J. Biol. Chem.-1944-Fraenkel-Conrat-239-46

8/14/2019 J. Biol. Chem.-1944-Fraenkel-Conrat-239-46

1/8

THE USE OF DYES FOR THE DETERMINATION OF ACID ANDBASIC GROUPS IN PROTEINS

BY HEINZ FRAENKEL-CONRAT AND MITZI COOPER(From the Western Regional Research Laboratory,* Albany, California)

(Received for publication, March 4, 1944)Only a few methods are available for the determination of acid and basicgroups of proteins. Titration curves have been most generally used forthis purpose. Their application is restricted, however, to proteins whichare either soluble over a wide pH range (1, 2) or completely insoluble(3, 4) ; in addition, considerable amounts of material are needed. Theusefulnessof titration curves is also limited by difficulties of interpretation.Metaphosphoric acid has recently been suggested as a reagent for thedetermination of basic protein groups with which it combines stoichio-metrically in acid solution (5).Acid and basic dyes are known to combine with protein groups of oppo-site ionic charge. The pioneer work of Loeb (6) was carried further by

Chapman, Greenberg, and Schmidt (7-9) who measured the amounts ofacid and basic dyes bound by proteins at various pH values and correlatedtheir findings with those of titration studies. The present paper reportsanalytical methods for the determination of the total acid and basic groupsof proteins based upon this ability to combine with dyes in buffered alkalineor acid solutions. The technique is rapid and simple and is applicable toboth soluble and insoluble proteins; it is based on the photoelectric deter-mination of the uncombined dye and therefore can be carried out withprotein samplesas small as the accuracy of weighing permits.

EXPERIMENTALReagents-Dye solutions; 0.1 per cent orange G (Coleman and Bell) and 0.2 percent safranine 0 (National Aniline) in water.The orange G, while labeled as of 78 per cent dye content, was found tobe of constant chromogenic value and nitrogen and sulfur content after

fractionation and recrystallization. Air-dried samples of both dyes con-tained approximately 10 per cent of water. Standard solutions were pre-pared from material dried to constant weight at 70 regardless of the dyecontent indicated on the labels.* This is one of four regional research laboratories operated by the Bureau ofAgricultural and Industrial Chemistry, Agricultural Research Administration,United States Department, of Agriculture.

239

byguest,onDecember7,2009

www.jbc.org

Downloadedfrom
http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/


2/8

246 ACID AND BASIC PROTEIN GROUPSBuffers; pH 2.2, 980 ml. of 0.1 M citric acid and 20 ml. of 0.2 M disodium

phosphate (10). pH 11.5,250 ml. of 0.2 M disodium phosphate and 200 ml.of 0.1 N sodium hydroxide, water to 1000 ml. (11).Determination of Orange G Bound by Proteins (Basic Groups)-To each of

four 15 ml. test-tubes, suitable for use in an angle head centrifuge, areadded 5 mg. of the protein, 1 ml. of pH 2.2 buffer, increasing am0unt.s(1, 2, 3, or 4 ml.) of 0.1 per cent orange G, and two glass beads. If theprotein dissolves in the buffer, it will be reprecipitated by the dye. Thesuspensions are shaken metihanically for 20 to 24 hours. They are thencentrifuged, and aliquots of the supernatant solutions are diluted lOO-fold.The color intensities are determined by means of a photoelectric calorimeter(Klett-Summerson), with a blue filter (Corning No. 038 + Pyrex NQ. 554).The dye concentration is read from a standard curve prepared from dataobtained with appropriate dilutions of the orange G stock solution. Thiscurve is a straight line in the range of 0.5 to 10 mg. of dye per liter. Thereadings are not affected by the presence of the acid buffer.

The dye bound by the protein is determined by subtracting the excessfound in the supernatant from the amount added. Saturation of tjheprotein with dye is indicated when no more dye is bound by the samplesto which greater amounts of dye are added. For routine analyses a seriesof three samples is regarded as sufficient if the results agree within 10 percent. If greater exactness is desired (as in al l determinations listed inTables I and II), six to twelve samples are prepared. More dye is used if adefinite trend in the first series indicates that saturation has not beenreached. The average of al l values for the maximal amount of dye bound,in mg., multiplied by a conversion factor of 8.85 yields the results in termsof acid equivalents of dye bound per gm. of protein X 104.2An alternate technique applicable to proteins soluble at pH 2.2 permitsthe analysis of only about 1.5 mg. of protein. For example, 1 ml. of a 0.75per cerit solution (in distilled water, dilute salt solution, or 0.01 N hydro-chloric acid) is diluted with 1.5 ml. of the pH 2.2 buffer and four 0.5 ml.aliquots of this mixture are treated with 0.5 to 2.0 ml. portions of dye solu-tion. The resulting precipitate is centrifuged off the following day andthe supernatant treated as described above.

Determination of Xafranine Bound by Proteins (Acid Groups)-The pro-cedure is the same as that for the basic groups except for the use of the1 The factor represents the valenc e of the dye (2), X 10, divided by the mo lecula r

weight of the dye (452) X the amount of protein used (0.005 gm.). The emp iricalformula of orange G is C,eII1(IN20(503)2=N a2.+.

2 Analyse s on proteins containing few bas ic or acid groups are necess arily inexact(MO per cent) sin ce they are derived from the differen ces between large figures .

With p roteins of typical comp osition, the averages of three or four sam ples cangenerally be reproduced within 5 per cent.


www.jbc.org

Downloadedfrom


3/8

13. FRAENKEL-CONRAT AND M. COOPER 241pH 11.5 buffer and the 0.2 per cent safranine solution. After 24 hoursof shaking and subsequent centrifugation, the solutions are diluted lOO- or200-fold and read with the same blue filter used for orange G. The stand-ard curve is a straight line for 0 to 10 mg. of dye per liter. The factorfor calculating the results in terms of moles (= base equivalents) of dyebound per gm. of protein X lo4 is 5.62. Saturation of the protein withdye is again indicated by the lack of a progressive trend in the amounts ofdye bound by samples treated with increasing amounts of dye solution.The results of at least three samples, agreeing within 10 per cent, areaveraged.2Of proteins soluble at pH 11.5, samples of only 1.5 mg. are needed, witha technique corresponding to that described above.Effect of Variations in Experimental Conditions on Amounts of DyeBound by Proteins. pH-For purposes of convenience and practicability,the protein-dye combination was carried out in buffered solutions, thuscircumventing the need for pH measurements and adjustments. Buffersof pH 2.2 and 11.5 were selected because he work of Chapman, Greenberg,and Schmidt (7-9) had indicated that complete dissociation of proteingroups in the presence of dyes was approached at these points. It did notseem advisable to use more strongly acid or alkaline buffers, since the ad-vantages of more complete dissociation of protein groups might be over-shadowed by the disadvantages of protein breakdown. Thus proteinstreated with safranine at pH 12.0 bound considerably larger amounts ofdye which increased with the excess added. This phenomenon whichmay be due to protein breakdown or to physical adsorption of the dye underthese more extreme conditions was not further investigated.4

Concentration-When proteins were treated with an excess of orange G,the amount bound was independent of protein or dye concentration withinthe limits used (0.06 to 0.2 per cent protein and 0.05 to 0.1 per cent dye).Of safranine, slightly smaller amounts were bound at lower than at higherconcentrations; thus the values obtained with solutions of 0.06 to 0.12 percent protein concentration were 5 to 10 per cent lower than those obtainedwith 0.1 to 0.2 per cent protein concentration (dye concentration 0.12 to0.16 per cent in both series).When less han equivalent amounts of the dyes were added to dissolved3 Safranine 0 is a mixture of homologous monoacid bases; the molecular weightsof the two main constituents, (C20Ht~N4)+Cl-and (C~IH~IN&U-, are 350.5 and 364.5.Calculations were based on a value o f 355.4 Studies are in.progress aiming at differential determination of the most stronglybasic and acid groups by equilibrating proteins with the dyes in buffers less acid orbasic than pH 2.2 or 11.5. A method for the determination of the approximateisoelectric point of insoluble proteins through measurement of the dyes bound atvarious pH levels will be described elsewhere.


www.jbc.org

Downloadedfrom


4/8

242 ACID AlUD BASIC PROTEIN GROUPSproteins, the resulting protein-dye complexes were not completely pre-cipitated. This resulted, paradoxically, in higher calorimetric readings inthe soluble phase than when sufficient dye for saturation of the proteinwas added. This finding confirmed similar observations of Rawlins andSchmidt (9).

Time, Shaking, State of Protein-For routine analyses, a 20 hour periodwas found practical. That equilibration was completed during that timewas indicated by the fact that no greater amounts of dye were bound byproteins after 48 hours of interaction.6 A shorter period may be sufficientfor many purposes, since egg albumin was found to bind within 10 minutes90 and 86 per cent of the maximal amounts of orange G and safranine,respectively.

Efficient shaking was essential for equil ibration when protein sampleswere treated in test-tubes with more than 2 ml. of dye solution.

The state of dispersion of the protein did not greatly affect its capacityfor the dyes. Thus insoluble proteins (keratins) of 60 to 80 mesh part,iclesize bound only 10 to 20 per cent less dye than material which passed a200 mesh screen. Also proteins which were insoluble in neutral solutionbut soluble in the buffers used bound the same amount of dye when dis-solved in the buffer before addition of the dye as when solution was pre-vented by adding the dye before the buffer.

To investigate any possible effect due to incipient denaturation by theacid or alkaline buffer, samples of egg albumin were denatured by heatingfor 5 minutes to 70 at pH 2.5 or 11.7, and were subsequently analyzed fortheir dye-binding capacity. This was found to be unaffected by suchtreatment. That the heat treatment, but not the pH alone, led to appre-ciable denaturation was evident upon neutralization of aliquot samples.Signijkance of Dye-Binding Capacity of Proteins-The amounts of thetwo dyes bound by proteins, expressed in terms of acid and base equivalents,have been regarded as indicative of the number of basic and acid proteingroups dissociated under the conditions of the test at pH 2.2 and 11.5. Toascertain which types of groups were thus determined, several proteinswere analyzed. Two of these, egg albumin and ,&lactoglobulin, can beregarded as approximately pure proteins, inasmuch as the preparationshad been repeatedly recrystallized and had been electrodialyzed. Lyso-zyme and insulin were crystalline preparations not electrodialyzed.6 Theothers were commercial protein samples.

6 Wool, in contrast to other proteins studied including keratins and silk fibroin,did not reach an equilibrium with the dye solution within 20 to 24 hours.6 The preparations of crystalline egg albumin, ,%lactoglobulin, and lysozyme werekindly placed at our disposal by Dr. F. E. Lindquist, Dr. E. F. Jansen, and Dr. H. L.Fevold, respectively; crystalline zinc insulin was supplied by Eli Lilly and Company.


www.jbc.org

Downloadedfrom


5/8

H. FRAENKEL-CONRAT AND M. COOPER 243A comparison of the number of basic groups which bind orange G with

that number as determined by the methods of the literature (1,2,5,12,13)is summarized in Table I. The good agreement between correspondingvalues for the well characterized proteins may be regarded as evidence forthe reliability of the proposed method. The approximate proportionalityof results obtained with the crude proteins contributes additional support.It is concluded from these data that the number of basic groups bindingorange G at pH 2.2 represents the sum of the guanidyl, imidazole, and amino(a- and E-) groups of proteins.

TABLE IComparison of Ba sic Groups of Proteins As Determined bw Various Methods

Basic residues per g m . protein X 104Protein Dye-bindingcapacityt

Egg albumin ... ...,%Laetoglobulin ...Casein ............Fibrin ............Gelatin ...........Gliadin. ..........Insulin. ...........Lysozyme. ........Zein. .............

8.8$11.616.812.06.04.3$9.4t11.511.9

Titration(1, 2, 12, 13)m&$$ Analysis or

capacity (5) isolation (12, 13)

8.0- 8.711.57.6- 9.08.9- 9.63.410.111.7-12.81.8- 2.1

7.3- 8.0

10.3-11.0

9.011.58.0-9.4813.1510.754.4#9.512.21.5

* Egg albumin and ,%lactoglobulin were electrodialyzed; all proteins were cor-rected for moisture content. Casein, gelatin, gliadin, and zein were commercialpreparations. Two casein preparations gave identical values. See foot-note 6concerning the other proteins.t Moles of orange G bound at pH 2.2, X2.$ These analyses represent averages of data obtained with 5 and 1.5 mg. proteinsamples, with protein concentrations ranging from 0.1 to 0.2 and from 0.06 to 0.15per cent, respectively. Results of the two techniques agreed within 5 per cent.8 Determined by nitrogen distribution.

A similar comparison of the groups binding safranine at pH 11.5 withthe carboxyl and with the total acid (i.e. carboxyl + phenol + thiol)groups of several proteins is listed in Table II. The literature values givenfor most proteins must be regarded as rough approximations, inasmuch asthey were calculated from incompletely confirmed determinations, by isola-tion, titration, and calorimetry, of the content of dicarboxylic amino acids,tyrosine, cysteine, and amide nitrogen. Only the data for egg albuminand @-lactoglobulin appear reliable, since they were obtained by recentimprovements in analytical methods (14), supported by titration data


www.jbc.org

Downloadedfrom


6/8

244 ACID AND BASIC PROl!EIN OflOUPS(1, 2). These two proteins showed a capacity to bind safranine in anamount which corresponded to their total acid groups, a finding which wassupported by the data on most of the other proteins studied. It thusappears that the number of acid groups binding safranine 0 at pH 11.5comprises the sum of the carboxyl, phenol, and thiol groups of proteins.Uses and Applications--The described micromethods for the routinedetermination of acid and basic protein groups were developed primarilyfor use n studies of protein derivatives. A considerable number of deriva-tives (of keratins, gluten, egg albumin, /34actoglobulin, casein, etc.) hasrecently been prepared in this Laboratory by treatment with epoxides,

TABLE IIComparison of Iotal Acid and Carboxyl Groups of Proteins with Their Capacityto Bind Sajranine

Protein* Dye boundtEgg albumin. .... .... .... .... .. 13.58@-Lactoglobulin ................. 17.65Casein ......... ......... ....... 19.4Gelatin .... .... .... .... .... .... . 12.7Glisdin. ......... ......... ...... 5.5Insulin ......... ......... ....... 17.5Zein ............................ 5.5

* See the corresponding foot-note to Table I.

To tal acid groupsS Carboxyl group st

13.8 10.417.5 14.516 1310 107 513 66 3

t Moles of safranine 0 bound at pH 11.5 by lo4 gm. of protein (range of proteinconcentrations, 0.1 to 0.2 per cent).$ Of lo4 gm. of protein; calculated from analyses for glutamic and aspartic acids,amide N, tyrosine, and cysteine as summarized by Cohn and Edsall (12), Chibnallet al. (14), and Brand and Kassell (15), and as amended for glutamic acid of insulin,gelatin, and gliadin by Olcott (16). The unknown number of terminal carboxylgroups of the polypeptide chains was disregarded.5 In more dilute solution (protein concentrations 0.06 to 0.15 per cent), 12.4 and16.5 moles of dye were bound by egg albumin and @-lactoglobulin, respectively.aromatic isocyanates, anhydrides, aldehydes, nitrous acid, and combina-tions of these reagents. Epoxides were found to combine with both theacid and the primary amino groups of proteins (17). The esterificationwas demonstrated by the decreases n the number of acid groups of the

7 Before suitable conditions for the determination of total acid groups had beenrecognized, a relative measure of the acidity of various proteins and derivatives ivasobtained from their tendency to bind dyes when the amounts added were less thanthose needed for saturation. For these studies both safranine and methylene bluewere used. The amounts of these dyes most readily bound yielded comparative dataon the acidity of proteins and derivatives which have since been confirmed bydeterminations of their total acid groups.


www.jbc.org

Downloadedfrom


7/8

H. FRAENKEL-CONRAT-AND M. COOPER 245

derivatives, as estimated by dye methods, particularly with those of sub-maximal combination which are believed to measure primarily the carboxylgroups. On the other hand, the introduction of alkoxy residues in theamino groups did not remove their basic nature. Actually a slight increasein the basic groups was found in many epoxide-treated proteins, a phe-nomenon which is not yet understood.

Intensive treatment of certain proteins with phenyl isocyanate led tointroduction of the reagent to the extent of up to 30 per cent (by weight)of the protein.8 Dye methods have now yielded information which maycontribute to an accounting for this extent of interaction. A few typicalresults obtained with cattle hoof powder and rennet casein are listed in

TABLE IIIEfects of Various Reagents on Ba sic and Acid Groups of Proteins

Protein and treatmen t*

Hoof powder, more than 200 mesh, untreated.. ... ... ... .I 40-60 mesh, untreated .... .... .... .... .... .Phenyl isocyanate$. ...................................Phthalic anhydride ...................................Propylene oxide ......... ......... .......... ......... .

Same, followed by phenyl isocyanate. .... .... .... ....Formaldehyde ........................................Nitrous acid. ......... ......... .......... .......... ...Casein,untreated ......................................Phenyl isocyanate. ...................................Propylene oxide ......... ......... .......... ......... .-

-

-

Basicresiduest

8.88.00.03.69.20.07.13.86.80.06.4

Total acidresiduest

10.48.02.113.38.71.711.213.020.34.07.2 -* See foot-note 8 for the methods of preparation and treatment.t Per gm. of protein X 104. Determinations based on the capacity to bind orangeG and safranine.0 at pH 2.2 and 11.5, respectively.$ All hoof derivatives were powders which passed a 200 mesh screen.

Table III. Treatment with phenyl isocyanate was found to cause a loss ofalmost all basic groups and of a considerable proportion of the acid groupsof the prot.eins. On the other hand, phthalic anhydride was found to reactonly with part of the basic groups. In contrast to phenyl isocyanate,phthalic anhydride increased the number of acid groups, as would beexpected from introduction of phthalic acid residues. With both reagentsthe observed decrease n the basic groups corresponded to the loss n aminonitrogen.8 The increase in the acid groups of the deaminated protein mayhave been due to nitration of the phenol residues. Treatment with formal-dehyde caused only minor changes n both acid and basic groups. This is8 Fraenkel-Conrat, H., and Olcott. H. S., in preparation for press.


www.jbc.org

Downloadedfrom


8/8

240 ACID AND BASIC PROTEIN QROUPSin contrast to a marked decrease in the primary ammo nitrogen, indicatingthat the N-methyl01 or N-methylene groups resulting from the interactionof amino groups with aldehydes retain sufficient basicity to bind orange Gat pH 2.2.

SUMMARYMicroanalytical methods were developed for the estimation of the number

of acid and basic groups of proteins. These were based on the tendency ofthe polar groups to bind dyes of the opposite charge, resulting in a precipi-tation of the protein-dye complex. The acid dye, orange G, combinedstoichiometrically with basic protein groups in a buffer of pH 2.2. Thebasic dye, safranine 0, reacted with acid groups at pH 11.5, but the extentof combination was in this case slightly affected by other factors, such asprotein concentration.

The number of protein groups binding these dyes corresponded well tothe total number of basic (guanidyl, imidazole, ammo) and acid (carboxyl,phenol, thiol) groups of crystalline egg albumin and /34actoglobulin and,approximately, to those of several crude proteins studied.

The proposed micromethods were applicable to both soluble and insolubleproteins. They have proved useful tools in the interpretation of the actionof various chemical agents on proteins.

The valuable suggestions and criticisms of H. S. Olcott of this Laboratoryare gratefully acknowledged.

BIBLIOGRAPHY1. Cannan, R. K., Kibrick, A. C., and Palmer, A. H., Ann. New York Acad. SC.,

41,243 (1941).2. Cannan, R. K., Palmer, A. H., and Kibrick, A. C., J. BioZ. Chem., 142,803 (1942).3. Steinhardt, J., and Harris, M., J. Res. Nat. Bur. Standards, 24,335 (1940).4. Theis, E. R., and Jacoby, T. F., J. Biol. Chem., 146,163 (1942).5. Perlmann, G. E., J. Biol. Chem., 137,707 (1941).6. Loeb, J., Proteins and the theory of colloidal behavior, New York and London,2nd edition, 36 (1924).7. Chapman, L. M., Greenberg, D. M., and Schmidt, C. L. A., J. Biol. Chem., 72,707 (1927).8. C. Rawlins, L. M., and Schmidt, C. L. A., J. Biol. Chem., 82, 709 (1929).9. C. Rawlins, L. M., and Schmidt, C. L. A., J. Biol. Chem., 88,271 (1930).10. McIlvaine, T. C., J. Biol. Chem., 49, 183 (1921).11. Kolthoff, J. M., and Vleeschhouwer, J. J., Biochem. Z., 189,191 (1927).12. Cohn, E. J., and Edsall, J. T., Proteins, amino acids and peptides, AmerioanChemical Society monograph series, New York (1943).13. Abraham, E. P., Biochem. J., 33, 622 (1939).14. Chibnall, A. C., Rees, M. W., and William, E. F., Biochem. J., 37, 372 (1943).15. Brand, E., and Kassell, B., J. BioZ. Chem., 146, 365 (1942).16. Olcott, H. S., J. BioZ. Chem., 163, 71 (1944).17. Fraenkel-Conrat, H., J. BioZ. Chem., 164, 227 (1944).


www.jbc.org

Downloadedfrom

Documents

J. Biol. Chem.-1944-Fraenkel-Conrat-239-46