaccharide units, suggests that the sulphate distribution is regularand that the polysaccharide chain is more flexible than in hyaluronic acidbecause the repeating period is shorter.61 The similarity in optical rotatory6 1 B. Wortman, Biochem.Biophys. Acta, 1964, 83, 288.5% M. B. Mathews and J. A. Cifonelli, J . BioZ. Chem., 1968, 240, 4140.53 N. %no, K. Meyer, B. Anderson, and P. Hoffman, J. BioZ. Chem., 1966,240,1005.54 P. T. Grant and J. L. Simkin, Ann. Reports, 1964, 61, 491.6 5 A. H. Olsvesen and E. A. Davidson, J. BioZ. Chem., 1966, 240, 992.58 S. A. Barker, C. N. D. Cruickshank, and T. Webb, Carbohydrate Bee., 1965,1, 62.57 C. Tanford, E. Marler, E. Jury, and E. A. Davidson, J . BbE. Chem., 1964, 239,4034.68 B. N. White, M. R. Shetlar, H. M. Shurley, and J. A. Schilling, Bhchem. Biophys.Acta, 1965, 101, 97.6 8 N. A. Aronson and E. A. Davidson, J . BWZ. Chem., 1965,240, PC3222; F.Eutterer,Fed. Proc., 1965, 24, 557; D. Fisher, P. W. Kent, and P. Pritchard, Biochem. J., 1966,98, 46P.6 O C. de Duve, Fed. Proc., 1964, 23, 1045.@IF. A. Bettelheim, Biochim. Biophys. Ado, 1964, 83, 350REES : CARBOHYDRATB SULPRATES 475dispersion curves between certain dermatan sulphate and chondroitinsulphate derivatives, is interpreted to mean that the L-iduronic acid Unitin dermatan sulphate exists in the Reeves C-l conformation, with thecarboxyl group axial.62 The absorption characteristics of the metachromaticcomplexes formed by Toluidine Blue with hyaluronio acid and with chon-droitin 6-sulphate suggest that the polysaccharide molecules have rigidc o ~ o ~ a t i o ~ in the complexes, in which the carboxyl groups are closetogether .63 Interactions between ~ e i g h b o ~ ~ g dye molecules bound toheparin give rise to optical rotation, similar to the effects known for poly-peptides and polynucleotides.6~Only brief mention is possible of recently published work on the naturaloccurrence and distribution of glycosaminoglycan sulphates.A dis-cussion of the r6le of mast cells in storing, synthesising, and dischargingheparin into the circulation, is contained in a recent book.G5 Non-sulphatedchondroitin has been isolated from squid skin.6s The glycosaminoglycansulphates from the following sources have been inyestigated : urine,"?fracture callus,68 rat E d n e ~ s , ~ ~ rat skin,f9 aortic tissue of the aortasof swine raised on a copper-deficient diet," cattle retina,72 sclera from bovineeyes,73 tadpole tail fin and back skin,'* human heart valves,T5 human um-bilical human teeth,77 human fetal ~kin,7~ and avian oviduct,egg, skin, and comb.79Other carbohydrate sulphates from aninacrE tissues, A rare example of anaturally occurring sdphate of a reducing oligosaccharide, is the trisacchar-ide derivative present in water extracts of lactating-rat mammary glands.It has been assigned the structure O-or-~-acet~~euraminyl-(2-+3)-0-B-~-(galactopyranosyl tj-sulphate)-( 1 -+4)-~-glucose, on the basis of its hydrolysisby neuraminidase, the resistance of the galactose unit to periodate oxidation,and the identification of lactose and galactose 6-sulphate as partial hydrolysisproducts. so The desulphated derivative occurs in urine.The relateddisaccharide sulphate, O-/~-D-( ga~actop~ano~yl 6-sulphate)-( 1 +t)-D-g~UCOSe,62 E.A. Davidson, Biochim. Biophys. Acta, 1965, 101, 121.esM. D. Schoenberg and R. D. Moore, Biochim. Bwphgs. Ada, 1964, 88, 42.8 p A. L. Stone, Fed. Proc., 1964, 23, 282.65 H. Selye, " The Mast Cells," Butterworths, Washington, 1965.K. Anno, Y. Kawai, and N. Seno, BiOChim. Bivphgs. Acta, 1964, 83, 348.67 6, S. Berenson and E. R. Dalferes, Biochim. Biophgs. Actu, 1965, 101, 183.8s C. A. Antonopoulos, B. Engfeldt, S. Gaxdell, S . - 0 . Eljertquist, rmd K. Solheirn,6s D. Allalouf, A. Ber, and N. Sharon, Biochim. Biophys. Acta, 1964, 83, 278.7o 0, V. Sirek, 8. Schiller, aad A. Dorfman, Biochim. Bioph?/s. Acta, 1964,88, 148.71 A. Linker, W. F. Coulson, and W.H. Cames, J. BioZ. Chem., 1964, 239, 1690.7 2 E. R. Berman, Bbchim. Biophys. Ada, 1965, 101, 358.78 J. Rotstein and J. Seltzer, Bioclzim. Biophys. Acta, 1965, 101, 273.7 4 M. J. Lipson and J. E. Silbert, Bbchim. Biophy#. Acta, 1965, 101, 279.7f, S. Torii, R, I. Bashey, and K, Nakao, Biochim. Biophys. Ada, 1965, 101, 285.7 s I . Danishefsky and A. Bella, Fed. Proc., 1965, 24, 355.?7 3%. D. Clark, J. C. Smith, and E. A. Davidson, Biochim. Bwphys. Acta, 1965,78 J. G. Smith, R. D. Clark, and E. A. Davidson, Fed. Proc., 1965, 24, 558.79 D. F. Wood and P. A. Anastassiadis, Curd. J. Biochem., 1965, 43, 1839.81 J. K. Huttunen and T. A. Miettinen, Acta Chem. Xcarzd., 1965, 18, 1486.Biochirn. Biophys. Actcc, 1965, 101, 160.101, 267.L. C. Ryan, R.Carubelli, R. Caputto, ar,d R. E. Trucco, Biochim. Biophys.Ada, 1965, 101, 252476 BIOLOUICAL CHEMISTRY(lactose sulphate), has also been isolated from lactating-rat mammaryglands.82Other carbohydrate sulphates of mammalian origin are incompletelycharacterised a t present, but enough is known to indicate that some havenew types of structure. A mixture of products from human brain has beenfractionated into a sulphated xylose-rich component and a less acidic ara-binose-containing component. Both contained polypeptide, hexosamine,hexuronic acids, and other sugars. 83 Polysaccharide sulphates from humangastric juice and the gastric wall of the dog gave galactose, fucose, glucos-amine, galactosamine, and sialic acids on hydrolysis. Hyaluronic acid anddermatan sulphate were also present in the latter source.84 Bovine lenscapsule contains a sulphated polysaccharide which has units of galactose,mannose, glucosamine, fucose, and a neuraminic acid derivative.85 Thesulphated glycoproteins from sheep colonic mucin 86 and dog submaxillarygland 87 are particularly interesting because of their possible similarity toknown glycoproteins in function and in general structure.Molluscan polysaccharide sulphates and the associated enzymes, parti-cularly those from the marine gastropod Chronia kcmpas, have been studiedby Japanese workers for many years.A glucan sulphate (named charonin-sulphuric acid) from the mucous glands of this organism has been fraction-ated into (i) a sulphate-poor fraction which is partly hydrolysed by a- and/3-amylases and which gives an iodine coloration similar to glycogen, and(ii) a sulphate-rich fraction which contains ~-1A-linked glucose units.88The sulphate-rich fraction has now been shown to act as an acceptor insulphate transfer from PAPS by an enzyme extract from the mucous glands.89A sulphated glycoprotein, for which the name horatinsulphuric acid is pro-posed, has been isolated from the liver of the same organism.g0 The hypo-branchial gland of the whelk Buccinum undatum L.secretes a mucin whichcontains a glucan sulphate apparently bound covalently to peptide and inloose association with glycoprotein. 91 Sulphsted polysaccharide is similarlysecreted by the hypobranchial gland of another mollusc, Neptunia antiqua ;the component sugars in this case are glucosamine, galactosamine, glucose,and fucose.92 The jelly coat of starfish eggs contain a complex sulphatedpolysaccharide which has galactose and fucose as the major sugar units.93The Linkage of Glycosaminoglycan Sulphates to Protein.-Further evi-8 2 H.8. Barra and R. Caputto, Biochim. Biophys. Acta, 1965, 101, 367.83 Z. Stary, A. H. Wardi, D. L. Turner, and W. S. Allen, Arch. Biochem. Biophys.,84 I. Hiikkinen, K. Hartiala, and T. Terho, Acta Chem. Scand., 1965, 19, 797, 800.8 8 Z. Dische and G. Zelmenis, Fed. Proc., 1964, 23, 319.86 P. W. Kent and J. C. Marsden, Biochem. J., 1963, 87, 38P; P. W. Kent, ibid.,8 7 C. Bignardi, C. Aureli, G. Balduini, and A. A. Castellani, Biochem. Biophys.85 K , Iida, J .Biochem. Japan, 1963, 54, 181; F. Egami, T. Asahi, N. Takahashi,8 9 H. Yoshida and F. Egami, J . Biochem. (Japan), 1965, 57, 215.90 S. Inoue, Biochim. Biophys. Acta, 1965, 101, 16.g1 S. Hunt and F. R. Jevons, Biochim. Bwphys. Acta, 1965, 101, 214; Biochem. J.,Q2 J. Doyle, Biochem. J., 1964, 91, 6P.93 T. Muramatsu, J . Biochem. (Japan), 1965, 57, 223.1965, 110, 388.1964, 90, 1P.Res. Comm., 1964, 17, 310.S . Suzuki, S. Shikata, and K. Nishizawa, Bull. Chem. SOC. Japan, 1955, 28, 685.1966, 98, 522REES : CARBOHYDRATE SULPHATES 477dence to that discussed in last year’s Report,54 points to the existence ofglycosidic linkages between several polysaccharides and hydroxyamino-acidresidues in the protein. The amino-acid units that remained attached tochondroitin 4- and 6-sulphates after digestion of the whole tissue withproteolytic enzymes, were liberated in cold alkali with formation of a-aminoacrylic acid units that were characterised chromatographically asalanine after hydrogenation.This suggests that the amino-acid-containingmaterial had been linked through the hydroxyl group of serine, and wasreleased by #?-elimination. 94 The possibility that it was released by hydro-lysis of an ester link to glucoronic acid was excluded by the failure of thepolysaccharide-protein complex to form a non-dialysable hydra~ide,~~ aswell as by earlier evidence. Three distinct glycosaminoglycan-protein sub-fractions, each of which contains part of the total chondroitin sulphate,are recognisable in bovine nasal cartilage.Two are essentially insolubleand give a product that is very similar to the third (soluble) sub-fractionon very mild treatment with aqueous hydroxylamine or an extraction withpotassium ~yanide.~5 The soluble sub-fraction is electrophoretically hetero-g e n e o u ~ . ~ ~ In an approach to the characterisation of the chondroitin4-sulphate-protein linkage in the soluble sub-fraction, acid hydrolysis ofglycopeptide fragments produced by successive action of testicular hyalu-ronidase and papain gave substances with the properties expected of 0-xylopyranosylserine, O-/3-D-galactopyranosyl- ( 1 +3)-O-~ylopyranosylserine 7and 0-B-D- (glucopyranosyluronic acid)-( 1 -+3)-~-galactose.~8 Together withanalytical data for the glycopeptides, this suggests that the sequence ofunits in the region of the linkage to protein is D-glucuronic acid+2-acet-smido-2-deoxy-~-galactose 4-sulphate -+D-glucuronic acid +D-galactose +D-galactose+xylose-+serine.The model compound O-P-D-xylopyranosyl-L-serine has been synthe~ised,~~ and a xylosylserine has been isolated fromhuman urine.loO Analysis for the N-terminal amino-acids of the residualpeptide chains attached to chondroitin 6-sulphate isolated from sharkcartilage after treatment with proteolytic enzymes, gave no evidence for asingle type of chain.94 Sialic acid is present in connective tissue, and someof it would appear t.0 be bound to the chondroitin sulphate-protein complex,although the exact location is unknown.101The conclusion that chondroitin sulphate is joined to protein through a‘‘ linkage region ” containing sugar units which do not otherwise occur int’he polysaccharide, raises some interesting questions about biosynthesis.The synthesis of chondroitin sulphate by minced embryonic chicken cartilageis inhibited by puromycin, whether the synthesis is measured by incorpor-94 B.Anderson, P. Hoffman, and K. Meyer, J . BioZ. Chem., 1965, 240, 156.95 S. Pal and M. Schubert, J . Biol. Chern., 1965, 240, 3245.g6 T. A. Mashburn, P. Hoffman, B. Anderson, and K. Meyer, Fed. Proc., 1965,97 H. D. Gregory, T. C. Laurent, and L. Rodh, J . Biol. Chem., 1964, 239, 3312;g8 L. RodBn, Fed. PTOC., 1964, 23, 484; L. Roden and 0. Armand, J . Biol. Chern.,sQ B. Lindberg and B.-G.Silvander, Acta Chem. Xcand., 1965, 19, 530.24, 606.L. Roden and U. Lindahl, Fed. Proc., 1965, 24, 606.1966, 241, 65.loo F. Tominaga, K. Oka, and H. Yoshida, J . Biochern. (Japan), 1965, 57, 717.lol A. J. Anderson, Biochem. J . , 1962, 82, 372478 BIOLOGICAL CHEMISTRYation of labelled serine, acetate, or sulphate into the chondroitin sulphatefraction isolated after papain digestion. Puromycin probably interfereswith the synthesis of the protein part of the chondroitin sulphate-proteincomplex. This and other evidence suggests that the biosynthesis involvesaddition of carbohydrate units to preformed protein.102Heparin prepared under mild conditions contains combined amino-acids, with serine as the main component; galactose and xylose units are alsopresent, even after exhaustive purification.lo3 Only traces of amino-acidsare present after more drastic isolation conditions.Partial acid hydrolysisgave products with the properties expected of 0-D-xylosylserine, O-D-galactosyl-0-D-xylosylserine, and glucuronosylgalactose, suggesting that thecarbohydrate-protein linkage is similar to that in the chondroitin sulphates.Serine is a major amino-acid in a heparan sulphate-peptide fraction fromhuman aorta, suggesting that heparan sulphate also occurs bound to proteinthrough a serine unit.lW A similar fraction has been isolated from the liversof patients with the Hurler syndrome, a hereditable disease of connectivetissue (gargoylism) .lo5Dermatan sulphate would also appear to occur normally in the form ofa protein complex.1o6, It can be liberated by treatment with alkalior by exhaustive proteolytic digestion.From the tissues and urine ofpatients with the Hurler syndrome, however, this polysaccharide can bereadily isolated almost free from protein. This suggests that there is ametabolic defect associated with this disease, which results in the failureof dermatan sulphate to be fixed normally in connective tissue.48, lo' Therewould seem to be a similar defect in the linkage of heparan sulphate to proteinin the liver.lo5Recent work suggests that keratan sulphate is linked to protein indifferent ways in different types of tissue (the possibility that the polysac-charide chain might differ in structure from one source to another, has beenmentioned above).After treatment with proteolytic enzymes, differentsamples of keratan sulphate from skeletal tissues usually show similar amino-acid patterns with high contents of glutamic acid, proline, and threonine,whereas aspartic acid is the major amino-acid combined in corneal keratans ~ l p h a t e . ~ ~ , 53 Treatment with alkali has been used to obtain informationabout the carbohydrate-protein linkages,63 as in studies on chondroitinsulphates. Elimination of the 6-sulphate from 1 +&linked glucosamine units(compare refs. 6 and 7) is probably only a minor side-reaction under the mildconditions that are used. The destruction of both serine and threonineunits in skeletal keratan sulphate, with the formation of products that couldbe converted by hydrogenation into alanine and a-aminobutyric acid, in-dicated that some serine and threonine hydroxyl groups were substituted.l o a A.Telser, H. C. Robinson, and A. Dorfman, Proc. Nat. Acad. Sci. U.S.A., 1965,103 U. Lindahl, J. A. Cifonelli, B. Lindahl, and L. Rodbn, J . Biol. Chem., 1965,loC 8. Jacobs and H. Muir, Biochem. J., 1963, 87, 38P.105 J. C. Knecht and A. Dorfman, Fed. Proc., 1965, 24, 606.106B. P. Toole and D. A. Lowther, Biochim. Biophp. A&, 1965, 101, 361.107 A. Dorfman, Biophys. J . , 1964, 4, 155.54, 912.240, 2817; U. Lindahl and L. Roden, ibid., p. 2821REES : CARBOHYDRATE SULPHA‘I‘ES 47 9The production of reducing sugar groups in the reaction suggested that thesubstituents were glycosidic.However, some peptide still remained boundto the polysaccharide, and other types of carbohydrate-peptide linkage aretherefore likely to be present. The relative importance and relative r6lesof these different types is undecided. In contrast, the amino-acid contentof corneal keratan sulphate was not decreased by alkali treatment. If allthe polysaccharide chains of corneal keratan sulphate were linked to peptide,and bylinkages of a single type, analysis after papain treatment andextensivepurification would indicate that aspartic acid is the only amino-acid presentin sufficient amount to give the necessary number of attachment points.58There is some chemical evidence consistent with the presence of glycosyl-amine linkages to asparagine in corneal keratan ~ulphate.~~ For skeletalkeratan sulphate, similar arguments from analytical data 52 suggest thatthreonine, proline, or glutamic acid, but not aspartic acid, could providethe attachment points.Non- covalent interactions between glycosaminoglycans and protein arealso important from a biological point of view.Such interactions withcollagen, which are probably largely electrostatic, might have a r61e in theorganisation of connective tissue. Free-solution electrophoresis has shownan apparent gain in stability of the collagen complex with increasing mole-cular weight of the polysaccharide, suggesting a parallel alignment of chainsin the complex. The polysaccharide side-chains of the chondroitin sulphate-non-collagenous protein macromolecule might therefore lie parallel withcollagen in connective tissue.lo8 Chondroitin sulphate-protein macro-molecules also complex with plasma proteins and with glycoproteins ofsufficiently low sialic-acid content. These effects are also explained in termsof electrostatic interactions.logPolssaccharide Sulphates from Seaweeds.-Most classes of marine algaeseem to contain polysaccharide s u l p h a t e ~ .~ ~ ~ ~ The discussion here willbe confined to recent progress in the polysaccharide sulphates from theRhodophyceae (red seaweeds) and the Chlorophyceae (green seaweeds) be-cause it is with these that most recent work has been done. Other types ofpolysaccharide sulphate have been isolated from the Phaeophyceae (brownseaweeds),llO and from a marine diatom.ll1Red seaweeds.These sulphated polysaccharides are usually galactansof a well-defined structural type, which nevertheless have many remarkablediEerences.l12 The molecule commonly has galactose units linked a-l+3and ,!I-1 4 in alternating sequence. Among the variations are the presenceof the 4-linked galactose units as the D- or the L-enantiomorph, as the 3,6-anhydride of either enantiomorph, or as the 6-sulphate or 2,6-disulphate.The 3-linked unit is usually D-galactose which may be 2- or 4-sulphated, orlo* M. B. Mathews, Biochem. J., 1965, 96, 710.loo A. J. Anderson, Biochem. J . , 1965, 94, 401; ibid., 1965, 97, 333.logs S. Peat and J. R. Turvey, Portschr. Chem. org. Natwrstofle, 1965, 23, 1.R. H. CBt6, J . Chem. Soc., 1959, 2248, and references given there; B.Larsen,A. Haug, and T. J. Painter, in “Proceedings of the V International Seaweed Sym-posium,” eds, E. G. Young and J. L. Mclachlan, Pergamon, 1966, p. 287.ll1 C. W. Ford and E. Percivd, J . Chem. SOC., 1965, 7042.112 N. S. Anderson, T. C. S. D o h , and D. A. Rees, Nature, 1965, 205, 1060; andreferences given there480 BIOLOGICAL CHEMISTRY6-O-methylated. Some of the known polysaccharides of this type areshown in Table 2. There are some polysaccharides, such as the mucilageof DiEsea edulis (= D. carnosa),6, lI3 which might not fit this pattern. Forseveral of these polysaccharides, quantitative comparison of the productsof partial fragmentation of the suitably modified polysaccharides with thosefrom disaccharide 115 together with other evidence,114 has shownthat there can be little if any deviation from a strictly alternating arrange-ment of 1+3 and 1 4 linkages. This alternating arrangement of thesugar units may however be masked by units within the 3-linked series andwithin the 4-linked series being present as different derivatives.For ex-ample, different samples of porphyran (1) have the component units (D-galactose, 6-O-methyl-~-galactose, 3,6-anhydro-~-galactose, and L-galactose6-sulphate) present in widely varying proportions, and the alternating struc-ture is thus overlaid by substitution and modification. The alternatingCH2OH H CHzOMe H0‘ H OM? H OMe ‘ “-‘O H(2)struct,ure was proved by conversion into niethylated agarose (2) by alkalineelimination of the 6-sulphate followed by methy1ati0n.l~~ The structureof agmose, its methylated derivative, and various fragments derived fromeach, had previously been established by Araki’s classical work.l16The carrageenans are examples in which the 4-linked units are present113 1’.C. Barry and J. E. McCormick, J. Chem. Soc., 1957, 2777.114 N. S. Anderson and D. A. Rees, J . Chem. SOC., 1965, 5880.115 N. S. Anderson and D. A. Reesin in“ Proceedings of the V International SeaweedSymposium,’’ eds. E. G. Young and J. L. McLachlan, Pergamon, 1966, p. 243.1l6 C. Araki and S . Hirase, Bull. Chem. SOC. Japan, 1960, 33, 597, and referencesgiven there: see also G. Araki in “ Proceedings of the V International Seaweed Sym-posium,” eds.E. G. Young and J. L. McLachlan, Pergamon, 1966, p. 3AgaroseTABLE 2 Structural variations on the thewe [P-Galp-(l--+ 4)-ac-Galp-( 1 --+ 3)Derivatives of the Derivatives of the 4-3-linked galactose iiriit galactose unitD-Galactose 3,6-Anhydro-L-galactosePorphyrenK- Carrageenann -Galactose, 6-0 -meth yl-D -galactosen - Gnlac t osc 4 -sulphat,eA-Carrageenan D -Galactose ;Third component of D Galactose ; n-galactoseChondrus carrageenan 4-sulphateFur cellarnn D-Galactose (some as branchpoints through position 6);D -galactose 4-sulphateD -Galactose 2 -sulphate3, 6-Anhydro an galactose ; L-galactose3,6-Anhydro-~ -galactose ; D -galactosegalactose 2,6-disulphate ; (probably)galactose 2-sulphateD-Galactose 2,6-disulphateD -Galactose 6 -sulphate ; (probably)galactose ; (perhaps) D-galactose3,G-anhydro -D -galactose 2 -sulphate3,6-Anhydro-u-galactose ; D-galactoseD -galactose 2,6-disulphate ; (perhaps)galactose 2-sulphateNotes.References are givcn in the test. There are probably many other polysaccharides of this type, thoughthose in this Table (see ref. 112 for discussion of this point)482 BIOLOGICAL CHEMISTRYas D-enantiomorphs. These polysaccharides are usually obtained byfractionation of Chondrus crispus extracts, but it is probable that there arestructurally similar polysaccharides in some other species. Methylationanalysis of hamageenan, before and after desulphation, established thatthe 3-linked units are mainly 2-sulphated with some being 4-sulphated andsome non-sulphated, and that the 4-linked units occur mainly as D-galaCtOSe2,6-di~ulphate.~~ After alkali-catalysed conversion of the 4-linked unitsinto 3,6-anhydrogalactose (mostly in the form of its 2 sulphate), the productwas separated into two fractions.All the 3-linked galactose 2-sulphatewas present in one fraction (“ alkali-modified 1-carrageenan ”), and all the4-sulphate in the other ( ‘ I alkali-modified third component of Chondruscarrageenan ”). The 4-linked units are less extensively 2-sulphated in thethird component derivative. Both fractions contain non-sulphated 3-linkedunits, and both have structures corresponding closely to a perfect alternationof 1 4 and 1 4 linkages.l15, 117 It is suspected that the variable propor-tions of 3,6-anhydrogalactose often found in crude A-carrageenan 7,118 isassociated with the native third component.K-Carrageenan also has aperfect, or near-perfect, alternating structure.ll5 Methylation l 1 7 showsthat the 3-linked units occur virtually completely as D-galactose 4-sulphatewhereas the 4-linked units occur as 3,6-anhydro-D-galactose and probablyto some extent as its 2-sulphate. The presence of relatively minor amountsof 4-linked galactose 6-sulphate and 2,g-disulphate has been shown byalkaline elimination of the 6-sulphate coupled with the use of periodateoxidation and partial fragmentation.ll5 Part of the K- carrageenan moleculeis resistant to bacterial K-carrageenase, but alkali treatment of the residue,which results in a rise in the proportion of 3,6-anhydrogalactose7 rendersit susceptible.11s This behaviour is readily explained in terms of the structureput forward.Further conhation of this structure is the paper-chromato-graphic identification of the expected monosaccharide sulphates after partialhydrolysis.l20 A similar polysaccharide to K-carrageenan, but which differsin being branched and less sulphated, is extracted from Furcellaria fmti-giata.112~ 121 The recent confirmation 122 of the structure of carrabiose (3,6-anhydro-4-O-/3-~-galactopyranosyl-~-galactose) 123 means that the 1 4 -linkage in the carrageenan family of polysaccharides is more firmly estab-lished. A painstaking survey has provided very useful information aboutthe distribution and variation of carrageenan-type polysaccharides.ll* Thedegree of the variation suggests that the carageenans are very variable withrespect to their content of particular structural features.It is possible thatfurther distinct fractions will be recognised in the future.11’ T. C. S. Dolan, Ph.D. Thesis, Edinburgh, 1965.118 W. A. P. Black, W. R. Blakemore, J. A. Colquhoun, and E. T. Dewar, J. Sci.Food Agric., 1965, 16, 573.llS J. Weigl, J. R. Turvey, and W. Yaphe in “ Proceedings of the V InternationalSeaweed Symposium,” eds. E. G. Young and J. L. McLachlan, Pergamon, 1966, p. 329;J. Weigl and W. Yaphe:‘Canczd. J . Microbiol.. in the press.leo T. J. Painter, in Proceedings of the V International Seaweed Symposium,”eds. E. G. Young and J. L. McLachlan, Pergamon, 1966, p.305.121 J. Christensen and A. Haug, personal communication.la2 T. J. Painter, J. Chem. Soc., 1964, 1396.12* C. Araki and S. Hirase, Bull. Chem. SOC. Japan, 1956, 29, 770REES : CARBOHYDRATE SULPHATES 483Only limited information is available on the biosynthesis of red seaweedgalactan sulphates. The nucleotides present in Porphyra perforata, an algafrom which porphyran (1) may be extracted, include uridine diphosphateD-galactose and guanosine diphosphate ~-galactose.l~* The latter may arisefrom guanosine diphosphate D-mannose, which was also isolated, becausetransformations similar in some respects to those required for this con-version are known to occur elsewhere. Porphyra extracts contain an enzymewhich catalyses the conversion of L-galactose 6-sulphate units into 3,6-anhydro-L-galactose a t the polysaccharide level.125 Porphyran is a muchbetter substrate than oligosaccharides containing L-galactose 6-sulphateunits.These facts, together with the variation in porphyran compositionwhich is very wide indeed, but which apparently occurs without abolishingthe perfect alternation of D- and L-units, suggest that 3,6-anhydride form-ation and perhaps also 6-O-methylation occur after formation of the poly-saccharide chain.126 The most likely route for porphyran biosynthesiswould therefore be synthesis of a DL-galactan from uridine diphosphateD-galactose and guanosine diphosphate L-galactose, perhaps by way of anintermediate disaccharide derivative,l26a followed by sulphation, partialmethylation, and 3,6-anhydride formation from some of the sulphatedunits. The last three steps could, of course, occur in a different order.Itmay be noted that nucleotide derivatives of 3,6-anhydrogalactose and 6-O-methylgalactose were not detected in P. ~erforata.1~~ Attempts to demon-strate the enzymic transfer of methyl groups from X-adenosylmethionine toporphyran by Porphyra extracts have met with no In Lomentariabaileyana, there is evidence that the Golgi apparatus has a function in thesynthesis and deposition of a substance into the cell walls that might be agalactan sulphate.128The close structural relationships between the three co-occurring carra-geenans suggest that they are interconverted to some extent in &vo, or thatthey have a common precursor such as a galactan or a galactan 6-sulphatewhich gives rise to the known polysaccharides by further sulphation and, inthe formation of K-carrageenan only, 3,6-anhydride formation.In con-nexion with the first possibility, the interconversion of K- and A-carrageenanswould require extensive rearrangement of sulphate groups in addition toother changes, and the conversion of either of these into the ‘‘ third com-ponent ” would require rearrangement of sulphate groups or opening of the3,6-anhydro-ring. It is therefore perhaps likely that both K-carrageenan andA-carrageenan are (relatively) end-products of biosynthesis, which might,however, undergo metabolic alteration in wivo to a small extent. Never-theless, the structure of the ‘‘ third component ” suggests that it might verywell be a precursor of K-carrageenan. An earlier suggestion 7 that A- might1 2 p J.C. Su and W. Z. Haasid, Biochemistry, 1962, 1, 474.125 D. A. Rees, Biochem. J., 1961, 81, 347.lZ6 D. A. Rees and E. Conway, Biochem. J., 1962, 84, 411.Compare J. M. Weiner, T. Higuchi, L. Rothfmld, M. Saltmarsh-Andrew, M. J.Osborn, and B. L. Horecker, PTOC. Nut. Acad. Sci. U.S.A., 1965, 54, 228; A. Wright,M. Dankert, and P. W. Robbins, {bid., p. 235.127 D. A. Rees, unpublished work.l a s G. B. Bouck, J . Cell Biol., 1962, 12, 553484 BIOLOGICAL CHEMISTRYbe a precursor of K-carrageenan was made before the distinction betweenA-carrageenan and the third component became clear.Green seaweeds. Polysaccharide sulphates from these sources are complexand no overall picture is yet available of their structures.There has, how-ever, been substantial recent progress in locating the sulphate esters. Methy-lation 129 has shown that the sulphated polysaccharide from Enteromorpha(probably representative of a group of polysaccharides all composed mainlyof xylose, rhamnose, and glucuronic acid units, as well as having other featuresin common 5 9 I3O) has a complicated structure, because the neutral sugarunits each occur in several different structural situations. The sulphatedpolysaccharides of Codium f r q i l i contain galactose and arabinose as themajor units. The nature of some of the glycosidic linkages and the locationof some of the sulphate has been shown by the isolation of 3-O-/?-~-arabino-pyranosyl-L-arabinose, 3 -O-/?-D -galactopyranosyl-~ - galactose, D - galactose.Q-sulphate, and D-galaCtOSe 6-sulphate as partial hydrolysis products.l31The water-soluble sulphated polysaccharides from Cladophora rupestris,Chaetomorpha liaum, and Ch.cupilhris are similar to each other.132 Theysuperficially resemble those from Codium in containing galactose and ara-binose as the major sugar components and ingiving 3-O-~-~-galactopyranosyl-D-galactose and galactose 6-sulphate on partial hydrolysis. Differencesare shown, however, by the isolation of further products not given by theCodium polysaccharides, namely 6-O-~-~-galac~opyranosy~-D-ga~ac~ose, L-arabinose 3-sulphateY and an O-L-a.rabinopyranosy1-L-arabinosc 3-sulphatethat was either 1 4 - or 1+5-linked.Another difference was that 3-0-8-~-arabinopyranosyl-L-arabinose and galactose 4-sulphate were not isolated.The presence of L-arabinose 3-sulphate units which must be unsub-stituted on C-2 in the Cladophora polysaccharide, was confirmed by theisolation of 2-O-methyl-L-xylose, formed by ring-opening of an intermediate2,3-epoxideY after treatment with sodium rnethoxide followed by acidhydrolysisCarbohydrate Sulphates in Lipids.-Sulphatides. A brief summary hasbeen published of the chemistry and natural distribution of these glycolipidscontaining sugar sulphate units,133 and this discussion is confined to subse-quent work. Further confirmation that cerebroside sulphate from bovinebrain contains a 3-sulphated rather than a 6-sulphated galactose unit, isthat the natural product is not identical with a sulphation product of cere-broside in which the galactose was shown by methylation t o be 6-~u1phated.l~’Sulphate is solely a t this position in brain sulphatides from normal humanadult, normal infant, and from the infantile form of metachromatic leuco-dystrophy-a disorder in which there is an accumulation of ~u1phatides.l~~There are differences in fatty-acid composition between cmebroside sulphateslZe C.C. Gosselin, A. Holt, and P. A. Lowe, J . Chem. Soc., 1964, 5877.130 J. P. McKinnell and E . Percival, J . Chem. Soc., 1962, 2082; J. J. O’Donnelland E. Percival, ibid., 1959, 2168.131 J. Love and E. Percival, J . Chem.SOC., 1964, 3338.132 Sir Edmund Hirst, W. Mackie, and E. Percival, J . Chenz. Soc., 1965, 2958.13s A. N. Davison, Biochem. J., 1964, 91, 3P.13* T. Taketomi and T. Yamakawa, J . Biochem. Japan, 1964, 55, 87.135 M. MaIone and P. Stoffyn, Biochim. Biophys. Aeta, 1965, 98, 218REES : CARBOHYDRATE SULPHATES 485from different S O U ~ C ~ S . ~ ~ ~ , 136 Cerebroside, but not cerebroside sulphate, isoxidised by the galactose oxidase of Polyporus eircinatus (= Dactyliumdendroides), presumably at C-6 of the galactose ~ n i t . l 3 ~ It has been con-firmed that there is a cerebroside sulphate in human kidney which is similarto that of the brain.138 Methods for the fractionation of brain sulphatideshave been discussed in relation to the isolation of ~u1phatides.l~~Suggestions based on tracer evidence 14O that cerebrosides are the pre-cursors of cerebroside sulphates are now supported by the formation ofcerebroside sulphate by sulphate transfer from PAPS, apparently to cere-broside, by enzymes from sheep 141 and rat 142 brains.Transfer by the sheep-brain extract requires a protein-bound acceptor.141This sulpholipid occurs in all or almost all greentissues and is entirely different from the sulphatides of animal origin.143 Prooftha,t it is a derivative of 6-sulpho-6-deoxy-a-~-glucopyranosyl-( 1 + l ) - ~ -glycerol has been derived from its reacti0ns,l4~ partial synthesis,144 andX-ray analysis of the rubidium salt of the deacylated derivative.l45 Thefatty-acid composition is similar to the phosphatides, but contains a higherproportion of saturated acids than most other leaf 1 i ~ i d s .l ~ ~ The compoundmight have a function in photo~ynthesis.~~~ Plant tissues are unable tosynthesise a selenium-containing analogue, and it has been argued thatthis 148 and other evidence 14Sa might suggest that the biosynthesis involvesPAPS. Biosynthesis of the sulphodeoxy-sugar probably occurs at or beforethe level of the sugar nucleotide, and is followed by transglycosylation toform the li~id.1~~7 148a The bacterial degradation does not appear to involveprior desulphonation, becausg sulpho-acetic acid is formed, and this maythen be metabolised further.149The Synthesis and Reactions of Simple Carbohydrate Su1phates.-Thepreparation and properties of sugar sulphates have been thoroughly re-~ i e w e d .~ Direct sulphation of the primary hydroxyl groups of sugarderivatives is usually specific enough to be a convenient preparative method.136 J. S. O’Brien, D. L. Fillerup, and J. F. Mead, J . Lipid Res., 1964, 5, 109; J. S.O’Brien and G. Rouser, ibid., p. 339.13’ R. M. Bradley and J. N. Kanfer, Biochim. Biophys. Acta, 1964, 84, 210.13* A. Makita, J. Biochem. Japan, 1964, 55, 269; A. Makita and T. Yamakewa,ibid., p. 365.lS0 S. Yokoyama and T. Yamakawa, Jap. J . Exp. Med., 1964, 34, 29; J. Dittmer,Biochim. Biophys. Acta, 1965, 106, 425.140 N. S. Radin, F. B. Martin, and J. R. Brown, J . Biol. Chem., 1957, 224, 499;I. H. Goldberg, J. Lipid Res., 1961, 2, 103; G. Hauser, Biochim. Biophys. Acta, 1964,84, 212; Y.Kishimoto, W. E. Davies, and N. S. Radin, J. Lipid Res., 1965, 6, 525.141 A. S. Balasubramian and B. K. Baehhawat, Biochim. Biophys. Acta, 1965, 108,218.la2 G. M. McKhann, R. Levy, and W. Ho, Biochem. Biophys. Res. Comm., 1965,143 A. A. Benson, Adv. Lipid Res., 1963, 1, 387; Ann. Rev. Plant Physiot., 1964,144 J. Lehmann and A. A. Benson, J. Amer. Ckem. Soc., 1964,8$, 4469, and referencesla5 Y. Okaya, Acta Cryst., 1964, 17, 1276.146 J. S. O’Brien and A. A. BensQn, J. Lipid Res., 1964, 5, 432.14’ A. A. Rosenberg and M. Peeker, Biochemistry, 1964, 3, 254.148 P. Nissen and A. A. Benson, Biochim. Biophys. Acta, 1964, 82, 400.148a W. H. Davies. E. I. Mercer, and T. W. Goodwin, Biochem. J., 1966, $8, 369.14g H. L. Martelli and A. A.Benson, Biochim. Biophys. Acta, 1964, 95, 169.The Phnt sulpholipid.20, 109.15, 1.given there486 BIOLOGICAL CHEMISTRYThe products of direct sulphation of N-acetylglucosamine and N-acetyl-galactosamine have recently been characterised more fully as the 6-sulphates.150The 6-sulphates of 2 - acetamido-Z-deoxy- a-~-galactose 1 -phosphate, O-P-D-(glucopuvranosyluronb acid)-( 1 -+3)-2-acetamido-2-deoxy-~-galactose and(uridine diphosphate)-2-acetamido-2-deoxy- a-D -galactose, have been syn-thesised by direct sulphation.151 When the amino-function in aminodeoxy-sugars is not protected, selective N-sulphation can be achieved; this principlehas been used in the synthesis of the labelled compound 2-deo~y-[2-~~S]sulpho-amino-D -glucose .15 The product of direct sulphation of 5,6- O-isopropyli-dene-L-ascorbic acid, which contains no primary hydroxyl, is presumed tobe the 3-sulphate because the 3-hydroxyl is more acidic than the 2-hydroxylwhich is also available for sulphation. The rapid de-esterification of thiscompound by mild oxidising agents suggests an analogy with oxidativepho~phory1ation.l~~The action of sulphuryl chloride on carbohydrates gives a complex mixtureof products by reactions that have only recently begun to be understood.lNChlorosulphate esters are formed h t and can undergo several furtherreactions depending on the conditions. The stereochemistry of the chloro-sulphate ester is particularly important in determining the course of sub-sequent reactions.In an excess of pyridine, hexoside and hexitol derivativeswith vicinal chlorosulphate ester groups may react to give cyclic sulphateesters, except where the groups are diaxial, when the product is an epoxide.An epoxide can be formed from a diequatorial compound by the use of astronger base such as methoxide.The reaction of axial-equatorial vicinalchlorosulphate esters gives (in addition t o cyclic esters) keto-sugars whichare products of olefin-forming elimination followed by ketonisation. InH H CICI H Hcertain circumstances, the chlorosulphate groups may be converted intochlorodeoxy-groups with inversion of configuration. This nucleophilic dis-placement by chloride may be prevented by steric interactions which increasethe energy of the transition This displacement seems rarely tooccur at C-2 of the sugar ring, presumably because of electronic effects.The unsaturated disaccharide (3) is one of the products of the action ofsulphuryl chloride on arabinose, but the reactions leading to its formationare not fully understood.Is* E.Meezan, A. H. Olavesen, and E. A. Davidson, Biochim. Biophys. Acta, 1964,151 A. H. Olavesen and E. A. Davidson, Biochim. Biophy8. Acta, 1965, 101, 245.1 5 2 A. G. Lloyd, F. S. Wusteman, N. Tudball, and K. S. Dodgson, Biochem. J . ,ls3 E. A. Ford and P. M. Ruoff, Chem. Comm., 1966, 630.lS4 H. J. Jennings and J. K. N. Jones, Canad. J . Chm., 1966, 43, 2372, 3018, and15*5A. G. Cottrell, E. Buncel, and J. K. N. Jones, Ohm. and I&., 1966, 552.83, 256.1964, 92, 68.earlier papersR E E S : CARBOHYDRATE SULPHATES 487In the hydrolysis of certain hexitol cyclic sulphates by hot dilute acidthere is an initial rapid formation of anhydroalditol sulphate half-esterswhich indicates that the cyclic sulphate is opened by intramolecular nucleo-philic displacement with C-0 cleavage and inversion a t carbon.155 Thesulphate half'ester is then, as usual, hydrolysed relatively slowly with S-0cleavage.1ti6 J.S. Brimacombe, M. E. Evans, A. B. Foster, and J. M. Webber, J. Chm. Soc.,1964, 27355. PROTEINS AND PEPTIDESBy D. G. Smyth(National Institute for Medical Research, Mill Hill, London. N . W . 7)PROGRESS in the field of protein and peptide chemistry has again beenextensive. Striking similarities in the molecular behaviour of proteins havebeen noted in diverse biological fields.In Molecular Immunology, the struc-ture of antibody and the binding of antigen to antibody have found corollaryin Molecular Endocrinology with the structure of peptide hormone and itsbinding to receptor, and in Molecular Enzymology wit'h the structure ofenzyme and its binding and activation of substrate. At the molecular level,function is structure and change of structure. The rapid advances currentlybeing made in elucidating the amino-acid sequence and spatial conformationof complex protein and peptide molecules are therefore of fundamentalimportance.This year, the complete amino-acid sequence of sperm whale myoglobin 1(153 residues) has been presented, with definitive experimental daka.Thetotal sequence of ribonuclease-T, (ref. 2) (104 residues) and the almostcomplete sequence of papain (198 residues) have been elucidated. Preli-minary amino-acid sequences of two Type I Bence-Jones proteins 47 5 (212residues) have been proposed ; these, together with comparative studies onpartial sequences of other Bence-Jones proteins,6 provide a first insight intothe detailed structure of y-globulin. The #?-Lipotropic hormone (90 residues)has been isolated and its sequence determined.8 Thyrocalcitonin (86residues) and parathyroid hormone l o (ca. 60 residues) have been isolated;no detailed sequence studies have yet been reported. The peptide hormone,secretin (27 residues), has been isolated and much of its sequence deduced.11The structure of tyrocidine-C (10 residues) has been established.l2 The fieldis open for relating the physiological function of these molecules to theirchemical structures.y-Globulin.-It has been axiomatic that a protein should be availablein pure form before a detailed study of its structure is undertaken.Thusthe well-known heterogeneity of y-globulin has, in the past, diminished theincentive for investigation of its amino-acid sequence. In contrast Bence-A. B. Edmundson, Nature, 1965, 205, 883.K. Takahashi, J. Biol. Chem., 1965, 240, 4117.A. Light, R. Frater, J. R. Kimmel, and E. L. Smith, Proc. Xut. Acad. Sci.S. Hilschmann and L. C. Craig, Proc. Nut. Acad. Sci. U.S.A., 1965, 53, 1403.j K. Titani, E. Whitley, L. Avogardo, and F.W. Putman, Science, 1965, 149,C. Milstein, Nuture, 1966, 209, 370.Y. Birk and C. H. Li, J . BioE. Chem., 1964, 239, 1048.C. H. Li, L. Barnafi, M. ChrBtien, and C. Chung, Nature, 1965, 208, 1093.V. Mutt, S. Mapusson, J. E. Jorpes, and E. Dahl, Biochemistry, 1965, 4, 2358.lo M. A. Ruttenberg, T. I?. King, and L. C. Craig, Biochemistry, 1965, 4, 11.l1 A. Tenehouse, C. Amaud, and H. Rasmussen, Proc. Nut. Acad. Sci. U.S.A.,l2 H. Rasmussen, Y. Ling-Sze, and R. Young, J . Biol. Chenz., 1964, 239, 2852.U.S.A., 1964, 52, 1276.1090.1965, 53, 818SMYTH: PROTEINS AND PEPTIDES 489Jones proteins, which have been shown to be L-peptide chains derived frommyeloma y-globulin,l3-17 can be obtained in homogeneous form when iso-lated from a single individual.Studies on the amino-acid sequences ofBence-Jones proteins 4, 59 18? l9 are therefore practicable and are allowing afirst insight into the detailed structure of the L-chains of y-globulin.Dia.gramma.tic structure of Human y-GlobulinGlu S-S s-s1 1 t 1 , L-chain: K - or A-I sASP s (Bence-Jones protein)I H-chain: p, a-, or p - [ Is1 carbohydrateS ' H-chain I [ssAsp ' K- or il-chain i l s-s I IGlu S--SOn the basis of differing antigenic characteristics of Bence- Jones pro-tein~,22-~~ two distinct types of L-chain have been recognised: Type I orkappa, and Type I1 or lambda.25 The combination of two L-cha,ins K- or2-) with two H-chains (y, a-, or p) is currently considered to provide thefundamental skeleton of the 4-chain y-globulin molecule.(The diagram-matic structure shown is adapted from Fleischmann, Porter and Press; 2oEdelman and Benacerraf; 21 and Milstein 19).L-Chains. The two types of L-chain, studied as the kappa and lambdachains of Bence-Jones proteins, differ completely from each other in ter-minal NH,- groups, peptide maps, and composition of tryptic peptides;26, 27l3 G. M. Edelman and M. D. Poulik, J. Exp. Med., 1961, 113, 861.l4 G. M. Edelman and J. A. Gally, J. Ezp. Ned., 1962, 116, 207.l 5 X. Cohen, Biochem. J., 1963, 89, 334.l6 J. H. Schwartz and G. M. Edelman, J. Exp. Med., 1963, 118, 41.l7 F. W'. Putnam, Biochem. Biophys. Acta, 1962, 63, 539.l8 C. Milstein, J. Mol. Biol., 1964, 9, 836.lS C. Milshin, Nature, 1965, 205, 1171.2o J. B. Flekchmann, R.R. Porter, and E. M. Press, Biochem. J., 1963, 88, 220.21 G. M. Edelman and B. Benacerraf, Proc. Nut. Acad. Sci. U.S.A., 1962, 48, 1035.2 2 M. Mannik and H. G. Kunkel, J. Exp. Med., 1962, 116, 859.23 J. L. Fahey, J . Immzcnol., 1963, 91, 438.24 S. Migita and F. W. Putnam, J. Exp. Med., 1963, 117, 81.Z 5 World Health Organization, Bull. IVorld Health, 1964, 30, 447.26 F. W. Putnam and C. W. Easley, J. BWE. Chem., 1965, 240, 1626.27 K. Titani and F. W. Putnam, Science, 147, 1304400 BIOLOGICAL CHEMISTRYa K-chain thus possesses a totally different amino-acid sequence from a 2-chain. Furthermore, even within Bence-Jones proteins of the sameantigenic type-among K- chains for example-although there are basicsimilarities, considerable differences in primary structure occur.&hains,obtained as Bence- Jones proteins from many different individuals, presentcertain peptide areas of striking identity, certain areas of marked similarity210FIG. 1. Comparison of the amino-acid sequences of three Type I Bence-Jones proteins.Known sequences are indicated by joining together the abbreviations for the amino-acids by dashes. Areas of undetermined sequence are enclosed by a horizontal dashedline. In cases of known composition, the amino-acid residues are placed in the orderof their elution from the column of the analyser if the sequence is unknown for anyprotein; however, if the sequence is known for one protein, the residues enclosed bya dashed line (undetermined sequence) are ordered to correspond with the knownsequence.Regions of probable homologous or nonhomologous interchanges are em-phasiaed by enclosure in a box with a solid line. Regions of possible interchange areenclosed with a vertical dashed box. Data for specimens Roy and Cu are derivedfrom the report by Hilschmann and Craig,4 and data for specimen Ag from Titani,Whitley, Avogardo, and P ~ t n a r n . ~[Reproduced, with permission, from K. Titani, E. Whitley, L. Avogardo, andF. W. Putman, Scimce, 1965, 149, 1090.SMYTH: PROTEINS AND PEPTIDES 491containing identical or homologous amino-acids, and certain areas notablefor the lack of common structure. The question mark over the areas ofdisparity is whether their amino-acid sequences represent the molecular basisthat is critical for the functioning of different antibody molecules, or whetherthere is a degree of freedom a t these positions when the molecule is func-tionally active, making the necessity for precise structure unimportant.The answer to these problems will be found when the amino-acid sequence of alarge number of Bence- Jones proteins from different individuals is obtained,and structure related to differing function.The total amino-acid sequence of a Bence-Jones protein has not yet beenachieved, but almost complete sequences have been presented from twolaboratories (Figure 1).In the studies of Hilschmann and Craig: and ofPutnam and his colleague^,^ preliminary amino-acid sequences have beenproposed for three different Type I Bence-Jones proteins [obtained fromindividuals designated Ag, Roy, and Cummings (Cu)].The protein from asingle individual was found to possess a characteristic amino-acid sequence.With the exception of an interchange of valine for leucine at position 189,an identical sequence of 105 amino-acid residues was found a t the C0,H-terminus in the proteins Ag and Roy, and also in the protein Cu for which lessdata is available. The NH,-terminal portion of the three K-chains, on theother hand, shows markedly less correspondence. The replacements involveboth homologous and non-homologous amino-acids, the differences beingdistributed in a complex manner along the peptide chain.Bence-Jones proteins have been shown to occur as stable monomers,stable dimers, and as dissociable dimers.28? 29 The relationship between thefirst two of these structures and the single chain of amino-acids shown inPigure 1 has recently been elucidated 6 s l9 (Figure a), and further evidencehas been obtained to show that the cystine residue a t the COOH-terminus ofthe K- or ).-chain of Bence-Jones protein is the same residue that is involvedin the linkage of L-chains to H-chains in the heterogeneous y-globulinmolecule.Methods have been developed for radioactive labelling of thecystine residues in Bence-Jones protein and for rapidly determining thesequence of amino-acids around the labelled residues.6 In this manner, theconstant sequence of 24 amino-acids a t the C0,H-terminus has now beenconfirmed for the K-chains from 8 individuals.Another area common to theK-chains from 5 individuals is the sequence from position 115 to 133 (c$Figure 1). The sequence from 84 to 101 was common to three proteins, butthree other proteins showed minor changes in this region, homologous exceptin one case for the replacement of glycine by proline. Finally, a high degreeof correspondence was observed in the sequence from position 17 t o position36.The widespread structural differences among K-chains are clearly notcompatible with their development from a common precursor by single pointmutations, as is believed to be the case with the abnormal haemoglobins. Toallow a clear understanding of the basis for the remarkable variability inthe L-chains of y-globulin, total amino-acid sequences of a large number of28 J.A. Gally and G-. M. Edelman, J. Exp. Med., 1964, 119, 817.G. M. Bernier and G. M. Edelman, Nature, 1963, 200, 223492 BIOLOGICAL CHEMISTRYTYP INH, 212I Stable monomer Lys*Ser *Phe*Asp*Arg*Gly-Glii*Cys1CYSNHZIStable dimer Lys*Ser.Phe*Asp*Arg*Gly.G;luCysI NHZI Lys*Ser *Phe*Asp -Arg*Gly -Glu-Cys212Lys *Ser *Phe*Asp*Arg *Gly *Glu*Cys1NH2Iy-GlobulinH-chainType IIStable monomer Lys *Thr .Val.Ala-Pro.Thr .Glu-Cys*SerIcys.Stable dimer Lys *Thr *Val*Ala*Pro *Thr*Glu*Cys*SerLys-Thr *Val*Ala*Pro -Thr *Glu*Cys*Ser.Lys -Thr *Val*Ala*Pro *Thr -Glu*Cys *Ser y -Globulin1H-chainFIG. 2. Proposed structure around the interchain disulphide bridgein Bence-Jones proteins and in human y - g l o b ~ l i n .~ ~Bence- Jones proteins of the two antigenic types will be required. Meanwhile,further rapid methods are needed for comparative study of the loci of changesand variations in sequence.H-Chain. Studies on the H-ohain of y-globulin are less advanced. TheH-chains of human y-globulin have been found to be electrophoreticallyheterogeneo~s.~~ The detailed studies of Tanford and his ~olleagues,~~however, have shown that a large part of the H-chain in rabbit y-globulinhas a constant structure, and that this structure is the same in antibodyy-globulin. This section of the H-chain comprises all, or nearly all, of thatportion of the H-chain which is present in the papain-produced ‘‘ Fragment111,” but some peptides of “ Fragment I ” are also included in the constantsection.The differences in amino-acid sequence between H-chains thusappear to arise in a relatively small portion of the chain. In this connexion3 0 S. Cohen and R. R. Porter, Biochem. J . , 1964, 90, 278.31 C. A. Nelson, M. E. Noelken, C. E. Buckley, C. Tanford, and R. L. Hill, Bio-chemistry, 1965, 4, 1418SMYTH: PROTEINS AND PEPTIDES 493it may be noted that only very slight differences could be detected betweenpeptide maps from rabbit antibody and the corresponding maps from non-specific rabbit y-gl~bulin.~~Studies leading to the formulation of a sequence in the Fragment I11portion of an H-chain from rabbit y-globulin are in progress in the laboratoryof R. L. Hill.33 A tripeptide, Pro*Val*Thr., has been isolated from an enzymicdigest of the H-chain obtained from a human pathological immun~globulin,~~and evidence was presented that this sequence is derived from the NH2-terminus; the peptide could not be detected, however, in normal humanimmunoglobulin.In studies on the C0,H-terminus of the H-chain ofrabbit immunoglobulin,35 application of the cyanogen bromide technique 36for specific cleavage a t methionine residues yielded an 18 amino-acid peptidein good yield:NH, NH2I I* * * (Met) ~His~Glu*Ala~Leu~His~Asp~His~Tyr~Thr~Glu*Lys~Ser~ileu~Ser*Arg~Pro *GlyCO,HAn almost identical octa,decapeptide was isolated from human myelomaH-chain,37 and also from normal human immunoglobulin.3~NH2INH!2I - a - (Met) *His*Glu *Ala*Leu -His*Xsp -His *Tyr .Thr *Glu-Lys*Ser *Leu*Ser *Leu*Ser .Pro-Gly C0,HThe areas of variable sequence in the L- and H-chains constitute regionswhich need not be critical to a specialised function.The amino-acids in-volved could act simply as spacers between regions that contribute struc-turally to the antibody site. More plausibly, however, it is anticipated thatthe key to the nature of antibody specificity to widely differing antigens willbe found in those regions of the y-globulin molecule that possess a variableamino-acid sequence. Nothing is yet known at the molecular level of the wayin which amino-acid sequence may control antibody function. “ The rulesof protein stereochemistry are very complex and may not become apparentbefore the structures of maiiy proteins have been solved in atomic detail.’’ 39Pragmentation of y-globulin.The four-chain structure of y-globulin hasbeen broken down by a variety of methods to yield sub-units and fractions.Reduction of interchain disulphide bonds in the absence of urea with subse-quent exposure to acid 40, *l or urea l3 leads to dissociation of L-and H-chains;32 D. Givol and M . Sela, Biochemistry, 1964, 3, 451.33 R. L. Hill, personal communication.34 R. R. Porter and E. M . Press, Biochem. J., 1965, 97, 32P.35 D. Givol and R. R. Porter, Biochem. J., 1965, 97, 32C.36 E. Gross and B. Witkop, J . Biol. Chem., 1962, 237, 1856.37 E. M, Press, P. 5. Piggott, and R. R. Porter, Biochenz. J., 1966, in the press.38P. J. Piggott and E. M. Press, Biochem. J., 1966, in the press.39 M.F. Perutz, in “ Proteins and Nucleic Acids,” Elsevier Publishing Co., London,40 R. R. Porter, in “ Basic problems in Neoplastic Disease,” ed. A. Gellhorn and*l J. B. Fleischman, R. H. Pain, and R. R. Porter, Arch. Biochem. Biophys., 1962,1962, p. 59.E. Hirschberg, Columbia University Press, New York, 1962, p. 117.suppl. 1, 174494 BIOLOGICAL ClHEMISTRYneither treatment alone is sufficient. Separated L-chains dimerise and arecapable of forming an interchain disulphide bond.14 Mild reduction of y-glo-bulin followed by treatment with hydrochloric acid leads to separation ofthe L-H pairs as half mole~ules.~~ Hydrolysis with papain causes limitedoleavage of some peptide bonds in the H-chains;43 after reduction of thedisulphide bond between these chains,4* “ S ” and “B”’ fragments arereleased.Digestion with pepsin provides fragments similar to the S frag-ments coupled by a disulphide bond.44, 45 Finally, the selective cleavageof rabbit antibody by cyanogen bromide takes place at methionhe residuesapparently within the papain “Fragment I11 ” portion of the molecule,46and results in retention of the antibody activity. This is the first use of achemica’l reagent to fragment y-globulin which has not resulted in destruc-tion of the biological activity.*‘Fro. 3. Fragmentation of y-globdin.[Reproduced, with permission, from G. M. E d e h n and J. A. Gally, Proc. Ha€.Acad, Sci. U.S.A., 1964, 51, 846.3Antibody active site. The location of the antibody combining site, theactive centre in the y-globulin molecule that combines with antigen, has beenthe subject of many investigations. The active site appears to be presentprincipally on the H-chain of y-globulin ;20 separated L-chain does no$ appearto bind antigen.However, interaction of H-chains with L-chains is necessaryfor maximum expression of antibody activity.48~ 49 Furthermore, differencesin antibody specificity are reflected by changes in the electrophoretic beha-43 5. L. Palmer, A. Nisonoff, and K. E. Van Eolde, Proc. Nat. Acad. Sci. U.S.A.,1963, 50, 314.43 R. R. Porter, Biochem. J., 1959, 73, 119.44 A. Nisonoff, 0. Markus, and F. C. Wissler, Nature, 1961, 189, 293.46 R. J. Cahmann, R. Amon, and M. Sela, J. Biol. Chm., 1965, 240, PC2762.4 7 G. M. Edelman and J.A. Cally, Proc. Nat. Acad. Sci. U.S.A., 1964, 51, 846.48 G. M. Edelman, D. E. O h , J. A. Gally, and N. D. Zinder, Prm. Nat. Acad.S. Utsumi and F. Harush, Biochemistry, 1965, 4, 1766.Sci. U.S.A., 1963, 50, 753.F. Franek and R. S. Nezlin, FoZia Nicrobiol., 1963, 8, 128SMYTH: PROTEINS AND PEPTIDES 495viour of the L-chains,so and evidence has accumulated that portions of bothL- and H-chains are situated a t or near the antibody combining ~ite.~1-54Therefore the combining site for antigen in the y-globulin molecule appearsto be centred on the H-chain 2* but profoundly influenced by the L-chain.47For each of the L-H pair of chains in the y-globulin molecule there is onecombining site for antigen, and thus there are two combining sites permolecule of y-globulin; the precise structure of the active centres remains tobe elucidated.The specificity characteristics of the wide variety of antibodymolecules for different antigens appear to be based on the interaction betweena large number of L- and H-polypeptide chains of different amino-acidsequence.55, 66The molecular basis for complementarity between the structure of anantigen molecule and the specific amino-acid sequence of the correspondingantibody presents one of the most challenging problems in protein chemistry.Ribonuc1ease.-Bovine pancreatic ribonuclease (RNAase) continues to bewidely used in studies on the mechanism of enzyme action at the molecularlevel ; its established sequence and folded structure round the disulphidebridges form a complex steric molecule which is the basis for current investi-gations on the relation of structure to function.Reactions of monofunctional reagents with specific and defined groups inthe enzyme molecule are enabling deductions to be made on the functionalimportance of single amino-acid residues.The alkylation of ribonuclease bya series of unbranched cc- and #I-halogeno-acids of chain length 3 to 6 carbonatoms has been in~estigated.~~ As has been demonstrated previously in thereaction of iodoacetate with ribonuclease, substitution occurs either a timidazole nitrogen 3 of His-12 58 or a t imidazole nitrogen 1 of His-119.58, 59The rates of the reactions and the sites of alkylation were found to be defer-mined by the structural characteristics of the reagent, including chainlength, optical configuration, the presence of additional functional groups,and the position of the halogen relative to the carboxyl group.DL-CC-Bromocaproate (a), /I-bromopyruvate (b), and /?-bromopropionate (c) reactedexclusively a t His-119. The D-enantiomorphs of cc-bromopropionate anda-bromo-n-butyrate, on the other hand, reacted preferentially a t His-12.(a) Me*[CH,],*CHBr*CO,H(b) CH,Br*CO*CO,H(c) CH,Br *CH,*CO,HG. M. Edelman, €3. Benacerraf, Z. Ovary, and M. D. Poulik, Proc. Nut. Acad.Sci. U.S.A., 1961, 4'7, 1751.51D. A. Roholt, G. Radzimski, and D. Pressman, Science, 1963, 141, 726.52D. E. Olins and G. M. Edelman, J. Exp. Med., 1963, 116, 635.63F. Franek and R. 8. Nezlin, Biokhimiya, 1963, 28, 193.6c M.Metzger and S. J. Singer, Science, 1963, 142, 674.6s C. E. Buckley, P. L. Whitney, and C. Tanford, Proc. Nut. A d . Sci. U.S.A.,56 0. Smithies, Nature, 1963, 199, 1231.67 R. L. Henrikson, W. H. Stein, A. M. Crestfield, and S. Moore, J. Biol. Chem.,58 A. M. Crestfield, W. H. Stein, and S. Moore, J. BioE. Chern., 1963, 238,59E. A. Barnard and A. Ramel, Nature, 1962, 195, 243.1963, 50, 827.1965, 240, 2921.2421496 BIOLOGICAL CHEMISTRYFurther experiments from the same laboratory have strengthened the hypo-thesis that both histidine residues are involved in the active site of RNAase.58The interesting hybridization experiments which were noted briefly lastyear,6o in which aggregation of inactive derivatives of RNAase led to theformation of active molecules, have now been presented in full detaiLglThese specific reactions and interactions are a reflection of the intricatestereochemistry and conformation of the active site of the enzyme, but preciseinterpretation of the results must await further knowledge of the tertiarystructure.In a biologically active molecule where function resides in a, critical3-dimensional configuration, certain amino-acid residues may be far apartin the linear sequence yet closely positioned in the steric structure. Thedistance of separation of these residues may be estimated by cross-linkingwith a bifunctional reagent.The size of the reagent introduced, connectingthe two amino-acid residues in the protein, is a measure of the distance ofseparation of these residues in the active molecular structure, providingactivity is retained, Following an earlier application of 175-difluoro-2,4-dinitrobenzene in the cross-linking of insulin,62 the reaction of this reagentwith RNAase under conditions favouring the production of monomeric,intramolecularly cross-linked products has been studied.63 Three cross-linked derivatives of RNAase were obtained, and enzymatic activity was tosome extent retained in each.I n one of the products, a cross link had formedbetween the e-NH, groups of Lys-7 and Lys-41; this derivative was enzy-mically active.A product of the reaction of the monofunctional reagent fluorodinitro-benzene with RNAase is also a derivative in which the lysine residue atposition 41 is modified;64 the absence of activity in this product contrastswith the substantial activity of the cross-linked derivative, although bothcontain lysine a t position 41 blocked by similar substituents.However,with the monosubstituted product, there was evidence that after the blockingreaction, the molecule had undergone a conformational change which affectedthe accessibility of Lys-7 and distorted the anion binding site a t the activecentre; the cross-linked derivative may not be able to undergo such a change.The considerable enzymic activity of the cross-linked derivative precludesan important functional r81e for Lys-41 in the active site of RNAase, andestablishes a limiting spatial relationship between Lys-41 and Lys-7 thatis consistent with the expression of enzyme activity.The maximumdistance between the two a-carbon atoms of Lys-7 and Lys-41 was calculatedto be 17.2 A. This informsttion, together with the additional spatial datawhich will be obtained by identification of the other cross-linked derivatives,will provide valuable information for the construction of a 3-dimensionalmodel of the active enzyme. It will also provide data essential to answeringthe question of whether the structure which will be obtained by X-ray60 D. G. Smyth, Ann. Reports, 1964, 81, 507.61 R. G. Fruchter and A. M. Crestfield, J . Biol. Chem., 1965, 240, 3868; ibid., 3875.62 H. Zahn and J. Meienhofer, Malcromol. Chem., 1958, 26, 126.63 P. S. Marfey, H. Nowak, M. Uziel, and D. A. Yphantis, J .BWZ. Chem., 1965,64 C. H. W. Hirs, Brookhaven Syinposia in Biology, 1962, 1'01. XI', p. 154.240, 3264S N Y T H : PROTEINS AND PEPTIDES 497diffraction of the crystalline protein is the same steric structure as that pos-sessed by the active enzyme in solution.The difference in points of attack by a proteolytic enzyme on nativeRNAase compared with the denatured molecule has been used to provideinformation on the accessibility of certain residues in the 3-dimensionalstructure. In the native state, RNAsse, is resistant to digestion by trypsinand chymotrypsin, but after physical denaturation it becomes susceptiblet o attack.s5* 66 Mild treatment of native RNAase with subtilisin results in aspecific cleavage of the molecule, forming RNAase-S, without loss of enzymeactivity; only the peptide bond between Ala-20 and Ser-21 is split.67 Pepsinis capable of degrading RNAase to provide inactive products.The firstcleavage eliminates the C0,H-terminal tetrapeptide and results in someunfolding of the protein.68 The resulting molecule can be cleaved further bypepsin;GQ evidence was presented that the sites of attack were at the CO,Hgroups of Phe-120, Met-79, Thr-45, GluNH,-55, and Phe-8. The residualmolecule, held together by disulphide bonds, retained some conformation,because Tyr-25 in this material appeared to be unavailable for reaction withiodine. Carboxypeptidase attacks RNAase with reluctance, releasing amino-acids from the CO,H-terminus.Go, 70 Elastase cleaves the peptide bondbetween Ala-19 and Ala-20 and also that between Ala-20 and Ser-21, releas-ing free alanine and two polypeptide fragments, 1 to 19 and 21 to 124, whichbind strongly to each the product was designated RNAase-E, byanalogy with RNAase-S.As expected from the results of Richards and ofHofmann, KiNAase-E retained full enzymic activity.The ease of hydrolysis of the peptide bonds a t positions 19 and 20 indi-cates that Ala-19 and Ala-20 are on the outside of the molecule and arefreely available to certain enzymes. Since the 2 slanine residues are bor-dered by 7 serine or threonine residues, it seems likely that the hydroxy-amino-acid residues may play a special r81e in influencing the structure ofthe molecule. Hydroxyamino-acids appear to be capable of interactingwith each other in aqueous solvents. Poly-I;-serine, for example, is verysoluble in water, 'i2 and the dipeptide L-seryl-L-serine is only sparinglysoluble.73 It has been reported that hydroxyamino-acids exert disruptiveeffects on helices,'* and in synthetic polypeptides favour the formation ofp-structures and not helices.A tentative 3-dimensional structure of bovine pancreatic ribonucleasehas been proposed 75 (Figure 4). The model is based on the lineax sgquenceof amino-acid~,~~ data on the active site, interpretations of the results of ion6 5 J. A. Rupley and H. ,4. Sheraga, Biochemistry, 1963, 2, 421.6 6 T. Ooi, J. Rupley, H. A. Sheraga, Biochemistry, 1963, 2, 432.6 7 F. M. Richards and P. J. Vithayathil, J . Biol. Chem., 1959, 234, 1459.6 8 C.B. Anfinsen, J. Biol. Chem., 1956, 221, 405.69 H. Fujioka and H. A. Scheraga, Biochemistry, 1965, 4, 2197.70 J. T. Potts, D. M. Young, C. B. Anfinsen, and A. Sandoval, J. Biol. Chevn.,71 W. A. Klee, J. Biol. Chem., 1965, 240, 2900.72 Z. Bohak and E. Katchalski, Biochemistry, 1963, 2, 228.73 J. S. Fruton, J. BioE. Chem., 1942, 146, 463.7 4 J. C. Kendrew, Brookhaven Symposia in Biology, 1962, 15, 216.75 H. A. Saroff, J . Theoret. Biol., 1965, 9, 232.713 D. 0. Smyth, JTT. H. Stein, and S. Moore, J. Biol. Chem., 1963, 238, 227.1964, 239, 3781498 BIOLOGICAL CHEMISTRYnFIG. 4. IIEzlstrath of a proposed three-dimewional structure of ribonwkase.[Reproduced, with permission, from H. A. Saroff, J . Thoret. BWZ,, 1965, 9, 234.1FIG.5. Amim-acid sequence of bovine ribonuclease arranged to present 6 clustersgiving rise to proton and anion binding anomaties.[Reproduced, with permission, from D. G. Smyth, W. H. Stein, and S. Moore,J . Bwl. Chem., 1963, 238, 227.1binding, and on chemical evidence implicating the juxtaposition of certainamino-acid residues not closely related in the linear sequence (Figure 5).While many constraints have been included in the model, the structure isoutlined only approximately in certain areas. It may be noted that Lys-7is remote from Lys-41 in the model, although sound chemical evidence hasnow established their proximity.6s Until further chemical and physico-chemical data is available, the model should be considered as a workinghypothesisSMYTH: PROTEINS AND PEPTIDES 499The complete amino-acid sequence of Ribonuclease-TI (takdiastase)is now reported in full.77 The molecule consists of a single polypeptidechain of molecular weight 11,085, with 2 disulphide bridges (Figure 6).C p c c P C P p PI'1 1 1 1 1 2 13 14 i s 7 1 6 17 w 1 : o ; z o : 2 1 a n l a l i a s i n a1~ S c r ~ S ~ r ~ A s p ~ V r I ~ S c r ~ T h r - + A l r ~ C I n ~ A I r ~ ~ l a ~ C I ~ ~ T , r ~ C l n ~ L c r H i r C I uClr 31C T C C Thr 32I 1 ' I I 'Val ,jPC c..+Cir+GIn+Tyr + A s n + A s n t T ! .r + L ~ s + } i i s ~ P r o ~ T ~ r ~ S c r ~ A s n + S c f ~ l ~41 46 45 44 U 42 41 4 0 3 I n 36 3534'C P P P P*J Asn I 5 6 l j S 7 i S j 3 9 i G 4 61 6 2 7 0 44 (6 66p67S Asn t - + P r o ~ T j r - T j r ~ C I u ~ T T P ~ P r o ~ IIc .. L e u - * S c r - + S c r - . G l ? ~ A ~ ~ ~ ~ ~ ~ lt+ A GIy96 SerI T i s p pc53 n si w I) I n w 05 u tt aa a: DO r3 n n x 75 14 nFIG. 6. The amino acid sequence of ribonuclease T,.The points of hydrolysis by trypsin and chymotrypsin in the performic acid-oxidisedprotein and by pepsin in the heat-denatured protein are marked by T, C, and P, re-spectively. The solid lines represent extensive or rapid hydrolyses, and the dashedlines, incomplete or slower hydrolyses.[Reproduced, with permission, from K. Takahashi, J. Biol. Chem., 1965, 240, 4117.1The sequence was determined by the quantitative methods of Hirs,Moore, and Stein, including cleavage of the disulphide bonds by oxidationwith performic acid, and cleavage of the peptide chain by digestion withtrypsin, chymotrypsin, pepsin, subtfiin, and papain.The peptides wereseparated by chromotography on Dowex 1 -X2, Dowex-50x2, and diethy-laminoethyl cellulose, with volatile buffers as eluents. The amino-acidsequence of each purified peptide was determined by chemical and enzymaticmethods including dinitr op hen yla tion, Edman degradation, h y drazinol ysis ,and with the use of leucine aminopeptidase and carboxypeptidase.Trypsin caused cleavage at the carboxyl groups of Tyr-24, and Tyr-68,in addition to the expected sites at lysine and arginine. Trmtment of thetrypsin with diphenylcarbamyl chloride eliminated its chymotrypticactivity against synthetic substrates but did not appear to increase itsspecificity in attacking RNAase-TI.Chymotrypsin attacked peptide bondsinvolving the carboxyl groups of histidine, asparagine, leucine, alanine, andglutamine, in addition to the expected cleavages at tyrosine and phenyla-lanine; acidic residues adjacent t o the aromatic residues reduced the rates ofcleavage. Pepsin, papain, and subtilisin exhibited rather broad specificities.The structure of RNAase-T, (Figure 6) is of particular interest for thepurpose of comparison with that of pancreatic RNAase,'*, 95 the two enzymes77 K. Takahashi, J. Biol. Chern., 1965, 240, 4117.7 8 B. F. Erlanger and F. Edel, Biochemistry, 1964, 3, 347500 BIOLOGICAL CHEMISTRYperforming similar functions and differing in specificity. The amino-acidsequence of RNAase-T, (104 residues, 2 disulphide bridges) is found to betotally different from that of the pancreatic enzyme (124 residues, 4 disul-phide bridges).Takahashi 7 7 has suggested that similarities may exist inthe secondary and tertiary structures of the two enzyme molecules.MyogIobin.-The difficult amino-acid sequence of myoglobin has beenelucidated by Edmundson,l who used chemical methods. The completeprimary structure of the protein is illustrated in Figure 7 . Reaction of theprotein with iodoacetate 7 9 was exploited to advantage in overcoming theproblem of insolubility exhibited by the ‘’ core ” of tryptic peptides. Thereaction occurs a t all of the histidine residues, slightly a t the a-NH, groupof the terminal valine residue, and not a t all a t the E-NH, group of thelysine residues.The peptides obtained by digestion of the carboxymethy-lated protein with trypsin or chymotrypsin remained soluble, and weresuccessfully isolated in pure form. Trypsin and chymotrypsin caused cleav-age at sites in addition to those expected from their principal specificities.I n addition to this difficulty, some hydrolysis of the amides of glutamicacid and aspartic acid was noted for several peptides which were isolatedboth in acidic and in neutral forms. The size of the myoglobin molecule(153 residues), the fact that the protein possesses a single chain, the finding ofmultiple forms of peptides, and the occurrence of major problems withsolubility, combined to make the determination of this sequence a considerablefeat.The 3-dimensional structure of myoglobin in the crystalline state massolved in the outstanding experiments of Kendrew and his colleagues.80The amino-acid sequence presented above, established by chemical proce-dures, confirms and extends the results of current X-ray studies 81 in whichan atomic resolution of 1.5 A is being used to define unequivocal sequencesof amino-acids.In view of the difficulty in attaining this resolution withother crystalline proteins, it appears that determinations of amino-acidsequence will continue to be per€ormed by chemical methods and the se-quence thus obtained will be organised into a 3-dimensional structure byX-ray-crystallographic analysis. In this way a complete 3-dimensionalstructure of lysozyme has recently been elucidated in atomic detail by theX-ray-crystallographers a t the Royal Institution,82 the total amino-acidsequence having previously been established by chemical methods.83 Itshould be noted, however, that recent developments in the techniques ofX-ray diffraction have made it possible to detect the loss of as little as oneoxygen atom from oxymyoglobin, 84 the structure of the parent moleculebeing known at a resolution of less than 2 8.The new X-ray techniqueswere recommended for study of the interactions between proteins and small79 L. J. Banaszak, and F. R. N. Curd, J. BioE. Chem., 1964, 239, 1836.so J. C. Kendrew, R. E. Dickerson, B. E. Strandberg, R. G. Hart, D. R. Davies,D. C. Phillips, and V. C. Shore, Nature, 1960, 185, 422.s1 J.C. Kendrew, H. C. Watson, D. C. Phillips, and others, to be published.8a C. C. C. F. Blake, D. F. Koenig, G. A. Mair, A. C. T. North, D. C. Phillips, andV. R. Sam&, Nature, 1965, 206, 757.s8 R. E. Cenfield, J. BioE. Chem., 1963, 238, 2698.84C. L. Nobbs, H. C. Watson, and J. C. Kendrew, Nature, 1966, 209, 339r c p t i d c 3 - -- I------ -r132 I33 134 135 136 137 138 139 IM 141 142 143 I44 145 146 147 148 149 I50 $.Asn.Lys.Alo.Ltu.CIu.Leu.P~c.Arq,Lys.Asp,Ilc.Ala.Alo.Lys Tyr Lys.Glu.Lcu.Gly.Tyr 8 6 b 0I I 2 3 4 5 6 7 1 9 ;O I I I 2 I3 14 I S I6 17 16 19 2 0 21 22 23 24 25 26 2739 4 0 41 42 43 44 45 46 47 40 49 50 51 52 53 54 55P e p t i d r ?56 57 58 59 60 61 62 63 01 65 66 67 60 69 70 71 72 73 74 75 76 77 78 7 9 0 0 81 82.Lyi.Alaber~lu.AspL.u.Lyi.LyrHlr.GJ y.Val.Thr.Vol.Leu.Thr.Alo.Lcu G I y.Al0.11 r.bu.Lys Lya Lyi.GI y.HI s.HI r.Glu. b h b b8, 6 6 $ 6 g 694 9S 96 97 96 99 100101 102 103 0 4 105 lC% I07 IW 109 I10 I l l I12 113 114 115 116 I17 118 119 I2OIZIAla .Thr .kyr.Wls .Lye .I1 e .Pro .XI# .Lyr .Tyr .L.u.Clu .me .I I #.Sir .Glu Alo.11 c .I1 r.HI i .Pro.GI Q 6 6 6 602 BIOLOG TCAL CHEMISTRYmolecules, for example enzymes and coenzymes, provided suitable crystallinederivatives can be obtained.81, 85The binding of a fluorescent reagent by a nonpolar site in a protein opensthe possibility of obtaining information about regions in the molecule thatare not directly affected by the usual nucleophilic reagents. l-Anilino-8-napthalene sulphonate ( A N S ) binds stoicheiometrically to a specific site onapomyoglobin, but not to myoglobin itseKs3, Measurements of fluor-escence polarization and optical rotat'ory dispersion indicate that ANS com-bines a t the hydrophobic site occupied in myoglobin by hemin, and that theANS apomyoglobin complex exhibits a similar steric structure to nativemyoglobin.The reagent thus acts as a fluorescent probe for nonpolarbinding sites in proteins.Papain.-A previously reported sequence a t the NH,-terminus of papain 8 7has been corrected,SS, 89 and the suggestion that a large peptide fragmentcan be removed from the NH,-terminus without loss of the enzyme activityhas required re-examinati~n.~~ Error in the sequence work had arisen fromthe use of leucine aminopeptidase (LAP) which liberated amino-acids fromthe interior of the molecule, and from the use of the Edman degradationwith acetic acid-hydrochloric acid as the cyclizing agent.Both techniqueshad previously been shown to be suspect when applied in the determinationof the primary structure of ribonuclease, the misleading results obtainedNH,-Ileu-Pro-Glu-Tyr-Val-Asp -Trp-Arg-Gln-Lys-Gly- Ala-Val-Thr-Pro-Val-Lys-10Asn- Gln- Gly -Ser -Cys -G1 y -Ser -Cys-Trp/ /Ma-Phe // (Ileu) //Arg-Asn-Thr -Pro-20 30Tyr -Tyr -Glu-Gly -Val-Gln- Arg -Tyr -Cys - Arg- Ser - Arg - Glu-Lys-Gly -Pro -Tyr -Ala -40 50Ala-Lys-Thr-Asp-Gly-V~l-Arg-GIn-Val-G;ln-Pro-Tyr-A~~-Gln-Gly-Gly-Ala-Leu-Leu-Tyr -Ser-Ileu-Ala-Asn-Gln-Pro -Ser-Val-Val-Leu-Gln-Ala-&-Gly-Lys-Asp -Phe-Gln-Leu-Tyr-Arg-Gly-Gly-Ileu-Phe-Val-Gly-Pro-Cys-Gly-Asn-Lys-Val-Asp -His-Ala-Val- Ala - Ala -Val- G1 y -T yr - Asn -Pro -GI y -T yr -1leu -Leu-Ileu-Ly s - Asn- S er -Trp-Gly-Thr -Gly-Trp-Gly-Glu- Asp -GIp-Tyr-Ileu-Arg-Ileu-Lys-Thr-Gly-Asn-6070 8090 100110 120130Leu-Asn-Gh-Tyr -Ser-Glu-Gln-GIu-Leu-Leu-Bs~-Cys-Asp-Arg-Ar~-Ar~-Ser-Tyr -Gly -Cys -Tyr -Pro -Gly -Asp -G1 y -Trp / / Ser - Ala-Leu / /Val-Ala-Gln-Tyr -Gly -160 170Ileu-His-Tyr-Arg-Gly-Thr-Gly-,4sn-Ser -TUvr-Gly-Val-Cys-Gly-Leu-Tyr-Thr-Ser -140 150180 190Ser -Phe -Tyr -Pro -Val-Lys-Asn-CO,HFIG.8. Tentative amino-acid sequence of p a p a k 3Amino-acid residues placed in sequence are separated by dashes ; slant lines indicate-wiped peptides whose relative positions are not yet established.L.N. Johnson and D. C . Phillips, Nature, 1965, 206, 761.A. Light and E. L. Smith, J. Biol. Chem., 1960, 235, 3151.86L. Stryer, J. Mol. Biol., 1965, 13, 482.e8A. Light and J. Greenberg, J . Biol. Chem., 1965, 240, 258.89R. Frater, A. Light, and E. L. Smith, J. Biol. Chem., 1965, 240, 253SMYTH: PROTEINS AND PEPTIDES 503by these methods having necessitated a comprehensive re-investigation of theamino-acid sequence.76, 95The extensive studies currently reported on the structure of papainestablish firmly the almost complete amino-acid sequence (Figure 8).3, 8 * ,The disulphide bridges have been located between cystine residues a t posi-tions 22 and 159,43 and 152, and lOOand 186. The remaining single cysteine,curiously, is liberated only after reduction of the protein and appears to bedirectjly involved in the activity of the enzyme.Similar findings have beenreported in a detailed study of Streptococcal proteinase: a single reactivesulphydryl group again was liberated only after reduction of the precursorprotein. The amino-acid sequence of a large peptide containing the SH-groupwas identified. 91By designing an inhibitor related structurally to a specific substrate ofpapain and by inserting into the inhibitor a, reactive site with an affinity forSH-groups, inhibition of the enzyme might be accomplished and simul-taneously the active cysteine residue become specifically modified. Reactionof reduced papain with two equivalents of the chloroketone (I), related toN-tosylglycine methyl ester, resulted in inhibition of the papain activity andenabled identification of Cys-25 as the reactive cysteine a t the active site.92Enzymic hydrolysis of the peptide containing the modified cysteine residueled to its isolation as a dihydro-1,4-thiazine derivative (11).CHJCI S/ \CH 'CH2II IC CH*CO,H\ /A similar heterocyclic structure is formed when the addition product ofcysteine and N-ethylmaleimide (NEM) is exposed to neutral or alkalineCOlH CO,HI II - oc s0 - C ? ,s \ IOC-CH2 I /=, H HOIC - CHINC' Hcondition^,^^ the a-NH, group of the cysteine adduct undergoing infra-molecular cyclization with a ketonic group of the succinimide ring.Cytochrome C.-The complete amino-acid sequence of hog heart cyto-chrome C (104 residues) has been determined,94 and may be compared withDo J.R. Kimmel, H. J. Rogers, and E. L. Smith, J . BWZ. Chem., 1965, 240, 266.Dl T. Y. Liu, W. H. Stein, S. Moorey and S. D. Elliott, J. BWZ. Chem., 1965, 240,*3 D. G. Smyth, A. Nagamatsu, and J. S. h t o n , J . Amer. Chm. Soc., 1960, 82,D4 J. W. Stewart and E. Margoliash, C a d . J. Hochem., 1965, 43, 1187.1143; T. Y. Liu and S. D. Elliott, Ndure, 1965, 208, 33.4600.S. H. Hussain and G. Lowe, Chern. C o r n . , 1965, 345504 BIOLOGICAL CHEMISTRYthat of cytochrome C from other species (noted in reference 95). A strikingcorrespondence exists between hog protein, the horse protein, and the humanprotein. In the sequences of the hog and the human protein, only threeamino-acids differ, the human protein possessing threonine a t Ser-47 ,glycine at Lys-60, and glycine at Thr-89 (see Figure 9, in which an asteriskAcetyl-Gly -Asp -Val-Glu-Lys-Gly -L ys-Lys- Ileu -Phe -Val+ -GluNH, + -Lys -CyS-Ah+-GluNH,-CyS -His-Thr-Val-Glu-Lys-Gly-Gly-Lys-His-Lys-Thr-Gly-10I I 20Pro -AspNH,-Leu-His-Gly-Leu-Phe-Gly-Arg-Lys-Thr-Gly-GIuNH,-Ala-Pro -30 40Gly -Phe* -Ser *-T yr -Thr -Asp f -Ah-AspNH,-Lys - AspNH, -Lp- Gly- Ileu-Thr + -Try-Gly*-Glu-Glu+-Thr-Leu-Met-Glu-Tyr-Leu-Glu-AspNH,-Pro-~ys-L~s-5060 70Tyr -1leu -Pro -Gly -Thr -Lys-Met -1leu -Phe -Ah+ - Gly -1leu -Lys-Lys -Lys -Gly * 9 i- -80Glu-A rg -Glu+ -Asp -L eu-Ileu-Ala-Tyr-Leu -Lys-Lys-Ala-Thr-AspNH,-GluC0,H90 100 104FIG. 9.Amino-acid sequence of hog heart cgtochrome C.95The basic residues are in italics, and the hydrophobic residues are in bold-facedtype.Residues marked with an asterisk are those that vary from residues in correspond-ing positions in the horse heart protein. Residues marked with a + sign differfrom the corresponding amino-acids in human cytochrome.denotes residues varying from those in corresponding positions in horseheart protein,96 and a + sign those differing from the corresponding amino-acids in human cytochrome 9 7 ) . These findings extend the pioneer experi-ments of Tuppy and his colleagues 98, Q9 which revealed that the functionalheme peptide in hog heart cytochrome C is identical to that in the horse andbeef proteins. The high degree of correspondence exhibited by cytochromeC from species widely differing in the evolutionary scale indicates that theproteins derive as the product of a single gene which has become modifiedonly slightly during development (see Table 1, ref.95). In this connexion itmay be noted that cytochrome C isolated from the hearts of the EasternDiamond Back rattlesnake ‘‘ Grotalus adamanteus ” possesses a very similarsequence to that of homologous vertebrate proteins.lO0 In addition to thespecies similarities exhibited by cytochrome C, cytochrome proteins isolatedin good yield from brain, kidney, liver, and skeletal muscle exhibit an iden-tical amino-acid sequence, g4 despite the wide differences in embryologicalorigin and function of these tissues.l*l The biosynthesis of cytochrome Cin the various tissues of a single species may therefore be considered to becontrolled by a single structural gene.The search for specific amino-acid residues involved in co-ordination of95D.G. Smyth, Ann. Reports, 1963, 60, 468.96 E. Margoliash, E. L. Smith, G. Kreil, and H. Tuppy, Nature, 1961, 192, 1125.97 H. Matsubara and E. L. Smith, J. Biol. Chem., 1963, 238, 2732.98 H. Tuppy and G. Bodo, Monatsh., 1954, 85, 1182.99 S . Paleus and H. Tuppy, Acta Chern. Xcand., 1959, 13, 641.l o o Om. P. Bahl and“E. L. Smith, J. BkE. Chem., 1965, 240, 3585.101 A. H. Leninger, The Mitochondrion,” Benjamin, Inc., New York, 1964SMYTH: PROTEINS AND PEPTIDES 505the iron atom in cytochrome C has been narrowed. The classical view is thatco-ordination occurs through the imidazole nitrogen of two histidine resi-dues.102 However, some evidence was obtained that the ligands in questionwere an imidazole group and the E-NH, group of a lysine residue; firstLys-13 (ref. 103) and then Lys-25 (ref.104) was implicated. It has nowbeen found that all of the lysine residues in cytochrome C can be guanidinatedwithout significant change in structure, and with complete retention of thea~tivity.10~ An involvement of lysine in one of the co-ordination positionsa,bout the heme iron of cytochrome C is therefore precluded.The problems involved in relating the function of a protein to its struc-ture are increased when the protein is capable of existing in a number offorms that retain activity. The usual preparative methods for cytochrome Cresult in the isolation of polymerised material, in addition to the monomericmolecule;106 the polymeric forms exhibit substantial, though reduced,electron transfer activity compared with the monomer. The monomer itselfappears to have two alternative configurations in alkaline solution,lO7 andthese can be distinguished by their ability to react with reducing agents.Multiple forms exist also in yeast cytochrome C lo8 but in this case the freesulphydryl group of a cysteine residue participates in dimer formation in amanner similar to the formation of dimers from monomers of Bence- Jonesprotein.Peptide hormones.-In view of the great interest in structure-functionrelationships among polypeptides, the development of rapid methods ofpeptide synthesis has assumed considerable importance.The method ofsolid-phase peptide synthesis,10g which involves stepwise assembly of thepeptide chain anchored covalently at one end to an insoluble particle, hasbeen adapted for performance by an automatic machine;llO the total syn-thesis of bradykinin (9-residues) was achieved in 32 hours. An alternativemethod has been presented in which the peptide is synthesised on a polymericsupport not in the solid phase but in solution.ll1 The resin employed waspolystyrene substituted by chloromethylation to provide attachment sitesfor the peptide. Since this resin is soluble in organic solvents and insolublein water, the whole range of synthetic techniques currently used in the fieldcan be applied to the resin-bound peptide, while removal of excess reactantsand side products is facilitated simply by precipitation in aqueous medium.High-molecular-weight carriers have been elegantly employed in thesynthesis of cyclic peptides.ll2 The NH,-protected peptide is coupled throughlo* H.Theorell, J . Amer. Chem. SOC., 1941, 63, 1820.loa E. Margoliash, N. Frohwirt, and E. Weiner, Biochem. J., 1959, 71, 559.lo* E. Margoliash, in Enzyme models and Enzyme Structure, Brookhaven Symposialo6 T. P. Hettinger and H. A. Harbury, Proc. Nat. Acad. Sci. U.S.A., 1964,52,1469.lot) E. Margoliash and J. Lustgarten, J . BioZ. Chem., 1962, 237, 3397.lo' C. Greenwood and G. Palmer, J . Biol. Chern., 1965, 240, 3660.lo* K. Motonaga, E. Misuka, E. Nakajima, S. Veda, and K.Nakanishi, J . Biochem.loS R. B. Merrifield, J . Amer. Chena. Soc., 1963, 85, 2149.110 R. B. Merrifield and J. M. Stewart, Nature, 1965, 207, 522.ll1 M. M. Shemyakin, Yu, A. Ovchinnikov, A. A. Kinyushkin, and I. V. Kozhevni-112 M. Fridkin, -4. Patchornik, and E. Katchalski, J . Anzer. Chem. SOC., 1965, 87,in Biology, 1962, No. 15, p. 266.(Japan), 1965, 57, 22.kova, Tetrahedron Letters, 1965, 2323.4646.506 BIOLOGICAL CHEMISTRYQ N 0 2?p Pc=oINHZI OH 0IC=G y C = OI IPe2 0 PepI r w : AHits C0,K-group to the OH-group of poly-4-hydroxy-3-nitrostyrene, or to theOH-group of a branched copolymer of DL-lysine and 3-nitro-~-tyrosine.The intramolecular acylation reaction takes place in good yield, liberatingthe anticipated products. It should be noted that previous methods, whichinvolve cyclization a t high dilution, have provided cyclic peptides onlyin poor yield.The total synthesis of the A- and B-chains of insulin by the classicalmethods of peptide synthesis, reported last year (cf.ref. 60), has now beendescribed in full detail,l13 though the problem of correctly coupling the disul-phide bridges remains to be solved. An important advance has been madeby Zervas and his colleagues,114 who have successfully synthesized a frag-ment of the A-chain containing the intrachain disulphide bridge. Theremaining difficulty of correctly coupling the interchain disulphide bridgesmay be overcome by an adaptation of the peptide cyclization methodpresented by the Weizmann Institute group.l12 Further recent achieve-ments in the field of peptide synthesis include the total synthesis of the S-peptide (20 residues) of ribonuclease l l .5 and the synthesis of the fragments ofglucagon (29 residues) .l16Angiotemh-The availability of a specific antibody directed against apeptide of small size and known amino-acid sequence should facilitate studieson the structure of antibody y-globulin and lead to an understanding of theexact requirements and relationships of antigenic determinants in the peptidemolecule. Angiotensin has been coupled through its terminal NH 2-group,and also through its terminal C02H-group, to poly-I;-Iysine (Figure 10) ;l17both polymers were effective in eliciting the production of antibody withconsiderable affinity for angiotensin. The specificity of the binding wasllS H. Zahn, H. Bremer, W. Svoka, and J. Meienhofer, 2. Naturforsch., 1965, 20b,646; R. Zabel and H. Zahn, ibid., p. 650; H. Zahn, H. Bremer, and R. Zabel, iM.,p. 653; J. Mcienhofer and E. Schnebel, ibid., p. 661 ; H. Zahn, 0. Brinkhoff, J. Meienhofer,E. F. Pfeiffer, H. Ditschunett, and C. Gloxhuber, ibid., p. 666; Niu Chin-i, Kung Yueh-ting, Huang Wei-teh, Ke Lin-tsung, Chen Chan-chin, Chen Yuan-chug, Du Yu-cang,Jiang Rong-qing, Tsuo Chen-lu, Hu Shih-chuan, Chu Shang-quan, and Wang Keh-zhen, Sci. Sinica, 1965,14, 1386; Wang-Yu, Hsu Je-Zen, Loh Jen-yung, Huang Hang,and Huang Jing-jian, ibid., p. 1284.114 K. Zervas, I. Photaki, A. Cosrnatos, and D. Borovas, J. Amer. Chem. SOC., 1965,87, 4922.116 K. Hofmann, R. Schmiechen, R. D. Wells, V. Wolman, and Yanaihara (positions1-7); and other references, J . Amer. Chem. SOC., 1965, 87, 611.1l8 E. Wunsch and A. Zwick, Chem. Ber., 1964, 97, 2497; ibid., pp. 3298, 3305,3312; E. Wunsch and G. Wendelberger, ibid., p. 2504; E. Wunsch, F. Drees, and J.Jentsch, Chem. Ber., 1965, 98, 797, 803; K. Lubke and E. Schroder, Annalen, 1966,681, 231, 241, 250; E. Schroder, ibid., 1965, 088, 250.117 E. Haber, L. B. Page, and G. A. Jacoby, Biochemistry, 1965, 4, 693SMYTH: PROTEINS AND PEPTIDESI I[Lys-NHd,,I Lys-N-C-Phe-Pro-H is-Val-Tyr-Val-Arg-Asp-NH2I NH? 1 A&[Lys-NH?),I Lys-N-C-Phe-Pro-His-Val-Tyr-Val-Arg-Asp-NH2; I l l Ii H O NH2- A-(Phe)-PL507[tys-NH& 0Lys-N-C-N-C-CsH 4-C-N-C-NH-Asp-Arg-VaI-Tyr-VaI-His-Pro-Phe-C I HOHH2 H:HOI 0I // Lys-N-C-N-C-C~HcC-N-C-NH-Asp-Arg-Val-Tyr-Va1-His-Pro-Phe-Ci I I I I I I 1 llI //NH2 1 ‘OHNH2 1 ‘OHI l l 1 1 I I IIILys-NH&: H O H H a HcHOPL-(Am)-AFIG. 10. Schematic representation of angktensin poly-~-lysine wpolymers.[Reproduced, with permission, from E. Haber, L. B. Page, and G. A. Jacoby,Biochemistry, 1965, 4, 693.1demonstrated by displacement of 14C-labelled angiotensin from the antibodyby unlabelled angiotensin and by the inability of excess of bradykinin,vasopressin, and other peptides to compete in the binding process. Thus,although itself incapable of eliciting antibody formation, angiotensin iscapable of binding to an antibody produced to an angiotensin polymer. Thespecificity and the high affinity of the binding between angiotensin and theantiserum recommend an application of this peptideprotein interaction inan immuno-assay of the hormone.The polymer containing angiotensin coupled through its NH,-group wasreported to exhibit no biological activity, whereas the polymer coupledthrough the angiotensin C0,H-group retained substantial bioactivity.These findings in both cases codict with the currently held views on theimportance of these groups in the activity of the hormone. Furthermore,the C0,H-coupled angiotensin polymer exerted the same biological activityon smooth muscle when tested in the presence or absence of its specscantiserum. The affinity exhibited by the angiotensin polymer for its anti-body, though high, must be considerably less than its affinity for the physio-logical receptor. It may be noted that antibodies formed against extendedangiotensin and bradykinin molecules have been reported from two otherlaboratories; 118, 119 in these cases the carrier protein was itself antigenic andthe precise points of attachment between the peptide and the carrier proteinwere not determined.118 S. D. Deodhax, J. E x p Med., 1960, 111, 429.T. L. Goodfriend, L. Levine, and G. D. Fasman, Science, 1964, 144, 344508 BIOLOGICAL CHEMISTRYBrabkinh-The peptide hormone, bradykinin, is derived physiologicallyas a fragment of an inactive precursor polypeptide. It is therefore to beexpected that the coupling of amino-acids to the NH,-terminus of bradg-kinin, by peptide synthesis, would result in the formation of inactive or lessactive analogues. Thus, the naturally occurring hendecapeptide, methionyl-lysylbradykinin, exhibits one third of the activity of bradykinin,120 and thisMet *Lys.Arg*Pro*Pro*Gly*Phe-Ser *Pro*Phe*Argpeptide with t-butyloxycarbonyl blocking substituents attached at thefree NH,-groups possesses 1/150th the activity of bradykinin.121 Similarly,the addition of one oxygen atom to Met. Lys- Bradykinin, forming the sul-phoxide, was without effect on bioactivity, whereas the addition of twooxygen atoms, forming the sulphone, resulted in a 10-fold reduction inactivity.121There have been two reports on the isolation of almost pure brady-kininogen,122, 123 the precursor protein from which bradykinin is released byenzyme action in serum. Difficulty was encountered because smooth musclecontracting substances, presumably bradykinin and its relatives, werereleased during the purification procedure and severely depleted the yieldof bradykininogen. Perfunctory evidence was presented that trypsin andvenom enzyme acted on bradykininogen to release bradykinin, whereas hogpancreatic kallikrein released lysylbradykinin ( kallidin).124 A polypeptideinhibitor of kallikrein has recently been isolated from bovine lung. Theestablishment of its amino-acid sequence (58 residues) 124 (see Figure 11)revealed that its structure is identical with that of trypsin inhibitor fromborine p a n ~ r e a s . ~ ~ ~ - l ~ 'NH,Arg-Pro *Asp*Phe*Cys*Leu*Glu *Pro *Pro *Tyr *Thr*Gly -Pro *Cys *Lys *Ala*Arg*30NH220 ITleu.Ileu*Arg*Tyr*Phe*Tyr.Asp*Ala*Lys*Ala,*Gly*Leu*Cys *Glu *Thr -Phe-50H,N NH, NH240 I 1 IVal*Tyr *Gly -Gly*Cys -Arg*Ala*Lys *Arg *Asp 'Asp *Phe *Lys *Ser *Ala-Glu *AspCys-Met *Arg*Thr *Cys *Gfy *Gly *Ala.CO,H58FIG. 1 I. Amino-acid sequence of EaEliErein inhibitor from bovine120 D. F. Elliott and G. P. Lewis, Biochem. J . , 1965, 95, 437.121 V. Uhlinger, Experientia, 1965, 21, 271.lZ2 E. Habermam, W. Klett, and G. Rosenbusch, 2. physiol. Chem., 1963, 322,123 T. Suzuki, Y. Mizushima, T. Sato, and S. Iwanago, J . Biochem. (Japan), 1965,124 F. A. Anderer, 2. Naturforsch., 1965, 206, 462; ibid., p. 499.125 J. Chauvet, G. Nouvel, and R. Acher, Biochim. Biophys. Acta, 1964, 92, 200.lZ6 B. Kaasell, M. Radicevic, 31. J. Ansfield, and M. Laskowski, Biochem. Biophys.1 2 7 V. Dlouha, D. Popsilova, B. Meloun, and F. Sorrn, CoZZ. Czech. Chem. Comm.,121.57, 14.Res. Comm., 1965, 18, 255.1965, 30, 1311SMYTH: PROTEINS AND PEPTIDES 509This finding adds doubt to the current view that certain serum enzymes,such as kallikrein, bradykininase, and angiotensinase, are characteristicproteins with the specific functlion implicit in their titles. It seems probablethat these enzymes are in fact trypsin, carboxypeptidase, and leucineaminopeptidase, which exert their proteolytic activities against a broadspectrum of substrates.p-Lipotropic Hormone.-Several polypeptides have been demonstratedto possess in vitro lipotropic activity. These include growth, adrenocor cico-trophic, and a- and P-melanocyte-stimulating hormones. An additionallipotropic peptide has now been isolated from sheep pituitary glands, andits complete amino-acid sequence has been determined 8 (Figure 12).NH2IKH ,-Glu-Leu-Gly-Thr-G1u-Arg-Leu-Glu-Glu-Ala-Arg-Gly-Pro-Glu-Ala-1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5Ala-Glu-Glu-Ser-Ala-Ala-Ala-Ala-Arg-Ala-Gl~~-Leu-Glu-Tyr-Gly-Leu-Val-1G 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31NH,IAla-Glu-~~Glu-Ala-Ala-Glu-Lys-Lys-~p-Ser-Gly-Pro-T~-Lys-Met -32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47Glu-His-Phe-Arg-Try-Gly-Ser-Pro-Pro-Lys-Asp-Lys-~g-Tyr-Gly-Gly-45 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63NH,IPhe-~~et-Thr-Ser-Glu-Lys-Ser-Glu-Thr-Pro-Leu-Val-Thr-Leu-Phe-Lys-64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79NH2 NH2 NH2I 1 IAsp-Ala-Ileu-Lys-Lys-Asp-His-Ala-Lys-Gly-Glu-CO,H80 81 82 83 84 85 86 87 88 89 90FIG. 12. The antino-acid sequence of the sheep pituitarJ fi-LPH.aThe sequence Met*Glu*His*Phe*Arg*Try*Gly* is common to ACTH andMSH peptides from a variety of species; the sequence 37-58 in @-LPH isidentical t o that of human a-MSH except that Ser-42 and Lys-46 are re-placed by Glu and Arg respectively in the human hormone