ANALYTICAL BIOCHEMISTRY 141, 397—40l (1984}
`
`Sodium Sulfite as an Antioxidant in the Acid Hydrolysis
`
`of Bovine Pancreatic Flibonuclease A
`
`J. K. SWADESH, T. W. THANNHAUSER, AND H. A. SCHERAGA'
`
`Baker Laboratory of Chemistry, Cornell University, Ithaca. New York E4853
`
`Received February 24, 1984
`
`Treatment of hydrochloric acid with sodium sulfite prior to the acid hydrolysis of bevine
`pancreatic ribonuclease A has been found to suppress the oxidation of cystine, methionine,
`and tyrosine without adversely affecting the recoveries of other amino acids. Statistical analysis
`of the results indicated that the assumption of the independence of the mean and the variance,
`an assumption commonly used in the evaluation of the effects of various treatments. may not
`be valid in evaluating antioxidants used in the acid hydrolysis of proteins.
`KEY WORDS: amino acid analysis: antioxidant; sodium sulfite', hydrochloric acid; hydrolysate.
`
`Bfihlen and Schroeder (l) have noted the
`
`need for a simple procedure for the acid
`
`hydrolysis of proteins and peptides that does
`not require a separate determination of tryp-
`tophan, tyrosine, methionine, cysteine, and
`
`cystine. These authors used thioglycolic acid
`to prevent the oxidation of the first three of
`
`these amino acid residues, but required a
`
`separate analysis under oxidizing conditions
`for the determination of the sum of cysteine
`and cystine as cysteic acid. Inglis and Liu (2)
`used posthydrolytic conversion of cysteine,
`cystine, and cysteic acid to S-sulfocysteine as
`a strategy in amino acid analysis, and lnglis
`(3) has coupled this method with prehydro—
`
`lytic alkylation of cysteine as a procedure to
`differentiate between cystine and cysteine
`
`including
`and to recover all amino acids,
`tryptophan. Other authors (4—6) have used
`phenol as an antiOxidant, and dimethyl sulf-
`
`0xide has been used (7') to convert cysteine
`or cystine to cysteic acid during hydrolysis.
`Each of these methods has advantages and
`disadvantages.
`In particular, prehydrolytic
`derivatization of cysteine with 4-vinylpyridinc
`or iodoacetate (3) requires the removal of
`the alkylating agent prior to hydrolysis, and
`
`' To whom request for reprints should be addressed.
`
`would therefore be diflicult to apply to very
`large or very small samples or samples of
`low molecuiar weight. For reasons of conve—
`nience, hydrochloric acid is often preferred
`as the hydrolytic agent, but extensive oxida—
`tion is sometimes observed. The present work
`
`explores the use of sodium sulfite to improve
`the recoveries of cystinc, methionine, and
`tyrosine in the acid hydrolysis of bOvine
`pancreatic ribonuclease A, a protein that
`does not contain cysteine or tryptophan. The
`practical benefits conferred by pretreatment
`with sodium sulfite are discussed, and the
`general requirements for a statistical study of
`the effects of an antioxidant are explored.
`
`MATERIALS AND METHODS
`
`Sodium sulfite (NaESO3) was purchased
`
`from Fisher Scientific Company. Bovine pan-
`
`creatic ribonuclease A, type l-AS, was pur—
`
`chased from Sigma, and was purified by ion-
`exchange chromatography on carbcvxymeth-
`
`yloellulose (8) purchased from Whatman.
`Hydrochloric acid {12 M) was either pur-
`chased from Mallinckrodt and purified by
`simple or fractional distillation by a procedure
`
`designed to produce a constant-boiling frac-
`tion (9), or was purchased from Baker Chem-
`ical Company (Ultrex) and diluted to 6 M.
`
`MAIA Exhibit 1038
`
`MAIA V. BRACCO
`
`IPR PETITION
`
`397
`
`-397-
`
`0003-2697184 $3.00
`Copyright c I934 by Academic Press. Inc.
`All rights of reproduction in any form reserved
`
`-397-
`
`
`
`
`MAIA Exhibit 1038
`MAIA V. BRACCO
`IPR PETITION
`
`

`

`398
`
`SWADESH. THANNHAUSER, AND SCHERAGA
`
`The Mallinckrodt HCl (lot KMPE) exhibited
`an elevated boiling point during the latter
`part of the simple distillation, and complete
`destruction of cystine, methionine, and ty-
`rosine was observed with the use of this
`
`distilled acid in the absence of pretreatment
`With N32S03.
`Eleven groups of replicate hydrolysates,
`each containing 4-6 samples, were prepared.
`The first four groups of replicates were pre-
`
`In the following section, groups of rcpli-
`cates have been labeled with the symbols M
`or U to indicate Mallinckrodt or Ultrex acid.
`with the symbols 5 or f to indicate that the
`HCl was distilled by simple or fractional
`distillation, and with the symbol He if the
`HCl had been purged with helium. The
`symbol (,6 indicates that no sulfite was added,
`and the symbol S indicates the addition of
`N82803.
`
`pared as 0.25-ml aliquots of simply or frac-
`tionally distilled Mallinckrodt acid which had
`been treated with 1 mg/ml of Nagsog, or
`not treated, and into which ribonuclease A
`had been dissolved at a concentration of 1
`
`Statistical analysis was performed according
`to the technique of variance analysis (ll).
`This technique apportions the variance that
`exists between groups of replicates to random
`error or to the effects of the various treat-
`
`mg/ml. The fifth group of replicates was
`prepared as
`lO-pl aliquots of fractionally
`distilled Mallinckrodt acid containing protein
`at a concentration of 27 mg/ml. The next
`two groups of replicates were prepared as 10-
`pl aliquots of Ultrex acid which had been
`treated with 7 mg/ml of N32803, or not
`treated, and into which the protein had been
`dissolved at a concentration of 10 mg/ml.
`The final four groups of replicates were pre-
`pared as lO-pl aliquots of Ultrex acid with a
`protein concentration of 10 mg/ml. In two
`of these groups of replicates,
`the acid had
`been purged for l h with helium gas prior to
`sample preparation and, to two of the groups
`of replicates, 10 mg/ml of Nazsog had been
`added to the HCl prior to the addition of
`
`protein.
`Hydrolyses were performed in 8-mm-i.d.
`glass tubes. The solutions were degassed by
`the freeze—pump—thaw method (10) at 0.03
`mm Hg. The freeze—pump—thaw cycle was
`repeated three times. The additional precau-
`
`tion of decreasing the pressure in the tube to
`about 2 mm Hg prior to freezing the sample
`in liquid nitrogen was observed for the final
`four groups of replicates; this should prevent
`the condensation of oxygen. All samples were
`hydrolyzed at 110°C for 24 h, and the HCl
`was removed under vacuum. Samples were
`analyzed on a Technicon TSM amino acid
`
`autoanalyzer.
`
`ments, i.e., to the method of distillation, to
`the effect of purging with helium, or to the
`effect of addition of Nagsog. In the technique
`of variance analysis, the statistical significance
`of the effect of a particular treatment
`is
`determined by an F test. To evaluate the
`effects of various treatments in the first four
`
`groups of replicates, a two-sided F test was
`used,
`i.e.,
`it was assumed that addition of
`
`1432803 could either increase or decrease the
`
`yields of the oxidizable residues. The results
`demonstrated unambiguously that Na2803
`does not decrease the yields; hence, one-sided
`F tests were used in all subsequent analyses
`of variance. One cannot combine groups of
`replicates for the purpose of variance analysis
`unless it is known that the mean and the
`
`variance are approximately independent of
`one another. For the purpose of investigating
`the dependence of the variance on the mean,
`a phenomenon formally known as hetero-
`scedasticity, Bartlett's test (1 l) was used. A
`simpler, but qualitative technique involves
`graphing the variance or standard deviation
`against the mean.
`
`RESULTS AND DISCUSSION
`
`The principal results of this work are pre-
`sented in Table 1. Under conditions of high
`sample dilution, extensive destruction of the
`oxidizable residues occurred in the absence
`
`-398-
`
`-398-
`
`
`
`
`

`

`SODIUM SULFITE AS ANTIOXIDANT [N RIBONUCLEASE A ACID HYDROLYSIS
`
`399
`
`Hvonowses OF BoleE PANCREATIC RIBONUCLEASE IN MALLINCKRODT HCI
`
`TABLE 1
`
`No Nagso, added
`
`NafiO, added
`
`Method of
`treatment
`
`Amino acid
`residue
`
`Theoretical
`
`Replicate group
`number and
`mnemonic"
`
`Number
`of residues
`recovered"
`
`Replicate group
`number and
`mnemonic
`
`Number of
`residues
`recovered
`
`Simply
`distilled HCl
`
`Fractionally
`distilled HCl
`
`Halvays
`Met
`Tyr
`
`Half-Cys
`Met
`Tyr
`
`s
`4
`6
`
`3
`4
`6
`
`1 Msd
`
`3 M58
`
`0.0 [—)
`0.0 {—)
`0.0 (—l
`
`2 Mfo
`
`4 MfS
`
`1.48 {1.23)
`LS9 [0.96)
`1.166133}
`
`4.34 (0.51)
`3.25 (0.45)
`l.88 (1.50)
`
`5 .07 (c. 19)
`3.56 (0.18)
`3.3! 0.31)
`
`" See text for the nomenclature of the replicate groups. Each group contained five replicates. All samples were
`prepared with a protein-to-HCI molar ratio of 1:80.000. in samples M58 and MfS, the protein-to-Na2803 molar ratio
`was 1:100. In a fifih group of four replicates, the protein-to—HC] molar ratio was increased to 13,000, but no sulfite
`was added. The observed recoveries were; Cys 5.45119 = 0.97], Met 3.86 (a = 0J2), and Tyr 4.61 (a i 0.42] residues.
`” Values reported are the number of residues recovered. based on an assumed recovery of Ala plus Giu equal to
`24. Values in parentheses are standard deviations.
`
`of N32303. In a footnote to Table 1, it is
`
`noted that, at a higher concentration of pro-
`tein, fair recoveries of the Oxidizable residues
`were observed. Hence, although the quality
`of the HCl as purchased was undoubtedly
`poor, perhaps due to the length of time
`between production and purchase, it is evi-
`dent that recoveries of the oxidizable residues
`
`are improved by distillation of the acid, by
`reduction of the ratio of HCl to protein, and
`
`by the addition of P432803. Two-factor anal-
`ysis of the variance, using a two-sided F test
`as described in the previous section, showed
`that addition of Na1803 increased the recov—
`cry of the oxidizable residues with a confi-
`dence level greater than 99%. Fractional dis-
`tillation of the HC] also improved the re-
`coveries with a confidence level greater
`than 97.5%.
`
`Additional tests, using a higher grade of
`acid, were also performed, and these results
`are contained in Table 2. Groups 6 and “i
`could not be combined with the remaining
`groups because of heteroscedasticity. A one-
`
`sided F test on groups 6 and 7 demonstrated
`that
`the addition of sulfite improved the
`
`recovery of methionine at a confidence level
`
`of 99.9%, and the recovery of cystinc at a
`confidence level of 97.5%. The confidence
`
`level on the increased recovery of tyrosine is
`below 90%.
`
`Two-factor analysis of the variance of
`groups 8—1] was not conclusive because of
`the limiting effect of the protein composition;
`
`as one approaches quantitative yield in the
`absence of any treatment, increases in per-
`centage yield due to the effects of the various
`treatments 11le become small. The
`
`improvements of the yields of cystine and
`methionine, attributable either to purging of
`the HCl with helium or to the addition of
`
`N32803, are not statistically significant, al-
`though it is germane to note that addition of
`N32803 qualitatively seems to improve the
`
`yields of these residues. For tyrosine, however,
`a one-sided F test
`indicated at
`the 90%
`
`confidence level that the recovery is increased
`by purging with helium, by addition of
`
`-399-
`
`-399-
`
`
`
`
`

`

`SWADESH, THANNHAUSER, AND SCHERAGA
`
`TABLE 2
`
`HYDROLYSES or BOVINE PANCREATIC RIBONUCLEASE A [N ULTREx ACID
`
`Replicate group
`number and
`mnemonic“
`
`Number of
`replicates
`in group
`
`Molar ratio of
`protein:HCl:Na;SOa
`
`Number of
`
`half-cystine
`residues
`
`Number of
`
`methionine
`residues
`
`Number of
`
`tyrosine
`residues
`
`(theory = 8]
`
`(theory = 4)
`
`(theory = 6)
`
`Mean
`
`SD
`
`Mean
`
`SD
`
`Mean
`
`SD
`
`6 Up“
`? US’
`
`8 Up
`9 US
`10 Uc-He
`ll US-He
`
`6
`6
`
`5
`5
`5
`5
`
`l:6800:0
`l:8200:80
`
`1:5400:0
`1:5400:72
`1:5400:0
`1:5400:‘.-'2
`
`6.26
`7.12
`7.57
`7.64
`7.68
`7.83
`
`0.99
`0.40
`0.41
`0.21
`0.21
`0.19
`
`3.61
`3.76
`4.02
`4.00
`4.04
`4.14
`
`0.03
`0.14
`0.16
`0.14
`0.13
`0.06
`
`4.61
`5.05
`5.22
`5.40
`5.48
`5.?0
`
`0.12
`0.09
`0.22
`0.2?
`0.16
`0.30
`
`a See text for nomenclature. Up’ and US' differ from Us and US, respectively, because. in the former, the pressure
`inside the hydrolysis tubes was not reduced prior to freezing the sample in the freeze—pump—thaw cycle. Statistiml
`analysis, described in the text, precludes the combination of Ud’ with Us or US’ with US.
`
`Nazsog, and by a synergistic interaction of
`these treatments. Additional experiments, not
`presented in this paper, demonstrated that
`the addition of up to 40 mg/ml of Na2803
`to the hydrolysis acid did not reduce the
`yields of the other amino acids of ribonu-
`clease.
`
`We have not been able to establish the
`
`mechanism by which pretreatment of HCl
`used in acid hydrolysis imprOves the yields
`of cystine, methionine, and tyrosine. 1n acidic
`solution, Nazso,
`is converted to SD; gas,
`most of which is slowly evolved from solution
`
`prior to addition of the sample, and the
`remainder of which is presumably removed
`as the pressure in the hydrolysis tube is
`reduced under vacuum. The apparent syn-
`ergism of purging with helium and adding
`Na2803, observed in the recoveries of tyrosine
`in groups 8—1 1, suggests that NaZSO3 reduces
`trace oxidants present even in freshly prepared
`acid, and does not simply purge the I-ICl of
`dissolved molecular oxygen. One interesting
`observation, perhaps relevant to the elucida-
`tion of the mechanism by which antioxidants
`function, is the linear relationship obtained
`if the mean recovery of a given amino acid
`in the absence of blast); is plotted against
`
`the mean recovery in the presence of sulfite,
`i.e., if the result of Msd is plotted against
`that of M53, that of Mfd against that of MfS,
`and so on. Hence, although we do not know
`the antioxidative mechanism,
`the effect of
`addition of Na2803 appears to be regular and
`
`predictable.
`Successful amino acid analysis depends on
`many factors, including the composition of
`the protein, the molar ratio of HCl to protein,
`the freshness of the hydrolysis acid used, and
`the experimental details of sample prepara-
`tion. The present work demonstrates that
`improved yields of the oxidizable amino acids
`can be obtained by pretreatment of the HCl
`with sodium sulfite. The practical applications
`of this observation are many. First, in some
`applications, it may be possible to substitute
`a lower grade of hydrolysis acid. Second, it
`may be possible to extend the effective shelf
`life of hydrochloric acid used in amino acid
`analysis. Third, Nazso, may serve as a re-
`placement for phenol, which is difficult to
`purify. Finally, as amino acid analysis is
`extended to samples of smaller size, reduction
`of the volume of hydrolysis acid becomes
`increasingly difficult; hence, the use of an
`antioxidant becomes necessary. In many ap-
`
`-400-
`
`-400-
`
`
`
`
`

`

`SODIUM SULFITE AS ANTIOXIDANT IN RIBONUCLEASE A ACID HYDROLYSIS
`
`401
`
`plications, N32303 may prove to be the an-
`tioxidant of choice.
`
`ACKNOWLEDGMENTS
`
`This work was supported by research grants from the
`National Institute of General Medical Sciences (GM-
`l4312}, and from the National Science Foundation
`{PCM79-20279). JKS was an NIH Postdoctoral Fellow
`([932—1984}. We thank M. E. Benton. Y. Konishi, and
`C. A. McWherter for valuable discussions, and G. T.
`Montelione. S. H. Lin. and E. A. Cznrylo for helpful
`comments on the manuscript.
`
`REFERENCES
`
`l. Béhlen. P., and Schroeder. R. {1982Mnai Biochem.
`126, 144-152.
`Inglis. A. 3.. and Liu. T. Y. (1970) J. Bid. Chem.
`245. “2—116.
`
`2.
`
`3.
`
`J.
`
`Inglis. A. S. {1983) in Methods in Enzymology (Hits.
`C. H. W., and Timashefl'. S. N., eds), Vol. 9!.
`Part I. pp. 26-36, Academic Press. New York.
`Tsugita. A. and Schefl‘ler,
`.I.
`.l.
`(1982) Eur.
`Biochem. 124. 585-588.
`. Tsugita. A.. and Scheffler, J. I. {1982) Proc. Japan.
`Acad. Ser. B 58, 1—4.
`Sanger, F.. and Thompson, E. 0. P. {1963} Biochim.
`Biophys. Acta 71, 468—47].
`. Spencer. R. L., and Wold. F. {1969} Anal. Biochem.
`32, 185-190.
`Taborsky. G. (1959) J. Bio}. Chem. 234, 2652—2656.
`Skoog, D. A., and West, I). M. [1976) Fundamentals
`of Analytical Chemistry. 3rd ed., p. 3'30. Holt.
`Rinehart & Winston, New York.
`Moore, 5., and Stein. W. H. in Methods in Enzy-
`mology (Colowick. S. P.. and Kaplan. N. 0..
`eds). Vol. 6. pp. 819—831. Academic Press. New
`York.
`
`.05
`
`10.
`
`. Sokal, R. P... and Rohlf, F. 1. {I969} Biometry. pp.
`155—428. Freeman, San Francisco.
`
`-401-
`
`-401-
`
`
`
`
`