Sorption of Water Vapour by some Derivativesof Bovine Serum AlbuminBY COLIN H. ROCHESTER* AND A. VALERIE WESTERMANChemistry Department, The University, Nottingham NG7 2RDReceived 12th May, 1976Isotherms for the sorption of water vapour by succinyl, acetyl, amidino, methyl and carbodiimidederivatives of bovine serum albumin at 298 K have been determined gravimetrically. The effects ofthe specific chemical modifications on the uptake of water by B.S.A. are discussed by considerationof the groups in the protein which are possible sites for water sorption. Adsorption onto both sidechain polar or ionic groups and main chain peptide groups occurs. Release of main chain peptidegroups from the a-helical conformation enhances their capacity to sorb water.A study of the sorption of water vapour by five samples of bovine serum albuminisolated at different pH values in the range 2.0-10.2 has suggested that the sequenceCOO- > NH; > NH2 > COOH reflects the relative amounts of water sorbed at aparticular relative vapour pressure by the side chain groups in B.S.A.I The existenceof -COO-.+H3N- intermolecular salt bridges in the " dried " native proteindecreased water uptake.Water was also sorbed onto main chain peptide groupsparticularly when the latter were not involved in the hydrogen bonding interactionswhich lead to a-helical structure. The present paper reports a further attempt toseparate the effects of side chain groups, peptide bonds and structure on the sorptionof water by B.S.A. A series of chemically modified derivatives of B.S.A.have beenprepared and their interactions with water have been studied gravimetrically.EXPERIMENTALThe preparation of chemically modified derivatives of B.S.A. (source as before)' wasfollowed by dialysis against deionized water and freeze drying in a Chemistry LaboratoryInstrument SB4 freeze dryer. The products were stored over phosphorus pentoxide at273 K. Preparation details were as follows.SUCCINYLATED B . s . A . ~B.S.A. (10 g) was dissolved in water (500 cm3) and the pH was adjusted to 7.5 by additionof aqueous NaOH solution (0.2 mol dm-3). Succinic anhydride (25 g) was added in aliquotsover 2 h. The pH was maintained at 7.5 throughout the experiment by addition of aqueoussodium hydroxide (5 mol dm-3).The product is referred to as succinylated B.S.A. I. Fourother succinyl derivatives of B.S.A. were prepared by a similar procedure but with differingratios of B.S.A. to succinic anhydride as follows :mol anhydride/derivative mol lysinesuccinylated B.S.A. I 26.7succinylated B.S.A. IT 26.7succinylated B.S.A. IV 4.0succinylated B.S.A. I11 12.0succinylated B.S.A. V 1 .o1-2 33weight of weight of succinicB.S.A.lg anhydridejg10 25.05. 12.55 5.615 1.875 0.4634 WATER-I-B.S.A. DERIVATIVESACETYLATED B.S.A.3B.S.A. (10 g) in water (300 cm3) was cooled to 277 K and the pH adjusted to 7.5 by theaddition of aqueous sodium hydroxide (1 rnol dm-3). Acetic anhydride was added sIowIyover 2 h and the pH maintained at 7.5 with aqueous sodium hydroxide (1 rnol dm-3).AMID IN ATE D B.S.A.4B.S.A.(5 g) in water (500 cm3) was cooled to 273 K and the pH was adjusted to 8.3 withaqueous sodium hydroxide (5 rnol dm-3). Ethyl acetimidate (80 g) was added over 2 h thepH being kept within the range 8.3-8.6 with aqueous sodium hydroxide (5 rnol dm-3).Owing to its instability ethyl acetimidate was prepared immediately before use.METHANOL ESTERIFIED B.s.A.~B.S.A. (10 g) was stirred at room temperature for 3 weeks with a mixture of methanol(1 dm3) and aqueous hydrochloric acid (10 cm3, 35 % wlv). The solid product was separatedby filtration, washed with ether, dried and slurried with water (400 cm3) and the pH of thesolution adjusted to 6.0 by addition of aqueous sodium hydroxide (2 rnol dm-3).Theresulting derivative of B.S.A. after dialysis and freeze drying is referred to as methanolesterified B.S.A. I. Three further derivatives were prepared by altering the concentration ofcatalyst and time of reaction as follows :concentration of time ofderivative HCl catalyst/mol cm-3 reactionlweekmethanol esterified B.S.A. I 0.10 3methanol esterified B.S.A. I1 0.15 1methanol esterified B.S.A. I11 0.15 3methanol esterified B.S.A. IV 0.15 4CARBODIIMIDE DERIVATIVE OF B.S.A?Ethylene diamine (15 g) was added slowly to B.S.A. (5 g) in water (250 cm3) followed byethyl dimethylaminopropyl carbodiimide hydrochloride (5 g). The pH was kept at 4.7 withaqueous hydrochloric acid (3 rnol dmW3). The reaction was stopped after 2% h by theaddition of an equimolar mixture (I rnol dm-3) of aqueous sodium acetate and acetic acidat pH 4.7.HYDROXY-ACETYLATED B.s.A.*Acetic anhydride (6.7 cm3) was added to B.S.A.(5 g) in trifluoroacetic acid (70 cm3) at273 K and after 20 min the solution was poured into ice cold water. The precipitate waswashed with acetone (3 x 150 cm3) and ether (3 x 150 cm3), dried in vacuo and suspended inwater (300 c1n3). The pH of the solution was adjusted to 7.0 with aqueous sodium hydroxide(2 rnol dm-3).The a-helix, sodium ion and chloride ion contents of the B.S.A. derivatives were deter-mined as b e f ~ r e . ~ During the chloride analyses some of the derivatives gave a precipitateafter addition of the ferric alum and mercuric thiocyanate solutions. The precipitates wereremoved by centrifugation for 10 min.The extents to which B.S.A.had been modified during the preparation of derivatives wereestimated by n.m.r. analysis of hydrolysates of the derivatives.21 The n.m.r. spectrum ofB.S.A. in water consists of broad bands. The modified proteins were, therefore, firsthydrolysed because it was known that hydrolysis would decrease band widths and thusfacilitate analysis and interpretation of the relevant band positions and intensities. Thederivatives (1 g) were hydrolysed by refluxing (3 h) with concentrated aqueous HCI (6 cm3).Spectra were recorded with a Varian HA-100 spectrometer locked onto the water signal asthe reference standard. The analysis of the spectra is exemplified by the results for hydro-lysed B.S.A.and hydrolysed acetylated B.S.A. (fig. 1). The bands between 8.82 and 9.77may be assigned to the methyl groups of leucine, isoleucine, and valine, and are sufficientlC. H. ROCHESTER AND A. V. WESTERMAN 35well resolved from the rest of the spectrum to act as an internal intensity standard. Thereare 0.1911 mol of leucine, isoleucine and valine per mole of residues in B.S.A. Each aminoacid has six methyl H-atoms and, therefore, the intensity of the band between 8.8 z and 9.7 zwill be equivalent to 1.15 H-atoms per residue in B.S.A. Comparison of the spectra in fig. 1shows increases in absorption intensity between 6 2 and 72 in the spectrum of acetylatedB.S.A. due to the appearance of bands which may be assigned to H-atoms in acetyl groups.7FIG.1 .-N.m.r. spectra of (a) hydrolysed B.S.A. and (b) hydrolysed acetylated B.S.A.Measurement of the integrated intensities IA, IB, Ic, and ID in regions A, B, C and D re-spectively of the spectra (see fig. 1) enabled calculation [eqn (l)] of A1 the increase in intensityof bands due to the presence of acetyl groups.Hence the number of moles of acetylated residues per mole of total residues in B.S.A. wasgiven by (AI/3)(1.15/IB).The uptake of water vapour by the B.S.A. derivatives at 298 K was determined gravi-metrically as bef01-e.~A1 = Ic- IA(IDIIB). (1)RESULTSDetails of the characterization of the derivatives of B.S.A. are given in table 1.Molecular weights of the products were estimated as b e f ~ r e . ~ The modifications toB.S.A.which were carried out were chosen because of their specificity towards certainside chain functional groups. The extents of reaction are, therefore, quoted withrespect to the particular amino acid residues whose side chain groups were beingmodified. The number of residues of each amino acid in one molecule of B.S.A.was evaluated from the data of Spahr and Edsall lo and a molecular weight of 66 000for unmodified B.S.A.I36 WATER+ B.S.A. DERIVATIVESTABLE DETAILS OF THE CHARACTERIZATION OF THE DERIVATIVES OF B.S.A. THE QUOTEDSODIUM AND CHLORIDE CONTENTS ARE THE NUMBER OF IONS PER MOLECULE OF PROTEINB.S.A. derivative residues modified?bovine serum albumin -succinylated B.S.A. I 100 % Lys70 % Tyrsuccinylated B.S.A. I1 100 % Lys100 % Tyrsuccinylated B.S.A.111 100 % Lys70 % Tyrsuccinylated B.S.A. IV 95 % Lyssuccinylated B.S.A. V 20 % Lysacetylated B.S.A. 100 % Lys100 % Tyramidinated B.S.A. 85 % Lysmethanol esterified B.S.A. Imethanol esterified B.S.A. 11methanol esterified B.S.A. 111methanol esterified B.S.A. IVcarbodi-imidederivative of B.S.A.hydroxy-acetylated B. S. A.22 % Asp+ Glu16 % Asp+Glu3 1 % Asp+ Glu31 % Asp+Glu78 % Asp+ Glu15 % Ser+Thr100 % Tyrt f. -2 :<.Nafcontent7137731161013044105343111c1-content0000000003750521160molecularweight66 00075 30076 00075 30073 20067 40071 30068 50069 00068 80069 00069 10074 30067 500% helicalcontent54113214175331511922201616i** Not sufficiently soluble for O.R.D.analysis.The equilibrium uptake of water by the B.S.A. derivatives as a function of therelative vapour pressure of water (Po = 3159.5 N m-2 at 298 K) is given in table 2.The percentage regain equals one hundred times the weight of water sorbed dividedby the weight of “ d r y ” protein. The data are also converted to the number ofwater molecules sorbed per molecule of protein.TABLE 2.-GRAVIMETRIC DATA FOR THE SORPTION OF WATER VAPOUR BY B.S.A. AND THIRTEENDERIVATIVES OF B.S.A. AT 298 KR.V.P. %regain -ELFEZ, mol protein R.V.P. %regain -ELEE- R.V.P. %regain mol protein mol protein(a) succinylated B.S.A. I0.026 1.75 730.060 3.41 1430.148 5.31 2220.181 5.96 2490.278 7.71 3230.353 9.26 3870.516 15.09 63 10.628 19.37 8100.709 24.56 10270.774 32.10 13430.843 46.94 19640.928 71.90 3008(b) succinylated B.S.A. I10.051 1.09 460.082 1.97 830.091 2.33 980.196 4.63 1960.395 7.22 3050.499 9.74 4120.665 13.28 5610.730 15.74 6650.829 20.51 8660.903 24.84 10490.977 45.40 1917(c) succinylated B.S.A.1110.030 1.88 790.098 4.31 1800.156 5.35 2240.288 8.39 3510.428 12.04 5040.601 17.23 7210.700 20.84 8720.737 25.26 10570.788 29.84 12480.839 35.42 14820.944 63.80 266C. H. ROCHESTER AND A. V. WESTERMAN 37TABLE 2.-contd.R.V.P. %regain -EELEEL mol protein(d) succinylated B.S.A. IV0.019 1.29 530.048 2.46 1000.072 3.47 1410.099 4.42 1800.205 6.12 2490.351 8.81 3580.431 11.28 4590.575 15.30 6220.646 18.87 7670.726 22.41 91 10.762 23.82 9690.789 27.08 11010.816 29.35 11930.942 56.57 2296(9) amidinated B.S.A.0.065 3.06 1240.106 4.07 1550.214 6.28 2390.307 6.73 2560.404 8.36 3180.474 10.32 3930.601 12.57 4780.690 13.78 5240.755 16.91 6440.816 23.58 8970.884 27.64 1052( j ) methanol esterified0.0230.0570.1210.1650.2960.3870.4880.5420.5800.6570.7780.8340.8500.8750.9080.942B.S.A. I111.44 552.18 842.80 1073.53 1355.67 2176.50 2498.80 3379.63 3 6910.53 40412.54 48116.00 61318.65 71 520.27 77723.78 91228.40 108931.50 1208(m) hydroxy-acetylated B.S.A.0.017 0.94 350.052 2.19 820.093 3.33 1250.194 4.86 1820.319 6.74 253R.V.P.%regain mol protein(e) succ0.0500.0960.1690.3090.4590.5640.6520.7040.7580.8310.8810.952:inylated B.S.A. V1.07 402.64 993.73 1406.69 2519.54 35712.41 46515.28 57217.50 65519.86 74422.54 84425.92 97147.36 1773(h) methanol esterified0.072 1.87 720.101 2.85 1090.209 4.51 1730.279 5.64 2160.456 7.89 3020.467 8.44 3240.638 12.02 4610.711 14.08 5400.848 19.54 7490.870 22.04 8450.885 23.28 8920.952 28.52 1093B.S.A. I(k) methanol esterified0.0370.0800.1290.1840.2980.4070.4590.6180.7170.7930.8670.9390.9640.3770.5240.6380.6930.816B.S.A. IV1.36 522.38 913.45 1324.28 1 645.59 2157.30 2808.58 32910.89 41812.82 49215.51 59518.91 72622.60 86828.93 11127.87 29510.99 41212.84 48214.75 55318.92 710R.V.P.%regain - mol protein(f) acetylated B.S.A.0.026 1.59 630.056 2.85 1130.146 4.44 1760.257 6.30 2500.333 7.30 2890.411 9.42 3730.556 11.78 4670.663 14.76 5850.788 19.75 7820.894 28.95 11470.942 36.70 1454(i) methanol esteB.S.A. I10.046 1.760.090 3.580.219 5.320.323 6.560.407 8.160.530 10.720.611 12.560.698 14.650.785 17.600.825 20.480.884 23.630.898 26.580.919 29.98:rified6713720325 131241048056067378390310161146(I) carbodi-imide derivative of0.0210.0530.1270.1440.2290.3340.4370.5270.6010.7330.8270.9380.8190.8780.9100.943B.S.A.1.28 532.88 1194.04 1674.69 1945.69 2357.50 3109.57 39511.78 48614.62 60418.72 77325.21 104146.08 190319.19 72021.74 81524.18 90727.63 10338 WATER -k B.S.A.D ERI VAT1 VESDISCUSSIONMETHANOL ESTERIFIED B.S.A.Brodersen et aE.I2 reported that methylation of the carboxyl side chain groups ofhuman serum albumin had little effect on the extent of water uptake by the protein,and concluded that water was sorbed through interactions with main chain peptidegroups rather than with polar side chains. In contrast, Watt and Leeder found thatmethylation of keratin decreased water uptake, as would be expected if water sorption,at least in part, involves interactions between water molecules and specific polar sidechain groups in the protein. In the present work methylation of B.S.A.also causeddecreases in water uptake for all four methanol esterified derivatives over the entirerange of relative vapour pressure which was studied. The isotherms for two of thederivatives are compared with that for B.S.A. in fig. 2. The isotherms for the othertwo derivatives were similar but are omitted from the figure for clarity.Trelative vapour pressureFIG. 2.-Isotherms for the sorption of water vapour by (a) B.S.A.l, (b) methanol esterified B.S.A. 11,and (c) methanol esterified B.S.A. IV, at 298 M.The four methanol esterified derivatives of B. S. A. had similar a-helical contentswhich were appreciably less than the a-helical content of the unmodified protein(table 1). The release of main chain peptide groups from the a-helical conformationshould lead to an increase in water uptake l4 if sorption by the peptide groups wasthe predominant influence on the water sorption isotherm.This is the reverse of theexperimental result. The decrease in water uptake observed must be primarily dueto the replacement of charged -COO- side chain groups in the glutamic and asparticacid residues by uncharged -COOMe groups. Measurements of isotherms for thesorption of water by sodium poly-L-glutamate and by methanol esterified poly-L-glutamic acid have shown that the uptake of water by -COO- groups is much greateC. H. ROCHESTER AND A. V. WESTERMAN 39than that by -COOMe groups over the entire water vapour pressure range.9Detailed comparison of the results for B.S.A. and for the poly-L-glutamic acidderivatives shows that the decreases in water uptake per side chain group modified inpassing from B.S.A.to methanol esterified B.S.A. were similar in magnitude to thecorresponding decreases in passing from sodium poly-L-glutamate to methanol-esterified poly-L-glutamic acid. It must be concluded as before 1 9 that the inter-action between water and ionic or polar side chain groups in proteins has a majorinfluence on the total water sorption isotherm.CARBODIIMIDE DERIVATIVE OF B.S.A.The carbodiimide modification of B.S. A. using ethylene diamine as reagentresulted in the conversion of carboxyl side chain groups to -CO . NH . CH,CH,NHigroups which contain both a peptide group and a protonated amino group.' Anappreciable loss of a-helical structure also occurred.The replacement of one ionicside chain group by another, the release of main chain peptide groups from the helicalconformation, and the generation of side chain peptide groups, would probably beexpected to lead to the increase in water uptake which is observed (fig. 3) providingwater sorption onto both side chain and main chain peptide groups contributed to theoverall isotherm. Previous studies of water sorption by polypeptides and by B.S.A.derivatives have shown that side chain -NHZ groups do not adsorb as much waterat a given vapour pressure as do -COO- groups. The increased uptake of water byrelative vapour pressureFIG. 3.-Isotherms for the sorption of water vapour by (a) carbodiimide derivative of B.S.A., (b) 0acetylated B.S.A., 0 B.S.A., and (c) hydroxy-acetylated B.S.A., at 298 K40 WATER+B.S.A.DERIVATIVESthe carbodiimide derivative of B.S.A. may have arisen because the high extent ofmodification led to an increase in the total number of ionic sites as a result of thedisruption of intramolecular -COO-. +H3N- salt bridges. However, it is morelikely that the predominant influence arises from the generation of side-chain peptidegroups which, as before, act as adsorption sites contributing appreciably to theisotherm particularly at moderate and high humidities. It would be illogical tosuggest that side chain peptide groups act as adsorption sites, whereas those mainchain peptide groups which are not involved in any intramolecular hydrogen bondinginteractions do not.Adsorption of water onto main chain peptide groups, particularlyisolated peptide groups,14 must, therefore, contribute to the overall uptake of waterby proteins.HYDROXY-ACETYLATED B.S.A.Acetylation of all the tyrosine residues and 15 % of the serine and threonineresidues in B.S.A. led to a decrease in the uptake of water by the protein at all vapourpressures (fig. 3) and a product which was insoluble in water. The O.R.D. spectrumof the derivative could not, therefore, be determined and its a-helical content isunknown. However, comparison of the isotherms [fig. 3(b) and 3(c)] suggests thatside chain acetate groups in hydroxy-acetylated B. S. A. are weaker sorption sites thanthe corresponding unmodified hydroxy-groups in B.S.A.Interactions between waterand side chain hydroxy-groups must contribute to the overall isotherm for thesorption of water by the protein. Adsorption of water onto both aliphatic (serine andthreonine) and aromatic (tyrosine) hydroxy-groups has previously been characterizedby analogous studies involving keratin l 5 to those described here involving B.S.A.The results from the two studies are similar.concluded that at 50 % relative humidity each aliphatic and aromatic hydroxy-groupin keratin was associated on average with 0.34 and 1.0 water molecules respectively.The loss of all the tyrosine OH-groups and 15 % of the serine + threonine OH-groupsfrom B.S.A. would be expected, if these figures were applicable for B.S.A. as well askeratin, to give a decrease in water uptake of 23 molecules per molecule of protein.The experimental isotherms [fig.3(b) and (c)] differed by 27 molecules of water permolecule of protein. This comparison can only be qualitative, as it is a gross over-simplification, but it does show that there is a measure of agreement, at least up tomoderate humidities, between the present results for B.S.A. and previous data forkeratin.15 At high humidities the decrease in water uptake caused by the hydroxy-acetylation of B.S.A. was considerably greater than the losses caused by modificationof the hydroxy-groups in keratin. Perhaps in B.S.A. the presence of side chainhydroxy-groups promotes the build up of aggregates of water molecules at high vayourpressures of water.For example, Watt and LeederA MIDI N ATE D B.S.A.The amidination reaction4 led to 85 % replacement of the -NH2 groups oflysine residues by -NH .C(=NH)CH3 groups which are more strongly basic.16The helical, sodium and chloride contents of the product were similar to those for theunmodified protein. The slightly increased uptake of water at low vapour pressures(fig. 4) suggests that the primary adsorption of water onto -NH.C(=NH,+)CH,sites is more favourable than adsorption onto -NH; groups. At moderate vapourpressures the effect was reversed by up to a maximum decrease in water uptake at70 % relative humidity of - 1.3 water molecules per lysine residue modified. Thismight have arisen because the tetrahedral structure of the -NHZ ionic group waC.H. ROCHESTER A N D A. V. WESTERMAN 41more likely than the planar structure of the =NHZ group to favour the build up ofaggregates of water molecules around the ionic sites.3relative vapour pressureFIG. 4.-Isotherms for the sorption of water vapour by (a) B.S.A. and (6) amidinated B.S.A., at 298 K.ACETYLATED B.S.A.N.m.r. spectra of the hydrolysate of acetylated B.S.A. showed that the acetylationreaction had produced both N-acetylated lysine and 0-acetyl tyrosine in 100 % yield.Some di-acetylated lysine or 0-acetylated serine and threonine were probably alsoformed. The 44 sodium ions present in acetylated B.S.A. balanced the charge ofionized carboxylic acid groups. Despite the high extent of modification produced bythe acetylation reaction the water sorption isotherm for acetylated B.S.A.was identicalto that for B.S.A. over the entire water vapour pressure range (fig. 3).Brodersen et a1.12 found that acetylated and unmodified human serum albumingave similar water sorption isotherms, and concluded that water was bound to mainchain peptide groups rather than to polar side chain groups in the protein. However,water uptake by proteins was decreased by acetylation of collagen,17 silk fibroinand keratin l 3 and by benzoylation of casein.18 The decreases have been ascribedto changes involving hydration of the main polypeptide chain1' or of side chaingroups. 3 9 A study of water sorption by poly-L-lysine, poly-L-lysine hydrobromideand acetylated poly-L-lysine showed that side chain -NH .CO . CH3 groups interactless strongly than either -NH2 or -NH: groups with water.g However, atmoderate and high vapour pressures appreciable adsorption of water occurred ontoboth main chain and side chain peptide groups in acetylated poly-L-lysine. Thepresent results for B.S.A. probably arise because of several opposing effects. Theacetylation of tyrosine side chain hydroxy-groups decreases the uptake of water at allvapour pressures [fig. 3(c)]. Replacement of -NH, or -NHi groups (the latter arezwitterionic with --COO- groups in the dry protein) with -NH. CO. CH3 groupsshould decrease water sorption at low hun~idities,~ The elimination of -NH2 group42 WATER-I-B.S.A. DERIVATIVESwill also prevent zwitterionic interactions between -NH2 and --COOH groups and,therefore, give a decrease in water uptake due to the conversion of some -COO- to-COOH groups. This effect will be opposed by the presence of 44 sodium ionswhich balance the charge of 44 -COO- groups in acetylated B.S.A.Also-COO-.+H3N- salt bridges 19* 2o will be destroyed to give two hydrophilic sidechain sites, -NIP. CQ . CH3 and -COOH (or -COO-). The loss of helical content(table 1) caused by acetylation generates isolated main chain peptide groups l4 whichcan act as sites for water sorption particularly at moderate and high vapour pre~sures.~Thus, although the result for acetylated B.S.A. is apparently simple, there is ampleevidence l* that the explanation is probably complex. To discuss the result solelyin terms of water-protein interactions involving main chain peptide groups is in-adequate.Adsorption of water at both side and main chain groups must occur if theisotherm is to be satisfactorily rationalized.SUCCINYLATED B.S.A.The isotherms for the five succinylated derivatives of B.S.A. also require explana-tions involving several effects. Two derivatives (11, V) sorbed less, or about the sameamounts, of water than B.S.A. whereas three derivatives (I, 111, IV) sorbed more waterover the entire vapour pressure range. Some crossing of isotherms occurred as therelative abilities of the derivatives to sorb water differed slightly at low, moderate andhigh humidities. A detailed discussion of the results is unnecessary as it would onlyreiterate at length concepts which have already been presented in this and previousdiscussions.' 9 However, certain general implications of the data are worth noting.Succinylation resulted in the conversion of side chain -NH2 to-NH. CO.CHZCHZCOOHgroups (or their charged counterparts) and of tyrosine aromatic hydroxy-groups to-0. CQ . CH2CH2COOH groups. The relative abilities of the derivatives to sorbwater gave no obvious correlation with the numbers of lysine or tyrosine side chaingroups which had been modified. However, correlations were observed betweenwater uptake and both the sodium ion contents and the a-helix contents of the succinylcompounds. The relative amounts of water sorbed by the compounds were in thesequences I > I11 > IV > I1 > V at low humidities, Z 21 111 > IV > I1 V atmoderate humidities, and I > I11 > IV > V > I1 at high humidities.The sodiumcontents were in the similar sequence I > I11 > IV > I1 > V whereas the a-helicalcontents were the exact reverse I < 111 < IV < I1 < V. The isotherms, therefore,give further support for the conclusions 1* that side-chain charged carboxyl groupsconstitute strong adsorption sites for water and that water sorption is also enhancedby the loss of helical content in a protein. Main chain peptide groups constitutewater-sorption sites particularly when they are not involved in intramolecular hydrogenbonding interactions. l 4The authors thank the S.R.C. and Unilever Ltd for financial assistance andDr. B. Rossall for helpful discussions.C. H. Rochester and A. V. Westerman, J.C.S. Faraday I, 1976, 72,2498.I. M. Klotz, Methods EnzyrnoL, 1967, 11, 576.J. F. Riordan and B. L. Vallee, Methods Erzzymol., 1967, 11, 565.L. Wofsy and S. J. Singer, Biochem., 1963,2, 104.S. M. McElvain and J. W. Nelson, J. Amer. Chern. SOC., 1942, 64, 1827.K, L. Carraway and D. E. Koshland, Methods Enzymol., 1967, 11, 616.ti H. Fraenkel-Conrat and H. S . Olcott, J. Biol. Chern., 1945, 161, 259C. H. ROCHESTER AND A. V. WESTERMAN 43* J. Bello and J. R. Vinograd, J. Amer. Chem. SOC., 1956, 78, 1369.C. H. Rochester and A. V. Westerman, J.C.S. Faraday I, 1976, 72,2753.P. G. Squire, P. Moser and C. T. O’Konski, Biochemistry, 1968, 7,4261.lo P. F. Spahr and J. T. Edsall, J. Bid. Chem., 1964,239,850.l2 R. Brodersen, B. J. Haugaard, C. Jacobsen and A. 0. Pedersen, Acta Chem. Scand., 1973, 573.l3 I. C. Watt and J. D. Leeder, Trans. Faraday Suc., 1964,60, 1335.l4 C. €3. Baddiel, M. M. Breuer and R. Stephens, J. Colloid Interface Sci., 1972, 40,429.l5 I. C. Watt and J. D. Leeder, J. TextiZe Inst., 1968, 59, 353.l6 M. J. Hunter and M. L. Ludwig, J. Amer. Chem. Suc., 1962, 84, 3491.l 7 R. W. Green and K. P. Ang, J. Amer. Chem. Sue., 1953, 75, 2733.l 8 E. F. Mellon, A. H. Korn and S. R. Hoover, J. Arner. Chem. Suc., 1947, 69, 827.l9 C. Tanford, S. A. Swanson and W. S. Shore, J. Amer. Chem. Suc., 1955, 77, 6414.2o C . Tanford, J. G. Buzzell, D. G. Rands and S. A. Swanson, J. Amer. Chem. Suc., 1955,77,6421.21 C. J. Clemett, personal communication.(PAPER 6 /908