Inhibition of the reactivation of acid-dissociated lactate dehydrogenase isoenzymes by their aminoterminal CNBr fragments

Peptides, Vol. 8, pp. 773-778. ©PergamonJournals Ltd., 1987. Printedin the U.S.A.

0196-9781/87$3.00 + .00

Inhibition of the Reactivation of Acid-Dissociated Lactate Dehydrogenase Isoenzymes by Their Aminoterminal CNBr Fragments H E I N Z D O B E L I , .1 D I E T E R G I L L E S S E N , t , W I L L I A M L E R G I E R , ? A N D R I ~ V A N D I J K * A N D G U I D O A. S C H O E N E N B E R G E R *~

*Research Department, Kantonsspital, CH-4031 Basel and tCentral Research Units, F. Hoffmann-La Roche & Co. Ltd., CH-4002 Basel R e c e i v e d 9 F e b r u a r y 1987 DOBELI, H., D. GILLESSEN, W. LERGIER, A. VAN DIJK AND G. A. SCHOENENBERGER. Inhibition of the reactivation of acid-dissociated lactate dehydrogenase isoenzymes by their aminoterminal CNBrfragments. PEPTIDES 8(5) 773-778, 1987.--The catalytic activity of lactate dehydrogenase isoenzymes (LDH) depends on their tetrameric structure. Stabilization of this quaternary structure is achieved by interaction of the N-terminal part of one subunit with the C-terminal region of the other subunit. The N-terminal peptides from pig M-LDH and H-LDH which are responsible for this stabilization were obtained by CNBr-fragmentation and purification on reversed-phase HPLC. The effect of these peptides on the formation of the quaternary structure of LDH-isoenzymes was investigated by monitoring the reconstitution of the catalytic activity after acid-dissociation. Low concentrations of the N-terminal peptides led to an increased, and high concentrations to a decreased yield of reconstituted LDH activity. The effects of these two peptides were isoenzyme specific. The 32 residue peptide derived from M-LDH showed the highest effect when tested with M-LDH as target enzyme but only a poor effect with H-LDH. On the other side the 33 residue peptide generated from H-LDH showed a moderate effect with both isoenzymes. The effects of the N-terminal LDH peptides are antagonized by the coenzymes NAD÷ and NADH. The most significant influence was observed with NAD ÷ in the M-LDH peptide-M-LDH enzyme system. Comparison of the properties of the reactivation antagonists isolated from human origin [2] with the N-terminal CNBrpeptides of LDH revealed identity in all essential properties, suggesting that the former peptides are generated by degradation of LDH. Lactate dehydrogenase isoenzymes Quaternary structure

Aminoterminal CNBr fragments

L A C T A T E dehydrogenases (LDH) of vertebrates are tetrameric enzymes. Denaturing conditions such as low pH or chaotropic agents lead to dissociation into monomers and defolding. This is paralleled by a loss of the catalytic activity. Removal of the denaturing conditions induces refolding and reassociation of the monomeric units and reconstitution of the catalytic activity [7]. Using an assay system which is based on this dissociation-association behaviour, we were able to isolate peptides from human liver and urine which inhibit the recovery of the catalytic activity. Since no effect was observed when these peptides were incubated with LDH in their native tetrameric associated form, it was concluded that the

Reactivation agonists

interaction occurs during the reconstitution process [2, 3, 14, 15, 18]. Attempts to characterize the chemical structure of these peptides revealed a blocked N-terminus and evidence of multiple species with molecular weights ranging from 300010000 daltons. As we were unable to purify these peptides to homogeneity and to obtain N-terminal sequence analyses by the conventional w a y - - d u e to the blocked N-terminus--and because we speculated that these inhibitors might be proteolytic fragments of LDH itself, we decided to prepare fragments of LDH in order to f'md out whether they would match the properties of the naturally occurring inhibitors. The published information [4-6, 13] that the N-terminal part

tPresent address: Central Research Units, F. Hoffmann-La Roche & Co. Ltd., CH-4002 Basel. ZRequests for reprints should be addressed to G. A. Schoenenberger, Medical Center Mariastein, CH-4115 Mariastein, Switzerland.

773

D()BELI ET AL.

774 ABBREVIATIONS b Dns-Cl DSIP DTE EDTA GAPDH HPLC LDH NADt NADH SDS-PAGE H-peptide M-peptide

5-dimethylamino- 1-naphthalene sulphonyl chloride delta sleep inducing peptide dithioerythritol ethylenediamine tetraacetic acid glyceraldehyde-3-phosphate dehydrogenase high-performance liquid chromatography lactate dehydrogenase /3-nicotinamide-adenine-dinucleotide, oxidized form /3-nicotinamide-adenine-dinucleotide, reduced form sodium dodecylsulfate polyacrylamide gel electrophoresis 1-33 CNBr-fragment from pig H-LDH'~ 1-32 CNBr-fragment from pig M-LDH*

*Ammoacidsequence: Ac-ATLKDQLIHNLLKEEHVPHNKITVVGVGAVGM. tAmino acid sequence: Ac-ATLKEKLIAPVAQQETTIPNNKITVVGVGQVGM.

of LDH plays a crucial role in the stabilization of the tetramer, in combination with our own findings that the isolated inhibitory peptides were also blocked at their N-terminus like LDH itself, prompted us to prepare and investigate the N-terminal CNBr-fragments of M- and H-LDH. METHOD

Enzymes and Peptides LDH-isoenzymes (E.C.1.1.1.27) were obtained from Boehringer, FRG (pig muscle and pig heart-type LDH) and from Sigma, U S A (human M4-LDH from placenta). Leucine aminopeptidase from bovine lens (E.C.3.4.11.1) was from Serva, FRG, and tested for specificity with deltasleep inducing peptide (WAGGDASGE) and thymosin a, (Ac-SDAAVDTSSEITTKDLKEKKEVVEEAEN). Peptides used as molecular weight standards or reference substances for structural analysis were purchased from Serva, FRG, or synthesized in our laboratory.

Amino Acid Analysis The instrumentation consisted of a Varian 5500 HPLC equipped with a o-phthalaldehyde post-column derivatization device and a fluorescence detector. The separation of the amino acids was performed on a 4 mm x 15 cm MicroPak ion exchange column. The peptides were hydrolyzed with constant boiling HCI (Pierce) for 24 hr at 105°C. For quantification ovalbumin from chicken was hydrolyzed and taken as standard.

Determination o f the N-Terminal Amino Acids The peptides were derivatized with 5-dimethylamino-1naphthalene-S-sulphonyl chloride (Dns-Cl) according to [19] and hydrolyzed with constant boiling HCI for 18 hr at 105°C. The liberated dansylamino acids were identified by HPLC using an ultrasphere ODS column from Beckman (USA) and 0.01 M ammonium acetate pH 4 as mobile phase. Elution was performed with an acetonitrile gradient and detection by fluorimetry (hem=368 rim, bern-----500 nm).

too u c u

-7

75 ',

o 2

,&

25

I /

0

30

60 ' r e t e n t i o n t i m e , rain

3'0

6'0

FIG. 1. Purification of the N-terminal CNBr-peptides on HPLC. Soluble fragments obtained after CNBr cleavage were separated on a 4.6x250 mm RP-18 HPLC column (Spheri 5, Brownlee Labs) using as mobile phase 1.0 M acetic acid/0.9 M pyridine in water, pH 5 and as eluens 1.0 M acetic acid/0.9 M pyridine in 4(1% (v/v) I-propanol. The gradient is indicated by the dashed line. The flow rate was 50 ml/hr. Primary amines were monitored with fluorescamine (continuous) according to [ 1]. The position of the M-peptide (a) and the H-peptide (b) is indicated by dark peaks.

Assay of the LDH-lnhibitors The assay system is based on a reversible dissociation of the tetrameric L D H molecule. The dissociation step turned out to be a prerequisite for observing any effect of the peptides [15]. In order to allow detection of the inhibitor in lower quantities the test system described in [2] was slightly modified: 0.63 ~g L D H in 25/zl 0.01 M sodium phosphate buffer, pH 7.5 (+ 1 mM DTE + 1 mM EDTA) was incubated with 25 izl 0.1 M glycine/H3PO4 buffer, pH 2.5 (+ 1 mM DTE + 1 mM EDTA) for 10 min at 22°C. This led to the dissociation and complete deactivation of LDH [7]. Reactivation was induced by adding 200 ftl of 0.25 M sodium phosphate buffer, pH 7.64 (+ 1 mM DTE + mM EDTA). The effect of the inhibitor sample was investigated by measuring the recovered L D H activity (pyruvate ~ lactate reaction) after 18 hr incubation at 22°C. The dissociation and association reaction was performed in polypropylene tubes. The potency of the inhibitor (IC 50) was expressed by the quantity of material necessary to inhibit the reactivation of L D H activity by 50% in this assay system. The influence of the coenzymes on the recovery of L D H activity in the presence of the M-peptide was performed according to [2,3].

Preparation of the N-Terminal LDH Fragment LDH from pig muscle, pig heart and human placenta were dialysed against distilled water and lyophilized, Cyanogen bromide cleavage was performed in 70% formic acid with a 150 fold excess of CNBr per r e s i d u e of methionine (24 hr/22°C). The N-terminaJ peptides were obtained in pure form by (1) dilution with H20 and lyophilization of the reaction mixture to remove solvent and reagent, (2) dissolving in 1 M acetic acid/0.9 M pyridine pH 5.0 and centrifugation, and (3) reversed-phase HPLC (Fig. 1). The fragments from pig muscle and pig heart were identified by amino acid composition (Table 1). The fragment from human placenta which

INHIBITION OF REACTIVATION OF LDH

775

TABLE 1 IDENTIFICATIONOF THE M-PEPTIDE AND H-PEPTIDE BY AMINO ACID ANALYSIS M-peptide

H-peptide

Amino Acid

Theoretical

Experiment

Theoretical

Experiment

Asp + A s n Thr Ser Hse Glu + Gin Pro Gly Ala Cys Val Met Ile Leu Tyr Phe Lys His Arg

3 2 0 1 3 1 3 2 0 5 0 2 4 0 0 3 3 0

2.5 1.9 0.2 * 3.3 ? 3.5 2 0 3 0 1.1 4.2 0 0 3.3 2.2 0

2 4 0 1 5 2 3 3 0 5 0 3 2 0 0 3 0 0

1.8 3.8 0 * 5.4 -t 3 3 0 2.8 0 1.8 2 0 0 3.2 0 0

*Homoserine was detected in both samples but not quantified. ?Proline does not react with o-phthaldehyde and escapes detection. was available only in trace amounts was identical to the corresponding pig-muscle fragment by the following criteria: (1) RP-18 H P L C , (2) lack of free N-terminal amino groups, (3) SDS-PAGE.

Preparation of the Reactivation Inhibitor From Human Origin Purification from urine. Portions containing 400--500 ml freshly collected human urine from healthy persons were lyophilized. The residue (5-15 g) was dissolved in 0.43 M formic acid and centrifugated at 1000xg for 20 min. The supernatant was chromatographed in 0.43 M formic acid on a Bio-Gel P-2 column (100 × 4 cm). The fractions containing the reactivation inhibitor were collected and lyophilized to yield 100 mg of crude inhibitor. Preparative H P L C was performed on a 9 mm x 50 cm RP-18 column (Partisil M9 10/50 ODS 2, Whatman Springfield Mill, Kent, GB) with a binary solvent system of (a) 1 M acetic acid/0.9 M pyridine p H 5.0 and (b) 60% 1 M acetic acid/0.9 M pyridine p H 5.0 + 40% l-propanol (v/v). Per run -~ 33 mg o f crude inhibitor was applied, yielding - 400/zg of enriched inhibitor. Final purification was achieved by chromatography on a 7.5x600 mm TSK-G2000 SW column (LKB, Sweden) using 1 M formic acid/pyridine p H 3 at a flow rate of 12 ml/hr. Per mg o f applied enriched inhibitor --- 50-100/zg of peptide was recovered. Isolation of the inhibitor from human liver. Crude extract was prepared by homogenization of liver slices, trichloroacetic acid precipitation, extraction with diethylether, UM 10 and U M 05 ultraf'dtration as described in [15] with the following modifications: the U M 05 desalting step was performed at p H 2.0 (instead of 8.0) and the Sephadex G-25 column was replaced by a Bio-Gel P-2 column o f identical dimensions as described in the previous section. Further purification was achieved by H P L C as described for the inhibitor from urine.

RESULTS

Identification and Characterization of the N-terminal CNBr-Fragments From Pig LDH-Isoenzymes The peptides obtained after cyanogen bromide cleavage and purification on reversed-phase H P L C (Fig. 1) were lyophilized and subjected to the inhibition assay. Two peptides having a significant effect on the reactivation of L D H were identified by amino acid analysis to be the N-terminal 1-32 fragment from M - L D H and the 1-33 N-terminal fragment from H - L D H (Table I). Further proof of the identity was obtained by end-group analysis. Dansylation followed by acid hydrolysis did not liberate any amino acid except E-Dns-Lys. This is in accordance with the presence o f acetylated N-termini.

Effect of the N-terminal CNBr-Fragments on the Reactivation of LDtt-lsoenzymes The assay system used to investigate the effect of the peptide fragments was designed to monitor the purification procedure o f the inhibitors from human liver and urine. Therefore it was optimized more for accuracy of inhibitor detection than for recovery of L D H activity [2]. Due to this the yield of recovered L D H activity dropped to 30%. As a consequence of the suboptimal reconstitution conditions the assay system allowed in addition the detection of effector peptides with an activating effect. Surprisingly both peptides showed a biphasic doseresponse curve. At low concentration an increased reconstitution and at high concentration a decreased reconstitution were observed (Fig. 4). However, the potency of the peptides was significantly different, depending on the isoenzyme used as target system. The investigation ofaU 4 combinations revealed the relationship given in Fig. 2.

776

c

DOBELI ET AL. 150

150

o

.~

a ;>

"5

O

I00

100 -rn ._]

ZZZ

:k 50

~ 50 >

mM c 150,

NAD +

FIG. 3. Reversibility of inhibited M-LDH reactivation by NAD ÷. The reactivation of dissociated M-LDH was inhibited by the presence of 100 p.g/ml (1-32) M-LDH peptide and reversibility was tested by simultaneous addition of increasing concentrations of NAD + (x). Parallel incubations with the respective NAD+ concentrations without peptide served as controls for M-LDH reactivation (100%) (O). Results are the mean-+S.E.M, of three experiments. Significant differences between incubations with and without M-peptide are indicated by: ***p<0.001, **p<0.01, *.o<0.05.

o

b

"~ 100. I r~ _.J

4

50"

200

g 0

50

~60

150

Pg inhibitor/ml

260

2~0

J

-

FIG. 2. Effect of M-peptide (x) and H-peptide ((3) on the reactivation of M-LDH (a) and H-LDH (b). The dissociation-association experiment was performed as described in the Method section using a reassociation time of 2 hr. LDH reactivation in the absence of the inhibitors was taken as 100% level. The IC 50's of the inhibitors can be derived from the X-axis. This leads to the following relationship:

"6

15o

-r C3

lOO

~

o"

Z"

"

.%~ <, ,,

A

t

o

Effector peptide

target enzyme

IC 50

M-LDH M-LDH H-LDH H-LDH

75 t~g/ml 160 p.g/ml 250/zg/ml no inhibition

o 1

M-peptide H-peptide H-peptide M-peptide

Effect o f N A D + and N A D H The coenzymes NAD + and N A D H do not influence the recovery of acid-dissociated M- or H - L D H at concentrations below 10 and 3.5 raM, respectively [2,3]. However, when they are present in combination with the M-peptide during the reconstitution process of M-LDH they are able to antagonize the inhibition. The most significant antagonistic effect was observed with NAD + in the M-peptide-LDH system (Fig. 3). On the other side NADH exerted only a marginal effect at a concentration of 3.5 raM, Higher concentrations of N A D H were not tested due to their inhibitory effect on M-LDH reconstitution. DISCUSSION

The goal of the present study was to test whether the reactivation inhibitors isolated from human u.~ae [14,18] or human liver [2, 3, 15] could be N-terminal fragments of

l

[

2

3

/,

concentration (arbitrary units)

FIG. 4. Comparison of the dose-response curve of the M-peptide with LDH-inhibitors obtained from human urine and liver at different purification steps. In order to compare the shape of the curves the absolute concentrations used were transformed for a relative scale using the intercept of the curve with the 100% reactivation level as unity (dashed line). The symbols represent: (*) M-peptide; (A) human urine inhibitor after Bio-Gel P-2 column; ((3) human urine inhibitor after RP-18 HPLC; (r-l) human urine inhibitor after TSK-G 2000 SW HPLC; (O) human liver inhibitor after RP-18 HPLC.

LDH. The comparison of these inhibitors with the N-terminal CNBr-peptide of M-LDH revealed very similar properties by the following criteria: (a) No effect on the catalytic activity was observed when the inhibitors from urine or liver were incubated with native tetrameric LDH [15]. The same is true for the M-peptide with M-LDH as target enzyme. (b) The coenzymes N A D H and NAD + are antagonists of the human inhibitors [2,3] as well as of the M-peptide. (c) During our attempts to isolate the inhibitors from urine and liver in homogenous form we observed a c e ~ degree of isoenzyme specificity [2, 14, i5, 18]. This specificity was

INHIBITION OF REACTIVATION OF LDH

777

dependent on the origin o f the starting material and the purification step. (d) Like the M- and H-peptide the inhibitors from urine and liver showed a biphasic dose-response curve. The shape of this biphasic dose-response curve was preserved during the whole purification procedure (Fig. 4). This behaviour, together with the inhomogenous chain length of the inhibitors, led to much confusion during our attempts to purify and characterize them. (e) Inhibitors obtained from urine after rigorous fractionation procedure were apparently blocked at their N-termini by the following criteria: (1) Dansylation followed by hydrolysis liberated only ~-dansyl-lysine. (2) Leucine aminopeptidase failed to liberate any detectable amino acid. (3) Three cycles of an Edman-degradation followed by dansylation and hydrolysis o f the rest-peptide did not liberate any c~-dansylamino acid. This experiment was performed to exclude the possible peptide configuration Trp-X which is not susceptible for the dansyl endgroup method or the configuration X-Pro which is not cleaved by leucine aminopeptidase. (f) In order to cut off the segment which is responsible for the different chromatographic behaviour purified inhibitor with specificity for M - L D H was cleaved with CNBr. Analysis by RP-18 H P L C and T S K - G 2000 SW HPLC and comparison with the M-peptide from pig revealed peaks at identical positions (data not shown). As a further control the corresponding human M-peptide was prepared, This 32residue peptide had the same retention time on RP-18 HPLC, despite four amino acid replacements in positions 9 (His Tyr), 16 (His Gin), 17 (Val Thr) and 19 (His Gin) [17]. From these observations we concluded that the inhibitor peptides isolated from human liver and human urine are most

likely amino terminal fragments of L D H with varying chain length. The amino terminal part of L D H represents one of the main differences between the vertebrate and the bacterial enzymes. X-ray structure analysis of the vertebrate molecule and construction of a space filling model revealed that the 20 extra residues clamp one subunit to the neighbor [9,12]. F r o m this observation it was suggested that the amino terminal " a r m region" is important in the stabilization of the tetramer [10]. This is corroborated by the fact that bacterial L D H ' s which lack this amino terminal segment dissociate much easier [11]. In the light of these facts the inhibitor mechanism may be explained as a competition of associating subunits with amino terminal fragments for the " a r m region" binding site. Analogous effects exerted by artificially generated small peptides are observed in the aggregation of fibrin monomers during blood coagulation and in the association of cytosolic enzymes to membrane proteins. In the first case a series of tri- and tetra-peptides which bear the amino acid sequence of the association domain inhibit the appearence of fibrin gels in a thrombirdfibrinogen mixture [8]. The second example describes the association of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) to the anion transport protein of erythrocytes (Band 3 protein). By binding to the polyanionic amino terminal segment of this membrane protein the catalytic activity of G A P D H is inhibited. A 23 residue peptide with the amino acid sequence corresponding to the amino terminus of the Band 3 protein also inhibits the catalytic activity of G A P D H and in addition is able to displace the enzyme from the membrane binding site. As in the example with L D H the coenzyme N A D ÷ is an antagonist of the 23 residue peptide [16].

REFERENCES 1. B6hlen, P., S. Stein, J. Stone and S. Udenfriend. Automatic monitoring of primary amines in preparative column effluents with fluorescamine. Anal Biochem 67: 438-445, 1975. 2. D6beli, H., H. J. Tobler and G. A. Schoenenberger. Isoenzyme specific inhibition of the reactivation of in vitro dissociated lactic dehydrogenase isoenzymes by two different peptides isolated from human liver. Peptides 1: 167-174, 1982. 3. D6beli, H. and G. A. Schoenenberger. Regulation of lactate dehydrogenase activity: Reversible and isoenzymespecific inhibition of the tetramerisation process by peptides, Experientia 39: 281-282, 1983. 4. Eventoff, W., M. G. Rossmann, S. S. Taylor, H. J. Torff, H. Meyer, W. Keil and H.-H. Kiltz. Structural adaptations of lactate dehydrogenase isoenzymes. Proc Natl Acad Sci USA 74: 2677-2781, 1977. 5. Girg, R., R. Rudolph and R. Jaenicke. Limited proteolysis of procine--muscle lactic dehydrogenase by thermolysin during reconstitution yields dimers. E u r J B i o c h e m 119: 301-305, 1981. 6. Holbrook, J. J., A. Liljas, S. J. Steindel and M. G. Rossmann. Lactate dehydrogenase. In: The Enzymes, Vol II, edited by P. D. Boyer. New York: Academic Press, 1975, pp. 191-292. 7. Jaenicke, R. Proteinfaltung and Proteinassoziation. Angew Chem 96: 385-454, 1984. 8. Landano, A. P. and R. F. Doolittle. Synthetic peptide derivatives that bind to fibrinogen and prevent the polymerisation of fibrin monomers. Proc N a t l A c a d Sci USA 75: 3085-3089, 1978. 9. Li. S. S.-L., R. J. Feldmann, M. Okabe and Y.-C. E. Pan. Molecular features and immunological properties of lactate dehydrogenase C4 isozymes from mouse and rat testes. J Biol Chem 258: 7017-7028, 1983.

10. Li, S. S.-L., W. M. Fitch, Y.-C. E. Pan and F. S. Sharief. Evolutionary relationship of vertebrate lactate dehydrogenase isozymes A4 (Muscle), B4 (Heart), and C4 (Testis). J Biol Chem 258: 7029-7032, 1983. 11. Mayr, U., R. Hensei and O. Kandler. Factors affecting the quaternary structure of allosteric L-lactate dehydrogenase from Lactobacillus casei and Lactobacillus curvatus as investigated by hybridisation and ultracentrifutation. Eur J Biochem ll0: 527-538, 1980. 12. Musick, W. D. L. and M. G. Rossmann. The structure of mouse testicular lactate dehydrogenase Isoenzyme C4 at 2.9 A resolution. J Biol Chem 254: 7611-7620, 1979. 13. Rossmann, M. G., A. Liljas, C.-I. Branden and L. J. Banaszak. Evolutionary and structural relationship among dehydrogenases. In: The Enzymes, Voi II, edited by P. D. Boyer. New York: Academic Press, 1975, pp. 61-102. 14. Schoenenberger, G. A. and W. E. C. Wacker. Peptide inhibitors of lactic dehydrogenase (LDH). II. Isolation and characterization of peptides I and II. Biochemistry 5: 1375-1379, 1966. 15. Schoenenberger, G. A., S. Buser, L. Cueni, H. Drbeli, D. Gillessen, W. Lergier, G. Sch6ttli, H. J. Tobler and K. Wilson. Peptides isolated from human liver with specific inhibitory effects on reassociation/reactivation of in vitro dissociated lactic dehydrogenase (LDH-M4 and -H4) isoenzymes. Regul Pept 1: 223-244, 1980. 16. Tsai, I.-H., S. N. P. Murthy and T. L. Steck. Etlect of red cell membrane binding on the catalytic activity of giyceraldehyde3-phosphate dehydrogenase. J Biol Chem 257: 1438--1442, 1982. 17. Tsujibo, H., H. F. Tiano and S. S. -L. Li. Nucleotide sequence of the cDNA and an intronless pseudogene for human lactate dehydrogenase-A isoenzyme. Eur J Biochem 147: 9-15, 1985.

778 18. Wacker, W. E. C. and G. A. Schoenenberger. Peptide inhibitors of lactic dehydrogenase (LDH). I. Specific inhibition of LDH-M4 and LDH-H4 by inhibitor peptides I and II. Biochem Biophys Res Commun 22: 291-296, 1966.

D O B E L I ET AL. 19. Zanetta, J. P., G. Vincendon, P. Mandel and G. Gombos. The utilisation of 1-dimethyl aminoaphthalene-5-sutphonyl chloride for quantitative determination of free amino acids and partial analysis of primary structure of proteins..! Chromatogr 51: 441-458, 1970.