INFORMATION TO USERS The most advanced technology has been used to photo graph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are re produced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. These are also available as one exposure on a standard 35mm slide or as a 17" x 23" black and white photographic print for an additional charge.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
UMI University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600
Order Number 8907271
Chemical modification of amino groups in proteins: Part I. Deamination of lysine residues in RNAse A. Part II. N-cyanomethylation of amino groups in RNAse A
Monera, Oscar D., Ph.D.
The Ohio State University, 1988
Copyright ©1989 by Monera, Oscar D. All rights reserved.
UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106
CHEMICAL MODIFICATION OF AMINO GROUPS IN PROTEINS
PART I. DEAMINATION OF LYSINE RESIDUES IN RNAse A.
PART II. N-CYANOMETHYLATION OF AMINO GROUPS IN RNAse A.
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
Oscar D. Monera, B.S., M.S.
*****
The Ohio State University
1988
Dissertation Committee: Approved by
Dr. Gary E. Means
Dr. Edward J. Behrman
Dr. Richard P. Swenson Adviser
Department of Biochemistry To My Wife Linda, Daughters Roslyn and Rosmari
And
To My Parents
ii ACKNOWLEDGMENTS
I am deeply grateful to my adviser, Dr. Gary E. Means, whose invaluable guidance, support and encouragement made this study possible.
My sincere appreciation to Dr. Edward J. Behrman, who literally helped and guided me "from the very start to the very end" of my academic pursuit here at the Ohio State
University; and to Dr. Richard P. Swenson and Dr. Harold
Schecter for their guidance, suggestions and critical evaluations.
I would like to especially express my thanks to Mr.
David Chang of the Campus Chemical Instrumentation Center for his unselfishness, patience, and invaluable help in the interpretation of the mass spectral data.
My sincere appreciation to my fellow graduate students,
K. Ampon, J-W. Park, Y.Q. Song and K. Miller, for their encouragement, delightfulness and humor when nothing seemed to come out right.
To the Filipino Student Association, whose graduation gift came a year too early.
To my whole family, for their inspiration and support. VITA
October 11, 1952 .... Born in Looc, Plaridel, Misamis Occidental, Philippines.
1969-1973 ...... B. S. Chemistry, Silliman University Dumaguete City, Philippines
1973-1977 ...... Researcher/Instructor, Department of Chemistry, Silliman University
1977-1982 ...... Instructor (Dept. Head, 1981-82) Dept, of Ag. Chem. & Food Science Visayas State College of Agriculture Leyte, Philippines
1978-1980 ...... M. S. Biochemistry, University of the Philippines at Los Banos College, Laguna, Philippines
1982-1983 ...... EDPITAF-World Bank Fellow Ohio State University Columbus, Ohio
1983-88 ...... Teaching and Research Associate Department of Biochemistry Ohio State University Columbus, Ohio
PUBLICATIONS
Monera, O.D., M.K. Chang, J-W. Park and G.E. Means. (1986) The Deamination of Lysine Residues of Proteins. Fed. Proc., 45, 1540. Abstract of the ASBC/BDC-ACS Convention, Washington, D.C.
Monera, O. D. and E. J. del Rosario. 1982. Physico-chemical Studies on the Stabilization of Coconut Milk Emulsion. Annals of Tropical Research 4:47-54. TABLE OF CONTENTS
DEDICATION ...... ii
ACKNOWLEDGEMENTS ...... iii
VITA ...... iv
LIST OF TABLES ...... viii
LIST OF FIGURES ...... ix
LIST OF ABBREVIATIONS ...... xiii
PART I. DEAMINATION OF LYSINE RESIDUES IN RNAse A. PAGE
Introduction ...... 1
Materials ...... 17
Methods
Reaction of n-octylamine with sodium nitrite ... 19
(a) Effect of pH ...... 19
Identification and analysis of products . 20
(b) Effect of the length of reaction ...... 22
(c) Effect of temperature ...... 23
(d) Effect of sodium nitrite concentration .. 23
Reaction of n-octylamine with sodium nitro- prusside ...... 24
(a) Effect of p H ...... 24
(b) Effect of temperature ...... 25
Reaction of Na-CBZ-L-Lysine and sodium nitro- prusside ...... 26
v HC1 hydrolysis and amino acid analysis ...... 27
Derivatization for GC-MS ...... 27
Reduction of the products of the reaction of Na-CBZ-L-Lysine and SNP by Pc^-H ...... 28
(a) Preparation of palladium black column ... 28
(b) Pd2-H r e d u c t i o n ...... 29
Purification of RNAse A ...... 30
Preparation of nitroprusside-modified RNAse A .. 31
Separation of the reaction products ...... 31
RNAse A assay ...... 32
Gel filtration of native and nitroprusside- modified RNAse A ...... 33
Amino acid analysis ...... 34
Results
Reaction of n-octylamine with sodium nitrite.... 35
(a) Effect of p H ...... 43
(b) Effect of the length of reaction ...... 48
(c) Effect of temperature ...... 48
(d) Effect of sodium nitrite concentration .. 52
Reaction of n-octylamine with sodium nitro- prusside ...... 54
(a) Effect of p H ...... 54
(b) Effect of temperature ...... 57
Reaction of sodium nitroprusside with Na-CBZ-L-Lysine ...... 58
vi Reactions occuring under the conditions similar to those of the HCl hydrolysis of proteins .. 66
Time-course for the reaction of HCl with 2-amino-5-hexenoic a c i d ...... 76
Reaction of sodium nitroprusside with RNAse A .. 80
Discussion
The deamination reactions ...... 91
Reactions occuring under the conditions similar to those of the HCl hydrolysis of proteins .. 99
Reaction of sodium nitroprusside with RNAse A .. 108
PART II. N-CYANOMETHYLATION OF THE AMINO GROUPS IN RNAse A
Introduction ...... 113
Materials ...... 117
Methods
Reaction of n-butylamine with CH2 O and NaCN ... 119
N-Cyanomethylation of RNAse A ...... 120
Gel filtration of native and N-cyanometylated RNase A ...... 121
Performic acid oxidation of RNAse A ...... 122
Trypsin digestion ...... 122
Peptide mapping and isolation by HPLC ...... 123
Preparation of dinitrophenyl derivatives ...... 123
Separation of DNP derivatives by HPLC ...... 124
Carboxymethylation of Ne-DNP-lysine ...... 125 Results
Reaction of n-butylamine with CH 2 O and NaCN .... 126
N-cyanomethylation of RNAse A ...... 129
Gel filtration ...... 132
Identification of the modified lysine residue .. 138
Discussion ...... 144
List of References...... 152
viii LIST OF TABLES
TABLE PAGE
1 Summary of the gas chromatographic and mass spectral data employed to identify the products of the reaction n-octylamine and sodium nitrite ...... 37
2 Product ratios of the reaction of sodium nitrite and n-octylamine at different pH values ...... 46
3 Product ratios of the reaction of sodium nitroprusside and n-octylamine at different pH values ...... 57
4 Elution positions of homologous lactones from the amino acid analyzer ...... 72
5 RNAse A activities of the different fractions from the reaction of RNAse A and sodium nitroprusside ...... 83
6 Comparison of the amino acid composition of native RNAse A and the different products of its reaction with sodium nitroprusside ...... 89
7 Product distribution of the reaction of sodium nitroprusside and Na-CBZ-L-Lysine ...... 101
8 Distribution of products from 2-amino-5- hexenoic acid after hot HCl treatment ...... 103
9 Estimated number of lysine residues converted to their DNP derivatives ...... 140
10 Residual activities and lysine contents of native and the different modified RNAse A ...... 142 LIST OF FIGURES
FIGURE PAGE
1 The tertiary structure of RNAse showing the amino acid residues in the active site .... 12
2 A general model of the mechanism of RNAse A catalysis ...... 14
3 A typical gas chromatogram of the ether extract of the products of the reaction of n-octylamine and sodium nitrite ...... 36
4 Mass spectra of 1-octylnitrite and 2-octyl- nitrite produced from the reaction of sodium nitrite and n-octylamine ...... 40
5 Product distribution of the reaction of n-octylamine and sodium nitrite in the presence of NaCl ...... 42
6 Effect of pH on the reaction of n-octylamine and sodium nitrite ...... 44
7 Time-course of the conversion of the alcohols into nitrites at conditions similar to those in the reaction of n-octylamine and sodium nitrite ...... 45
8 Changes in the 1-octylnitrite/l-octanol and 2-octylnitrite/2-octanol ratios with pH ...... 47
9 Effect of the length of reaction of sodium nitrite and n-octylamine ...... 49
10 Effect of an ether layer on top of the sodium nitrite-n-octylamine reaction mixture on the conversion of alcohols into corresponding nitrites...... 50
11 Effect of temperature on the reaction of sodium nitrite and n-octylamine ...... 51
x 12 Effect of sodium nitrite concentration on the reaction of sodium nitrite and n-octylamine .... 53
13 Nitrite/alcohol ratios of the products at different sodium nitrite concentrations in the reaction mixture ...... 55
14 Effect of pH on the product distribution of the reaction of sodium nitrite and sodium nitroprusside ...... 56
15 Effect of temperature on the reaction of n-octylamine and sodium nitroprusside...... 59
16 Reverse-phase FPLC elution profiles of (A) Na-CBZ-L-Lysine (B) the aqueous phase and (C) the ether phase after ether extraction of the acidified reaction mixture ...... 60
17 Mass spectra of Pd2H-reduced products from the reaction of Na-CBZ-L-Lysine and sodium nitroprusside ...... 62
18 Elution profiles of the different derivatives in the amino acid analyzer ...... 63
19 Mass spectra of the Na-trifluoroacetylamino acid methyl ester of the most abundant material that elutes at 54 minutes in the amino acid analyzer ...... 67
20 Mass spectra of the Na-trifluoroacetylamino acid methyl ester of the material giving rise to the peak at 22 minutes in the amino acid analyzer...... 70
21 Amino acid chromatogram after hydrazine treatment of the dried residue from the reaction of hot HCl and 2-amino-5-hexenoic a c i d ...... 73
22 Mass spectra of the Na-trifluoroacetylamino acid methyl ester of the material giving rise the peak at 53 minutes in the amino acid analyzer ...... 75
xi 23 Time-course of the reactions of 2-amino-5- hexenoic acid in 5.8M HCl at 105°C ...... 77
24 Time-course of the reactions of 6-hydroxy- norleucine in 5.8M HCl at 105°C ...... 79
25 Cation-exchange chromatogram of (A) commercial RNAse A and (B) after purification ...... 81
26 Cation-exchange chromatogram of the reaction mixture of RNAse A and sodium nitroprusside .... 82
27 Cation-exchange chromatographic profiles of nitroprusside-modified RNAse A (A) without heating and (B) after heating at 65°C for 15 minutes ...... 84
28 Gel filtration of the products of the reaction of RNAse A and sodium nitroprusside ...... 86
29 Calculation of molecular weight of native and modified RNAse A from gel filtration data ..... 87
30 Amino acid profile of the nitroprusside- modified RNAse A after HCl hydrolysis ...... 90
31 Proposed reaction pathways and intermediates of the deamination of amines with nitroprusside. 94
32 Proposed reactions of water with the carbonium ion intermediate ...... 96
33 Possible routes for the formation of the lactone during treatment of 2-amino-5-hexenoic acid with hot HCl ...... 105
34 Schematic representations of the (A) SNP- mediated deamination of Na-CBZ-L-Lysine and the chemical transformations during incubation of (B) 6-hydroxynorleucine and (C) 2-amino- 5-hexenoic acid with hot HCl ...... 107
35 Gas chromatogram of the ether extract of the reaction solution containing n-butylamine, CH20 and NaCN ...... 127
xii 36 Reaction of n-butylaminoacetonitrile with formaldehyde and NaCN ...... 128
37 Cation-exchange chromatograms of the reaction mixture containing RNAse A, formaldehyde and NaCN (A) without inhibitor (B) with excess phosphate and (C) with excess 3'-CMP ...... 130
38 Gel filtration chromatograms of (A) the dialyzed reaction mixture of RNAse A, formal dehyde and NaCN (B) native RNAse A and (C) modified RNAse A ...... 133
39 Estimated molecular weights of the native and modified RNAse A ...... 134
40 SDS-gel electrophoresis of the native and modified RNAse A ...... 136
41 Variation in retention times with pH during gel filtration of the native and modified RNAse A ...... 137
42 HPLC chromatograms of the different DNP- derivatives of (A) native and (B) modified RNAse A ...... 139
43 HPLC chromatograms of the tryptic digests of (A) native and of (B) modified RNAse A ..... 141
44 Lineweaver-Burk plots of the native and the modified RNAse A ...... 143
xiii LIST OF ABBREVIATIONS
A s n ...... Asparagine
A s p ...... Aspartic acid
AU ...... Absorbance Unit
° C ...... Degrees Celsius (centigrade)
CBZ ...... Carbobenzoxy- or Benzyloxy carbonyl-
Cl ...... Chemical ionization
c m ...... Centimeter
CM ...... Carboxy methyl-
cCMP ...... Cytidine-23’-cyclic monophosphate
Cys ...... Cysteine
DNP ...... Dinitrophenyl-
EI ...... Electron impact e V ...... Electron volt
FDNB ...... l-Flouro-2,4-dinitrobenzene
FPLC ...... Fast Protein Liquid Chromatography g ...... gram
GC ...... Gas chromatography
HEPES ...... [4-(2-hydroxyethyl)-1-piperazine ethane sulfonic acid
His ...... Histidine
HPLC ...... High Performance Liquid Chromatography
xiv K m ...... Michaelis' constant
Lys ...... Lysine
M ...... Molar m/e ...... Mass-to-charge ratio p ...... micro- , micron m ...... milli-
MS ...... Mass spectrometry
N ...... Normal n m ...... nanometer
# ...... number
Pd ...... Palladium
Pd 2 H ...... Hydrogenated palladium black
% (w/w) ...... Percent by weight
% (w/v) ...... Percent weight per unit volume
% (v/v) ...... Percent by volume
Phe ...... Phenylalanine
RNAse A ...... Ribonuclease A
RT ...... Retention time
RV ...... Retention volume
Ser ...... Ser
SN-1- ...... nucleophilic substitution, unimolecular
SN^ ...... nucleophilic substitution, bimolecular
SNP ...... Sodium nitroprusside rpm ...... revolutions per minute
xv rpm ...... revolutions per minute
TCD ...... Thermal conductivity detector
TEA ...... Triethanol amine
Thr ...... Threonine
TPCK ...... Tosyl phenylalanyl chloromethyl ketone
Tyr ...... Tyr
UV ...... Ultraviolet
Val ...... Valine
VM ...... Maximum velocity (enzymatic) PART I.
DEAMINATION OF LYSINE RESIDUES IN RNAse A
INTRODUCTION
The deamination of amines with nitrous acid was discovered by Piria in 1846 (In Ridd, 1961) when he showed that treatment of either aspartic acid or asparagine with nitrous anhydride gave malic acid. The action of nitrous acid on aliphatic and aromatic amines had since generated interest for a variety of synthetic reasons (Streitwieser,
1957; Ridd, 1961; Collins, 1971; Moss, 1971).
"Nitrous acid" in this thesis is a general term used to include HONO, H20-NO+, N0+ and N203, the species that are most frequently found in aqueous solutions. The rela tive proportion of each species varies with pH, concen tration and the presence of other ions in solution (Ridd,
1961).
The reaction of secondary amines with nitrous acid produces nitrosamines; with aromatic amines the reaction produces a relatively stable diazonium; but with aliphatic amines the reaction proceeds to form a wide variety of
1 2
deamination products (Ridd, 1961). Thus, the reaction of nitrous acid with n-butylamine in IN HCl produced not only n-butyl alcohol but also sec-butyl alcohol, n-butylchloride, sec-butylchloride, n-butenes, traces of butyl nitrites, and some high-boiling materials (Whitmore and Langlois, 1932).
The ability of nitrous acid to deaminate aliphatic amines was extended into the deamination of e-amino groups of lysine residues and free a-amino terminal groups of proteins (Philpot and Small, 1938; Herriot, 1947; Gertler and Hofmann, 1967; Schrimger and Hofmann, 1967; Kurosky and
Hofmann, 1972; Knowles et al., 1974; Malin et al., 1984).
However, its usefulness as a reagent for selective chemical modification of specific amino acid residues has not' really gained wide applicability due to a variety of problems associated with the reagent: (1) the reaction is usually done at low pH where very few proteins remain active or retain their native conformations at that pH. (2) the occurence of a number of side reactions such as the forma tion of 6-N-nitrosamine with tryptophan, o-nitrosotyrosine with tyrosine (Kurosky and Hofmann, 1972), and the oxida tion of -SH groups. (3) the multiplicity of reaction products which are usually difficult to separate, identify and determine quantitatively. When sodium nitroprusside^ (SNP) was shown by Maltz et
al. (1971) to also deaminate amines under neutral or
slightly alkaline conditions, this opened up a potential
alternative reagent for selective deamination of proteins.
Sodium nitroprusside [Na2 Fe(CN)gNO.2 H 2 O] is a hydrated inorganic salt that readily dissolves in water to form a brownish-red solution. The crystal has an orthorhombic structure and is ruby-red in color. The nitroprusside ion has an approximate C^v symmetry and all CN“ groups seem to be equivalent (Swinehart, 1967). The overall structure of the nitroprusside ion can be schematically represented as:
CN .CN
CN Fe CN
+O N ^ | CN
Aqueous solutions of sodium nitroprusside are rela tively stable when kept in the dark but sensitive to decomposition upon exposure to light (Wolfe and Swinehart,
^Sodium nitroprusside is also known as sodium nitroprussiate, sodium nitrosylpentacyanoferrate(II), sodium pentacyanonitrosylferrate(II), sodium nitroferri- cyanide, Nipride™ (Roche), and sodium pentakis(cyano-C)- nitrosyl ferrate. 4
1975; Frank et al., 1976; Vesey and Batistoni, 1977; van
Loenen and Hoffs-Kemper, 1979; Bisset et al., 1981; Arnold et al.. 1984; Mahony et al.. 1984). When exposed to light the sodium nitroprusside molecule is photoexcited which leads to subsequent cleavage of nitric oxide (NO*) moiety from [Fe-^-^CN^]2-, which upon aquation produces
[FeIII(cN)5H20]2“ (Wolfe and Swineheart, 1975). This decom position is noticed as a gradual darkening of the brownish- red solution and subsequently into a bluish green color.
The chemical mechanism and nature of the intermediates of the decomposition are not yet fully understood but the ultimate products include pentacyanoaquoferrate(III) ion, cyanide, nitric oxide and possibly others.
The chemical and pharmacological properties of sodium nitroprusside (see reviews by Johnson, 1929; Swinehart,
1967; Tuzel, 1974; Tinker and Michenfelder, 1976; Kreye,
1980; Leewenkamp et al.. 1984) have attracted considerable interest and controversy at various times since its biolo gical effect was first reported by Hermann in 1886 (In
Johnson, 1929).
For more than two decades sodium nitroprusside has been extensively used as a vasodilatory, hypotensive agent in the treatment of angina, acute congestive heart failure and hypertensive emergencies. Its main advantages lie on its high potency, rapid, brief and reversible action. It is 5 we11-documented that vasodilation and the consequent reduction in blood pressure occuring upon the pharmacolo gical administration of sodium nitroprusside results from a relaxation of the vascular smooth muscles (Schlandt et al.,
1977; Schultz et al., 1977; Gruetter et al., 1979, 1981;
Kadowitz et al., 1981; Ignarro et al.. 1981) and is accompanied by an increase in the activity of vascular smooth muscle guanylate cyclase (Katsuki et al., 1977;
Mittal and Murad, 1977; Ignarro et al., 1980a,b; Craven and
DeRubertis, 1983; Kamisaki et al., 1986). The activation of guanylate cyclase by sodium nitroprusside requires the presence of heme and has been attributed to the formation of a NO-heme complex, similar to that thought to be involved in guanylate cyclase activation by nitric oxide and other nitroso-compounds (Mittal and Murad, 1977; Craven and DeRubertis, 1978; Craven et al.. 1979; Ignarro et al..
1981, 1982; Ohlstein et al., 1982; Ignarro et al.. 1984).
However, the precise biochemical pathways and the nature of the intermediates have not yet been fully understood.
Concerns about the potential health risks of using sodium nitroprusside have been expressed recently (McDowell et al., 1974; Merrifield and Blundell, 1974; Greiss et al.,
1976; Park and Means, 1985). It should be pointed out that the biological effects of sodium nitroprusside were first observed in dead animals suspected of being poisoned by 6
cyanide (In Johnson, 1929). The sodium nitroprusside mole
cule contains five cyanide groups which can be potentially
released through photochemical decomposition during the preparation and storage of SNP solutions (Wolfe and Swine-
hart, 1975; Frank et al., 1976; van Loenen and Hoffs-
Kemper, 1979;, Arnold et al., 1984; Mahony et al., 1984;
Hartley et al.. 1985). After infusion into the blood, more
cyanide can be potentially released from the interaction
of SNP with specific components in the erythrocytes such
as sulfhydryl-containing compounds (Hill, 1942; Page et
al.. 1955), hemoglobin (Smith et al.. 1974; Nakamura et
al., 1977; Vesey et al., 1980), and quite possibly other components in the other tissues (Kreye, 1980; Tinker and
Michenfelder, 1976). Indeed, infusion of sodium nitroprus
side into laboratory animals and human patients has been reported to result in increased cyanide concentrations in the blood (Vesey et al.,1976; 1979; 1982). Sodium nitro prusside can also potentially react directly with amine- containing drugs (Park and Means, 1985) that are adminis tered simultaneously into the patient.
When the ability of sodium nitroprusside to deaminate amines was demonstrated by Maltz et al. (1971) almost 20 years ago, this generated a new area of research because many biologically important substances contain amine 7
groups in their structures. These include the proteins,
nucleic acids and a number of drugs. Modification of the
catalytically-important lysine residues in enzymes can
potentially alter its function (Park and Means, 1987).
Deamination of the amino groups of guanine, adenine and
cytosine can potentially lead to mutagenesis. Many drugs
also have secondary amines in their structures which,
under physiological pH, had been shown to give N-nitros-
amines (Park and Means, 1985) which are potential carcino
gens.
On the brighter side, the ability of sodium nitroprus
side to deaminate amines makes it a potential reagent for
site-specific lysine modification, especially in enzymes
that have lysine at or near their active sites. Such
selective modification can be a useful tool in studying
the mechanisms of enzyme action for the purpose of inhibi
ting or enhancing their activity. For example, angiotensin
I converting enzyme (EC 3.4.15.1) was rapidly inactivated when incubated with sodium nitroprusside (Park and Means,
1987). This inactivation was believed to be a result of
chemical modification of a specific lysine residue in or
near the active site since the presence of competitive
inhibitors during incubation suppressed the inactivation by SNP. 8
Sodium nitroprusside-mediated deamination and subse quent reactions can also be potentially important in the synthesis of other compounds. For example, McGarvey and
Kimura (1986) reported that the use of SNP in place of nitrous acid decreased the relative proportions of the unwanted olefinic side products during the conversion of nitrogen to oxygen at some saturated carbon centers of model compounds. Katho et al. (1984) and Beck et al.
(1984) showed that with some diamino acids the amino group in the side chain was preferentially deaminated and, where five- or six-membered ring formation was possible, formed a cyclic structure with the amino group. Thus, proline and pipecolic acids were synthesized from the reaction of SNP with ornithine and lysine, respectively.
Recent interpretation of the mechanism of deamination of amines by nitroprusside is understandably seen in the light of the long-known deamination reaction by nitrous acid. The [ (CN^FeNO]^" ion can be formally considered as an iron(II) complex which is a carrier of the nitrosonium ion (Swineheart, 1967). Both reactions are presumed to start with an attack by an electrophilic nitrosonium equi valent (NO+) on the amino group, followed by a series of rearrangement and elimination reactions (Whitmore and
Langlois, 1932; Brewster et al., 1950; Roberts and Mazur,
1951; Streitwieser and Schaeffer, 1952; Collins, 1971; 9
Moss, 1971; Maltz et al., 1971; Collins and Benjamin,
1972; Katho et al., 1984; Dozsa et al., 1984; White and
Field, 1975). Most of the postulated mechanisms involve
the formation of diazonium ion intermediate which, in
competing fashion, is either displaced by a nucleophile
(SN2 type) or undergoes unimolecular fission into a carbo- nium ion (SN-*- type). The carbonium ion can then further react with an appropriate nucleophile to form a product, eliminate a proton to yield olefin an or rearrange a hydrogen or a carbon function to give a new carbonium ion.
In all cases, aliphatic amines give a mixture of alcohols and olefins upon deamination by either nitrous acid or nitroprusside in aqueous solution.
In an effort to understand the mechanism of the reaction of SNP with amines, several workers have used a variety of model amines and amino acids. The rate of deamination depends on the basicity of the amine (Katho et al.. 1984) and increases slightly with pH (Dozsa et al..
1984). The nature of the final products also depends on the presence of internal nucleophiles that can potentially react with either the diazonium ion or the carbonium ion to form cyclic products (Katho et al.. 1984; Beck et al.,
1984; McGarvey and Kimura, 1986). 10
Sodium nitroprusside has an advantage over nitrous
acid, the traditional deaminating reagent, since deamina
tion can be carried out under basic conditions (Maltz et
al., 1971). Such reaction under neutral or slightly alka
line conditions especially suitable for protein studies
where irreversible denaturation can occur at low pH values,
the optimum range of nitrous acid reactivity. Furthermore,
nitrous acid has very little selectivity since, aside from
its reactivity towards e-amino groups of lysine and
a-amino terminal group, it can also react with the ring
nitrogen of tryptophan and, to a lesser extent, with the
aromatic ring of tyrosine (Herriott, 1947; Kurosky and
Hofmann, 1972). Indeed, the use of nitrous acid for the
modification of enzymes was limited to NH2-terminal modi
fication (Gertler and Hofmann,1967; Schrimger and Hofmann,
1967; Kurosky and Hofmann, 1972).
As of this writing, there has been only one report
(Park and Means, 1987) on the use of sodium nitroprusside
for selective modification of lysine residues in proteins,
thus the need for more investigations. In this study,
RNAse A (E.C.2.7.7.16) was used as a model protein because
it is probably one of the best characterized proteins (see
reviews by Richards and Wyckoff,1971; Blackburn and Moore,
1982). It has also at least one lysine residue thought to be involved in catalysis. It is a relatively small protein, 11
with 124 amino acids and a molecular weight of 13,700 daltons (Hirs et al., 1956). Its complete amino acid sequence had been determined (Smyth et al., 1963) and its tertiary structure is shown in Figure 1 (Cantor and
Schimmel, 1983).
The mechanism of enzyme catalysis has been illucidated mostly from X-ray crystallographic and chemical modifica tion studies. The preliminary crystal structure of the intact RNAse A has been determined in several laboratories
(Avey et al., 1967; Kartha et al., 1967; Carlisle et al..
1974) and a higher resolution more recently (Wlodawer,
1980; Wlodawer et al., 1982). Several amino acid residues have been postulated to be at or near the active site:
His-12, His-119, Lys-41, Asn-44, Thr-45, Asp-121, Ser-123,
Val-43, and Phe-120 (Richards and Wyckoff, 1971). However, most X-ray crystallographic data indicate that only His-12,
His-119 and Lys-41 are directly involved in catalysis.
The other side chains are assumed to mainly play structural roles.
Bovine pancreatic RNAse A breaks down RNA chains by cleaving the phosphodiester bond between the 3' and 5 1 positions of the ribose moieties in a stepwise manner. The first is hydrolysis of the phosphodiester bond between nucleosides to give an oligonucleotide terminating in a Figure 1. The tertiary structure of ribonuclease A showing the amino acid residues in the active site (Reproduced from Cantor and Schimmel, 1980). 13
pyrimidine 2':3'-cyclic monophosphate. The second is the
hydrolysis of the cyclic monophosphate to give a terminal
3'-phosphate (Roberts et al., 1969). These are summarized
in Figure 2. The two steps are distinctly separate and
cyclic phosphates, the mandatory intermediates (or pro
ducts of the first step) can be isolated in good yield.
Therefore, even if RNA is the natural substrate of RNAse A,
the use of synthethic cytidine- or uridine-2 ':3'-cyclic monophosphate is common for the assay of RNAse A activity.
This model (Figure 2) is a general acid-base catalysis
facilitated by His-12 and His-119. Lys-41 is thought to be
involved in stabilizing the pentacoordinated transition
state intermediate but is not thought to participate in
the acid-base catalysis (Wlodawer, 1983). Although this model may not be perfect, this is consistent with most published data.
Since His-12, His-119 and Lys-41 were long suspected to be involved in catalysis, several chemical modification studies on these specific residues have been done in an effort to understand the catalytic mechanism (Richards and
Wyckoff, 1971; Blackburn and Moore, 1982). Although there are 10 e-amino and one a-amino groups in RNAse A, the e-amino group of Lys-41 has been found to have greater reactivity towards a variety of chemical modification reagents. This is often attributed to its significantly 14
B o il CH,
HO OH NH NH*HN
119 B o si
© ...H jN
HO OH HN NH
B ase CH2OH 119
S'y H ,N 1 Al HOCH, 0 Bala
HO OH HN NH
119
B o te
NH HN /\
119
NH HN or" 119
8 b G..8*1
HN or O-H 1(9
HOCH.
Figure 2. A general model of the mechanism of RNAse A catalysis (Roberts et al., 1969). 15
lower pKa of 8 . 8 (Murdock, et al., -1966) which is at least
1.5 pH units below the pKa of the other 9 e-amino groups.
Although the selectivity towards Lys-41 is often not 100 percent, the fraction containing the modification at
Lys-41 can often be isolated, purified and studied. Thus,
RNAse A modified at Lys-41 by carboxymethylation (Hein- rickson, 1966), reaction with pyridoxal phosphate followed by reduction with borohydride (Means and Feeney, 1971;
Raetz and Auld,1972; Riquelme et al., 1975; Dudkin et al..
1975), arylation with 2-carboxy-4,6 -dinitrochlorobenzene
(Bello et al., 1979; Iijima et al., 1977), amidination
(Blackburn and Gavilanes, 1980; Reynolds, 1968; Blackburn and Jailkhani, 1979), guanidination (Richards and Wyckoff,
1971) or dinitrophenylation (Murdock et al.# 1966) led to complete or nearly complete inactivation. Different chemical modifications of the other lysine residues has generally produced much lower degrees of inactivation
(Richards and Wyckoff, 1971).
Similarly, both His-12 and His-119 have been found to have higher reactivity with a number of reagents compared to free histidine or histine in short peptides. Thus selective chemical modification of either His-12 or
His-119 has been accomplished and both products were essentially inactive. Crestfield et al. (1963a;b) 16
found that the reaction of RNAse A with iodoacetate produced a major l-CM-His-119-RNAse A and a minor 3-CM-His-
12-RNAse A which were both inactive towards RNA or cytidine-23'-cyclic monophosphate. The reaction was found to be mutually exclusive which led them to postulate that His-12 and His-119 are close to each other and that one histidine residue seems to orient the reagent to attack the other. The relative reactivities of His-12 and
His-119 vary depending on the structure of the modifying reagent as well as the pH of the reaction (Richards and
Wyckoff, 1971). Other methods of chemically modifying
His-12 and/or His-119 also resulted to either partial or total loss of RNAse A activity.
Although the larger objective of this research project is to explore the potential use of sodium nitroprusside as a reagent for selective modification of lysine residues in proteins, the specific aims of this study are:
(1) To identify the products and determine the product distribution of the nitrous acid- and SNP-mediated deami nation of lysine and other model amines under a variety of reaction conditions.
(2) To elucidate the transformations that occur to the primary deamination products during HC1 hydrolysis. MATERIALS
The following chemicals used in these experiments are listed with their corresponding catalog numbers in paren thesis. Purchased from Sigma Chemical Co., P.O. Box 14508,
St. Louis, Mo. 63178 were acetic anhydride (A-6404), cytidine-23'-cyclic monophosphate (C-9630), cytidine-3- monophosphate (C-1133), hydrazine hydrate (H-0883), iodo- acetic acid (1-6375), Na-CBZ-L-lysine (C-7130), Na-CBZ-L- lysine methyl ester (C-3377), Na-2,4-DNP-L-lysine hydro chloride (D-0380), N,N 1-di(2,4-DNP)-L-lysine (D-0255), ribonuclease A Type IIA (R-5000), sodium nitroprusside dihydrate (S-0501), trifluoroacetic anhydride (T-8258), and trypsin (T-8642).
Purchased from J.T. Baker Chemical Co., 222 Red School
Lane, Phillipsburg, N.J. 08865 were acetic acid (9507), acetonitrile (9017-3), citric acid monohydrate (0110), diethyl ether (9244-3), formic acid (5-0128), methanol
(9093-3), methylene chloride (9315), potassium perchlorate
(3220), sodium borohydride (V023) and sodium cyanide
(3662).
17 Purchased from Aldrich Chemical Co. Inc., 940 W. St.
Paul Ave, Milwaukee, Wi 53233 were 1-octanol (11,261-5),
1-octene (0-480-6), n-octylamine (0-580-2), and palladium chloride (28,360-6).
Purchased from Fisher Scientific Co., Fair Lawn, New
Jersey 07410 were ammonium bicarbonate (A-641), hydro chloric acid (A-144), sodium chloride (S-271), sodium hydroxide (S-318), sodium nitrite (S-347), and succinic acid (A-294).
Purchased from Mallinckrodt Inc., Paris, Kentucky 40361 were ammonium hydroxide (3256), hydrogen peroxide (5240) and phosphoric acid (2796).
2-Octanol (6 6 ) was from Eastman Organic Chemicals, P.O.
Box 92894, Rochester, New York 14650.
All chemicals used in these experiments were used without further purification, except ribonuclease A, the purification of which will be described under "Methods". METHODS
Reaction of n-octylamine with sodium nitrite
a) Effect of pH
Solutions containing 100 y,l of 1M n-octylamine in 800 y.1 of 0.2M succinic acid were mixed in each of a series of
16 x 150 mm test tubes. The pH values were adjusted to within a range of 2.0 to 6.0 at 0.5 pH intervals with either 6 N HC1 or 6 N NaOH. Then, 100 ul of 5M NaN02 was added to each tube (total volume = 1 . 0 ml) with vigorous shaking for about 5 seconds. The tubes were covered with parafilm and the reactions were allowed to proceed at room temperature (22°C) for 30 minutes with occassional shaking.
No pH adjustment was made throughout the course of the reaction.
One ml of diethyl ether and 0.5 ml of concentrated ammonium hydroxide were added to each reaction mixture, followed immediately by vigorous shaking (vortexed at high speed) for about 1 minute. The layers were allowed to sepa rate and the ether layer was pipetted into another test tube, immersed in an ice bath, and covered with parafilm.
19 20
Identification and analysis of products
Aliquots of the ethereal extract (typically 1-3 y.1) or
solutions of authentic standards (where available) were
injected into a Varian Model 3300 Gas Chromatograph equip
ped with a thermal conductivity detector. It was fitted
with a stainless steel column (6 1 x 1/8" I.D) packed with
3% OV-17 on Gas-Chrom Q (100-200 mesh). Helium flow rate
was set at a constant 30 ml per minute. In most analyses,
the initial column temperature was set at 50°C and was
programmed to increase at 20°C per minute to 150°C, with
one minute holding time before and after the temperature
gradient. Thermal conductivity detector (TCD) conditions
were: temperature, 250°C; filament temperature, 250°C; and
filament current, 178 mA. The injector port was maintained
at 260°C.
The products were identified by comparing their reten
tion times with those of the standards, whenever available.
Their respective amounts were calculated by comparing
their peak areas with those of the standards. After
correcting for differences in TCD sensitivities, their
relative abundance, expressed in percent, were calculated
based on the total amounts of products and remaining
n-octylamine. 21
The identity of the products were further confirmed by
subjecting both the samples and the available standards to
GC-mass spectrometry. In most cases, analyses were done
on a quadropole type GC-MS computer system (Finnigan Model
4021) fitted with a combined EI/CI ion source set at 250°C.
When necessary,the samples were directly injected into the
ion source with a heated direct inlet probe. Chemical
ionization in methane (99.999%) was used to determine the parent molecular ions and electron impact at 70 eV to differentiate the isomers.
Operational parameters and data processing were
controlled by a Nova data system that stores a National
Bureau of Standards library of mass spectra of standard compounds^. The identification of the samples was therefore facilitated by a computer library search to match the
fragmentation pattern of the samples with those of known standards.
When high resolution mass spectra were necessary, the
GC mass spectrophotometric analyses were done on a Hewlett-
Packard Model 5890 gas chromatograph coupled to VG 70-250S mass spectrometer. This made possible the determination of molecular formula from elemental analyses based on the exact masses of either parent or other significant peaks.
^All mass spectral analyses were done at the Campus Chemical Instrumentation Center, Ohio State University. 22
b) Effect of the length of reaction
A solution containing 800 p.1 of 1M n-octylamine and
6.4 ml of 0.2M succinic acid was mixed in a 16x150 mm test tube and the pH was adjusted to 3.5 with 6 N HC1. A 900 y.1 aliquot of this solution was placed into each of a series of test tubes. Into each tube was added 100 p.1 of 5M NaN02 and the mixture shaken vigorously for about 5 seconds. The tubes were covered with parafilm and incubated at room temperature. At different time intervals, 1.0 ml of ether and 0.5 ml of concentrated ammonium hydroxide were added to one of the tubes and immediately vortexed at high speed for about one minute. The layers were allowed to separate and the products of the reaction were quantified by gas chromatography as before.
In some separate treatments, 1.00 ml of ether was layered on top of each reaction mixture right after mixing with NaN02 - At similar time intervals, 0.5 ml of concen trated ammonium hydroxide was added to one tube and the products were extracted and analyzed as before. At longer time intervals, it was necessary to add ether to maintain its volume at about 1 . 0 ml. 23
c) Effect of temperature
Five hundred y.1 of 1M n-octylamine was added into 4 ml of 0.2M succinic acid. After brief mixing the pH of the
solution was adjusted to 3.5 with 6 N HCl or to 4.5 with 6 N
NaOH. A 900-p.l aliquot was transferred into each of 4
tapered test tubes and pre-incubated at the desired tempe
ratures for about 5 minutes. Then 100 p.L of 5M NaN0 2 was
added to each and shaken vigorously for about 5 seconds.
Immediately, each tube was sealed by fusing at the tapered portions and further incubated at their respective tempe ratures for 30 minutes.
At the end of the reaction period the tapered tips were cracked open and 1.0 ml of ether and 0.5 ml concen trated ammonium hydroxide were added. The mixtures were immediately vortexed at high speed for about 1 minute. As soon as the layers separated, the ether layer was pipetted out into another tube and the product composition was determined by gas chromatography as before.
d) Effect of nitrous acid concentration
One hundred p.1 of 1M n-octylamine were added into each of a series of 16 x 150 mm test tubes containing calculated volumes of 0 .2 M succinic acid to make the final volume of
1.00 ml after the addition of the required volume of 5M 24
NaNC>2 . The pH of the resulting solutions were adjusted to
3.5 with 6 N HC1. Then 20, 60, 100 and 200 \l1 of 5M NaN0 2 were added to the respective tubes with vigorous mixing for about 5 seconds. The tubes were covered with parafilm and the reaction was allowed to proceed for 30 minutes at room temperature (22°C). After the reaction, the products were extracted and analyzed by gas chromatography as before.
Reaction of n-octylamine with sodium nitroprusside
a) Effect of pH
Solutions containing 950 y.1 of 0.2M TEA and 50 y.1 of
1M n-octylamine were mixed in a series of test tubes. The solutions were adjusted to various pH values from 7.0 to
10 at 0.5 intervals by the addition of 6 N HC1. A similar set was prepared using 0.2M succinic acid at pH 5.0, 5.5,
6.0 and 6.5. The samples were then transferred into another set of tapered test tubes and 150 mg of sodium nitroprusside were added tp each. The mixtures were quickly shaken until the sodium nitroprusside completely dissolved. Then the tubes were sealed by melting at the tapered sections and the reaction mixtures were kept in the dark at room temperature (22°C) for about 24 hours with occasional shaking. No pH adjustment was made. 25
After the reaction period, 1.0 ml of ether and 0.5 ml of concentrated ammonium hydroxide were added to each tube and the mixtures were shaken vigorously (vortexed at high speed) for about 1 minute. The ether extract was pipetted out into another test tube, immersed in an ice bath, covered with parafilm and analyzed by gas chromatography as before.
b) Effect of temperature
Solutions containing 4.75 ml of 0.2M triethanol amine and 250 jj.1 of 1M n-octylamine were mixed in a 16 x 150 mm test tube. The pH was adjusted to the desired value (pH
7.5, 8.5 or 9.5) with 6 N HC1. A 1.0 ml aliquot of this solution was transferred into each of a series of tapered test tubes and pre-incubated at 0, 22, 37, and 50°C for about 5 minutes. Then, 150 mg sodium nitroprusside was added to each reaction mixture with vigorous mixing until the sodium nitroprusside had completely dissolved. The tubes were sealed by fusing at the tapered ends and further incubated at their respective temperatures for about 24 hours. After this reaction period, the glass seals were cracked open and the products were extracted and analyzed as before. 26
Reaction of Na-CBZ-L-Lysine and Sodium Nitroprusside
A 10-ml solution containing 2.98 grams of sodium nitroprusside dihydrate in 0.05M TEA, pH 8 . 6 was mixed with 280 milligrams of Na-CBZ-L-lysine in an equal volume of the same buffer. After 24 hours of incubation at 22°C in the dark the reaction mixture was brought to pH 2 with
IN HCl and the products of the reaction were twice extracted into 5.0 ml of diethyl ether. The pooled ethereal extract was washed with 1 0 ml of water and evaporated to dryness in a vacuum desiccator at 40°C. The residue was then taken into 2 ml of water-methanol-formic acid mixture (50:50:1) and injected in 200 ul portions into the FPLC (Model GP-250, Pharmacia Fine Chemicals) using a reverse phase column (Pro-RPC HR 5/10, Pharmacia
Fine Chemicals). The reaction products were eluted with a linear gradient of A (0.05% H 3 PO 4 in 0.01M KCIO 4 , pH
2.5) and B (A + CH3 CN, 1:3) solvents at a combined flow rate of 1 ml/min. The different peaks were collected separately and the pooled fractions were evaporated to dryness at 40°C in a vacuum desiccator. The residue was taken into 1.0 ml of methanol and the insoluble KC104 salt was removed by decantation.
The different methanol fractions were used separately in the subsequent steps. 27
HC1 Hydrolysis and Amino Acid Analysis
Aliquots of each methanol soluble fractions were transferred into separate hydrolysis tubes and dried in a vacuum desiccator at 40°C. Then, 0.5 ml of 6 N HC1 was added and hydrolysis was allowed to proceed for 2 2 hours at 105°C. After HCl hydrolysis the samples were evaporated to dryness at 40°C in a vacuum desiccator and the residues were dissolved in appropriate volumes of the starting buffer for amino acid analysis (Sodium citrate, pH 3.25,
Beckman Instruments). Appropriate aliquots were subjected to amino acid analyses performed on a Beckman 119 CL analyzer with a standard 0 . 6 x 2 0 cm column sulfonated poly-styrene resin, Type W-3H, Beckman 338038).
Derivatization for GC-MS
N-trifluoroacetylamino acid methyl ester derivatives were prepared from the different HCl-hydrolyzed products by the method of Knapp (1979). Separate aliquots (100 jj.1) of the HCl hydrolyzates were evaporated to dryness at 40°C in a vacuum desiccator and the residues were dissolved in
1.0 ml of 4M HCl in methanol (made by bubbling HCl gas into anhydrous methanol). The samples were heated at 65®C for 90 minutes in sealed and evacuated tubes. The samples were again evaporated to dryness at 40®C in a vacuum 28
desiccator. The residues were dissolved in 100 jil of
trifluoroacetic anhydride and allowed to stand for an hour
at 22°C. The samples were placed in an ice bath and
the excess reagent was evaporated to dryness through a
water vacuum. The residues were each dissolved in 200 ul
of methylene dichloride and subjected to GC-mass spectro
metry.
Reduction of the products of Na-CBZ-L-lysine and SNP reaction by Pdp-H.
a) Preparation of palladium black column.
Palladium black was prepared by the procedure of
El-Amin (1980). One ml of concentrated HCl was added to
one gram of palladium chloride in a 125-ml Erlenmeyer
flask. The flask was swirled gently to dissolve all of the
palladium chloride, forming a dark brown syrup. The
suspension was diluted with 50 ml of distilled water and
then heated in a water bath for about 5 minutes, after
which 0.3 ml of formic acid was added. While still in the
water bath, the pH of the solution was brought to about pH
9 (litmus paper) with 6 N NaOH. After about 2 more minutes,
the flask was removed from the water bath and the suspen
sion was neutralized with formic acid. The liquid was
decanted and the palladium black (precipitate) was washed 29 with another 50 ml water. The washing procedure was repeated 4 more times.
The freshly prepared suspension of palladium black was poured into a 1 cm x 10 cm (I.D.) column with a thin layer of glass wool at the bottom. When about 4 cm of the palladium black was packed into the column, the liquid was slowly drained until the liquid level was the same as that of the palladium black. Then 5 ml of 25 % formic acid was carefully added into the column and slowly allowed to drain. When excess formic acid was detected by litmus paper in the eluate, the stopcock was again closed and more formic acid was added to leave about 1 -cm space on top of the burette. The acid was allowed to stand in the column for about one hour and then drained. The now fully- hydrogenated palladium black was repeatedly washed with water until the eluate was neutral to litmus paper.
Finally, 5 ml of ^O-methanol-formic acid (50:50:1) solvent mixture was added to the column, then drained until the level of solvent was the same as that of the palladium black.
b) Pdp-H reduction
Appropriate aliquots (100-500 y.1) of the supernates from the reaction products between Na-CBZ-L-lysine and sodium nitroprusside were diluted to 3 ml with the H 2 O- 30
methanol-formic acid (50:50:1) solvent mixture. The sample was then carefully added into the freshly prepared palladium black column and the solvent slowly drained (0.5 ml/min.) until the solvent level was the same as that of the palladium black. Then the column was further washed with another 3 ml of the solvent mixture and the eluents were pooled. The completeness of the deprotection (and reduction) was checked by the absence of a characteristic peak at about 255 nm for the CBZ-group in the UV spectrum of the eluent. Aliquots of the eluents were then evapo rated to dryness at 40°C in a vacuum desiccator and the residues were taken into the starting buffer for amino acid analysis. Another set of aliquots were also evapo rated to dryness but the residues were dissolved in methanol for direct injection without derivatization into the mass spectrometer.
Purification of RNAse A
A 200 mg portion of commercial RNAse A (Sigma Type
IIA, R-5000) was dissolved in 20 ml water. Five hundred y.1 aliquots of the solution were injected into the cation exchange column (Mono S, Pharmacia Fine Chemicals HR 5/5) and eluted with buffer A (0.05M HEPES, pH 7.0) with a linear gradient of 0 to 40% buffer B (0.05M HEPES, pH 7.0 31 in 0.5M NaCl). The major fractions from several runs were pooled, dialyzed overnight with several changes of
1 x 10”3M HC1 and was finally lyophilized. To check for homogeneity, a 1.0 mg portion of the purified RNAse A was dissolved in 1 . 0 ml of water and rechromatographed under similar conditions. The rest of the purified RNAse A was stored in the freezer to be used in the subsequent steps.
Preparation of Nitroprusside-modified RNAse A.
A 13 mg portion of purified RNAse A was dissolved in
1.0 ml of 50 mM triethanolamine buffer, pH 8 .6 . This was followed by addition of 29.8 mg of sodium nitroprusside dihydrate. The reaction mixture was incubated in the dark at 22°C for 10 hours. The excess nitroprusside was removed by passing the reaction solution through anion-exchange column (1.5 x 5 cm, Biorad AG2 X- 8 ). The modified RNAse was eluted with water, pooled, dialyzed and then lyophi lized. Aliquots were taken for activity assay (Crook et al., 1960) and for HCl hydrolysis and subsequent amino acid analysis.
Separation of the Reaction Products
Portions of the eluted samples (500 y.1) were subjected to Fast Protein Liquid Chromatography (FPLC, Model GP-250) equipped with a cation-exchange column (Mono S, Pharmacia 32
Fine Chemicals). The sample was eluted with buffer A
(0.05M HEPES, pH 7.0) with a linear gradient of 0 to 40% buffer B (0.05M HEPES, pH 7.0 in 0.5M NaCl) at a combined flowrate of 1 ml per minute. The different peaks were collected separately and the same fractions from several runs were pooled.
Separate samples were heated at 65°C for 15 minutes before injection into the FPLC and chromatographed under similar conditions.
Each of the fractions were dialyzed for 3 hours against
1 x 10"3M HCl, which was changed every hour. Aliquots of the fractions were subjected to RNAse A activity assay
(Crook et al., 1963) and amino acid analysis. The rest of the isolated fractions were lyophilized and stored in the freezer for later use.
RNAse Assay
RNase activity was determined by an adaptation of the method by Crook et al. (1960). An 875 y,l solution of 0.05M succinate, pH 5.0 was placed in a 1-ml quartz cuvette
(1-cm path length). A 100-ml solution of 1.0 mg cytidine-
2 *:3'-cyclic monophosphate per ml of 0.05M succinate buffer, pH 5.0, was added into the cuvette and the solutions were mixed by carefully inverting the cuvette.
The cuvette was placed into the sample compartment of the 33
Cary 118C spectrophotometer. A reference solution was prepared containing the same components as in the sample, except that an equal volume of buffer was added in place of RNAse A solution. With the use of a 100-y.l glass syringe, 25 y.1 of RNAse A solution (1.0 mg RNAse A/ml buffer) was placed onto a teflon "applicator-stirrer". The enzyme solution was then added into the sample cuvette with quick mixing. The increase in absorbance was immediately monitored with time at 284 nm. Raw data from the absorbance/time slope were converted into specific activity units.
With the SNP-modified RNAse A fractions, adjustments in the operational parameters such as the amount of enzyme, chart speed and absorbance range were made to obtain good slopes within reasonable time. The percent activity was calculated by dividing the specific activity of the various SNP-modified RNase A fractions by the specific activity of the native RNAse A and multiplying the quotient by 1 0 0 .
Gel Filtration of the nitroprusside-modified RNAse A.
Aliquots of RNAse A solutions, SNP-modified RNAse A reaction mixture, or isolated fractions of its reaction products were injected into the GF-250 gel filtration 34 column (ZorbaxR bio-series, Du Pont Co.) connected to a
VarianR Model 5000 HPLC system. The samples were eluted with 0.2M PC>4 = buffer, pH 7.0 at 0.5 ml per minute.
Protein standards were used to calibrate the column.
In some cases the modified SNP-modified RNAse A were heated at 65°C for 15 minutes before injecting into the gel filtration column.
Amino Acid Analysis
All amino acid analyses were done on a BeckmanR Model
119 CL amino acid analyzer fitted with a 0.6 cm x 20 cm column packed with cross-linked, sulfonated polystyrene resin (Type W-3H, Beckman 338038). A 120-minute program was used and amino acid calibration standards were always run for every batch of samples. RESULTS
Reaction of n-Octylamine with Nitrous Acid
A typical gas chromatogram of an ether extract of the
products of the n-octylamine-nitrous acid reaction is
shown in Figure 3. The retention times for peaks #1, 5, 6
and 7 matched exactly with those of authentic 1-octene,
2 -octanol, n-octylamine and 1 -octanol standards, respec
tively. The identity of the products were further
confirmed by GC-mass spectrometric data, as summarized in
Table I.
The mass spectra of the materials in peaks #1 and #2
(Figure 3) both showed prominent M-l peaks at m/e 112
under chemical ionization in methane. However, after
fragmentation by electron impact at 70 eV only a very
small M-l peak at m/e 112 was observed. Comparison of
their fragmentation patterns with those stored in the
National Bureau of Standards computer library of the mass
spectrometer system indicated that peaks # 1 and # 2 in
Figure 3 were 1-octene and 2-octene, respectively.
The mass spectra of the material in GC peaks #5 and #7
in Figure 3 showed prominent M-l peaks at m/e 129 under
chemical ionization but very small M-l peaks under
electron impact. Their fragmentation patterns matched with
35 36
Solvent 1.0
0.8
P 0.6-
0.4-
0.2 -
l j
0 2 4 6 8 10 12 14 Retention Time, min.
Figure 3. A typical gas chromatogram of the ether extract of the products of the reaction of n-octylamine with sodium nitrite (100 mM n-octylamine and 500 mM sodium nitrite in 0.2M succinate buffer, pH 3.5, 22°C and 30 minutes). GC conditions were described under "Methods". 37
Table 1. Summary of the gas chromatographic and mass spectral data employed to identify products of the reaction between initrous acid and n-octylamine.
# Products M.W. R.T. Parent Peaks Other MS Data (min.) (Cl) (El)
1 ) 1 -octene 1 1 2 1.70 1 1 1 1 1 2 Computer library search (M-l) matched 1 -octene.
2 ) 2 -octene 1 1 2 1.90 1 1 1 1 1 2 Computer library search (M-l) matched 2 -octene.
3) 2 -octyl- 159 3.75 157 158 Exact mass at 158.1175 nitrite (M-2) (M-l) indicated CgH^g0 2 N
4) 1 -octyl- 159 4.30 157 158 Exact mass at 158.1186 nitrite (M-2) (M-l) indicated CgH^g0 2 N
5) 2 -octanol 130 4.45 129 129 Computer library search (M-l) (M-l) matched 2 -octanol.
6 ) 1 -octyl- 129 4.80 130 none None amine (M+l) 7) 1 -octanol 130 5.15 129 129 Exact mass at 130.1331 (M-l) (M-l) indicated CgH^gO; computer library search matched 1 -octanol.
2 -chloro- 148 4.15 None 1 1 2 Computer library search octane (M-HC1) matched 2 -chlorooctane.
1 -chloro- 148 4.65 None 148 Exact mass at 148.1002 octane indicated CoH^Cl computer library search matched 1 -chlorooctane.
The 1- and 2-chlorooctanes were found in significant amount only when NaCl was deliberately incorporated into the reaction mixture. (M.W. = molecular weight, R.T. = retention time, Cl = chemical ionization,and El = electron impact) 38
those of authentic 2 -octanol and 1 -octanol standards, respectively. A further comparison of their mass spectra with those in the computer library confirmed that the materials in peaks #5 and #7 were indeed 2-octanol and
1 -octanol, respectively.
GC peaks #3 and #4 in Figure 3 were suspected to be octylnitrites but because no authentic 1 - and 2 -octyl- nitrite standards were available and no mass spectra were stored in the computer library for comparison, a more detailed interpretation of the mass spectral data will be presented here to establish their identities. As indicated in Table 1, the chemical ionization mass spectra of the material in GC peaks #3 and #4 both showed small M-2 peaks at m/e 157, which seemed to suggest a loss of 2 hydrogens from an octyInitrite, MW = 159 and very prominent peaks were observed at m/e 129, corresponding to the loss of NO, also suggested that these products were nitrites. This ease of removal of NO is a known characteristic of nitrites (D'Or and Collin, 1953) and not of the nitrates, their structural isomers.
When the materials in GC peaks #3 and #4 were analyzed by electron impact mass spectrometry, much smaller parent ions (M-l) were detected at m/e 158 (Figure
4, inserts), so weak that they were detected only when a more sensitive (high resolution) mass spectrometer was 39 used (VG 70-250 S). Computer-based elemental analyses from the exact masses of both indicated the molecular formula + c 8h 16°2N' possibly from the formation of R-CH=0-N0 and -hp-NO R-C-CH^ from 1- and 2-octylnitrite, respectively. This was supported by the mass spectra of the material in GC peak #3 (2-octylnitrite) which showed its characteristic + peak at m/e 74, (Figure 4A) corresponding to CH3 CH=0 -N0
(D’Or and Collins, 1953). That of the material in GC peak
#4 (1-octylnitrite) showed the characteristic peak at m/e + 60 (Figure 4B), corresponding to CH2 =0 -NO. An interpreta tion of the fragmentation patterns from their respective mass spectra are shown as inserts in Figure 4. Finally, when 1 - and 2 -octylnitrite were prepared by another known procedure (Noyes, 1936) and subjected to GC-mass mass spectrometry (electron impact) they gave the same reten tion times and fragmentation patterns with those of the 1 - and 2 -octylnitrites obtained in this reaction.
When NaCl was present during the deamination reaction, the gas chromatogram of the products showed two additional peaks at the retention times of 4.15 (almost co-eluted with 1-octylnitrite) and 4.65 minutes (co-eluted with n-octylamine), corresponding to 2 - and 1 -chlorooctane, respectively. The identities of these products were confirmed by computer library search-and-match and, in the case of 1 -chlorooctane, also by computer-based elemental analysis of the exact mass of the parent molecular ion. At |— ” "i
CH3-CH2-CH2-CHj-CH2-CH2-CH-CHi! r ° i 40
. J J.I.I 40 60 80 100 120 140 i 160 '180 200
160B
^ 90‘ - 2 H o 80‘ -H ___ J -1 2 7 - g 7 0 .
• a 6 0 .
E 5 0 - < 4 0 . 0 > 3 0 .
200 m /e Figure 4. Mass spectra of the materials in (A) GC peak #3 (2-octylnitrite) and (B) GC peak # 4 (1-octylnitrite) from the reaction of n-octylamine and sodium nitrite reaction. Schematic representation of their fragmentation patterns are shown in upper inserts. Highly magnified parent peaks and their respective exact molecular masses are shown as lower inserts. 41 pH 3.5 and room temperature, the relative proportion of
1 -chlorooctane increased significantly with the increase in Cl” concentration (Figure 5). This increase in
1 -chlorooctane concentration was accompanied by a decrease in all the other products, most notably 1 -octylnitrite
(Figure 5B). In 5M NaCl, the highest Cl” concentration tested, 1 -chlorooctane became the dominant product, much higher that the combined proportions of 1 -octanol and
1-octylnitrite (labelled as "1-octanol" 3 in Figure 5B).
A small amount of 2-chlorooctane was also formed but, because it co-eluted with 1 -octylnitrite during gas chromatography, it was not quantitatively determined.
Consequently, the relative amount of 1-octylnitrite in
Figure 5A may actually include a small amount of 2-chloro octane. In one sample (in 5M NaCl) chromatographed under a different GC system, where the 2-chlorooctane peak was resolved from that of 1 -octylnitrite, the peak area for
1 -chlorooctane was calculated to be at least 1 0 times more than 2 -chlorooctane.
3In this thesis, "1 -octanol" refers to the combined amounts of 1 -octanol and 1 -octylnitrite, which was probably solely formed from 1-octanol. Similarly, "2-octanol" refers to the combined amounts of 2 -octanol and 2 -octylnitrite. 42
60 o— o 1—octene □— □ 1 —octylnitrite •— • 2-octene ■— ■ 2—octylnitrite ^ ♦ 50 ‘ a — a 1-o cta n o l x" •k 0) a — a 2 -o cta n o l c 40 1-chlorooctane 0 T5 30 <1
a <0 * 10
0 0 1 2 3 4 5
60 a. o— o l — octene a — a "1—octanol" — • 2—octene a — a "2-octanol" ♦ * 50- ♦ 1 —chlorooctane a) c 40- 0 x> 1 30- <
o> * 10-
0 t 1------1------1------1------1------1------1------1------1------r 0 1 2 3 4 5 [NaCl], Molar
Figure 5. Product distribution of the reaction between n-octylamine and sodium nitrite in the presence of NaCl. (pH 3.5, 100 mM n-octylamine, 500 mM sodium nitrite, 22°C, 30 minutes). (A) the products are plotted individually and (B) combined as "1-octanol" and as "2-octanol". 43
a) Effect of pH
The optimum pH of the n-octylamine-nitrous acid reac tion was around pH 3.5 (Figure 6 B), where under the standard conditions employed, about half of the reactant
(n-octylamine) was converted into products. The extent of reaction decreased at either side of the pH range and became insignificant above pH 6 . The major product at any pH was 1-octanol but a significant portion of it was further converted into 1 -octylnitrite especially at lower pH (Figure 6 B). A similar fraction of the 2-octanol was likewise converted partially to 2-octylnitrite. At pH 3.5, for example, it was independently observed that both
1 -octanol and 2 -octanol were readily converted into their corresponding nitrites upon incubation with sodium nitrite
(Figure 7). So, on the assumption that all of the nitrites were derived from their alcohol precursors, the amounts of
1 -octanol and 1 -octylnitrite were added together and plotted as "1-octanol" (Figure 6 B) to represent the total amount of 1 -octanol produced in the deamination reaction.
2 -0 ctanol and 2 -octylnitrite were also added together and plotted as "2 -octanol".
"1 -Octanol" predominated over the rest of the products by about 60% and the ratios of the various products did not vary significantly with pH (Table 2), suggesting that the reaction pathways were probably not affected by either 44
20 □— o 1—octylnitrite ■— ■ 2—octylnitrite 15 o— o 1—octene v •— • 2—octene u c a— a 1—octanol o ana — —a 2 —octanol X) c 10 3 XI A < 0) > Ip 5 o on0)
1.5 2.5 3.5 4.5 5.5 6.5 30 60 o 1—octene 25- • 2 —octene 50 a "1 —octanol" ecn Reaction Percent a "2—octanol" 8 20 - -40 c o 15 ♦ — ♦ % Rxn. -30 3 s 10 -20 £ _o 5 -I -10 onV 0 - 0
"—•I"— 1.5 2.5 3.5 4.5 5.5 6.5 pH
Figure 6 . Effect of pH on the reaction of n-octylamine with nitrous acid. (A) Individual products plotted separa tely and (B) combined as "1-octanol" and as "2-octanol". Values from pH 3 to 6 were averages of 4 independent experiments and those at pH 2.0 and 2.5 were averages of 2 experiments.(100 mM n-octylamine and 500 mM sodium nitrite in 0.2M succinate buffer at 22°C.) Percent nitrite 60 0 4 20 - alcohols into nitrites at conditions similar to those in those to similar conditions at nitrites into alcohols h -cyaiesdu irt rato (H35 22°C, 3.5, (pH reaction nitrite n-octylamine-sodium the 100 mM 1-octanol or 2-octanol with 500 mM nitrous acid). nitrous mM 500 with 2-octanol or 1-octanol mM 100 0
iue . iecus o te ovrin f the of conversion the of Time-course 7. Figure 1
ie Hours Time, 2
2-octylnitrite • — 1-octylnitrite • o o— 3
4
5 46
Table 2. Product ratios of the reactions of n-octyl amine with sodium nitrite at different pH values, as plotted in Figure 6 B. The ratios at pH 5.0, 5.5 and 6.0 were not included because the product concentrations were too low to give accurate values.
pH Product Ratio
2 . 0 2.5 3.0 3.5 4.0 4.5
"1 -Octanol" 3.6 4.0 4.0 3.7 4.1 4.8 1 -Octene
"1 -Octanol" 3.8 3.8 3.5 3.5 3.9 4.0 "2 -0 ctanol"
"1 -Octanol" 12.4 15.0 13.0 9.5 10.4 9.0 2-Octene
"2-Octanol" 1 . 0 1 . 1 1 . 1 1 . 0 1 . 0 1 . 2 1 -Octene
"2-Octanol" 3.3 4.0 3.7 2.7 2.7 2.3 2-Octene
H+ ion or OH” concentrations. However, the 1-octylnitrite/
1 -octanol and 2 -octylnitrite/2 -octanol ratios decreased with pH (Figure 8 ) suggesting that the conversion of the alcohols to corresponding nitrite analogs was pH dependent.
It was not known why the log of the 1-octylnitrite/l-octa- nol ratio deviated from lineartity at low pH while that of log 2 -octylnitrite/2 -octanol ratios did not. Log nitrite/alcohol - - 0 1 0.50 1 1 2-octylnitrite/2-octanol ratios with pH. (Values calculated (Values pH. with Figure ratios from 2-octylnitrite/2-octanol . . . . 00 0 5 0 5 0 0 - - Figure Figure - - - 8 6 Cags nte -cyntie1otnl and 1-octylnitrite/1-octanol the in . Changes . ) A a — a 1 —octylnitrite/1 —octanol 1 —octylnitrite/1 a — a a — a
2—octylnitrite/2—octanol pH 6
-O 48
b) Effect of the length of reaction
Knowing pH 3.5 as the optimum pH, the reaction was run at various lengths of time in identical conditions. After
5 hours of reaction only about 10% of the starting n-octyl amine remained unreacted (Figure YB). Most of the n-octyl amine was converted into products during the first hour.
Beyond the 15-minute period, the most abundant product was
1-octylnitrite. There seemed to be an initial surge in the amounts of 1 -octanol and 2 -octanol but were eventually superseded by their conversion products, 1 - and 2 -octyl nitrite, respectively. However, this conversion of the alcohols into nitrites was greatly reduced when ether was layered on top of the reaction mixture in the course of the reaction (Figure 9Z and 10). Presumably, this was due to the extraction of the alcohols into the ether layer once they were formed, making them less available for further reaction with excess nitrous acid.
c) Effect of temperature
The extent of reaction significantly increased with temperature and was expectedly greater at pH 3.5 than at pH 4.5 (Figure 11YB and 11ZB). At both pH values,
"1-octanol" was higher than any of the other products. It should be pointed out here, however, that there was no significant reaction at 0°C, even when the reaction time was increased to 24 hours (data not shown). x 2 e Relative A bundance, with ether and Z) ether was layered on top of the of top on layered was ether Z) and and "1-octanol" ether as with combined (B) and separately plotted n-octylamine and nitrous acid. (A) Individual products Individual (A) acid. nitrous and n-octylamine s 2otnl. Y Te ecinmxuewsnt layered not was mixture reaction The (Y) "2-octanol". as reaction mixture during the whole course of the reaction. the of course whole the during mixture reaction 10m -cyaie 50m oimntie p 35 22°C.) 3.5, pH nitrite, sodium mM 500 n-octylamine, mM (100 1 iue9 Efc f h egh frato between reaction of length the of Effect 9. Figure
egh fRecin Hours eaction, R of Length 2
3
• — ♦ P ercen t Rxn t ercen P ♦ — • a a a — a 1—octylnitrite 1—octylnitrite a — a —
2—octylnitrite 2 ■ — o — o 1—o ctan e e 1—o ctan o — o —octene 2 • — • a a a a — 1 a — " 2 -o c ta n o l'' l'' o n ta c -o 2 " — 1 " 4 a 1oct e n te c 1—o o oct e n te —o c 2 • oct l o n ta c -o 2 octanol ctanol o 0 2 10 0 3 0 2 40 10 ------0 — ene n te c 1—o o — o • a — a 1—octylnitrite —octylnitrite 2 ■ 1—octylnitrite ■— a — a ----- oct e n te —o c 2 • egho ecin Hours Reaction, of Length 2 a — a 1 -o c ta n o l l o n a ta c -o 1 a a — a 4 3 a a o — o 1—o ctane 1—o ctane o — o oct e n te c -o 2 e — e — — —octanol. 2 a a “ 1—octanol" 1—octanol" “ 2- anol" o n ta c -o "2
P ercen t Rxn Rxn t ercen P
S
-60 •40 ■20 Pret Reaction Percent . Nitrite/alcohol ratio 2 3 4 5 0 1 orsodn irts (Values9.)from calculated Figure nitrites. corresponding reaction mixture on the conversion of alcohols intoalcohols theconversion of on reaction mixture A & & A A —0 0— ' • • 0 i • i i | i" | | i i i • • • i •
Figure 10. Effect of an ether layer topoftheetheron an 10. of Effect Figure A -cyntielotnl (layered ether) with 1-octylnitrite/l-octanol 1 2 2 -octylnitrite/ -octylnitrite/ -octylnitrite/ 1
Length of reaction, HoursLengthreaction, of 1 2 2 2 otnl (layered ether)-octanol with -octanol -octanol
3
4
5
Relative A bundance, % Relative A bundance, 40 30 20 10 SO 40 0 ■— ■ 2 —octylnitrite —octylnitrite 2 ■ 1—octylnitrite a ■— — o a— a a l o 1—octan a a— — a oct , e n te c —o 2 • e n te c 1—o — • o — o oct e n te —oc 2 • — • a — a a e n te c ‘1—o -----o o ♦ cent , t n e rc e P ♦ — ♦ 0
lte sprtl ad B cmie a "-cao" n as and "1-octanol" as n-octylamine, combined (B) mM (100 and separately nitrite. plotted sodium and n-octylamine "2-octanol".(Y) reaction at pH 3.5 (Z) reaction at pH 4.5. pH at reaction (Z) 3.5 pH at reaction products Individual "2-octanol".(Y) minutes).(A) 30 nitrite, sodium mM 500 a 2 —octan o l l o —octan 2
1— ctanol' "1 o 2— "2 Rxn anol1 o n ta c o Figure 11. Effect of temperature on the reaction of reaction the on temperature of Effect 11. Figure T em p eratu re, *C re, eratu p em T Y 40 50 100 30 20 10 40 20 ■ ■ ■ o — o 1—o ctene ctene 1—o • o — o a a — a 2—octanol 2—octanol a — a 0 1—octylnitrite ■ a — o ------— a "2 —octanol" —octanol" "2 a — — —octene 2 e n te • c — -o 1 —o - ♦ P ercen t Rxn t ercen P ♦ - a • 2—o cten e e cten 2—o • a 2ocyl itrite ln cty 2-o ■
1- anol l" o n ta c -o "1 1—octanol 10 T em perature, *C perature, em T 20 ’ z j ' ' . 050 30 40
80 40 60 20 Pret eaction R Percent ' to 52
At pH 3.5, 1-octylnitrite was the major product. The proportion of 1 -octanol converted into 1 -octylnitrite and
2 -octanol converted into 2 -octylnitrite slightly increased with temperature (Figure 11YA). At pH 4.5, 1-octanol was the major product, as also earlier observed in the pH-dependence study. However, the proportions of 1-octanol and 2 -octanol converted into their nitrite analogs did not significantly increase with temperature (Figure 11ZA), in contrast to that at pH 3.5. There was also a slight but consistent increase in ratios between the rearrangement products (ie, 2 -octanol and 2 -octene) over the primary deamination products (ie, 1 -octanol and 1 -octene) as the reaction temperature was increased.
d) Effect of nitrite concentration
As expected, the extent of reaction increased with the concentration of NaN0 2 (Figure 12B). However, this also led to an increase in the proportions of 1 -octanol and 2 -octanol further converted by excess nitrite into
1-octylnitrite and 2-octylnitrite, respectively (Figure
12A). Smaller amounts of octylnitrites were formed at equimolar ratios of n-octylamine and NaN02 but the extent of reaction was almost insignificant. At 1M NaN02, the highest concentration tested, almost all of the n-octyl- amine was converted into products. 53
40 • n— a 1 —octylnitrite ■— ■ 2—octylnitrite ■ a — a 1 —octanol 2q . a — a 2 -o cta n o l o— o 1—octene . «— * 2—octene
T------1------1------1------1------1------1------1------1------1— 0.2 0.4 0.6 0.8 1.0
1 0 0 60 o— o 1-octene •— • 2-octene 80 a— a "1—octanol Reaction Percent a— a "1—octanol 40 60 ♦ — ♦ Percent Rxn 40 .1 20 20
0 0.2 0.4 0.6 0.8 1.0 [N aN 02], Molar
Figure 12. Effect of sodium nitrite concentration on the n-ocylamine-sodium nitrite reaction. (100 mM n-octyl amine, pH 3.5, 22°C, 30 minutes.) (A) Individual products plotted separately and (B) combined as "1-octanol" and as "2 -octanol” . 54
With increasing nitrite concentration, there was also a significant increase in the combined amounts of 1 -octa nol and 1 -octylnitrite (or "1 -octanol") relative to the other products, which increased only minimally (Figure
12B). The apparent discrepancy in the ratios of 1-octyl- nitrite/l-octanol and 2 -octylnitrite/2 -octanol at 1 M NaNC>2
(Figure 13) was probably due to the much lower concentra tion of the 2 -octanol relative to 1 -octanol at the same nitrite concentration, that is, the rate of reaction is a function of alcohol concentration at a given concentration of nitrite. Another possibility is steric hindrance on
2 -octanol against the attack of I^O-NO4".
Reaction of n-octylamine with sodium nitroprusside a) Effect of pH
The reaction of n-octylamine with sodium nitroprusside was generally slower than that with nitrous acid and has a substancially different pH dependence. While there is little reaction with either at pH 6 the reaction with the latter increased rapidly with pH (Figure 14) and that with the former increased at lower pH values (Figure 6 ). Appa rently, the extent of reaction was greatest at pH 10, but still about 30% of the n-octylamine remained unreacted.
The major product in all cases was 1-octanol, followed by
2-octanol, 1-octene and 2-octene. The product ratios were Nitrite/Alcohol Ratio 4 6 2 0
0
different nitrous acid concentrations. (Data taken from taken (Data concentrations. acid nitrous different iue 12A.) Figure T iue 3 Ntieachl aiso tepout at products the of ratios Nitrite/alcohol 13. Figure a a — — 0.2 T a a
1 —Octylnitrite/1 -Octanol 2-0ctylnitrite/2-0ctanol 0.4 [NaNO*]. Molar
0.6
0.8 T
1.0 A -
Relative Abundance, 5 3 0 3 40 5 2 20 15 10 0 5
t ------
- ♦ — A — A mM sodivun nitroprusside, 22°C, dark, 24 hours in screw- in hours 24 dark, 22°C, sodivun mM nitroprusside, mn ihsdu irpusd. 5 Mnotlmn, 500 n-octylamine, (50mM nitroprusside. sodium with amine apdtbs) aus rmp t 1 ee vrgs f 3 of single a averages from were 6.5 and 6.0 were 10 5.0, experiment. to 7pH at pH from those and Values trials tubes.) capped —a 2—octanol 2—octanol —a — 2—octene —• 1—octene —o ♦ ecn Rxn Percent ♦ - T 6 iue 4 Efc fp nterato f n-octyl- of reaction the on pH of Effect 14. Figure a 1—octanol 1—octanol “T" 7 pH T 8 —r 9 I— — 10
-.20 - - -
80 40
60 0 ecn Reaction Percent U cn 1 57
similar to those with nitrous acid and did not signifi
cantly change with pH (Table 3). It should be noted at
this point was that no octylnitrites were formed from the
reaction of n-octylamine with sodium nitroprusside, in
contrast to what was observed with nitrous acid.
b) Effect of temperature
At the three pH values tested, the extent of reaction
increased significantly with temperature (Figure 15). As
in the case of nitrous acid, the amounts of rearrangement
Table 3. Product ratios from the reaction of n-octyl amine with sodium nitroprusside, calculated from pH 7 to 10 only. Below pH 7, the product concentrations were too low to obtain accurate values.
pH Product Ratio 7.0 7.5 8 . 0 8.5 9.0 9.5 1 0 . 0
1-Octanol 4.8 4.2 4.1 4.0 4.0 4.2 4.5 1 -Octene
1 -Octanol 3.3 3.7 3.4 3.5 3.3 3.3 3.3 2-Octanol
1-Octanol 12.9 12.7 10.8 10.1 11.2 11.3 1 1 . 1 2-Octene
2-Octanol 1.5 1 . 1 1 . 2 1 . 2 1 . 2 1.3 1.4 1-Octene
2-Octanol 4.0 3.4 3.2 2.9 3.4 3.4 3.3 2-Octene 58
products (2 -octanol and 2 -octene) also slightly increased
with temperature. Unlike that of nitrous acid, the reac
tion of n-octylamine with sodium nitroprusside occured
readily even at 0°C, which is important in reactions with
heat-labile proteins.
Reaction of Sodium Nitroprusside with Na-CBZ-L-Lysine
After 24 hours of reaction with sodium nitroprusside,
about 40% of the Na-CBZ-L-lysine was converted into
products. When the reaction mixture was acidified to pH 2
and extracted with ether, the ether extract (Figure 16C)
did not contain any of the residual Na-CBZ-L-lysine, indi
cated as peak #1 in Figure 16A. All of the unreacted
Na-CBZ-L-lysine remained in the aqueous phase (Figure 16B).
The first step in identifying the products of the * deamination reactions was to remove the CBZ'-groups of the
different derivatives by passing them through a Pd2-H
column. In addition, the Pd2-H column also reduces any
isolated double bonds present, as represented in the
following equation:
O O Pd-H vv ii / i \ CH*-0-C-NH-CH-C-0H + 2 HCOOH ------► f / Y T 0 * 3 + 3 C 0 2 + H3N -C H -C -O H I 1 j
9H2 CHII c h 2 59
30 60 X Sc 20 40 » ■oo 3c 3 o 20 £ a o te.
0 10 20 30 40 50 50 100
■ 80
i 30 60 s* S 20 40 a
20 §
0 10 20 30 40 50
50 100
80
§ 30 60
§ 20 40 a es 2 0 g
0 10 20 30 40 50
Temperature, C
Figure 15. Effect of temperature on the distribution of the products of the reaction of n-octylamine with sodium nitroprusside reaction as a function of temperature. (50 mM n-octylamine, 500 mM sodium nitroprusside, 0.2M triethanol amine, 36 hours, dark, sealed tubes.) A) pH 7.5, B) pH 8.5 and C) pH 9.5. A A 1-Octanol, A --- A 2 -0 ctanol, o — o 1-Octene, • - • 2-Octene, + + Extent of Reaction. Absorbance, 2 54nm.' 0.75 1.00 0.50 0.25 R51 clm, hrai ie hmcl) Bfe was A Buffer Chemicals). Fine Pharmacia column, 5/10 HR xrcino teaiiid ecin ouino sodium of of extract solution ether (C) and reaction Na-CBZ-L-Lysine and nitroprusside acidified the of extraction h aiiidrato ihsdu irpusd. (Pro-RPC nitroprusside. sodium with reaction acidified the CH H 0.05% A N-B--yie B te qeu pae fe ether after phase aqueous the (B) (A) Na-CBZ-L-lysine 3 CN, 1:3 mixture. 1:3 CN, Figure 16. Reverse-phase FPLC elution profiles of profiles elution FPLC Reverse-phase 16. Figure 3 P0 4 n 0.01M in 90%B
0 SNP eeto tm, minutes time, Retention KClO 5 4' H25 Bfe Bws + A was B Buffer 2.5. pH 10 90%B .. L 15 0 5 10 90%B I -- 1
15 61
After the materials in the peaks in Figure 16 were
separately eluted from the Pd2 H column, an aliquot of each
was subjected to mass spectrometry and amino acid analysis.
Mass spectra were obtained by direct injections into the
mass spectrometer since these underivatized amino acids
have very low volatility and therefore could not be passed
through the GC column (MacKenzie, 1984). The mass spectra
(Figure 17A) of reduced peak #2 (from Figure 16) showed a
strong peak at m/e 148, which was assumed to be the proto-
nated parent peak (M + 1) since very little fragmentation
is expected under chemical ionization (Leclercq and Desi-
derio, 1973; Meot-Ner and Field, 1973). Besides, the M+29 peak was also visible, characteristic of (M + C 2 Hg)+
formed when methane gas is used in the ionization (Silver-
stein et al, 1981). The prominent M-18 peak at m/e 130 corresponds to the loss of H 2 0 , suggesting that the sample was an alcohol derivative. This was supported by its elution in the amino acid analyzer (Figure 18A) in the position characteristic for 6 -hydroxynorleucine (Kurosky and Hofmann,1972; Knowles et al., 1974; Davis et al..1972).
When the materials in peaks #4 and #5 (Figure 16C) were reduced with Pd 2 H and subjected to amino acid analyses the Pd 2 H-reduced products both eluted in the position of authentic norleucine (Figure 18B). Their identification as norleucine was also supported by their mass spectra (Figure 17B) wherein both products showed (M+1) 1H* It .i
0 NHgrCH-C-O.H
cri2 © o ^ h 2 c OB c h 2 n 3C 'CH2-O H •U A < I? .1 © > «♦«» W © a:
M|.l I7« . t %•', i i I'M .Ju * [ . ' ? ? , ft ? -.-Wr
(M +1)
O NH. -CH-C-OH 85 CH 2 © O c c h 2 W *D B c h 2 3C I •Q CHg < © > •C © DC
lift. I Hf.l 393.1 T , 'V JUZJL — j - 11 ■ i JZSri inti 359 m/e
Figure 17. Mass spectra of PcUH-reduced materials from (A) peak #2 and (B) peak #4 or #5 from Figure 16. This was done by direct injection of the sample into a Finnigan Model 4021 Quadrupole mass spectrometer under chemical ionization in methane gas. Other GC-MS conditions were described under "Methods". 63
Figure 18. Elution profiles of the following samples on a Beckman amino acid analyzer model 119 CL with a 6 x 20 mm standard column.
(A) the major alcohol product of SNP and Na-CBZ-L-Lysine (peak #2, Figure 16C) after Pd 2 H reduction.
(B) the second most abundant product (peak #4, Figure 16C) after Pd 2 H reduction.
(C) the major product of the reaction of SNP and Na-CBZ-L- Lysine (peak #2, Figure 16C) after standard HC1 hydrolysis.
(D) the second most abundant product (peak #4, Figure 16C) after HC1 hydrolysis.
(E) HCl-hydrolyzed Na-CBZ-L-lysine and
(F) amino acid standards.
The peaks in each chromatogram were later identified as (a) 5-hydroxynorleucine (b) 6 -hydroxynorleucine (c) nor leucine (d) 5-chloronorleucine and (e) 6 -chloronorleucine. 0.25 AU
E c o N- u?
0 o c cO XI 1_ Xa
i i i ' » '■ 25 50 7 5 100 Retention Time, min. 65 strong protonated parent peaks (M+1) at m/e 132 under chemical ionization. On the basis of these data it appears that norleucine was obtained by reduction of the terminal and of the penultimate olefins. However, since there was no direct and convenient method of removing the CBZ-groups without potentially affecting the double bonds, the iden tities of these products were deduced by comparison with the GC elution profiles and relative product distributions in the deamination of n-octylamine with nitroprusside. On this basis, peaks # 4 and #5 (Figure 16C) were assumed to be CBZ-2-amino-5-hexenoic acid (primary olefin) and CBZ-2- amino-4(t)-hexenoic acid (secondary olefin), respectively.
The former was later supported by the formation of mostly
5-chloronorleucine upon incubation with 6 N HCl at 105°C for 22 hours. Under such conditions, the latter would have produced both 5-chloro- and 4-chloronorleucine.
Peak #3 (Figure 16C) had not been isolated in suffi cient purity and quantity so no detailed analyses were made. However, by comparison again with the GC elution profiles and relative product distribution of the deamina tion products of n-octylamine with nitroprusside, it was assumed that peak #3 (Figure 16C) was CBZ-5-hydroxynor- leucine. 66
Reactions Occurinq Under the Conditions Similar to Those of the HCl Hydrolysis of Proteins
Amino acid analysis of the material resulting from incubation of the component giving rise to peak # 2 of
Figure 16C (i.e., 6 -hydroxynorleucine) with hot HCl (i.e.,
5.8M HCl, 22 hours at 105°C, as normally employed for protein hydrolysis) showed three major new peaks, with retention times of 22, 28 and 54 minutes (Figure 18C).
When the same products of the hot HCl treatment were derivatized to the Na-trifluoroacetylamino acid methyl esters and subjected to GC-mass spectrometry, three major peaks were also observed. The second biggest peak had mass spectra deduced to be those of Na-trifluoroacetyl-6 - hydroxynorleucine methyl ester. This therefore confirmed that the material eluting at 28 minutes (peak b, the second largest peak, Figure 18C) corresponded to 6 -hydroxynorleu cine.
The Cl mass spectra (Figure 19) of the material in the largest GC-MS peak showed an M+1 parent peak at m/e =276 and an M/M+2 ratio of 3, characteristics of compounds containing one chlorine atom (Silverstein et al, 1981).
A prominent peak at (M - CH2 =C1)+ suggests that chlorine was at the terminal position (Budzikiewicz et al, 1967). 67
o o M+1 II iee.e-i CF3 -C-NH-CH-C-0 -CH3 275.8 CHo + I 2 M-CH2=CI CHo 225.9
M-rtCI 239.9 M4.8
7 1.1 85 . 1 97.1 112-1 1.1 I lil. Jit 129-8 JSUm M A k 188 158 258 388 m/e
Figure 19. Chemical ionization mass spectra of the Na-trifluroacetylamino acid methyl ester of the most abundant material that elutes at 54 minutes in the amino acid analyzer. 68
On this basis, it was concluded that the most abundant product eluting from the amino acid analyzer at 54 minutes was 6 -chloronorleucine. 6 -Hydroxynorleucine is known to be readily converted to 6 -chloronorleucine during acid hydro lysis (Lent et al., 1969; Malin,et al., 1984). The elution position of this chlorine derivative in the amino acid analyzer has been reported previously and has been called the "pre-tyrosine peak" (Kurosky and Hofmann, 1972).
The peak at 22 minutes (peak a, Figure 18C) was exactly in the position of glutamate and therefore was not observed earlier in the HCl hydrolyzates of nitrous acid- treated proteins (Knowles et al., 1974; Kurosky and
Hoffman, 1972). The presence of this peak was first reported by Malin et al. (1984) but was not identified.
Therefore, further tests were conducted here to establish its identity.
A much larger proportion of the peak eluting at 22 minutes (peak a, Figure 18D) was obtained by hot HCl treatment (5.8M, 22 hours at 105°C) of the terminal olefin product of the reaction of SNP and Na-CBZ-L-Lysine (i.e.,
2-amino-5-hexenoic acid, peak #4, Figure 16C). So this hot
HCl-treated product was used to investigate the identity of the peak eluting at 2 2 minutes in the amino acid analyzer.
When the hot HCl-treated product was similarly treated with
4M HCl in methanol followed by trifluoroacetic anhydride to produce the corresponding Na-trifluoroacetylamino acid 69
methyl ester and then subjected to GC-mass spectrometry,
the fragmentation data (Figure 20) was very different from
those of the other derivatives. Mass spectra under
chemical ionization in methane (Figure 20A) showed a peak
for the protonated molecular ion (M+1) at m/e = 226, without significant fragmentation. Under electron impact
(Figure 20B) a very faint M + 1 ion was observed and a very prominent peak at m/e = 139, corresponding to the combined loss of CF 3 (Manhas et al.. 1970; MacKenzie,
1984) and CH 4 (Budzikiewicz, 1967; Porter, 1985). There was also a total absence of the peak at M-60, correspon ding to the combined loss of CC>2 and CH^ (Leimer et al.,
1977) and prominently observed in all other derivatives.
The peak at m/e=139 was also characteristic of this compound and was not present in all the other derivatives.
This suggests that there was no methyl ester formed during derivatization and therefore the parent compound did not have a free carboxyl group. Its reaction with ninhydrin during amino acid analysis and trifluoroacetic anhydride during derivatization for mass spectrometry indicated the presence of free amino group. These observations, taken together, suggest that the product that gave rise to the peak at 2 2 minutes in the amino acid analyzer was the lactone, 2 -amino-6 -caprolactone, which was presumably formed under the conditions of hot HCl treatment. Under similar conditions 4-hexenoic acid, for example, was (M+1) 70 226. N H - C - C F 3 0 o c
CM,
+-> iS a) cc 159
139.
a CD DC (M-60)
57 .«
159 m / e
Figure 20. Mass spectra of the Na-trifuoroacetylamino acid methyl ester of the material in the peak eluting at 22 minutes in the amino acid analyzer. (A) Chemical ionization in methane gas and (B) electron impact at 70 eV. 71
earlier reported to produce 6 -caprolactone as one of the major products (Linstead and Rydon, 1934). However, the much lower acidity during amino acid analysis was presumed to have induced the opening of the lactone ring, resulting in the elution of its open form, 5-hydroxynorleucine, from the amino acid analyzer at 22 minutes. This retention volume was much earlier than the expected elution position of 2-amino-6-caprolactone4. For example, homologous lactones like homocysteine thiolactone (Table 4) elutes at about 90 minutes (Sober,1968) and hydroxynorvaline lactone after 70 minutes (Jennings and Anderson, 1987). However, due to the absence of authentic 2 -amino-6 -caprolactone and
5-hydroxynorleucine standards, no direct comparisons were made.
To further prove the presence of a lactone structure, a drop of hydrazine was added into the dried HCl-treated residue, diluted with the amino acid analysis starting buffer and subjected to amino acid analysis. The new amino acid chromatogram showed that the peak at 2 2 minutes was almost totally gone, and a new peak appeared at around 82 minutes (Figure 21). Since hydrazine (Tsunokawa, 1975) or related compounds such as ammonia (Linstead, 1932) and hydroxylamine (Bruice and Bruno, 1961) are capable of
4The compound 6 -caprolactone appears in the literature in other names: 5-methyl-valerolactone, and 6 -n-hexolactone. 72
Table 4. Elution positions of homologous lactones from the amino acid analyzer.
Structure Name Rt, min.
Hydroxynorvaline lactone 70
J4H3
Homocysteine thiolactone 90
2-Amino-6-caprolactone Absorbance, 570 nm 1.00 50- 0 .5 0 0.75- 25- 5 .2 0 0 ih2aio5hxni ai. opr hswt iue 8 w 18D Figure with this Compare acid. 2-amino-5-hexenoic with ramn f h re eiu fo h ecino o C ^ HCl hot of reaction the from residue dried the of treatment for the same sample but without hydrazine treatment. hydrazine without but sample same the for iue 1 Aio cd hoaorm fe hydrazine after chromatogram acid Amino 21. Figure -O 25 U eeto time.min Retention s: 50 5-Hydroxynorleucine Hydrazide of of Hydrazide 75
100 74 opening lactone rings, this new peak was believed to be + + CH3 -CHOH-CH2 ~CH2 -CH(NH3 )-CO-NH-NH3 , the hydrazide deri vative of 5-hydroxynorleucine (Equation 2).
(2 )
As a dibasic amino acid derivative, this hyrazide derivative of the lactone would be expected to elute along with the basic amino acids, i.e., after 75 to 80 minutes.
This observation therefore supports the contention that the material eluting in the amino acid analyzer at 2 2 minutes was indeed 5-hydroxynorleucine derived from its lactone.
The treatment of the major olefin derivative (i.e.,
2-amino-5-hexenoic acid, peak #4, Figure 16C) with hot HCl
(5.8M, 22 hours at 105°C) not only showed the three peaks in significantly different proportions but also showed an additional peak at 53 minutes (Figure 18D), immediately preceeding that of the now-smaller 6 -chloronorleucine peak at 54 minutes. N-trifluoroacetylamino acid methyl ester derivatization and subsequent GC-mass spectrometry (Figure
2 2 ) indicate the presence of one chlorine (M/M+ 2 ratio was
3) and a m/e of 276 for the protonated molecular ion (M+1). (M + 1) 2 7 5 .8 O O 75 II II CF 3 -C-NH-CH-C-O-CH 3
C H o I * JP c h 2 0) o c a co ?H-CI •u c h 3 c 3 S3 < <0 > i3 £
71.1 85.1 iHii-
( M - 6 0 )
m CD O C ' <0 ~o c in x>5 *• < a> > ■H CO a> a: 136 1*3
156 m/e Figure 22. Mass spectra of the N-trifluoroacetylamino acid methyl ester of the material giving rise to the peak at 53 minutes in the amino acid analyzer: (A) Chemical ionization spectra in methane gas and (B) electron impact at 70 eV. 76
The M-60 peak (Figure 22B) indicates the combined loss of
CO 2 and CH4 (Leimer et al., 1977). The presence of a
prominent El peak at m/e 180, corresponding to the loss of
HCl (Budzikiewicz et al.. 1967; Silverstein et al., 1981),
suggests that the product is a secondary chlorine deriva
tive, 5-chloronorleucine.
Time-course for the Reaction of Hot HCl with 2-amino-5- hexenoic acid
When the major olefin product (i.e., CBZ-2-amino-5- hexenoic acid, peak #4, Figure 16C) was incubated with
5.8 M HCl at 105°C for different lengths of time and then subjected to amino acid analysis, 2 -amino-6 -caprolactone
(eluting in the amino acid analyzer as 5-hydroxynorleucine) was observed to increase during the first ten hours and levelled off after about 24 hours (Figure 23). This was accompanied by a decrease in 5-chloronorleucine, which was formed very early in the reaction. These observations suggest that addition of HCl to 2-amino-5-hexenoic acid to give 5-chloronorleucine is very rapid and most of the
2 -amino-6 -caprolactone was derived from it by a further slow intramolecular reaction, shown in equation 3. Under conditions existing during amino acid analysis, this lactone ring is presumably opened to give 5-hydroxynor- leucine. Some of the 5-hydroxynorleucine, however, may also be formed by direct solvolysis of 5-chloronorleucine, Relative peaks areas 200 150- 100 50- - based on peak areas on the amino acid chromatogram. acid amino the on areas peak on based -eeoc cd n .M C a 15C Cluain were Calculations 105°C. at HCl 5.8M in acid 5-hexenoic 0 iue 3 Tm oreo teratoso 2-amino- of reactions the of course Time 23. Figure 20 egh fhdoyi, Hours hydrolysis, of Length 40 6—Chloronorleucine 5-Chloronorleucine 5-Hydroxynorleucine 6—Hydroxynorleucine 080 60 0 0 1
78 as shown in equation 4.
(3)
^ CH3 >''h I"3 5-Hydroxynorleuclne
2-Amlno-5-hoxenolc a c id NH3
(4)
On the other hand, there was a slight but constant increase in 6 -chloronorleucine which was accompanied by a decrease in 6 -hydroxynoleucine (Figure 23). Under the conditions of HCl hydrolysis of proteins the conversion of
6 -hydroxynorleucine to 6 -chloronorleucine has been repor ted to occur quite readily (Lent et al., 1969; Malin et al., 1984). Indeed, when the major alcohol derivative of the reaction of SNP and Na-CBZ-L-lysine (i.e.,CBZ-6 -hydro- xynorleucine) was also incubated in 5.8M HCl at 105°C for various lengths of time, 6 -hydroxynorleucine was almost exclusively converted into 6 -chloronorleucine at a fast rate (Figure 24). Relative peak areas 1000 - 0 0 8 - 0 0 6 - 0 0 4 200 Figure 24. Tirae-course of the reactions of reactions the of Tirae-course 24. Figure - - 0 in the amino acid chromatogram. acid amino the in .MHla 0°. acltoswr ae npa areas peak on based were Calculations 105°C. at HCl 5.8M 20 Length of hydrolysis, Hours. hydrolysis, of Length 0 4 o 6 — Hydroxynorieucine Chloronorleucine — 6 □ Hydroxynorieucine — 6 o • 5 — Hydroxynorleucine Hydroxynorleucine — 5 • 0 8 0 6 6 -hydroxynorleucine in -hydroxynorleucine 0 0 1
VO 80
Reaction of Sodium Nitroprusside with RNAse A.
It was observed that the RNAse A from the bottle (Type
II-A, Sigma R5000) was not chromatographically homogeneous
(Figure 25A) so all RNAse A samples used in these experi
ments were further purified by FPLC using cation-exchange
column (Mono S, Pharmacia Fine Chemicals). Reinjection of
the purified RNAse A under the same chromatographic system
showed a single sharp peak (Figure 25B).
In order to determine if sodium nitroprusside reacts
with the lysine residues of proteins, ribonuclease A was
incubated with a 1 0 0 -fold molar excess of sodium nitro
prusside under the same conditions earlier used with the
model amine, Na-CBZ-L-Lysine. After 10 hours of incubation
in the dark under these conditions the mixture of diffe
rently-modified RNAse A in solution retained about 1/3 of
its original RNAse A activity. The typical chromatogram
of the reaction products is shown in Figure 26.
About 20 percent of the RNAse A remained unmodified
(Figure 26A, peak # 5) and this fraction was found to be
fully active (Table 5). Fraction #1 contained protein
and non-protein components which were not resolved even by
rechromatography under different conditions. Fraction #4
was a pool of several peaks that were also not effectively
separable by rechromatography under other conditions. When
fraction #2 was reinjected into the FPLC under identical 81
1.00 ftNAse A RNAse A
0.75 B
40%B 40%B CM © 0.50 /—I / ■ 1
•O
0.25
0 J ______L. 0 5 10 15 10 15 Retention volume,ml
Figure 25. Cation-exchange chromatogram of (A) commer cial RNAse A (Type II-A, Sigma R5000) and (B) after purifi cation. (Mono S column, Pharmacia Fine Chemicals.) Buffer A was 0.5M HEPES, pH 7.0 and buffer B was 0.5M HEPES, pH 7.0 in 0.5M NaCl at a combined flow rate of 1 ml/min. 82
5 1.00 RNAse
0.75
RNAse
c 0.50 XI 40%B 40%B
0.25
3 __/ J __ / i. I i --- \ I--- 10 15 0
Retention Volume, ml.
Figure 26. Cation-exchange chromatogram of the RNAse A-nitroprusside reaction: (A) without heating and (B) after heating at 65°C for 15 minutes. In both reaction mixtures, excess nitroprusside was previously removed by passing the reaction mixture through an AG2-X8 column (anion exchange). Chromatographic conditions were the same as those in Figure 25. 83
Table 4. RNAse A activities of the different fractions from the nitroprusside-RNAse A reaction.
Sample Percent Activity
Native RNAse A 1 0 0
RNAse A-nitroprusside reaction mixture 36
Fraction #1 18
Fraction #2 15
Fraction #3 11
Fraction #4 30
Fraction #5 97
conditions, it gave very broad peak and could not be
purified to a reasonable homogeneity.
Fraction #3 (Figure 26A) was the most prominent single peak. However, when the collected fraction was rechromato
graphed under similar conditions, it produced a broad peak
and a new RNAse A peak (Figure 21k). When what remained of
the rechromatographed fraction #3 was again rechromato graphed, more of the RNAse A peak appeared with a corres ponding decrease and broadening of peak #3. After the
fourth reinjection, the sample left was too small and the peak was so broad that it was no longer practical to continue the purification process. All of the modified Absorbance, 280 nm 0.25 0.50 0.75 1.00 iue 7 Cto-xhne hoaorpi poie of profiles chromatographic Cation-exchange 27. Figure etn t 5° o 1 mnts Crmtgahc condi Chromatographic minutes. 15 for °C 65 at heating in ee h sm a toei iue 25. Figure in those as same the were tions enetd rcin 3 a wtothaig n () after (b) and heating (a)without #3 fraction reinjected 0 5 eeto Vlm, ml. Volume, Retention 10 0 B 40% “I /“ I ___
5 10 5 0 i I 84 i 0 B 40%
85
RNAse A (fraction #3) used in subsequent experiments were
obtained only after a second chromatographic separation.
Fraction #3 seemed to be a heat-labile product. It
was observed that when the reaction mixture was heated to
65°C for 15 minutes prior to cation-exchange chromato
graphy, the peak corresponding to fraction #3 was signifi
cantly reduced (Figure 26B). This was accompanied by a
significant increase in peak #5, which corresponded to the
unmodified RNAse A. The other fractions did not seem to
significantly change in their chromatographic profiles
upon heating. Similarly, when the isolated fraction #3 was
heated for 15 minutes at 65°C, there was a significant
decrease in the amount of peak #3 which corresponded to a
big increase in peak #5 (Figure 27B). There was also an
appearance of a broad peak near the void volume but the
area of this peak did not seem to correspond to the disso
ciation phenomena and sometimes, for unknown reasons, it
did not even appear.
When fraction #3 was subjected to gel filtration
(Figure 28A) it gave a peak that corresponded to a
molecular weight almost twice that of native RNAse A
(Figure 29). However, when the same fraction was heated at
65°C for 15 minutes prior to gel filtration, the high
molecular weight peak was almost totally gone and a new
peak that corresponded to the molecular weight of native
RNAse A appeared (Figure 28B). 0.2 AU
CM
Retention Time, min. Figure 28. Gel filtration of the products of the nitroprusside-RNAse A reaction. (A) fraction # 3 (B) fraction #3 heated at 65°C for 15 minutes (C) SNP-RNAse A reaction mixture (D) SNP-RNAse A mixture heated at 65°C for 15 minutes. Retention Volume, ml. 12.000 1.750 - 1.750 1.500 - 1.500 -i— 0 0 0 . 1 - 1.250 4.100
Ns rmtegl itaindt (iue 28). (Figure data filtration gel the from A RNAse Figure 29. Calculation of molecular weight of modified of weight molecular of Calculation 29. Figure aie Ns (12,600) RNAseNative A Log of Molecular Weight Log of Modified RNAse A (17,800) 4.2004.150 Myoglobin P rpi InhibitorTrypsin 4.250
4.300 88
Similarly, when the nitroprusside-RNAse A reaction
mixture was subjected to gel filtration under the same
conditions, two peaks were observed, with one peak almost
twice the molecular weight of the other (Figure 28C). When
the same reaction mixture was heated at 65°C for 15 minutes
prior to gel filtration, the high molecular weight peak
was reduced while that which corresponded to the molecular
weight of native RNAse A increased (Figure 28D).
Amino acid analysis of the different fractions (Table
6 ) showed that fraction #5 had about the same number of
lysine residues as that of native RNAse A. The modified
RNAse A (fraction #3, Figure 26) seemed to have lost about
one lysine residue. The amino acid chromatogram of frac
tion #3 (Figure 30) showed the 6 -hydroxynorleucine (RT =
28.2 minutes) and unresolved 5-chloro- and 6 -chloronor-
leucine (RT = 53.8 minutes) eluting very close to tyrosine.
In addition, there were two small unidentified peaks, one eluting between methionine and isoleucine and the other
about 3 minutes before histidine.
Since repeated rechromatography of the modified RNAse
A (fraction #3,Figure 26) resulted in no further reduction
in the amount and broadening of the peak and with corres ponding recurrence of fraction #5, a significantly pure
fraction #5 was not succesfully isolated. Therefore, no
further studies was made. 89
Table 6 . Comparison of the amino acid composition of native RNAse A and the different products of its reaction with sodium nitroprusside. All calculations based on alanine at 1 2 .0 0 .
Amino Theor. Native Fract. Fract. Fract. Fract. Fract. Acid No. RNAse A 1 2 3 4 5
Asp 15 15.10 14.72 15.13 15.07 14.90 14.85
Thr 1 0 1 0 . 0 2 10.15 10.17 10.37 1 0 . 2 0 10.15
Ser 15 14.75 14.90 14.87 15.10 15.20 14.90
Glu 1 2 1 1 . 8 8 11.85 11.67 11.80 11.80 11.90 Gly 3 3.12 3.28 3.13 3.13 3.20 3.60
Ala 12 12.00 12.00 12.00 12.00 12.00 12.00
Cys 8 6.55 6.58 6.60 6.45 6.80 6.48
Val 9 7.62 7.52 7.90 7.47 7.90 7.58
Met 4 3.95 4.08 3.93 3.97 3.80 3.90
H e 3 1 . 6 8 1.92 1.77 1.73 2 . 1 0 1.90
Leu 2 2.38 2.92 2.67 2.67 2.60 2.90
Tyr 6 5.55 5.50 5.33 5.57 5.10 5.60
Phe 3 2.85 2 . 8 8 2.83 2.83 3.00 3.10
His 4 3.82 4.02 3.97 3.90 3.90 4.00
Lys 1 0 9.92 8 . 6 8 9.57 8.93 8.90 9.95
Arg 4 3.88 3.95 4.00 4.00 3.90 4.00
No. of separate trials averaged 4 4 3 3 1 4 Absorbance, 570 nm. .5 ■ 0.75 0.50 - 0.50 0.25 ■ 0.25 1.00 0 oiid Ns A fato 3 Fgr 6. mn acid Amino 26). Figure (fraction#3, A RNAse modified 6 nlssws oe na tnad ainecag column cation-exchange standard a on done analysiswas is identified. 06 mx 0 mclm. yeW3, eka 383) Pa # o #1 338038). Peak Beckman W-3H, Type column. cm 20 x (0.6 cm clr- n 5clrnrecn. ek n ee not aandwere Peaks b 5-chloronorleucine. and -chloro- 6 -hydroxynorleucine andpeak -hydroxynorleucine iue 0 Aio cd rfl o HCl-hydrolyzed of profile acid Amino 30. Figure CO e 25 > eeto ie min.Time, Retention 50 2 # isunresolved of mixture 75
100 DISCUSSION
The Deamination Reactions
Sodium nitroprusside has long been used as a color
test for the presence and quantification of ionizable -SH
groups and, for more than two decades, has been exten
sively used as a hypotensive drug. On the other hand,
nitrous acid has long been used to bring about the deami
nation of amines (Whitmore and Langlois, 1932). However,
its usefulness in deaminating amino groups in proteins
(Herriott, 1947) has been severely limited due to the
relatively harsh reaction conditions (pH below 5), the
occurence of a number of side reactions with tyrosine,
tryptophan and amino-terminal groups, and the formation of
several products. In 1970, Maltz et al showed that sodium
nitroprusside could also deaminate primary amines under
alkaline conditions to give products similar to those
obtained upon deamination by nitrous acid. This similarity
in products suggest a similarity in reaction mechanisms
and the possibility of replacing nitrous acid with sodium
nitroprusside for site-specific modification of lysine
residues in proteins.
91 92
It is generally assumed that the reactivity of nitrous acid (HONO) lies in its nitrosonium ion, NO+ . Similarly, the NO moiety in the nitroprusside ion can also be formally considered to be a coordinated N0+, that is,
(CN) tjFe-NO+) . Therefore, analogous reactions with primary amines can be written:
HjO O HiO . N, RNH2 + ON-OH ^ RNH-N —4- R-N=N - 4 - R+ ( 5)
H,0-Fe(CN)J‘ N, + II / + t RNH2 + ON-Fe(CN)5 RNH-N-Fe(CN)5 ------^-►R-N=N - 4 -R + (6)
The exact reaction pathway that leads to the formation of the carbonium ion with nitroprusside has not been clearly elucidated. Dozsa et al. (1984) suggested that the pH dependence of the overall reaction rate, measured from the rate of N 2 evolution, depended mainly on the protona tion constant of the amines and the rest of the reaction steps were thought to be very fast. This and the earlier observation (Maltz et al, 1971) that aromatic amines were almost inert to nitroprusside, are consistent with the widely-held belief that the first step involves the attack of the nucleophilic amino group on the electrophilic N0+ . 93
After the formation of the [R-NH2 _N-Fe(CN)tj]^ complex
there seems to be a controversy about which pathways and possible intermediates are involved, some of which are
summarized in Figure 31. Dozsa et al. (1984) suggested that the subsequent steps involves dehydration to form a
2 - dinitrogen complex [R-N=N-Fe(CN)5 ] and a subsequent
aquation to form an exclusively alcohol derivative of the
starting amine, in an SN2 type of reaction. The other product is diazoferricyanide complex which in turn is broken down into [FefCN^^O]2- and N 2 * This interpreta tion, however, is not totally consistent with a number of previous reports (Maltz et al., 1971; Garvey and Kimura,
1986) and the results from this study since olefins and secondary alcohols were also formed from the reaction of nitroprusside with aliphatic amines. These later products strongly suggest at least a transient carbonium ion forma tion, similar to that observed in the deamination with nitrous acid.
On the other hand, Maltz et al. (1971) previously suggested that the alkyldiazonium ion is formed which is then attacked by a hydroxide ion in an SN2 fashion.
However, the authors also expressed the possibility that at least some of the products might have come either directly from the dehydrated dinitrogen complex in an SN2 RNH2 + N-Fe(CN)|
RNH0-N-Fe(CN)2"
H-O-Fe(CN)*”
RNH R-N=N-Fe(CN)? R-NHo-NO
OH"
R-OH NEN-Fe(CN)? R-N=N
R-OH
R+ Carbonium ion
Figure 31. Proposed reaction pathways and interme diates of the deamination of amines with nitroprusside. (1) Dozsa et al., 1984 (2) Maltz et al., 1971. 95 fashion or by hydroxide attack of the transient carbonium ion intermediate.
The data from this study seem to suggest that the carbonium ion intermediate is probably highly reactive and reacts with any nucleophile in the reaction mixture. The most abundant nucleophile and hence most probable reactant is water, which forms a primary alcohol derivative (Figure
32). Water can also facilitate the abstraction of one hydrogen from the carbon adjacent to the carbonium ion with a consequent double bond formation. Furthermore, the carbonium ion can migrate into the adjacent carbon and, through the same reactions with water, forms secondary alcohol and olefins. To a much lesser extent, 4-alcohols and 4-olefins might even be formed but were not detected by the methods used in this experiment.
The hypothesis that water is involved in the formation of products from the carbonium ion intermediate was based on the observation that the distribution of the deamina tion products did not significantly change within a very wide range (10®) of OH” concentrations .covered by both nitrous acid and nitroprusside (Tables 2 and 3). This, however, does not preclude the possibility that a certain fraction of the primary alcohols were formed through SN'1 type of reaction, that is, the addition of water before the complete removal of the diazo complex. R+ ►-c h 2- c h - c h 3
- C H o - C - C H .
♦ I + - C - C H - C H o - c h 2- c h - c h 3 1 . H
- c h 2 c h - c h 3 -CH0-CH0-CH9-OH - c h =c h - c h 3 OH
VO CTi Figure 32. Proposed reactions of water with the carbonium ion intermediate. 97
In the presence of Cl“ even at molar concentration about 1 0 times less than that of water, 1 -chlorooctane became the most predominant product, more than 1 -octanol.
The relative predominance of 1-chlorooctane over 2-chloro- octane (at least 1 0 times higher) suggests that these chloride derivatives did not arise from the 1 -octene as an intermediate (Whitmore and Langlois, 1932). Under the conditions used, it was also inconceivable that these chloride-containing products were derived from their respective alcohols.
If it is correct to assume that the carbonium ion formed is highly reactive, then it would not descriminate between water and Cl“, just as it did not descriminate between water and OH“. However, Figure 5 clearly indicated that there was a sharper increase in the relative propor tion of 1 -chlorooctane over 1 -octanol than can be accoun ted for by the mole fraction of NaCl. This suggests a more descriminating intermediate than carbonium ion. It might be that Cl“ directly reacts with the diazonium ion before
N 2 is completely released, that is, an SN^ type of substi tution reaction. This seems to be a plausible explanation of the dependence of the relative proportion of the chlorine-containing products on the Cl” concentration in the mixture. This, however, does not also preclude the possibility that some of the chlorine-containing deriva- 98
tives were formed from direct attack of Cl” on both the primary and secondary carbonium ion.
The increase in temperature not only increased the overall extent of reaction but also seemingly increased the relative proportions of the rearrangement products, i.e., 2-octanol and 2-octene. This suggests that tempera ture enhances the overall rate of formation of the carbonium ion intermediates which, upon formation, have more tendency to rearrange before finally reacting with water to form different products. This is understandable since the activation energies for the reaction of the primary carbonium ion with water is conceivably different from the activation energy for the migration of the positive charge into the second carbon.
The obvious disadvantage in using nitrous acid for deamination is the further conversion of some of the deamination products (1 - and 2 -alcohols) into their nitrite derivatives especially when nitrous acid is in excess. Under slightly different conditions, these reac tions were even used for the synthesis of butylnitrite from butyl alcohol (Noyles, 1936) or 2-octylanitrite from
2-octyl alcohol (Kornblum and Oliveto, 1947). These nitrite derivatives, however, are not ideal final products for lysine modification in proteins because of the inherent 99
chemical instability of nitrites in general (Kabasakalian
and Townely, 1962). This particular side reaction was
absent when sodium nitroprusside was used in place of
nitrous acid under neutral or slightly alkaline conditions.
This may be attributed to the greater stability of the
-N0+ that is coordinated to iron in sodium nitroprus
side. This coordinated -N0+ is not sufficiently reactive
to react with the hydroxyl group of the alcohols under the
conditions employed. It may be also recalled that the
coordination of -N0+ by iron enhances the stability of
nitroprusside at high pH, a condition where nitrous acid would spontaneously have decomposed. Thus, in these
respects, sodium nitroprusside produces a "cleaner"
reaction compared to nitrous acid.
Reactions Occurinq Under Conditions Similar to Those Used
for the HCl Hydrolysis of Proteins
Since sodium nitroprusside is a potential reagent for
site-selective deamination of lysine residues in proteins,
it was deemed necessary to investigate the stability of the
deamination products under conditions similar to those used
for the hydrolysis of proteins for subsequent amino acid
analysis. The model compound selected was Na-CBZ-L-Lysine, where the CBZ-group both protects the a-amino group from
reacting and gives UV absorbance to monitor product separa 100
tion by HPLC. Furthermore, the CBZ-group is relatively easy to remove under the conditions similar to those in
HCl hydrolysis of proteins.
The reaction of sodium nitroprusside with Na-CBZ-L-
Lysine gave products (Table 7) analogous to with those obtained under the same conditions from n-octylamine.
Surprisingly, there was more of the primary olefin than
2 -alcohol derivative, as opposed to their reverse order in abundance observed with n-octylamine and sodium nitroprus side (Figure 14). The reason for this apparent discrepancy was not known or investigated.
During hydrolysis of Na-CBZ-L-lysine in 5.8M HCl, the
CBZ-group is easily removed by the reaction:
c h 2c i + co2 + NH3-CH-C-OH
(6)
Similarly, fast removal of the CBZ moiety would be expected for the deamination products listed in Table 6 .
As its removal under acidic condition is so fast (Morrison
and Boyd, 1983) it would probably be complete before a
significant amount of chemical transformation occurs in the side group. 101
Table 7. Product distribution of the reaction of sodium nitroprusside and Na-CBZ-L-Lysine. (5 mM Na-CBZ-L- Lysine, 50 mM sodium nitroprusside, in 0.05M TEA, pH 8 .6 , 22°C, 10 hours.)
Product Percentage
6 -hydroxynorleucine 60
2-amino-5-hexenoic acid 25
5-hydroxynorleucine 9
2-amino-4(t)-hexenoic acid 4
When 6 -hydroxynorleucine (Figure 16C, peak #2) was incubated under conditions commonly used for the complete hydrolysis of proteins (i.e., 5.8M HCl, 22 hours at 105°C), about 70% of it was converted into 6 -chloronor leucine
(Figure 18C, peak e). This reaction has been described
(Lent et al.. 1969; Malin et al., 1984) and the elution position of the resulting product, once known as "pre tyrosine" (Kurosky and Hofmann, 1972) was reported to be just prior to tyrosine. The reaction presumably involves a simple nucleophilic substitution (Morrison and Boyd, 1983).
When the major olefin derivative (2-amino-5-hexenoic acid, Figure 16C, peak #4) was subjected to the same condi tions of hot HCl treatment, however, it was completely 102
eliminated and gave rise to four major products upon amino
acid analysis (Figure 16D). The reaction involves addition
of HCl to the double bond which favor the "Markovnikov"
addition of the nucleophile into the second carbon (Morri
son and Boyd, 1983). The apparent absence of a 4-chloronor-
leucine derivative suggest that the intermediate was not a
carbonium ion. The presence of small amounts of 6 -hydroxy-
norleucine and 6 -chloronorleucine (Table 8 ) was somewhat
surprising but could be due to "anti-Markovnikov" addition
in the presence of oxygen in the reaction mixture. The presence of 5-chloronorleucine in the HCl hydrolyzate of nitrous acid-deaminated poly-L-lysine had earlier been
reported by Malin et al. (1984).
As indicated earlier, much effort was spent on the
identification of the 5-hydroxynorleucine peak since it was unidentified in previous reports due to apparent
"discrepancies" in its chemical nature. While it elutes in the amino acid analyzer as a linear 5-hydroxynorleucine, the evidences presented here suggest that it was derived for the most part from a cyclic intermediate, 2 -amino-6 - caprolactone. It should be noted in Figure 23 that even at the early stage of the hot HCl treatment, a significant amount of the lactone (shown in the graph as 5-hydroxynor-
leucine, its open form) was immediately formed. This suggests at least part of this lactone was formed by a 103
Table 8 . Distribution of products from 2-amino-5- hexenoic acid after hot HCl treatment. (5.8M HCl, 22 hours at 105°C).
Product Percentage
5-hydroxynorleucine 6 6
5-chloronorleucine 23
6 -chloronorleucine 7
6 -hydroxynorleucine 4
direct internal electrophilic addition reaction converting the olefin into a lactone (Figure 33A) when heated under highly acidic conditions (Roberts and Caserio, 1964). The formation of lactones from similar olefins under highly acidic conditions is well-documented in the literature
(Linstead, 1932; Linstead and Rydon, 1933; 1934; Boorman and Linstead, 1934). The reaction (Figure 33A) is favored to produce a six-membered lactone ring (Linstead, 1932;
Linstead and Rydon, 1933) at least partly because the terminal methyl group has some stabilizing effect on the lactone ring (Linstead and Rydon, 1934). It should be noted, however, that this six-membered lactone ring is still very susceptible to hydrolysis under neutral or alkaline conditions, as will be discussed later. 104
Figure 23 also suggests that the lactone was probably formed from internal substitution reaction of 5-chloronor- leucine (Figure 33B) since the consistent increase in the relative proportion of the lactone (graphed as 5-hydroxy- norleucine) was accompanied by a similar decrease in
5-chloronorleucine, which appeared to have been formed early in the incubation with hot HCl. This is not surpri sing since Cl“ is a good nucleophile and was present at relatively high concentration. Once formed,it might easily undergo internal nucleophilic substitution to form the lactone (Figure 33B).
Finally, it was also conceivable that a certain amount of 5-hydroxynorleucine was formed directly as shown in
Figure 33C and then immediately cyclized by internal este- rification to form the lactone. The chemistry of this reaction under the conditions of these experiments is also known (Roberts and Caserio, 1964; Fessenden and Fessenden,
1982).
The presumed opening of the 2-amino-6-caprolactone ring when it was diluted in the starting buffer (0 .2 M citrate, pH 3.25) or during an early stage of the amino acid analysis, is probably due to the overall susceptibi lity of lactones to hydrolysis under less acidic or alka line conditions (Hegan and Wolfenden, 1939; Hall et al.,
1958; Huisgen and Ott, 1959; Wheeler and Gamble, 1961; (5)
.n h 3 h c i r r -L.
qxj^£_rY- CH3 HO*,<,^’o O O CH? CH2 H O ' ^ O ch3 H (1) (2) (3) (5) OH- c
(4 ) (5)
Figure 33. Possible routes for the formation of the lactone during treatment of 2-amino-5-hexenoic acid with hot HCl (5.8M HCl, 105°C). (1) 2-amino-5-hexenoic acid (2) postulated carbonium ion intermediate (3) 5-chloronor- 105 leucine (4) 5-hydroxynorleucine and (5) 2 -amino-6 -capro- lactone. 106
Blackburn and Dodds, 1974; Kaiser and Kezdy, 1976; Edward et al., 1982). For example, 6 -n-hexolactone was reported to be 52% hydrolyzed into 5-hydroxyhexanoic acid at equilibrium in boiling water (Linstead and Rydon, 1934).
This susceptibility of 6 -membered lactones to ring opening
(Huisgen and Ott, 1959) was attributed to a considerable relief of steric strains upon the formation of the tetra hedral intermediate during hydrolysis of the lactone ring from the more strained chair conformation caused by the trigonal carboxyl carbon in the lactone (Brown et al.,
1954), although other factors are also presumed to be involved (Kaiser and Kezdy, 1976). Consistent with this view, the rate of hydrolysis of the six-membered lactone ring has been reported to be mainly dependent on the rate of addition of the OH” ion to the carbonyl carbon, the elimination step being very fast (Kaiser and Kezdy, 1976).
There is no information available on the effect of the
2 -amino group on the stability of the 2 -amino-6 -capro- lactone against hydrolysis but it is not expected to be a major factor considering that both the lactone formation and its hydrolysis occur very well below the pKa of the amino group.
In summary, the transformations during HC1 hydrolysis of the two major deamination products, 6 -hydroxynorleucine and 2-amino-5-hexenoic acid, are shown in Figure 34. 107
- C H j - C H j - C H - C I I j ( 9%) A OH
-CHj-CHj-CMICHj (25%) CBZ-L-LYS + SNP
-CHj-CHSCH-CHg ( 4%)
Ivt §**#*«♦•■ • *
NH, H -C -C H ,-C H , - CH, - CH,-CI ( 70%)
B COOH
NH* N Hj
H-C-CH,-CH, -CH i -CH j -OH ( 2 %) I CHX COOH
(27%) NHi Cl H -C -C H 2-C H j-C, - £ Hh -C- c H j (<<1%)
COOH
(66%) C Hj—
Hllj Cl H-I-CH j -CH j -CH-CH, 3 (23%) NH, COOII I H-C-CH,-CH,-CH=CH, I Nil,| OOOH ff-C-CHj-CI^-CHj-CHj-CI (7%)
COOII NH. I H-j-CHj-Cllj-Cllj-Cllj-OH (4% )
Figure 34. Schematic representations of the (A) SNP- mediated deamination of Na-CBZ-L-Lysine and the chemical transformations during incubation of (B) 6 -hydroxynorleu- cine and (c) 2-amino-5-hexenoic acid under conditions normally used in the hydrolysis of proteins for amino acid analysis. 108
Reaction of Sodium Nitroprusside with RNAse A
In an attempt to selectively modify Lys-41 in RNAse A,
its reaction with sodium nitroprusside was carried out at
pH 8 .6 , just slightly lower than the estimated pKa of
Lys-41 (pKa = 8 .8 , Murdock et al.. 1966) and about 2 pH
units below the pKa of the other e-amino groups. The a-amino terminal group (pKa « 7.5) might be also expected
to react because of its lower pKa value. At pH 8 . 6 and
room temperature (22°), about 80% of the RNase A was modi
fied after 1 0 hours.
The chromatogram of the reaction mixture, with the excess nitroprusside previously removed by anion-exchange chromatography, showed several minor peaks and one promi nent peak (peak #3, Figure 26A). The presence of minor products was expected each amine-nitroprusside reaction produces at least 4 products and there are 10 e-amino and one a-amino groups in RNase A, each of which is poten tially capable of reacting under the long incubation period of 10 hours. Therefore, the presence of one promi nent sharp peak was thought to be promising enough, at least initially. The next obvious step was to purify this major product for further characterization. 109
It turned out to be more complicated than that. The modified RNAse A (fraction #3) did not rechromatograph
cleanly. Instead, a new peak corresponding to native
RNAse A was observed and what used to be a sharp peak #3 was broadened in each of the rechromatographic step. After
the third rechromatography, very little of the material in
fraction #3 was left and it was impractical to further purify the product. In each rechromatographic step, no other significant peaks appeared to indicate the other part of the dimer, assumed to be the modified RNAse A.
When the reaction mixture was also subjected to gel
filtration, two peaks were observed: one corresponded to the retention time (hence molecular weight) of native
RNAse A and the other at almost twice the molecular weight of native RNAse A. Since it is known that native RNAse A is capable of forming dimers and even higher aggregates under certain conditions (Crestfield et al..1962; Fruchter and Crestfield, 1965a) it was immediately assumed that the high molecular weight-peak was that of a dimer.
The results of this study, however, strongly suggest that the dimer is most probably not of both native RNAse A even if repetitive chromatography continued to produce the native RNAse A peak. First, the modified RNAse A (fraction
#3, Figure 26A) showed only about 1/5 of its original
RNase A activity, in contrast to the full catalytic 110
activity of a native RNAse A dimer (Fruchter and Crest
field, 1965a; Crestfield et al., 1962). In the proposed structure of the native RNAse A dimer (Fruchter and Crest field, 1965a; Crestfield et al., 1962) the orientations of the two monomers are such that the active sites are unaffected and can undergo normal catalysis.
Secondly, the amino acid analysis of fraction #3
(Figure 26A) showed that about one lysine residue was lost
(Table 6 ), although actual values from several independent trials indicated a wide range from 8.4 to 9.6 of the 10 initial lysine residues. This was an unusually high deviation compared to most other amino acids. This might be a reflection of the different degrees of purification of fraction #3, that is, differences in the ratios of modified and native RNAse A in the dimer.
While the data clearly showed that at least one component of the dimer was unmodified RNAse A, it was not clear whether the dimer formation was between a native and a singly modified RNAse A such as proposed by Fruchter and
Crestfield (1965b) or a native and a doubly modified RNase
A. The expected impurities after a single chromatographic separation and the broadening of peak #3 after rechromato graphy made it difficult to get accurate quantitative data to compare the relative amounts of the unmodified and Ill modified RNAse A. Besides, it was observed that further rechromatography produced more of the native RNAse A, indicating the homogeneity was not achieved and therefore comparison of relative peak areas was meaningless.
The dimer was observed to be at least partially disso ciated by heating at 65°C for 15 minutes, the conditions that Fruchter and Crestfield (1965a) have shown to disso ciate dimers of native RNAse A. This further indicates that no covalent bond was formed between monomers. (It was indicated earlier that some part of the dimer continued to dissociate upon rechromatography.) When the reaction mixture was heated at 65°C for 15 minutes prior to chroma tography, peak #3 was almost totally lost (Figure 26B).
Similarly, when an isolated fraction #3 was heated prior to rechromatography, more of the dimer dissociated as shown by the larger RNAse A peak (Figure 27B).
The lability of the dimer to heating was also demons trated by gel filtration. When the reaction mixture was heated at 65°C for 15 minutes prior to gel filtration, the peak correpsonding to the dimer was significantly reduced (Figure 28D). A similar result was observed when an isolated fraction #3 was heated prior to gel filtration.
At this point, it can only be assumed that the structure of the dimer(s) is probably similar to that proposed by
Fruchter and Crestfield (1965) since Lys-41 or the a-amino 112 terminal, the likely targets of this chemical modification procedure, are not known to be critically involved in the dimer formation.
The above problems made the pursuance of further studies difficult. Until an effective method of stabili zing and separating the dimers is found, the signifi cance of this initial study will not be known. PART II
CYANOMETHYLATION OF THE AMINO GROUPS IN RNASE A
INTRODUCTION
In 1850, Strecker inadvertently obtained alanine from a mixture of acetaldehyde, ammonia and HCN after hydrolysis
(In Pichat, 1970). Since then, the "Strecker synthesis" and its many modifications have become a valuable method for the synthesis of a-amino acids and several other types of compounds. Aminoacetonitrile, for example, has been prepared by the Strecker synthesis (Ponseti et al., 1956), a reaction which would soon find its importance in protein chemistry.
h 2 so 4 CH20 + NaCN + NH 4 C1 ------► H 2 NCH2CN (8 ) EtOH
A number of enzymes have been found to be sensitive to
cyanide in the presence of their natural carbonyl-
containing substrates. This group includes fructose-1,6 -
diphosphatase (Cash and Wilson, 1966), transketolase
(Brand and Horecker, 1968), acetoacetate decarboxylase
(Autor and Fridovich, 1970), 2-keto-4-hydroxyglutarate
113 114
aldolase (Hansen et al., 1974), pyruvate dehydrogenase
(Stepp and Reed, 1985) and transaldolase (Brand and Hore-
cker, 1968). The mechanisms of inhibition were not under
stood for some time because most of the observed inhibi
tions were reversible by dialysis or dilution. When
irreversible or more stable adducts were later discovered
(Hansen et al., 1974; Stepp and Reed, 1985) this
allowed for the characterization of the obligatory
products mediates of the inhibitory reaction. It turned
out that the reaction involved the addition of cyanide
to a Schiff-base intermediate formed between the carbonyl-
containing substrates (or intermediates) and the e-amino
group of a lysine residue in the enzyme (Hansen et al.,
1974). Thus, this century-old reaction has resurfaced
again.
This finding opened the possibility of using cyanome-
thylation (or cyanoalkylation) as a means of selectively modifying lysine residues that are involved in enzyme
catalysis. Such selective lysine modification has already been done on enzymes with catalytically-active lysine
residues such as 2-keto-4-hydroxyglutarate aldolase
(Hansen et al., 1974) and anthranilate synthetase (Bower
and Zalkin, 1982). Aside from the aforementioned, several other enzymes are known to have catalytically active
lysine residues. These include liver alcohol dehydroge nase (Tsai et al., 1974; 1985), glucose-6 -phosphate dehy- 115
drogenase (Jentoft and Dearborn, 1979), and ribonuclease A
(Means and Feeney, 1968; Richards and Wyckoff, 1971;
Blackburn and Moore, 1982). Thus the respective active site lysine residues can be potentially cyanomethylated.
The addition of cyanide to a Schiff base may also have
a significance in the commonly used technique of methyla-
tion of lysine residues in proteins with formaldehyde and
NaCNBHg (Dottavio-Martin and Ravel, 1978; Jentoft and
Dearborn, 1979; MacKeen et al., 1979). Since cyanide is
liberated from NaCNBHj under the reaction conditions
(Jentoft and Dearborn, 1980) or is often present as an
impurity of the NaCNBH3 reagent, CN” can potentially
compete with NaCNBH3 for the Schiff base. Indeed, this
had been observed by Gidley and Sanders (1982). N-cyanome-
thylation of important lysine residues in proteins may
also have important application as a reversible lysine
modifier since most of the N-cyanomethyl groups in the
lysine residues of proteins can be removed by simple
dialysis at low pH (Gidley and Sanders,1982). The reaction
could therefore be used to protect important, usually the
more reactive, amino group while some other chemical modi
fications are carried out elsewhere in the protein. If so
desired, the N-cyanomethyl group can also be further
converted to an irreversible N-carboxymethyl group by mild
hydrolysis. 116
This study was therefore conducted with these specific aims:
(1) To explore the possible use of N-cyanomethylation as a technique of chemically modifying lysine residues in ribonuclease A.
(2) To investigate the reversibility of the N-cyanome thylation reaction and/or the stability of the N-cyanome- thylated product. MATERIALS
The following chemicals used in these experiments are
listed with their corresponding catalog numbers in
parenthesis. Purchased from Sigma Chemical Co., P.O. Box
14508, St. Louis, Mo. 63178 were bovine serum albumin
(B-2518), a,e-di-DNP-lysine (D-0255), 2,4-dinitrophenol
(D-7004), e-DNP-lysine (D-0380), 1-fluoro-2,4-dinitroben-
zene (D-1529), iodoacetic acid (1-6250), iodoacetic acid
(sodium salt) (1-6375), [4-(2-hydroxyethyl)-1-piperazine
ethanesulfonic acid (H-3375), (3-lactoglobulin (L-0130),
ovalbumin (A-3014), pepsin (P-7012), TPCK-treated trypsin
(T-8642), and trypsinogen (T-1143).
Purchased from J.T. Baker Chemical Co., 222 Red School
Lane, Phillipsburg, N.J. 08865 were acetic acid (9507),
acetonitrile (9017-3), diethyl ether (9244-3), formic acid
(5-0128), methanol (9093-3), sodium borohydride (V023) and
sodium cyanide (3662).
(Butylamino)acetonitrile (12,180-0) was from Aldrich
Chemical Co. Inc., 940-W. St. Paul Ave., Milwaukee, Wi.
53233.
Ammonium hydroxide (3256), formaldehyde (5016), and phosphoric acid (2796) were from Mallinckrodt Inc.,
Paris, Ky. 40361.
117 118
Some chemicals used in this study were also included
under "Materials" in Part I, pages 17-18. All chemicals used in this experiment were used without further purification. METHODS
Reaction of n-butylamine with CH^O and NaCN
To 2 ml of 0.05M succinic acid were added 15 y,l (0.2
mmole) of n-butylamine and 98 mg NaCN (2 mmole) with
vigorous mixing. The pH of the solution was adjusted to 5.0
with 6 M HC1. Then, 16.5 y.1 (0.2 mmole) of 37% formaldehyde
was added with further mixing. The test tube was covered
with parafilm and the reaction mixture was incubated at room
temperature for different lengths of time. The products of
the reaction were extracted twice with 5 ml portions of
diethyl ether. After the ethereal extract was evaporated to
dryness the residue was dissolved in 0.5 ml methanol.
Aliquots of typically 0.5-1.0 ill were injected into a Varian
model 3300 gas chromatograph equipped with a 3% oV-17 on
Gas-Chrom Q column and a thermal conductivity detector
(TCD). The inj ector port and the detector were both set at
250°C. Column temperature was held at 120°C for 1 minute,
then programmed to rise to 220°C at 20°C/min. The peaks
were identified by comparison of their retention times with
those of the standards.
Variation of the reaction conditions include different
pH, different formaldehyde concentrations and different
reaction times.
119 120
Cyanomethvlation of RNAse A
A 490 mg sample of NaCN (10 mmoles) was dissolved in
20 ml of 0.05M succinic acid. The pH of the solution was adjusted to 6.0 with 6 N HC1. Then, 126 mg (0.01 mmole) of the purified RNAse A was added and the solution was gently swirled to dissolve the protein. Finally, 167 p.1 of 0.36% formaldehyde (0.02 mmole) was added with mixing and the solution was allowed to stand at room temperature
(22°C) for about 24 hours. In some treatments, 10 times molar excess (relative to RNAse A) of phosphate or 3 '-CMP were incorporated into the reactive mixture.
After 24 hours, the sample was dialyzed against hourly changes of 10”% HCl for about 5 hours. An aliquot of the protein solution was taken every hour and assayed for RNAse A activity. The rest of the protein was lyophilized and the residue was redissolved in about 5 ml of 0.05M succinate buffer, pH 5.0. The solution was filtered through a 0.45jx Millex-HV4 filter (Millipore).
Then, 500 y.1 aliquots were injected into the FPLC equipped with a cation-exchange column (Mono S, HR 5/5, Pharmacia
Fine Chemicals) and eluted with 0.05M succinate buffer, pH 5.5 with a 0-0.2M NaCl gradient. The absorbance of the eluate was monitored at 280 nm. Four peaks were observed and their respective fractions were collected. The same fractions from several runs were pooled and dialyzed 121
against 3 hourly changes of 10“^M HCl. (A separate aliquot of the major product peak was dialyzed for 1 2 hours against hourly changes of 10”^M HCl and assayed for
RNase A activity before each change in solvent.) After dialysis, an aliquot was also taken and rechromatographed under the same conditions to check for homogeneity. The rest of the fractions were lyophilized and stored in the freezer for further use.
Gel Filtration
The following protein solutions were prepared at 3 mg/ml concentrations in water: native RNAse A, each of the two lyophilized major fractions, and selected protein standards such as 0 -lactoglobulin, pepsin, ovalbumin, trypsinogen and bovine serum albumin. Then, 50 y.1 of each protein solution was injected into a Varian Model 5000
HPLC equipped with a GF-250 column (9.4 mm x 250 mm I.D., particle size of 5u, zirconia-stabilized silica support,
Du Pont). The absorbance of the peaks was monitored at
280nm using a Spectroflow 773 detector (Kratus Analytical
Instruments). The samples were eluted with either 20 mM acetic acid at pH 3.2 or 0.2M phosphate at pH values from
4 to 8 and at the elution rate of 0.5 ml per minute. 122
Performic Acid Oxidation of RNAse A
The performic acid oxidation of RNAse A and modified derivatives of RNAse A were carried out by the procedure of
Hirs (1956). The performic acid solution was prepared by adding 100 y.1 of 30% hydrogen peroxide to 1.9 ml of 99% formic acid. The solution was mixed, covered with parafilm and allowed to stand for 2 hours in an ice-salt bath (-5 to
-10°C).
In a separate large test tube, 10 mg of native or modified RNAse A was dissolved in 500 ul of 99% formic acid, after which 1 0 0 y.1 of anhydrous methanol was added with stirring. The protein solution was also cooled in the same ice-salt bath for 30 minutes. Then, 0.5 ml of the performic acid solution was added to the protein solution, mixed briefly and the reaction was allowed to proceed at ice-salt bath temperature for 2.5 hours. The reaction mixture was diluted to 2 0 ml with ice-cold water, frozen and lyophi lized.
Trypsin Digestion
The performic acid-oxidized proteins were dissolved in
2 ml of 0.2M NH 4 HC03, pH 8.5, and 1% (w/w) TPCK-treated trypsin was added. The reaction mixture was incubated at
37°C for 2 hours, then another 1% (w/w) TPCK-treated trypsin was added and the incubation extended to 1 2 more hours. Finally, the trypsin hydrolyzates were lyophilized. 123
Peptide Mapping and Isolation by HPLC
The lyophilized tryptic digest was redissolved in 500 y.1 of 0.1% H 3 P0 4 in water and filtered through a 0.45y.
Millex-H4 filter unit (Millipore). The samples were then injected in 100 y.1 portions into a Varian 5000 liquid chromatograph, equipped with a 4.6 x 250 mm Lichrosorb
RP-18 column (10y. particle size, Brownlee) connected to a
Spectroflow 773 detector (Kratus Analytical Instruments) and a single channel chart recorder (Linear Model 1200).
The peptides were eluted at 1 ml/min with 0.1% H 3 P0 4 with a gradient of 0-70% CH3 CN. The UV monitor was set at 220 nm and the different peptide peaks were collected manually. The "same fractions" from several runs were pooled and were either lyophilized directly or diluted with water to reduce the CH3CN concentration before lyophilization.
The elution profiles between the native and modified
RNAse A peaks were compared and where there were noticeable differences, the fractions that differed bet ween the various samples were subjected to HCl hydrolysis and amino acid analysis.
Preparation of Dinitrophenyl Derivatives
The procedure was adapted from the general method of
Fraenkel-Conrat et al. (1955). To a 200-y.l solution con taining 2 mg native or modified RNAse A were added 50 y.1 124
of 4.2% NaHC0 3 and 400 y.1 of 5.3% FDNB in absolute ethanol. The suspension was incubated at room temperature for 1 hour with occassional shaking. Throughout the incubation period, the pH was checked and maintained at
8.5 by the addition of 4.2% NaHC03 . After one hour, 1.0 ml of water and 0.05 ml of 4.2% NaHC03 were added. The unreacted FDNB was extracted three times with 1 ml portions of ether and the extracts discarded. Then, the pH was adjusted to 1.0 with 6 N HCl and the suspension was centrifuged at 3,000 rpm. The liquid phase was decanted and 0.5 ml of 6 M HCl was added to the precipitate. The resulting mixture was vortexed at low speed to resuspend the precipitate. The resulting suspension was transferred to an amino acid hydrolysis tube, the tube was evacuated with a vacuum pump and sealed. Hydrolysis was carried out at 105°C for 24 hours. The HCl hydrolyzate was dried in a vacuum desiccator at 40°C. The residue was redissolved in 100 yl of CH3 CN, diluted with 300 y.1 of water and was centrifuged at 3000 rpm.
Separation of DNP Derivatives by HPLC
The solutions of DNP derivatives were diluted 100 times with 25 % CH3CN and aliquots (usually 5-10 y.1) were injected into a Varian 5000 chromatograph equipped with a reverse phase column ( 6 x 250 mm Lichrosorb RP- 8 , (10y. particle size, Brownlee) pre-equilibrated with 25% CH3CN 125
in 0.1% H 3 PO 4 . The samples were eluted at 1 ml/min with a
linear gradient of 25% to 75% CH3 CN in 0.1% H 3 PO 4 .
The peaks that appeared were identified by comparing
their retention times with authentic e-DNP-Lys, a,e-di-
DNP-Lys and a-DNP-Lys standards.
Carboxymethvlation of Nc-DNP-L-Lysine
To 1.0 ml solution containing 1.74 mg Ne-DNP-L-lysine-
HC1 (5 mmol) was added 1.04 mg (5 mmol) iodoacetic acid
(sodium salt) at pH 9.0. The reaction mixture was incu bated for about 2 hours at 22°C. Then, the reaction mixture was evaporated to dryness in a vacuum desiccator at 40°C. The residue was dissolved in 1.0 ml of 25% CH 3 CN.
Suitable aliquots were diluted 100 times, from which 10 y.1 portions were injected into the HPLC under the same conditions described in the preceeding paragraph. RESULTS
Reaction of n-butylamine with CHoO and NaCN
Typical gas chromatograms of the ether extract of the reaction mixture containing n-butylamine, CH20 and
NaCN at different reaction times are shown in Figure 35.
Peak #1 corresponds to the retention time of authentic butylaminoacetonitrile. Peak #2 corresponds to the peak produced when butylaminoacetonitrile was treated with
CH20 and NaCN at pH 5.0. Therefore, peak #2 was believed to be dicyanomethylated n-butylamine, that is, N,N-di-
(cyanomethyl)n-butylamine. Peaks a and b were not iden tified because of the absence of suitable standards.
Because both peaks a and b were no longer present at longer incubation periods, these peaks were pressumed to be unstable intermediates of the reaction.
These results show that at longer incubation periods
(20 hours or more) at pH 5.0, both monocyanomethylation and dicyanomethylation occured (Figure 35C). However, at pH 9.0 the monocyanomethylated product was exclusively formed (Figure 35F). This was confirmed by incubation of authentic n-butylaminoacetonitrile with CH20 and NaCN at pH 5.0 and pH 9.0 (Figure 36). At pH 9.0 n-butylamino acetonitrile was unchanged but at pH 5.0, almost all of
126 127
0 .9 -
CL 0 .3 -
“ 1 P 10 0 10 0 5 10 Retention Time, Min.
pH 0.9-
Cl 0.3-
0 - T T r T 0 5 o 6 0 5 10 Retention Time, Min.
Figure 35. Gas chromatogram of the ether extract of the reaction mixture containing n-butylamine (100 mM), formaldehyde (100 mM) and NaCN (1M) at different reaction times at: pH 5.0 (A) 5 minutes (B) 2 hours (C) 20 hours. pH 9.0 (D) 5 minutes (E) 2 hours (F) 20 hours. Relative Peak Intensity, TCD 0.9 0.6 0.3 0 0 ih omleye 10m) n NC (M atr 0 minutes 30 after (1M) NaCN and mM) (100 formaldehyde with incubation at room temperature (22°C. (A) pH 5.0 (B) pH 9.0 (B)pH 5.0 standard. (A) pH (22°C. n-butylaminoacetonitrile (C) and temperature room at incubation iue 6 Rato o nbtlrioctntie 10 mM) (100 n-butylarainoacetonitrile of Reaction 36. Figure 5 10 eeto Tm, min. Time, Retention r 1 0 o 10 5
10 128 129 the n-butylarainoacetonitrile was converted to the dicya- nomethyl derivative. (In the presence of excess formal dehyde and NaCN there was 100% conversion into the dicya- nomethyl derivative.)
When peak #1 (i.e., butylaminoacetonitrile, Figure 36) was incubated in 0.2M succinate at pH 5.0 for 2 hours, extracted into ether, the ether extract evaporated to dryness, dissolving the residue in methanol and injecting into the gas chromatograph, no new peak appeared and there was no change in the peak size. This suggests that although the monocyanomethyl derivative was preferen tially formed at pH 9.0, it was also stable at pH 5.0.
N-Cyanomethylation of RNAse A
Two major peaks were observed from the cation-exchange chromatography (Mono S, Pharmacia Fine Chemicals) of the reaction mixture at pH 5.5 (peaks #2 and #4, Figure 37A).
Since peak #4 was in the same position as that of native
RNAse A, peak #4 was immediately assumed to be residual
(unmodified) RNAse A. Rechromatography of peak #4 on the same Mono S column at either pH 5.5 (0.05M succinate) or pH 7.0 (0.05M HEPES) produced a clean RNAse A peak. This fraction was shown to be fully active (Table 8 ) towards cytidine-2 ':3'-cyclic monophosphate and had the same amino acid composition as native RNAse A. Absorbance, 280 nm 0.75- 1.00 0.50 0.25 0 5 hsht () ih1 tms oa xes f 'CP and 3'-CMP of excess molar times 10 (C)with phosphate reaction mixture containing RNAse A, formaldehyde and NaCN. and formaldehyde A, RNAse containing mixture reaction was set at 280 nm. (Mono S column, Pharmacia Fine Chemicals, Fine Pharmacia column, S (Mono nm. 280 at set was in 0.5M NaCl. Flow rate was at 1 ml/min and the UV monitor UV the and ml/min 1 5.5 at pH was rate Flow succinate, 0.05M was NaCl. B 0.5M in buffer and 5.5 pH succinate, Inc.) (A) without inhibitor (B) with 10 times molar excess of excess molar times 10 (B) with inhibitor without (A) D rcrmtgahd rcin 2 Bfe A a 0.05M was A Buffer #2. fraction rechromatographed (D) iue 7 Cto-xhne hoaorm o the of chromatograms Cation-exchange 37. Figure 10 5 0 15 «./ 5 10 n Volume, ml.on I * T 0 5 1 ------10
4 0 % B % 0 4 / --
1 130 131
However, when the major modified RNAse A fraction (peak
#2, Figure 37A) was rechromatographed at either pH 5.5 or pH 7.0, a large peak and a small peak corresponding to unmodified RNAse A appeared, suggesting that the fraction was not yet homogeneous. The pooled fraction #2 was then rechromatographed at pH 7.0. Further rechromatography of the eluted fractions at either pH 5.5 or 7.0 on the Mono S column produced a sharp single peak (Figure 37D), sugges ting homogeneity.
At least two other minor peaks were also observed from the reaction mixture, indicated as peaks #1 and #3 in
Figure 37. However, these minor modified-RNAse A fractions were not fully characterized because they did not rechromatograph cleanly and therefore, fractions of sufficient amount and purity were not obtained.
When the cyanomethylation reaction was carried out at neutral or higher pH, there was a progressive decrease in the amounts of product peaks. At pH 9.0, for example, the sizes of the peaks were relatively small and poorly resol ved even after 5 days of incubation. This might be due to increased number of different lysine residues modified.
Conversely, at increasingly acidic pH the product peaks decreased in number but increased in resolution after a 24- hour incubation period. The bulk of the preparations of the modified RNAse A was therefore carried out at pH 5.5 where one major new product peak was observed. Also, the 132 chromatographic profile of the reaction mixture at pH
5.0 was not significantly different from that at pH 5.5.
Quantification of the amounts of the individual product peaks was, however, difficult because of the degree of overlap of the peak bases.
Addition of increasing concentrations of formaldehyde increased the number and amount of product formed (based on the amount of residual RNAse A) but also resulted in loss of resolution and selectivity. There was also a progressive increase in the peak at the void volume, sug gesting the possibility of multiple-site modifications within a single RNAse A molecule. Another possibility is cross-linking of the proteins which is known to occur at higher formaldehyde concentrations (French and Edsall,
1945; Galembeck et al.,1977; Tome et al.. 1985). However, this observation was not investigated any further.
Gel Filtration
Gel filtration of the dialyzed reaction mixture, native and modified (cyanomethylated) RNAse A at pH 7.0 are shown in Figure 38. The gel filtration chromatogram of the reaction mixture (Figure 38C) showed two peaks, one corresponding to about twice the molecular weight of the other (Figure 39). Comparison of these peaks with those of the native (Figure 38A) and the modified
RNAse A (Figure 38B) showed that the latter gave rise to Absorbance, 280 nm. 0 2 . 0 0.15 0.10 0.05 0 RNAse A and (C) dialyzed reaction mixture. Elution was Elution mixture. reaction (C)dialyzed and A RNAse h clm ws F20 a icnasaiie silica zirconia-stabilized a GF-250, was column The done with 0.1M phosphate at the rate of 0.5 ml per minute. per ml 0.5 of rate the at phosphate 0.1M with done mx20m .. oun (DuPont). column I.D. mm 250 x mm support with particle size of about 5y. packed into a 9.4 a into 5y. about of packed size particle with support A ntv RAeA B ioae fato o modified of fraction isolated (B) A RNAse native (A) 5 iue 8 Gl itain hoaorm o the of chromatograms filtration Gel 38. Figure * 10 12.4 0 eeto Vlm, ml. Volume, Retention 10 5 -Tr- 10
11.0
J 12.4
15 133 Retention Volume, ml. 13 12 11 10 0 0 0 . 4 — H 9 column (GF-250, Du Pont). Du (GF-250, column n mdfe RAe b euin rm h gl filtration gel the from elution by A RNAse modified and
A ( A e s A N R Figure 39. Estimated molecular weights of the native the of weights molecular Estimated 39. Figure oiid A e s A N R Modified 23, ) 0 0 ,7 3 (2 0 0 2 . 4 13 o oeua t h g i e W Molecular of g o L , 200 dimer^v. v ^ r e m i d n i l u b o l g o t c a L - B ) 0 0 4 . 4 n e g o n i s p y r T n i m u b l a v O 0 0 6 . 4
135 the high molecular weight peak. These results suggest that the modified RNAse A underwent dimerization, a pheno menon earlier observed with the nitroprusside-modified
RNAse A. Heating of the modified RNAse A fraction at
65°C for 15 minutes prior to gel filtration gave barely noticeable reduction in the dimer peak and an appearance of only a small native RNAse A peak.
Further examination of this material by SDS-gel elec trophoresis showed that both native RNAse A and the material eluting at 12.4 ml (modified RNASe A) had iden tical migration, hence, both appear by this technique to have the molecular weight of RNAse A (Figure 40). This suggests that there was no covalent bond formation between the monomers of the dimeric N-cyanomethyl-modified
RNAse A.
It was observed, however, that there was a significant variation of retention volumes of both the native RNase A and the modified RNAse A dimer during gel filtration at different pH values (Figure 41). Below pH 6.0, the peaks were broader and showed extensive "tailing" and therefore calculations of their peak areas were not made. Below pH
3 most of the samples and standards did not elute from the gel filtration column. No further studies were made on the pH dependency of the retention volumes during gel filtration. Relative Mobility, c m . - 4 4.1 modified RNAse A. RNAse modified aieo oiidRNAseA e s A N R Modified or Native iue 0 SSgl lcrpoei o ntv and native of electrophoresis SDS-gel 40. Figure o oeua t h g i e W Molecular of g o L n e g o n i s p y r T n i l u b o l g o t c a L - B 4.34.2 4.4 . 4.6 4.5 Pepsin
136 o Native RNAse A • Modified R N A s e A
6.0 7.0 8.0 p H
Figure 41. Variation of retention times with pH during gel filtration of native and modified RNAse A. All other chromatographic conditions were the same as those in Figure 38. 138
Identification of the Modified Lysine Residue
When HCl-hydrolyzates • of FDNB-treated RNAse A were subjected to reverse-phase HPLC and the absorbance of the fractions monitored at 360 nm, two peaks were observed.
The retention times for peak #1 and #2 (Figure 42A) corresponded to authentic samples of e-DNP-Lys and a,e-di-DNP-Lys, respectively. However, the HCl hydroly- zate of the FDNB-treated RNAse A showed three peaks
(Figure 42B). The a,e-diDNP-Lys peak was reduced to almost nothing and a new peak appeared about 3.5 minutes ahead of e-DNP-Lys. The retention time of this third peak corresponded to that of authentic a-carboxymethyl- e-DNP-Lysine prepared by treating e-DNP-Lysine with iodoacetate (equation 6 a), suggesting that this new peak was a-carboxymethyl-e-DNP-Lysine (equation 6 b).
HOOC HOOC
< c h 2 >4 NH-DNP
O NH-CH2 -C -O H n h - c h 2- c n 5.8M HCl ■ilH + NH-OC-CH - HOOC (6 b) 10 5 °C ,2 2 H (CH2 )4
H-DNP NH-DNP Absorbance, 360 nm 100 - 0 5 5- 75 25- 0 -, 0 T
25%B eiaie o () aiead B mdfe Ns . The A. RNAse (B) modified and native (A) of derivatives 5 mID (Brownlee). I.D. mm 250 x ape ee ltdwt 01 hsht wt 2% to 25% a with phosphate 0.1M with eluted CH 75% were samples Figure 42. HPLC chromatograms of the different DNP different the of chromatograms HPLC 42. Figure 3 CN gradient. The column was Lichrosorb RP- Lichrosorb was column The gradient. CN 10 T
1 eeto Tm, min. Time, Retention ------75%B 75%B 20 T ,
0 T
5B / 25%B 10 T
8 , 4.6 1 ----- 75%B
20
T i 139 140
The relative abundance of the different peaks, corres
ponding to the number of DNP-lysine residues, are shown
in Table 9.
Comparison of the HPLC profiles of the tryptic digests
of native and modified RNAse A showed that only one peak was missing in the modified RNAse A (Figure 43B). Amino
acid analysis showed that this missing peak corresponded
to the Lys-^-Lys-y fragment of ENAse A. A new peak . also
appeared on the tryptic digest of the modified RNAse A but because it almost coeluted with another bigger peak,
it could not be isolated in sufficient amount and purity
for accurate amino acid analysis.
Table 9. Estimated number of lysine residues converted to their DNP derivatives. The absorption coefficient of the di-DNP derivative was assumed to be twice that of the mono-DNP derivatives. Values are averages of 3 separate trials.
No. of Lysine Residues
DNP Derivative Native RNAse A Modified RNAse A
e-DNP-Lysine 9.94 + 0.12 9.89 + 0.19
e,a-di-DNP-Lysine 1.06 ± 0.13 0.07 + 0.03
a-CM-e-DNP-Lys ine 0 1.04 ± 0.20 Absorbance, 220nm. Absorbance , 220nm. 1.00 of (A) native and (B) modified RNAse A. RNAse (B) modified and native (A) of 0 Figure 4§. HPLC chromatograms of the tryptic digests tryptic the of chromatograms HPLC 4§. Figure 10 10 20 20 eeto Vlm, ml. Volume, Retention eeto Volume,ml Retention Retention 050 30 30 40 40 50 070 60 60 141
70 142
The residual RNAse A activities of the different products are shown in Table 10. The RNAse A cyanomethy- lated at the amino terminal lost about 1/3 of its acti vity.
Comparison of the Lineweaver-Burke plots of both the native and modified RNAse A (Figure 44) showed that both the Km and Vm values were changed. The calculated Km value was consistent with previous data of 0.99, 0.75 and
0.60 mM, respectively (Hummel et al., 1958; Litt, 1961;
Nelson and Hummel, 1961).
Table 10. Residual activities and lysine contents of the different Mono S peaks. RNAse A activity was measured using cytidine-231-cyclic monophosphate as substrate and the percentages of RNAse A activities were calculated based on the alanine content of 12 residues. Values for peaks #1 and #3 were the average of 2 trials; those of peaks #2 and #4 were average of 5 trials.
Peak No % RNAse A Activity Lysine contents
1 61.3 ± 3.2 8 . 8 ±0.4
2 67.0 + 2.0 9.0 ± 0.1
3 75.8 ± 2.8 9 . 1 ± 0.2
4 1 0 0 .1 ± 1 . 8 9.9 ±0.1
RNAse A 100 10 143
1 6 0 0 n • N — cyanomethylated RNAse A Y = 0.113X + 61 K m = 1.85e-3 § 1200 V m = 1.64e-2
E 8 0 0 D $ 4 0 0 - o Native RNAse A Y = 0.087X + 108 K m = 8.06e— 4 Vm = 9.26e— 3
5 0 0 0 1 E 4 2 E 4 -1 1 / [ S o ] , M
Figure 44. Lineweaver-Burk plots of the native and modified RNAse A using cyclic-2’:31-cytidine monophosphate, [native RNAse A] or [modified RNAse A] = 2 pM, 22°C, 0.05M succinate, pH 5.0. 144
DISCUSSION
Amines have the ability to act as nucleophiles because the nitrogen atom has a lone pair of electrons.
On the other hand, the reactivity of the carbonyl carbon of aldehydes (or ketones) is primarily due to the partial positive charge on the carbon atom as a result of the higher electronegativity of the oxygen bonded into it.
The carbonylamine condensation therefore involves the attack of the nucleophilic nitrogen atom on the carbonyl carbon, forming an imine or a Schiff base (Feeney et al..
1975). The imine thus formed can readily hydrolyze back to the starting reactants under neutral pH. However, the protonated imine or iminium ion (pK ~ 7) can be further attacked by other nucleophiles present in solution to produce a more stable product. In this study, such a nucleophile is CN“. Thus, when formaldehyde and proteins are used as reactants, an N-cyanomethyl derivative of the e-amino groups of lysine and/or the a-amino terminal group can be produced.
h 2o CN" + R-NH2 + CH20 R-N=CH r -n h =c h 2 R-NH-CH2CN (7)
h 2o
144 145
The effect of pH on the distribution of products in the reaction between n-butylamine, CH2 O and NaCN was consistent with the above equation (Eq. 7). At both pH 5.0 and 9.0, there seemed to be enough concentration of unpro- tonated amine for fast condensation with formaldehyde. At both pH values, monocyanomethylated products were already formed even in the first few minutes (or may be seconds).
After monocyanomethylation at pH 5.0, the secondary amine could still condense with another molecule of formadehyde and the carbon in the resulting Schiff base would still be electrophilic enough to react with another CN” to finally produce the dicyanomethylated product.
At pH 9.0 condensation of the secondary amine with another CH20 may still occur but such secondary Schiff base may not be reactive towards CN”. The addition of CN” requires a protonated imine (Feeney et al., 1975) which would be less likely to exist under alkaline conditions.
Therefore, the reaction could not proceed any further and the principal product will be the monocyanomethylated derivative. Again, this was supported by the inertness of butylaminoacetonitrile when incubated with formaldehyde and NaCN at pH 9.0.
Ribonuclease A was again used as the model protein in this study because of the unusually reactive Lys-41 located in the catalytic site (Richards and Wyckoff, 1971; 146
Blackburn and Moore, 1982). It has been previously
observed that addition of even a small amount of formal
dehyde caused a rapid decrease in RNAse A activity (Means
and Feeney, 1968). The original objective of this study
was therefore to effect selective N-cyanomethylation of
Lys-41.
When reaction with CH20 and CN” was carried out at
different pH values, it was observed that the reaction
was very slow under neutral or slightly alkaline condi
tions. This might be due to the low concentration of
iminium ion that is believed to be the reactive species
(Feeney et al., 1975; Borch et al., 1971) towards the CN” .
At lower pH, the carbonyl-amine condensation was probably
rate-limiting but the addition of the CN” into the higher
concentration of iminium ion resulted in the production
of more cyanomethylated products.
When the reaction mixture was chromatographed after
various incubation periods, the product peaks increased
with time but the profile was otherwise essentially the
same. This suggests that there probably was no signi
ficant interconversions of the products.
Although the original objective was to modify Lys-41,
the results showed that most of the N-cyanomethylation of
RNAse occured at the a-amino terminal group which resulted
in the loss of 1/3 of its activity. This is based on the 147
observation that only the Lys1-Lys7 fragment was
significantly reduced in the tryptic digest (Figure 43B).
This was accompanied by the appearance of a new peak that
eluted earlier (RT - 5 minutes) but its resolution was
poor and could not be separated in enough quantity and
purity for accurate amino acid analysis.
The reverse-phase HPLC chromatogram of the HCl hydro-
lyzates of FDNB-treated modified RNAse A also supports
the contention that the a-amino group was the site of
N-cyanomethylation. Figure 42B showed the disappearance of the a,e-di-DNP-Lys derivative. Instead, a new peak corresponding to a-carboxymethyl-e-DNP-Lys was observed, which was quantitatively similar to that of the lost a,e-di-DNP-Lys. Since RNAse A has 10 e-amino groups and one a-amino group, extensive dinitrophenylation results in 9 e-DNP-Lys and one a,e-di-DNP derivatives. This was quantitatively observed in this study. However, if the a-amino group would be cyanomethylated, only the e-amino group of Lys-1 would be converted to e-DNP-derivative.
Upon HCl hydrolysis, this cyanomethyl group would be converted into a carboxymethyl group and would therefore elute as a-carboxy-methyl-e-DNP-Lys-1. These results are summarized in the schematic representations below: 148
NH2 NH-DNP I FDNB , iSSSN I 5.8M HCl 9 E-DNP-Lya CLys)g-^-CH ------(C-DNP-Lys)9-^-AH » + W I PH 8 '5 W I 105 C '22 H 1 a.E-dlDNP-Lya ( c h 2)4 NH2 NH-DNP NH-CH2-CN n h - c h 2- c n FDNB S.8 M HCl 9 E-DNP-Lya ays,9- ® f ^ DNp--ys)9-w-fH 105;c,22 H* , _ ++ a-CM-e-DNP-Lya Cch 2 )4 <9h 2U NH2 Ah - DNP Although tyrosine, histidine, arginine and cysteine have been reported to sometimes react with DNFB under certain conditions (Fraenkel-Conrat et al., 1955; Hirs, 1967) no other peak absorbing at 360 nm was observed in the reverse-phase chromatogram in significant amounts. It was not surprising that while the amino acid analysis showed a loss of one lysine residue in the Na-cyanomethylated RNAse A, no new peak in the amino acid chromatogram was observed. The Na-carboxymethylamino 149 group (from HCl hydrolysis of the Na-cyanomethylamino group) would not be expected to react with ninhydrin. The two minor peaks (#1 and #3, Figure 37A) are most probably due to the formation of N-e-cyanomethylated deri vatives of lysine residues located at various sites including the active site of RNAse A. Addition of phosphoric acid in the reaction mixture significantly reduced the size of peak #1 (Figure 37B) while addition of 3'-CMP reduced both peaks #1 and #3 (Figure 37C). The amino acid chromatogram of the HCl hydrolyzate of the materials in peaks # 1 and #3 also showed the presence of a small peak corresponding to e-CM-Lys. However, due to the difficulty of isolating fractions #1 and #3 in suffi cient amount and purity, no tryptic mapping was done. The observation that the a-amino terminal group was preferentially modified is not surprising. The a-amino terminal group (pKa= 7.5) has been reported to be usually reactive towards chemical modification reagents (Means and Feeney, 1968; Jentoft and Dearborn, 1979; Gerkin et al.. 1982; Means, 1984; Acharya et al., 1985) because of its lower pKa than that of Lys-41 (pKa « 8.8, Murdock et al., 1966). At the reaction pH of 5.5, the e-amino group of lysine 41 would be expected to be almost completely protonated, making it less reactive reactive. What was surprising was that when the reaction was carried out at 7.0 or 9.0, the formation of products was even slower 150 and there seemed to be no apparent change in selectivity. This is in contrast with the reductive methylation of proteins with formaldehyde and NaCNBHg at around the neutral pH where Lys-41 was preferentially methylated Jentoft and Dearborn, 1979). The loss of 1/3 of the activity upon the Na-cyanome- thylation RNAse A is somewhat surprising since the a-amino terminal group is not thought to have a vital role in its catalytic action. This might be attributed to conforma tional changes in the overall protein structure. Dialysis of the Na-cyanomethylated RNAse A derivative against several changes of 10“3M HCl or 0.01M acetate over a 24-hour period did not seem to significantly reverse the reaction back to RNAse A, CH2 O and CN”. While it has been reported that the formation of some N-cyanoalkyl-protein derivatives is readily reversible at slightly acidic pH (Gidley and Sanders, 1982), our results suggest that this is not the case with RNAse A. Other N-cyanomethyl-protein derivatives have also been observed to be more stable (Hansen et al.. 1974; Stepp and Reed, 1985). This may be either a consequence of the lower pKa of the a-amino group relative to the a-amino group of Lys-41 or to the presence of stabilizing groups around the modified residue. Similar to that of the SNP-modified RNAse A, the Na-cyanomethylated RNAse A also appeared to form a dimer. The latter was more stable than the former since heating 151 at 65°C for 15 minutes did not significantly dissociate the dimer. However, such dimer did not involve covalent bond formation since SDS-gel electrophoresis dissociated them into monomers. Formaldehyde is long known to cause cross-linking of proteins (French and Edsall, 1945; Galem- beck et al., 1977; Tome et al., 1985) but this does not seem to be the case here. The results of this study also supports the previous reports (Gidley and Sanders, 1982) that the presence of CN“ can reduce the efficiency of reductive methylation when using NaCNBH3 as the reducing agent. Cyanide is a common impurity of the NaCNBH3 reagent and, although it is stable in aqueous solutions (Borch et al., 1971) it is released as one of the by products from the incubation of NaCNBH3 under the conditions used for reductive methyla tion (Jentoft and Dearborn, 1980). REFERENCES ACHARYA, A.S., L.G. SUSSMAN and J.M. MANNING. 1985. Selec tivity in the modification of the a-amino groups of hemoglobin bn reductive alkylation with aliphatic carbonyl compounds. Biol. Chem. 260:6039-46. ARNOLD, W.P., D.E. LONGNECKER and R.M. EPSTEIN. 1984. Photodegradation of sodium nitroprusside: Biologic activity and cyanide release. Anaesth. 16:254-260. ARNOLD, W.P., C.K. MITTAL, S. KATSUKI and F. MURAD. 1977. Nitric oxide activates guanylate cyclase and increases guanosine 3’:5'-cyclic monophosphate levels in various tissue preparations. Proc. Nat * 1 Acad. Sci.,USA. 74:3202-3207. AUTOR, A. and I. FRIDOVICH. 1970. The interactions of acetoacetate decarboxylase with carbonyl compounds, hydrogen cyanide and an organic mercurial. Biol. Chem. 241:4290-2. AVEY, H.P., M.O. BOLES, C.H. CARLISLE, S.A. EVANS, S.J. MORRIS, R.A. PALMER, B.A. WOOLHOUSE and S. SHALL. 1967. Structure of Ribobonuclease. Nature (London). 213:557- 562. BECK, M.T., A.KATHO and L. DOZSA. Formation of heterocyclic amino acids in the reaction of a,6- and a,e-diamino acids with nitrosylpentacyanoferrate(II). Inorq. Chim. Acta 55:L55-L56. BELLO, J., H. IJIMA and G. KARTHA. 1979. A new arylating agent, 2-carboxy-4,6-dinitrochlorobenzene. Int. J. Peptide Prot. Res. 14:199-212. BISSET, W.I.K., A.R.BUTLER, C. GLIDEWELL and J.REGLINSKI. 1981. Sodium nitroprusside and cyanide release: Reasons for re-appraisal. Br. J. Anaesth. 53:1015-1018. BLACKBURN, G.M. and H.L.H. DODDS. 1974. Strain effects in acyl transfer reactions. Pt III. Hydroxide and buffer- catalyzed hydrolysis of small and medium-ring lactones. J. Chem. Soc., Perkin Trans. II. (4):377-82. 152 153 BLACKBURN, P. and B.L. JAILKHANI. 1979. Ribonuclease inhi bitor from human placenta: Interaction with derivatives of RNAse A. J;. Biol. Chem. 254:12488-12493. BLACKBURN, P. and J. G. GAVILANES. 1980. The role of Lys-41 of RNAse A in the interaction with RNAse A inhibitor from human placenta. J^_ Biol. Chem. 255:10959-10965. BLACKBURN, P. and S. MOORE. 1982. Pancreatic Ribonuclease. The Enzymes. Vol. 15, Sec. 12. Acad. Press. BOORMAN, E.J. and R.P. LINSTEAD. 1933. Investigation of the olefinic acids. Part X. The formation of lactones from a- and "-n-butenoic acid and P-pentenoic acids. J. Chem. Soc., London. 577-580. BORCH, R.F., M.D. BERNSTEIN and H.D. DURST. 1971. The cya- nohydridoborate anion as a selective reducing agent. J. Am. Chem. Soc. 93:2897-2904. BOWER, S. and H. ZALKIN. 1982. Modification of Serratia Marcescens anthranilate synthetase with pyridoxal 5'-phosphate. Arch. Biochem. Biophys. 219:121-7. BRAND, K. and B. HORECKER. 1968. Reaction of cyanide with the transaldolase-dihydroxyacetone complex. Arch. Biochem. Biophys. 123:312-6. BREWSTER, P., F. HIRON, E.D. HUGHES, C.K. INGOLD and P.A.D.S. RAO. 1950. Configurations of amino compounds and the steric course of deamination. Nature 166:179-180. BROWN, H.C., J.H. BREWSTER and H. SHECHTER. 1954. An interpretation of the chemical behavior of five- and six-membered ring compounds. Am. Chem. Soc. 76:467- 474. BRUICE, T.C. and J.J.BRUNO. 1961. The mechanism of the reaction of hydroxylamine with -butyro- and 6-valero- lactones. Examples of reactions kinetically both third and fourth order. J.Am. Chem. Soc. 83:3494-8. BUDZIKIEWICZ, H.,C. DJERASSI and D.H. WILLIAMS. 1967. Mass Spectrometry of Organic Compounds. Chapter 4. Esters and Lactones, pp.174-213. Holden-Day, Inc. CANTOR, C.R. and P.R. SCHIMMEL. 1980. Biophysical Chemistry p. 930. Freeman, San Francisco, U.S.A. 154 CARLISLE, C.H., R.A. PALMER, S.K. MAZUMBAR, B.A.GORINSKY and D.G.R. YEATES . 1974. The structure of RNAse at 2.5A resolution. J_j_ Mol. Biol. 85:1-18. CASH, D. and I. WISLON. 1966. The cyanide adduct of the aldolase dihydroxyacetone phosphate. J± Biol. Chem. 241:4290-2. COHEN, S.A., K. BENEDEK, Y. TAPUHI, J.C. FORD and B.L. KARGER. 1985. Conformational effects in the reversed- phase liquid chromatography of RNAse A. Anal. Biochem. 144:275-84. COLLINS, C.J. 1971. Reactions of primary aliphatic amines with nitrous acid. Accts. Chem. Res. 4:315-322. COLLINS, C.J. and B.M. BENJAMIN. 1972. Molecular rearrange ments. XXIX. Exo/Endo stereospecificity of substituted classical norbonyl cations. A reassessment of "hot" carbonium ions. J^ Org. Chem. 37:4358-4366. CRAM, D.J. and M.R.V. SAHYUN. 1962. Studies in stereo chemistry. XXXIV. Variation of carbonium ion structure with variation in origin. J_;_ Am. Chem. Soc. 85:1257- 1263. CRAVEN, P.A. and F.R. DeRUBERTIS. 1978. Restoration of the responsiveness of purified guanylate cyclase to nitro- soguanidine, nitric oxide, and related activators by heme and hemeproteins. Evidence of the involvement of the paramagnetic nitrosyl-heme complex in enzyme acti vation. J\ Biol. Chem. 253:8433-8443. CRAVEN, P.A. and F.R. DeRUBERTIS. 1983. Requirement for heme in the activation of purified guanylate cyclase by nitric oxide. Biochim. Biophys. Acta 745:310-321. CRAVEN, P.A., F.R. DeRUBERTIS and D.W. PRATT. 1979. Elec tron Spin resonance study on the role of NO.catalase in the activation of guanylate cyclase by NaN 3 and NH2OH. J_j. Biol. Chem. 254:8213-8222. CRESTFIELD, A.M. and R. FRUCHTER. 1967. The homologous and hybrid dimers of ribonuclease A and its carboxymethyl histidine derivatives. J_j_ Biol. Chem. 242:3279-3284. CRESTFIELD, A.M., W.H. STEIN and S. MOORE. 1962. On the aggregation of bovine pancreatic RNAse. Arch. Biochem. Biophys.. Suppl. 1:217-222. 155 CRESTFIELD, A.M., W.H. STEIN and S. MOORE. 1963. Alkylation and identification of the histidine residues at the active site of RNAse. Biol. Chem. 238:2413-2419. CROOK, E.M., A.P. MATHIAS and B.R. RABIN. 1960. Spectro- photometric assay of bovine pancreatic RNAse by the use of cytidine 2 3 ' -phosphate. Biochem. J. 74:234. DAVIS, N.R. and A.J. BAILEY. 1972. The chemistry of the collagen cross-links. Biochem. J. 129:91-6. D 1OR, L. and J. COLLIN. 1953. Comportment des nitrites aliphatiques sous 1'action du bombardment electronique. Bull. Soc. Roy. Sci. Liege. 22:285-304. DOTTAVIO-MARTIN, D. and J.M. RAVEL. 1978. Radiolabelling of proteins by reductive alkylation with [ C]formal dehyde and sodium cyanoborohydride. Anal. Biochem. 87: 562-5. DOZSA, L., V. KORMOS and M.T. BECK. 1984. Kinetics of the reaction of pentacyanonitrosylferrate(II) with aliphatic amines. Inorq. Chim. Acta 82:69-74. DUDKIN, S.M., L.V. KARABACHYAN, S.N. BORISOVA, S.V. SHLYAP- NIKOV, M.Y. KARPEISKY and T.G. GEIDAROV. 1975. Spectral properties of phosphopyridoxyl-Lys'*41)-RNAse A. Bio- chim. Biophys. Acta. 386:275-282. EDWARD, J.T., E. COOKE and T.C. PARADELLIS. 1982. Stereo chemical studies. XI.Equilibria between 5-hydroxyacids and 6-lactones in aqueous tetrahydrofuran. Relative stabilities of isomeric 6-lactones. Can. J. Chem. 60: 2546-53. EL-AMIN, B. 1980. Development and characterization of a palladium hydrogenation system. Part I. PhD Thesis. The Ohio State University. FEENEY, R. E., G. BLANKENHORN and H. B. T. DIXON. 1975. Carbonyl-amino reactions in protein chemistry. Adv. Protein Chem. 29:135-203. FESSENDEN, R.J. and J.S. FESSENDEN. 1982.Organic Chemistry 2nd Ed., Section 13.6. Willard Grant Press. FIELDS, R. 1971.The measurement of amino groups in proteins and peptides. Biochem. J. 124:581-90. 156 FRANK, J.M., J.B. JOHNSON and S.H. RUBIN. 1976. Spectro- photometric determination of sodium nitroprusside and its photodegradation products. Pharm. Sci. 65:44-48. FRAENKEL-CONRAT, H.J., I. HARRIS and A.L. LEVY. 1955. Recent developments in technique for terminal and sequence studies in peptides and proteins. Methods of Biochem. Anal. Vol. 11:359-425. FRENCH, D.and J.T. EDSALL. 1945. Reactions of formaldehyde with amino acids and proteins. Adv. Protein Chem. 2: 277-335. FRUCHTER, R.G. and A.M. CRESTFIELD. 1965a. Preparation and properties of two active forms of ribonuclease dimer. Biol. Chem. 240:3868-3874. FRUCHTER, R.G.and A.M. CRESTFIELD. 1965b. On the structure of ribonuclease dimer. Isolation and identification of monomers derived from inactive carboxymethyl dimers. J. Biol. Chem. 240:3875-82. GALEMBECK, F . , D.S. RYAN, J.R. WHITAKER and R.E. FEENEY. 1977. Reaction of protein with formaldehyde in the presence and absence of sodium borohydride. Agric. Food Chem. 25:238-45. GEOGHEGAN, K.F., D.M. YBARRA and R.E. FEENEY. 1979. Rever sible reductive alkylation of amino groups of proteins. Biochemistry. 18:5392-9. GERKIN, T.A., J.E. JENTOFT, N. JENTOFT and D.G. DEARBORN. 1982. Intramolecular interactions of amino groups in 13C-reductively methylated hen egg-white lysozyme. Ji Biol. Chem. 257:2894-2900. GERTLER, A. and T. HOFMANN. 1967. The involvement of the amino-terminal amino acid in the activity of pancreatic proteases. I. Effects of nitrous acid on elastase. J. Biol. Chem. 242:2522-2527. GIDLEY, M.J. and J.K.M. SANDERS. 1982. Reductive methyla- tion of proteins with sodium cvanoborohydride.Biochem. 203:331-4. GILLAT, P.N., R.J. HART and C.L. WALTERS. 1984. Suscepti bilities of drugs to nitrosation under standardized chemical conditions. J. Chem. Toxicol. 22:269-274. 157 GOLDSTEIN, J. 1967. Digestion of ribonucleic acid by an alkylated ribonuclease. Mol. Biol. 25:123-30. GREISS, L., N.A.G. TREMBLAY and D.W. DAVIES. 1976. The toxicity of sodium nitroprusside. Can. Anaesth. Soc. J. 23:480-485. GRUETTER, C.A., B.K. BARRY, D.B. McNAMARA, D.Y. GRUETTER, P .J .KADOWITZ and L.J. IGNARRO. 1979. Relaxation of bovine coronary artery and activation of coronary arterial guanylate cyclase by nitric oxide, nitroprus side and a carcinogenic nitrosamine. Cyclic Nucleo tide Res. 5:211-224. GRUETTER, C.A., D.Y. GRUETTER, P.J. KADOWITZ and L.J. IGNARRO. 1981. Evidence that cyclic GMP formation may mediate vascular smooth muscle relaxation induced by nitrates, nitrites, nitroso-compounds and S-nitroso- thiols. Fed. Proc. 40:688. GURD, F.R.N. 1967. Carboxymethylation. In Methods in Enzymology. XI. Enzyme Structure, pp. 532-541. Ed. by C.H.W. Hirs. HANSEN, B.A., R.S. LANE and E.E. DEKKER. 1974. Formaldehyde binding by 2-keto-4-hydroxyglutarate aldolase.J. Biol. Chem. 249:4891-6. HARTLEY, T.F., J.C. PHILCOX and J. WILLOUGHBY. 1985. Two methods for monitoring the photodecomposition of sodium nitroprusside in aqueous and glucose solutions. J. Pharm. Sci. 74:668-671. HALL, H.K., M.K. BRANDT and R.M. MASON. 1958. Hydrolysis rates and mechanisms of cyclic monomers. J_j_ Am. Chem. Soc. 80:6420-7. HEGAN, D.S. and J.H. WOLFENDEN. 1939. The kinetics of the alkaline hydrolysis of some -lactones. Chem. Soc., London. 508-10. HEINRICKSON, R.L. 1966. On the alkylation of amino acid residues at the active site or RNAse. Biol. Chem. 241:1393-1405. HERRIOTT, R.M. 1947. Reactions of native proteins with chemical reagents. Adv. Prot. Res. 3:169ff. 158 HILL, H.E.1942. A contribution to the toxicology of sodium nitroprusside. I. The decomposition and determination of sodium nitroprusside. Aust. Chem. Inst. J. Proc. 9:89. HIRS, C.H.W. 1967. Reactions with reactive aryl halides. In Methods in Enzymology XI.Enzyme Structure, pp. 548- 555. Ed. by C.H.W. Hirs. HIRS, C.H.W., S. MOORE and W.H. STEIN. 1956. Peptides obtained by tryptic hydrolysis of performic acid- oxidized RNAse. Biol. Chem. 219:623. HONKANEN, E., T. MOISIO and P. KARVONEN. 1965. The mass spectra of some aliphatic lactones. Acta. Chem. Scand. 19:370-374. HUISGEN, R. and H. OTT. 1959. Die konfiguration der carbo- nestergruppe und die sondereigenschaften der lactone. Tetrahedron 6:253-67. HUMMEL, J.P., M. FLORES and G. NELSON. 1958. Anionic polymers. II. Inhibition of ribonuclease. J^ Biol. Chem. 233:717-21. IGNARRO, L.J., B. BALLOT and K.S. WOOD. 1984. Regulation of guanylate cyclase activity by porphyrins and metal- loporphyrins. Biol. Chem. 259:6201-6207. IGNARRO, L.J., B.K. BARRY, D.Y. GRUETTER, J.C. EDWARDS, E.H. OHLSTEIN, C.A.GRUETTER and W.H. BARICOS. 1980a. Guanylate cyclase activation by nitroprusside is related to formation of nitrosothiol intermediates. Biochem. Biophys. Res. Commun. 94:93-100. IGNARRO, L.J., J.N. DEGNAN, W.H. BARICOS, P.J. KADOWITZ and M.S. WOLIN. 1982. Activation of purified guanylate cyclase by nitric oxide requires heme. Comparison of heme-deficient, heme-reconstituted and heme-containing forms of soluble enzyme from bovine lung. Biochem. Biophys. Acta 718:49-59. IGNARRO, L.J., J.C.EDWARDS, D.Y. GRUETTER, B.K. BARRY and C.A. GRUETTER. 1980b. Possible involvement of S-nitro- sothiols in the activation of guanylate cyclase by nitroso-compounds.FEBS Lett. 110:275-278. 159 IGNARRO, L.J., P.J. KADOWITZ and W.H. BARICOS. 1981. Evidence that regulation of hepatic guanylate cyclase activity involves interactions between catalytic site -SH groups in both substrate and activator. Arch. Biochem. Biophys. 208:75-86. IGNARRO, L.J., H. LIPTON, J.C. EDWARDS, W.H. BARICOS, A.L. HYMAN, P.J.KADOWITZ and C.A. GRUETTER. 1981. Mechanism of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: Evidence for the involvement of S-nitrosothiols as active intermediates. J^ Pharmacol. Exp. Ther. 218: 739-49. IIJIMA, H., H. PATRZYC and J. BELLO. 1977. Modification of amino acids and bovine pancreatic RNAse A by ketoxal. Biochim. Biophys. Acta. 4:305-315. JENNINGS, M.L. and M.P. ANDERSON. 1987. Chemical modifi cation and labelling of glutamate residues at the stilbenedisulfonate site of human red blood cell Band 3 protein. J^. Biol Chem. 262:1691-7. JENTOFT, N. and D.G. DEARBORN. 1979. Labelling of proteins by reductive methylation using sodium cyanoborohydride. J^ Biol. Chem. 254:4359-65. JENTOFT, N. and D.G. DEARBORN. 1980. Protein labelling by reductive methylation with sodium cyanoborohydride: Effect of cyanide and metal ions on the reaction. Anal. Biochem. 106:186-90. JENTOFT, N. and D.G. DEARBORN. 1983. Protein labelling by reductive methylation. Meth. Enzymol. 91:570-9. JENTOFT, J .E., T.A. GERKEN, N. GENTOFT and D.G. DEARBORN. 1981. [I3]CMethylated ribonuclease A. J^ Biol. Chem. 256:231-6. JENTOFT, J.E., N. JENTOFT, T.A. GERKEN and D .G.DEARBORN. 1979. ^3C NMR studies of ribonuclease A methylated with [13C] formaldehyde. J_;_ Biol. Chem. 254(11): 4366-70. JOHNSON, C.C. 1929. The actions and toxicity of sodium nitroprusside. Arch. Int. Pharmacodyn. Ther. 35:480- 496. 160 KABASAKALIAN, P. and E.R. TOWNELY. 1962. Photolysis of nitrite esters in solution. I. Photochemistry of n-octylnitrite. J_;_ Am. Chem. Soc. 84:2711-24. KADOWITZ, P.J., P.NANDIWADA, C.A. GRUETTER, L.J. IGNARRO and A.L. HYMAN. Pulmonary vasodilator responses to nitroprusside and nitroglycerin in the dog. J.Clin. Invest. 67:893-902. KAISER, E.T. and F.J. KEZDY. 1976. Hydrolysis of cyclic esters. 5.Hydrolysis of lactones. Prog. Bioorq. Chem. 4:254-64. KAISER, F.E., C.W.GEHRKE, R.W. ZUMWALT and K.C. KUO. 1974. Amino acid analysis. Hydrolysis, ion-exchange, clean up, derivatization and quantitation by gas liquid chromatography. Chromatoqr. 94:113-133. KAMISAKI, Y., S.A.WALDMAN and F. MURAD. 1986. The involve ment of catalytic site thiol groups in the activation of soluble guanylate cyclase by sodium nitroprusside. Arch.Biochem. Biophys. 251:709-714. KARSUKI, S., W. ARNOLD, C. MITTAL and F. MURAD. 1977. Stimulation of guanylate cyclase by sodium nitroprus side, nitroglycerin and nitric oxide in various tissue preparations and comparison to the effects of sodium azide and hydroxylamine. Cyclic Nucleotide Res. 3: 23-35. KARTHA, G., J. BELLO and D. HARKER. 1967. Tertiary struc ture of RNAse. Nature(London). 213:862-865. KATHO, A., Z. BODI, L. DOZSA and M.T. BECK. 1984. Kinetics of the reactions of pentacyanonitrosylferrate(II) with mono- and diamino acids. Inorg. Chim. Acta 83:145-150. KNAPP, D.R. 1979. Handbook of Derivatization Reactions. John Wiley & Sons. KNOWLES, M.E., D.J.McWEENY, L. COUCHMAN and M. THOROGOOD. 1974. Interaction of nitrite with proteins at gastric pH. Nature(London) 247:288-289. KORNBLUM, N. and E.P. OLIVETO. 1947. 2-Octylnitrite. J^. Am. Chem. Soc. 69:465-6. 161 KUROSKY, A. and T.HOFMANN. 1972. Kinetics of the reactions of nitrous acid with model compounds and proteins, and the conformational state of N-terminal groups in the chymotrypsin family. Can.J. Biochem. 50:1282-1296. KREYE, V.A. 1980. Sodium nitroprusside. In Pharmacology of Antihypertensive Drugs. Ed. by A.Scriabine. pp.373-396. Raven Press. LECLERQ, P.A. and D.M.DESIDERIO. 1973. Chemical ionization mass spectra of amino acids and derivatives. Occurence and fragmentation of ion-molecule reaction products. Org. Mass Spectrom. 7:515. LEEWENKAMP, O.R., W.P. VanBENNEKOM, E.J. VanDER MARK and A.BULT. 1984. Nitroprusside, antihypertensive drug and analytical reagent.Review of (photo)stability, pharma cology and analytical properties. Pharm. Weekbl., Sci. Ed. 6:129-140. LEIMER, K.R., R.H. RICE and C.W. GEHRKE. 1977. Complete mass spectra of N-trifluoroacetyl-n-butyl esters of amino acids. JN Chromatogr. 141:121-144. LENT, R.W., B. SMITH, L.L. SALCEDO, B. FARIS and C. FRANZ- BLAU. 1969. Studies on the reduction of elastin. II. Evidence for the presence of a-aminoadipic-6-semialde- hyde and its aldol condensation product. Biochem. 8: 2837-45. LINSTEAD, R.P. 1932. Investigations of the olefinic acids. Part VI. Lactonization and allied additive reactions. Part I. The system a-p-acid-0- -acid- -lactone. J. Chem. Soc.. London. 115-127. LINSTEAD, R.P. and H.N. RYDON. 1933. Investigations of the olefinic acids. Part XI. The formation of lactones from some -unsaturated acids as an example of ring- chain (lacto-enoic) tautomerism. Chem. Soc..London. 580-5. LINSTEAD, R.P. and H.N. RYDON. 1934. Investigations of the oleifinic acids. Part XIV. Preparation and additive reactions of °-n-hexenoic acid. An unusual isomeric change in the three-carbon system. J;_ Chem. Soc., London. 1995-2001. LITT, M. 1961. The kinetics of ribonuclease action on cytidine-2',3'-cyclic phosphate. Biol. Chem. 236: 1786-90. 162 MacKENZIE, S.L. 1981. Recent developments in amino acid analysis by gas-liquid chromatography. In Methods in Biochemical Analysis. 27:1-88. MacKENZIE, S.L. 1984. Amino acids and Peptides. In Gas Chromatography-Mass Spectrometry. Applications in Micro biology. Chapter 5.p.157. Ed. by Odham et al. Plenum Press. MAHONY, C., J.E. BROWN, W.W. STARGEL, C.P. VERGHESE and T.D. BJORNSSON. 1984. In vitro stability of sodium nitroprusside solutions for intravenous administration. J. Pharm. Sci. 73:838-839. MALIN, E.L., R. GREENBERG, E. PIOTROWSKI, T. FOGLIA and G. MAERKER. 1984. Deamination reactions at the e-amino group of lysine. Biochemistry 23:3376. MALININA, L.V., Y.N. KOSAGANOV and Y.S. LAZURKIN. 1978. Effect of formaldehyde on enzymatic activity of RNAse. Mol. Biol. 12:356-64. MALTZ, H., M.A. GRANT and M.C. NAVAROLI. 1974. Reaction of nitroprusside with amines.J. Org. Chem. 36:363-364. MANHAS, M.S., R.S. HSIEH and A.K.BOSE. 1970. Mass spectral studies. Part VII. Unusual fragmentation of some N-trifluoroacetylamino acid methyl esters. Chem. Soc. Sec. C, 116-119. MATHESON, R.R. Jr., and H.A. SCHERAGA. 1979. Steady-state kinetic study of action of RNAse A involving confor mational change between 30°C and 40°C. Biochemistry. 18:2446-2450. MATTA, M.S. and D.T. Vo. 1986. Proton inventory of the second step of ribonuclease catalysis. Am. Chem. Soc. 108:5316-5318. MCDOWALL, D.G., N.P. KEANY, J.M. TURNER, J.R. KANE and Y. OKUDA. 1974. The toxicity of sodium nitroprusside. Br. J. Anaesth. 46:327-332. MCFADDEN, W.H., E.A. DAY, and M.J.DIAMOND. 1965. Correla tions and anomalies in mass spectra. Lactones. Anal. Biochem. 37:89-92. MCGARVEY, G.J. and M. KIMURA. 1986. Sodium nitroprusside mediated substitution of oxygen for nitrogen at satu rated carbon centers. J\_ Org. Chem. 51:3913-3915. 163 MEANS, G.E. 1984. Reductive alkylation of proteins. J. Prot. Chem. 3:121-30. MEANS, G.E. and R.E. FEENEY. 1968. Reductive alkylation of amino groups in proteins. Biochemistry 7:2192-2201. MEANS, G.E. and R.E. FEENEY. 1971. Affinity labelling of pancreatic ribonuclease. J_j_ Biol. Chem. 246:5532-5533. MEOT-NER, M. and F.H.FIELD. 1973. Chemical ionization mass spectrometry. XX. Energy effects and virtual ion temperature in the decomposition kinetics of amino acids and amino acid derivatives. Am. Chem. Soc. 95:7207-7211. MERRIFIELD, A.J. and M.D. BLUNDELL. 1974. Toxicity of sodium nitroprusside. Br. J. Anaesth. 46:324. MITTAL, C.K. and F. MURAD. 1977. Properties and oxidative regulation of guanylate cyclase. Mini-review.J. Cyclic Nucleotide Res. 3:381-391. MOHRIG, J.R., K. KEEGSTRA, A. MAVERICK, R. ROBERTS and S. SWELLS. 1974. Bimolecular displacement of nitrogen (SN2) in an alkanediazonium ion. J_;_ Chem. Soc., Chem. Commun. 780-781. MORRISON, R.T. and R.N. BOYD. 1983. Organic Chemistry. 4th ed. Allyn and Bacon, Inc. MOSS, R.A. 1971. Deamination chemistry. Chem. Enq'q News 49:28-36. MURDOCK, A.L., K.L. GRIST and C.H.W.HIRS. 1966. On the dinitrophenylation of bovine pancreatic RNAse A. Arch. Biochem. Biophys. 114:375-90. NAKAMURA, S., T. SHIN. Y. HIROKATA and A.SHIGENATSU. 1977. Inhibition of mitochondrial respiration by sodium nitroprusside and the mechanism of cyanide liberation. Br. J. Anaesth. 49:1239-44. NELSON, C.A. and J.P. HUMMEL. 1961. The inhibition of pancreatic ribonuclease by 2'-cytidilic acid. J.Biol. Chem. 236:3173-6. NOYES, W.A. 1936. n-Butylnitrite. Organic Synthesis 16:7-8. 164 OHLSTEIN, E.H., K.S. WOOD and L.J.IGNARRO. 1982. Purifica tion and properties of heme-deficient hepatic soluble guanylate cyclase: Effects of heme and other factors on enzyme activation by NO, NO-heme and photoporphyrin IX. Arch. Biochem. Biophys. 218:187-198. PAGE, I.H., A.C. CORCORAN, H.P. DUSTAN and T. KOPPANYI. 1955. Cardiovascular actions of sodium nitroprusside in animals and hypertensive patients. Circulation 11: 188-198. PARK, J.W. and G.E. MEANS. 1985. Formation of nitrosamines from sodium nitroprusside and secondary amines. New England J. Med. 313:1547-1548. PARK, J.W.and G.E. MEANS.1987.Inactivation of angiotensin- converting enzyme by sodium nitroprusside. Biochem. Biophys. Res. Commun. 145:1048-1053. PHILPOT, J.S.L.and P.A. SMALL. 1938. The action of nitrous acid on pepsin. Biochem. J. 32:542-551. PICHAT, L. 1970. The Chemistry of the Cyano Group. Chap. 13. Synthesis and uses of isotopically-labelled cyani des. p.759. Ed. by Z. Rappoport. Interscience Pub. PONSETI, I.V., S. WAWZONEK, W.E. FRANKLIN, R.E. WINNICK and T. WINNICK. 1956. Distribution of C14-labelled aminoacetonitrile in tissues of rat, metabolism and mode of elimination. Proc. Soc. Exp. Biol. Med. 93:515-9. PORTER, Q.N. 1985. Mass spectrometry of heterocyclic com pounds. Chap. 6.Lactones and cyclic anhydrides. 261-7. Second Ed. RAETZ, C.R.H and D.S.AULD. 1972. Schiff bases of pyridoxal phosphate with active center lysines of RNAse A. Biochemistry. 11:2229-2236. REYNOLDS, J.H. 1968. Acetimidation of bovine pancreatic RNAse A. Biochemistry. 7:3131-3135. RICHARDS, F.M. and H.W. WYCKOFF. 1971. Bovine pancreatic RNASe A.The Enzymes 4:647-806. 3rd ed. Acad. Press. RIDD, J.H. 1961. Nitrosation, diazotisation, and deamina tion. Quart. Rev. 15:418-441. 165 RIQUELME, P., W.E. BROWN and F. MARCUS. 1975. Modification of bovine pancreatic RNAse A with pyridoxal-5'-phos phate. Isolation and identification of derivatives. Int. J. Peptide Prot. Res. 7:379-387. ROBERTS, G.C.K., E.A. DENNIS. D.H. MEADOWS, J.S. COHEN and O.JARDETZKY. 1969. The mechanism of action of ribonuc lease. Proc. Nat. Acad. Sci. 62:1151-1157. ROBERTS, J.D. and M.C. CASERIO. 1964. Basic Principles of Organic Chemistry. Sec. 16-10. W. A. Benjamin, Inc., New York. ROBERTS, J.D. and R.H. MAZUR. 1951. Small-ring compounds. IV. Interconversion reactions of cyclobutyl, cyclo- propyl carbinyl and allylcarbinyl derivatives. J_j_ Am. Chem. Soc. 73:2509-2516. SCHLANDT, R.C., T.S. TSAGARIS and R.J.ROBERTSON. 1962. Studies on the acute cardiovascular effects of intra venous sodium nitroprusside. J_j_ Am. Cardiol. 9:51-59. SCHULTZ, K.D., K. SCHULTZ and G. SCHULTZ. 1977. Sodium nitroprusside and other smooth muscle-relaxants increase cyclic GMP levels in rat ductus deferens. Nature(London) 265:750-751. SCRIMGER, S.T., and T. HOFMANN. 1967. The involvement of the amino terminal amino acid in the activity of pancreatic proteases. II. The effects of nitrous acid on trypsin. J_;_ Biol. Chem. 242:2528-2533. SILVERSTEIN, R.M., G.C. BASSLER and T.C. MORRILL. 1981. Spectrophotometric Identification of Organic Compounds. 4th ed. John Wiley & Sons., New York. SMITH, R.P. and H. KRUSZYNA. 1974. Nitroprusside produces cyanide poisoning via a reaction with hemoglobin. J. Pharmacol. Exp. Ther. 191:557-563. SMYTH, D.G., W.H. STEIN and S. MOORE. 1963. The sequence of amino acid residues in bovine pancreatic ribonuc lease: Revisions and confirmations. J^ Biol. Chem. 238:227-234. SOBER, H.A. (Ed.) 1968. Handbook of Biochemistry. Sec. B. Amino Acids, pp. B53-B54. 166 STEPP, L.R. and L.J. REED. 1985. Active-site modification of mammalian pyruvate dehydrogenase by pyridoxal 5'- phosphate. Biochemistry 24:7187-91. STEWART, T.D. and C.H.LI. 1938. The kinetics and mechanism of a-aminoisobutyronitrile formation. Jj_ Am. Chem. Soc. 60:2782-9. STREITWIESER, A.Jr.1957. An interpretation of the reaction of aliphatic primary amines with nitrous acid. J. Org. Chem. 22:861-866. STREITWIESER, A. Jr. and W.D. SCHAEFFER. 1956. Stereo chemistry of the primary carbon. VI. The reaction of optically-active 1-aminobutane-l-d with nitrous acid. Mechanism of the amine nitrous acid reaction. Am. Chem. Soc. 79:2888-2893. SWINEHART, J.H. 1967. The nitroprusside ion. Coord. Chem. Rev. 2:385-402. SZASZ, J., R. BURNS and H. STERNLICHT. 1982. Effects of reductive methylation on microtubule assembly. J. Biol. Chem. 257:3697-3704. TANAKA, T., S. IKEDA and J.F. SCHEISS. 1983. Temperature dependent release of cyanide from sodium nitroprus side. J\ Univ. Occup. Environ. Health 5:201-206. TINKER, J.H. and J.D. MICHENFELDER. 1976. Sodium nitro prusside: Pharmacology, toxicology and therapeutics. Anaesth. 45:340-354. TOME, D., A. KOZLOWSKI and F. MABON. 1985. Carbon-13 NMR study on the combination of formaldehyde with bovine serum albumin. J^ Aqric. Food Chem. 33:449-55. TSAI, C.S., D.J. SENIOR, J.H. WHITE and J.L. PITT. 1985. Functional and structural responses of liver alcohol dehydrogenase to lysine modifications. Bioorqanic Chem. 13:47-56. TSAI, C.S., Y.H. TSAI, G. LAUZON and S.T. CHENG. 1974. Structure and activity of methylated horse liver alcohol dehydrogenase. Biochemistry 13:440-3. TSUNOKAWA, T. 1975. Hydroxyhydrazides. Japan Kokai 75 95,220. In Chem. Abstr. 84:16790c. 1976. 167 TUZEL, I.H. 1974. Sodium nitroprusside: A review of its clinical effectiveness as a hypotensive agent. Clin. Pharmacol. 14:494-503. VAN LOENEN, A.C. and W. HOFFS-KEMPER. 1979. Stability and degradation of sodium nitroprusside. Pharm. Weekblad Sci. Ed. 1:52-64. VESEY, C.J. and G.A. BATISTONI. 1977. The determination and stability of sodium nitroprusside in aqueous solu tions. Ck. Clin. Pharmacol. 2:105 VESEY, C.J., P.V. COLE and P.J. SIMPSON. 1976. Cyanide and thiocyanate concentrations following sodium nitroprus side infusion in man. Br. J. Anaesth. 48:651-660. VESEY, C.J., P.V. COLE and P.J. SIMPSON. 1982. Sodium nitroprusside and cyanide release. Br. J. Anaesth. 54: 791-792. VESEY, C.J., J.R. KRAPEZ and P.V. COLE. 1980. The effects of sodium nitroprusside and cyanide on hemoglobin function. J_;_ Pharm. Pharmacol. 32:256-261. VESEY, C.J., P.J. SIMPSON, L. ADAMS and P.V. COLE. 1979. Metabolism of sodium nitroprusside and cyanide in the dog. Br. J. Anaesth. 51:89-97. WALKER, E.J., G.B. RALSTON and I.G. DARVEY.1975. An allos- teric model for ribonuclease. Biochem. J. 147:425-33. WHEELER, O.H. and D.S. GAMBLE. 1961. Saponification of methyl-substituted a-butyrolactones. J^ Org. Chem. 26:3221-3. WHITE, E.H. and K.W. FIELD. 1975. Deamination of aliphatic amines in ethanol. Am. Chem. Soc. 97:2148-2153. WHITMORE, F.C. and D.P. LANGLOIS. 1932. Rearrangements involved in the action of nitrous acid with butylamine. J. Am. Chem. Soc. 54:3441-3447. WLODAWER, A. 1980. Studies of RNAse A by X-ray and neutron diffraction. Acta Crvstallogr. Sect. B. 36:1826-31. WLODAWER, A., R. BOTT and L. SJOLIN. 1982. The refined structure of RNAse A at 2.0A resolution. J^ Biol. Chem. 257:1325-32. 168 WLODAWER, A., M. MILLER and L. SJOLIN. 1983. Active site of RNase: Neutron diffraction study of a complex with uridine vanadate, a transition-state analog. Proc. Natl. Acad. Sci. USA. 80:3628-3631. WOLFE, S.K. and J.H. SWINEHART. 1975. Photochemistry of penta-cyanonitrosylferrate(II), nitroprusside. Inorq. Chem. 14:1049-1053.