Biochem. J. (1965) 97, 513
Lactate Dehydrogenase Isoenzymes of Human Semen
By J0RGEN CLAUSEN AND BJARNI 0VLISEN University Department of Biochemi8try and University Department of Obstetrics and Gynaecology, Rigshospitalet, Copenhagen, Denmark
(Received 16 October 1964)
1. A lactate dehydrogenase isoenzyme present in human spermatozoa and semen isolated and characterized biochemically in term of its pH for optimum activity and by means of Km values for lactate, NAD+ and NAD analogues. The results were compared with those obtained with the human heart-type and the liver-type lactate dehydrogenase isoenzymes. 2. The enzyme was characterized by its resistance to digestion with different proteolytic enzymes. The time for 50% digestion in terms of residual dehydrogenase activity was compared with
times obtained for the H4- and M4-types.
All human and other animal tissue cells contain lactate dehydrogenase isoenzymes, which are tetramers of two polypeptide chains: the H- and the M-chain. Thus five different isoenzymes with different electrophoretic mobilities and different composition occur, corresponding to the tetramers H4, H3M, H2M2, HM3 and M4 (Dewey & Conklin, 1960; Wieland & Pfleiderer, 1962). Further, in semen and extracts of testes from different manunalian species a sixth lactate dehydrogenase isoenzyme has been described with a mobility between the isoenzymes H2M2 and HM3 (Zinkham, Blanco & Kupchyk, 1963; Blanco & Zinkham, 1963; Goldberg, 1963). This isoenzyme (LDHX) has not been characterized in detail, and it is therefore the aim of this study to provide data on its enzymological properties.
Extracts of human liver and human heart tissue were prepared as described by Gerhardt, Clausen, Christensen & Riishede (1963). These were used for isolation of LDH* isoenzymes H4 and M4 as described below. Unless otherwise stated the chemicals used were those of highest purity obtained from British Drug Houses Ltd.
METHODS Estimation of LDH activity. For routine assay the reaction mixture contained 1001J. of suitably diluted sperm extract, 1 0mm-pyruvate, 0 33mm-NADH and
50mm-phosphate buffer, pH7-5, in a final volume of 3ml. (Bergmeyer, Bernt & Hess, 1962; Clausen & Gerhardt, 1963). Changes in extinction at 366 m,u were measured in silica cells of 1 cm. path length at 250 in an Eppendorf or a Vitatron photometer equipped with a mercury lamp as light-source. The concentration of enzyme was adjusted to give a change in extinction of about 0 030/sec. On the basis of the extinction change at 25, the dilution factor of the sample, and a molar extinction coefficient of NADH at 366m,u of 3.3cm.2/,Lmole (Hohorst, 1956), the number of LDH enzyme units present in lml. of the original sample was expressed as jtmoles of substrate transformed/min. From the total protein content or total DNA P content of the sample the specific LDH activity was expressed as ,umoles of substrate transformed/min./mg. of protein or per ,ig. of DNA P. The effect on LDH activity of variations in pH and in concentrations of substrate and of NAD+ (NADH) were estimated in the above system or by dehydrogenation of lactate in the presence of NAD+. The lactate reduction system was also used for estimation of the Km values of the individual isoenzymes. For determination of Km values 0-20M-L-lactate and 8-3mM-coenzyme were used. These values are higher than the optimum ones estimated as described below. Buffers used in lactate systems of pH> 8-0 were made with BOimM-glycine and NaOH.
* Abbreviation: LDH, lactate dehydrogenase. Bioch. 1965, 97

MATERIALS

Human semen was obtained from the Laboratory of The Copenhagen Health Insurance Society. The samples were collected in a glass tube during coitus interruptus or by masturbation. The samples were stored until use at 4. Only samples free of leucocytes and bacteria were used. Each sample was treated singly as follows: the spermatozoa were isolated by centrifuging for 15min. at 2000g and 40 followed by five successive resuspensions of the cells in an equal volume of 0.9% NaCl. The isolated spermatozoa were mixed with 0-5vol. of 0 05M-KH2PO4-Na2HPO4 buffer, pH7-5, containing 5% of Triton X-100 (alkylphenoxypolyethoxyethanol; Rohm and Haas Co. lot no. 5330) in a plastic centrifuge tube fitted with a closely fitting Teflon piston. Disintegration was carried out at 00 by rotating the piston for 5min. at 40rev./min. Under these conditions 70% of the spermatozoa were destroyed. After centrifugation at 10OQOg for 20min. the supernatant fluid was stored at 4 until required. When extracts were required for micro-electrophoresis Triton X-100 was not used because it interferes with the formazan staining. 17

J. CLAUSEN AND B. 0VLISEN
Reactions of the isoenzymes with NAD analogues
Protein content. Protein of sperm extracts and sperm plasma was determined by the Lowry method as modified by Lous, Plum & Schou (1956), with tyrosine used as standard. Determination of DNA. Sperm extracts and sperm plasmas were analysed as described by Ceriotti (1952) and modified by Glick (1963). The phosphorus content of the DNA standard was determined by the method of Fiske & Subbarow (1925).
Microelectrophoresis on agar gel
Preparation of agar-get 8lides. The microscope slides were first coated with a 1mm. layer of 1% agar (Difco Special Noble Agar) dissolved in water. This layer was dried before the electrophoretic agar-gel medium was applied, as a 1 mm. homogeneous layer, as described by Wieme (1959). The agar coat on the glass surface prevented migration of the protein solution between the glass surface and the agar gel. Distortion of the agar layer during the electrophoretic run was prevented by sealing the trough with a layer of melted electrophoretic medium. Etectrophoretic procedure. Electrophoresis of seminal plasma and of extracts was performed as described by Wieme (1959) at pH8-6 in 0-05M-barbitone buffer containing 1% of Difco Special Noble Agar. The running time was 28min. The separated LDH isoenzymes were made visible by formazan formation (Gerhardt et al. 1963). The relative proportions of enzyme units in the different isoenzyme bands were estimated by scanning the slides at 546m,u in the Vitatron photometer equipped with a scanning device, an automatic recorder and an integrator. Beer's law was valid in the scanning procedure. Thus for each isoenzyme in an extract a straight-line relationship was observed between the area of the peak in the recorder tracing and the reciprocal of the dilution, provided that the number of enzyme units applied to the electrophoretic slide exceeded 0-02unit. The number of enzyme units in an individual isoenzyme band was calculated by multiplying the fractional activity in the band by the total units applied to the slide. Separation of LDH isoenzymes. A lmm.-broad trough across the coated slide was made by forcing a lmm. thick filter paper (Whatman no. 17) through the agar. The extract (50p1.) was then placed in the trough. Eight to ten slides were run simultaneously and one slide was stained for LDH activity. On the basis of the distribution of the LDH bands in this slide the remaining agar strips were cut into pieces each containing only one LDH band. The separated isoenzymes were eluted from the agar gel with 0.05M-phosphate buffer (pH7.5) containing 5% of Triton X-100. Buffer (lvol.) was added to 2vol. of pooled agar pieces and the mixture was homogenized as described above. After centrifugation for 20min. at 10000g the supernatant solution was examined for enzymic homogeneity by agargel electrophoresis. By this procedure the predominant isoenzyme of human heart tissue (H4), of human liver tissue (M4) and of human spermatozoa (LDHX) were isolated. The yields were 48-6% for the H4-form, 21-7% for the M4-type and 53.0% for the LDH. By this electrophoretic isolation the specific LDH activities were increased from 1.33units/mg. of protein to 7-18units/mg. of protein for H4 and from 2.06units/mg. of protein to 7-90unitsfmg. of protein for M4.

The activities of the isoenzymes were determined in the presence of 3-thionicotinamide-adenine dinucleotide, 3pyridinealdehyde-adenine dinucleotide and nicotinamidehypoxanthine dinucleotide as described by Cahn, Kaplan, Levine & Zwilling (1962) and Kaplan & Ciotti (1961). These analogues, which were all used in the oxidized forms, were obtained from Pabst Laboratories, Milwaukee 5, Wisconsin, U.S.A. They were used at the same concentration as that of NAD+ (853mM). The following solutions were mixed in the cuvette: 2-85ml. of 0-05M-glycine-NaOH buffer (pH9-00) containing 0-20M lithium lactate, 50bu. of glycine buffer containing 0-5m-NAD analogues and 100,l. of electrophoretically purified isoenzyme solution.

Michaelis constants

The constants were determined for lactate in the presence of about 10 times the saturation limit for NAD+ and NAD analogues (8-3mM). The saturation limits were estimated by measuring the minimum concentration of coenzyme (or substrate), which, under the experimental conditions used, was required for maximum LDH activity. Further, constants for the NAD+ and NAD analogues were determined in the presence of 0-20M-lithium lactate. Km values were calculated by the method of Lineweaver & Burk (1934). Km values for lactate are results of duplicate analyses. The deviations from the means were at most + 10%. Km values for the nucleotides are results of single determinations.
Treatment with proteolytic enzymes To ivol. of isolated enzyme fraction was added an equal volume of a solution of proteolytic enzyme. The mixtures were incubated at 370 and the LDH activity was estimated at zero time and at every 10 to 15min. as long as activity was present. Proteolytic digestion was performed with trypsin, papain, peptidase, carboxypeptidase and a-chymotrypsin. Digestion with trypsin (Novo Terapeutisk) was performed with a solution containing 250mg. of trypsin dissolved in 10ml. of 0-05M-phosphate buffer (pH7.5) containing 5% of Triton X-100. Digestion with papain was performed by using a solution of the proteolytic enzyme containing 7.88mg. of cysteineHCI, 18-6mg. of EDTA and 25mg. of papain (Merck 1:350)/lOml. of phosphate buffer (0-05M; pH7.5). Digestion with a peptidase from hog intestinal mucosa (Sigma lot no. 93 B-1960) was performed with a solution containing 7-25ml. of 0-05m-veronal buffer (pH8-6)+ 2-5ml. of O-OlM-Co(NO3)2+0-500ml. of peptidase in veronal buffer as described by Wiist (1962a). Carboxypeptidase was used as a solution containing 4-8ml. of 0-05M-tris buffer, pH7-8, 0-lOOml. of 0-25M-CaCl2 solution and 0-lOOml. of carboxypeptidase from bovine pancreas (Sigma lot no. 23 B-1590; 50mg. of protein/ml.) (Wiust, 1962b). Digestion with a-chymotrypsin was carried out at pH7-25 as a compromise between the pH optimum of 9-0 and the optimum of stability at pH4-0. a-Chymotrypsin (50mg., Sigma lot no. 113 B-0310), with an activity of 9-Ounits/mg., was dissolved in 2ml. of 0-05M-phosphate buffer, pH7-25, containing 5% of Triton X-100.

VOl. 97

SEMEN LACTATE DEHYDROGENASE ISOENZYMES

RESULTS

LDH activities of human 8emen, seminal pla8ma and of extracts of i8olated 8permatozoa
515 DNA phosphorus or 0 061-units/mg. ofprotein. The values for seminal plasma were 0-00670.148unit/,ug. of DNA phosphorus or 0-001590.255unit/mg. of protein.
The LDH activity of semen is in the range 0 9714.55units/ml. (mean of 25 samples: 3.68). The Properties of the LDHX isoenzyme from 8permatozoa LDH activity of lml. of total extract of isolated Stability of the enzyme. When the isolated LDHX
spermatozoa from 1 ml. of total sperm varied from 0-22 to 1.82units/ml. in 35 samples. In another series of experiments (30 samples) the corresponding values for seminal plasma were in the range 0 974.37units/ml. Electrophoresis of semen, seminal plasma and spermatozoal extracts revealed in all cases the intermediate LDH isoenzyme situated between H2M2 and H1M3. This fraction accounts for 6.8-57 % of total activity of semen (mean %), but 80-100% of activity of spermatozoal extract. The specific LDH. activities of semen were in the range 0.0036-0.566unit/,tg. of DNA phosphorus or 0.034-1.18 units/mg. of protein. The corresponding values of extracts were 0.011-0.132unit/pg. of
fraction was stored at - 20 enzyme activity disappeared within 10 days. At 40, however, the activity decreased only 40% in 10 days. Effect of pH on activity. The pH for maximum activity was 7-5 for the forward reaction and 8-for the back reaction. The pH optima are distinctly different from those of isoenzyme H4, but not significantly so from those of M4 (Table 1). Relation8hip of activity to 8ub8trate concentration. The optimum concentration of pyruvate in the forward reaction of LDH1, with NADH as coenzyme, is 09 mm. This value is higher than those for H4 and M4. The optimum concentration of lactate for the back reaction is 0-16m with NAD+
Table 1. Kinetic data obtained at 25 on lactate dehydrogena8e i8oenzymWe from human ti88Ue8 (A) Standard conditions for forward reaction: loopJ. of suitably diluted LDHX from human spermatozoa, M4 from human liver or H4 from human heart, 2-850ml. of 50mM-phosphate buffer (pH7.5), containing 1OmMpyruvate, and 0 33mm-NADH in a total volume 30Oml. In the standard assay of the back reaction the enzyme was mixed with 2-850ml. of 5OmM-glycine buffer (pH9 0) containing 0 2M-lithium lactate and 0 83mM-oxidized nucleotide. The saturation limit is defined as the minimum concentration of substrate or coenzyme, which, under the present experimental conditions, is required for maximum LDH activity. (B) Km (nucleotide) was determined in glycine-NaOH buffer (pH9.0) and 0 2M-lactate. (C) Km (lactate) was determined in glycine-NaOH buffer (pH9.0) and 8 3mM-coenzyme. (A) LDH1 H4 M4 10-75 8-75-9-00 9-75 Optimum pH for lactate oxidation (back reaction) 7-25 8-00 7-50 Optimum pH for pyruvate reduction (forward reaction) 0-16M 0*20M 0.16M Saturating conen. of lactate in the assay method 0.5mM 0 9mM 0-25mM Saturating concn. of pyruvate in the assay method Nucleotide saturation limits (mm)

NAD+ (back reaction) 3-Thionicotinamide-adenine dinucleotide (back reaction) 3-Pyridinealdehyde-adenine dinucleotide (back reaction) Nicotinamide-hypoxanthine dinucleotide (back reaction) NADH (forward reaction) (B) NAD+ (back reaction) 3-Thionicotinamide-adenine dinucleotide (back reaction) 3-Pyridinealdehyde-adenine dinucleotide (back reaction) Nicotinamide-hypoxanthine dinucleotide (back reaction) Nucleotide: NAD+ (back reaction) 3-Thionicotinamide-adenine dinucleotide (back reaction) 3-Pyridinealdehyde-adenine dinucleotide (back reaction) * Broad maximum. t Narrow maximum.
0*83 0*83 0.83* 1*67 0.83*

0.50 0-21

1*25 0*83

0-83 0-83

1*25 0*17t

Km (nucleotide) (mM)

0.83* 0*83 0.33*

0*109 0*125

0.50 2-9 4-1 15

Km (lactate) (mm)

as coenzyme, identical with that of M4 but lower Table 2. Inactivation of isoenzymes H4, M4, LDHX than the value for H4 (Table 1). A distinct substrate by proteolytic enzymes inhibition was found in the forward, but not in All reactions were and rates of the back, reaction. With the NAD analogues as expressed as the timefirst-order inactivation. proteolysis are for 50% Experimental coenzymes for the back reactions the saturation conditions are given in the text. limits were in all cases below 0 16M-lactate. Time (min.) Relationship of activity to nucleotide concentration. ' The saturation limits of the different nucleotides H4 M4 LDH1 were identical (0.83mM) in the back reaction of Trypsin 22 LDH,. The value is also identical with those found Carboxypeptidase 114 for H4 and M4 with the exception of higher values a-Chymotrypsin 13 found for nicotinamide-hypoxanthine dinucleotide Papain 69 and NAD+ as coenzymes for M4 (Table 1). The 3-pyridinealdehyde-adenine dinucleotide saturation curve was characterized by a broad maximum because higher concentrations of the analogue gave 10 subjects showing aspermia although the other rise to inhibition of the back reaction. This pheno- five LDH isoenzymes were invariably present (B. menon was not observed in the back reactions 0vlisen, unpublished work). It is thus probable catalysed by the other nucleotides, but was even that LDHX originates from sperm cells. Zinkham more pronounced for NADH in the forward et al. (1963) found that LDH. was the predominant LDH isoenzyme in human spermatozoa, and this reaction (Table 1). Michaelis constants of nucleotides and of lactate in is confirmed by the present finding that 80-100% the back reaction. These are given in Table 1. The of LDH activity of sperm cells is of the LDH. type. values of Km (nucleotide) of LDH. in its reaction with lWTe are not able to say whether the high LDHX NAD+, 3-thionicotinamide-adenine dinucleotide content of seminal plasma is due to outward diffuand 3-pyridinealdehyde-adenine dinucleotide are sion of the isoenzyme from sperm cells or to sponlower than those of H4 and of M4 in contrast with taneous destruction of the cells. The LDHX isoenzyme from semen or testes conthe value for nicotinamide-hypoxanthine dinucleotide+, which is similar to that of M4. The Km (lactate) sists of sub-units different from the H- and M-chains values of LDH. are higher than those of H4 and of present in LDH isoenzymes from other mammalian M4 when NAD+ is coenzyme, but intermediate tissues (Markert & M0ller, 1959; Zinkham et al. between the values of the two other isoenzymes 1963). The lactate oxidation rates catalysed by when 3-thionicotinamide-adenine dinucleotide or LDH. in the presence of different nucleotides differ 3-pyridinealdehyde-adenine dinucleotide are co- from those obtainable with LDH fractions made up from H- or M-sub-units or both (Zinkham et al. enzymes in the back reaction. Relationship of activity to proteolytic treatment. 1963). Our results confirm and extend these The inactivation of LDH activity by proteolytic reports. In terms of pH optima, and of the lactate digestion proceeds logarithmically. With the ex- concentration required to saturate the enzyme, ception of the peptidase, which did not inactivate LDHX seems more related to M4 than to H4. The saturation limits and the values of Km (nucleotide) any of the isoenzymes, the inactivation proceeded until enzyme activity was eliminated. It was not that we find show significant differences in the possible by means of electrophoresis during the kinetic properties of H4, M4 and LDHX. Values proteolysis to detect any active sub-units with for Km (nucleotide) published by Nisselbaum, Packer & mobilities different from the isoenzymes under- Bodansky (1964) are not comparable with ours going digestion. The t j values for the digestion of because these authors used different pH, substrate LDH., H4 and M4 LDH isoenzymes by the proteo- and temperature. Values for Km (py-vate) are not lytic enzymes are given in Table 2. The rate of given in this paper because of the inhibition inactivation of LDH1 in the trypsin, carboxy- obtained with higher pyruvate concentrations and peptidase and papain experiments is intermediate because of the depression of LDH.activity at higher between the rates for the H4- and M4-isoenzymes. NADH concentrations. However, such Km values On the otherhand oc-chymotrypsin inactivates LDHx for H4 and M4 have been presented by Dawson, Goodfriend & Kaplan (1964) and by Nisselbaum more rapidly than either H4 or M4. et al. (1964). We, like Pesco, McKay, Stolzenbach, Cahn & DISCUSSION Kaplan (1964), have been unable to split the isoThe LDH1 isoenzyme is present in seminal enzymes into sub-units either by freezing solutions plasma and in extracts of spermatozoa. However, containing high concentrations of salts (Zinkham this isoenzyme could not be detected in semen from et al. 1963) or by means of/-mercaptoethanol (Fritz

Vol. 97

& Jacobson, 1963). We attempted to obtain enzymically active sub-units from isolated isoenzymes by proteolytic digestion, but electrophoresis of isoenzymes during digestion revealed no such active fragments. The different rates of inactivation of isoenzymes by trypsin, carboxypeptidase, a-chymotrypsin and papain suggest that LDH. is structurally distinct from H4 and M4.
These investigations have been supported by King Christian X's Fund, The Fund for Promotion of Medical Science and from The Technical Research Fund, Copenhagen, Denmark. We are grateful to Dr R. Hammen, of the Laboratory of The Copenhagen Health Insurance Society, for obtaining supplies of semen.

REFERENCES

Bergmeyer, H.-U., Bernt, E. & Hess, B. (1962). In Methoden der enzymatischen Analyse, p. 736. Ed. by Bergmeyer, H.-U. Weinheim: Verlag Chemie. Blanco, A. & Zinkham, W. H. (1963). Science, 139, 601. Cahn, R. D., Kaplan, N. O., Levine, L. & Zwilling, E. (1962). Science, 136, 962. Ceriotti, G. (1952). J. biol. Chem. 198, 297. Clausen, J. & Gerhardt, W. (1963). Acta neurol. scand. 39, 305. Dawson, D. M., Goodfriend, T. L. & Kaplan, N. 0. (1964). Science, 143, 929. Dewey, M. M. & Conklin, J. L. (1960). Proc. Soc. exp. Biol., N.Y., 105, 492.
Fiske, C. H. & Subbarow, Y. (1925). J. biol. Chem. 66, 375. Fritz, P. J. & Jacobson, K. B. (1963). Science, 140, 64. Gerhardt, W., Clausen, J., Christensen, E. & Riishede, J. (1963). Acta neurol. scand. 39, 85. Glick, D. (1963). Quantitative Chemical Techniques of Histo- and Cytochemistry, vol. 2, p. 222. New York: Interscience Publishers". Wiley and Sons Inc. Goldberg, E. (1963). Science, 139, 602. Hohorst, H. J. (1956). Biochem. Z. 328, 509. Kaplan, N. 0. & Ciotti, M. M. (1961). Ann N.Y. Acad. Sci. 94, 701. Lineweaver, H. & Burk, D. (1934). J. Amer. chem. Soc. 56, 658. Lous, P., Plum, C.-M. & Schou, M. (1956). Nord. Med. 55, 693. Markert, C. L. & M0oler, F. (1959). Proc. nat. Acad. Sci., Wash., 45, 753. Nisselbaum, J. S., Packer, D. E. & Bodansky, 0. (1964). J. biol. Chem. 239, 2830. Pesco, A., McKay, R. H., Stolzenbach, F., Cahn, R. D. & Kaplan, N. 0. (1964). J. biol. Chem. 239, 1753. Wieland, T. & Pfleiderer, G. (1962). Angew. Chem. 74, 261. Wieme, R. J. (1959). Agar-Gel Electrophoresis (Thesis). Brussels: Arscia Uitgaven N.V. Wuist, H. (1962a). In Methoden der enzymatischen Analyse, p. 824. Ed. by Bergmeyer, H.-U. Weinheim: Verlag Chemie. Wiust, H. (1962b). In Methoden der enzymatischen Analyse, p. 828. Ed. by Bergmeyer, H.-U. Weinheim: Verlag Chemie. Zinkham, W. H., Blanco, A, & Kupchyk, L. (1963). Science, 142, 1303.

Proc. Nat. Acad. Sci. USA Vol. 70, No. 6, pp. 1790-1794, June 1973
Aminoacid Sequence of Dogfish M4 Lactate Dehydrogenase
(glyceraldehyde-3-phosphate dehydrogenase/lactic acid yeast and liver alcohol dehydrogenases)
SUSAN S. TAYLOR, SUSANNA S. OXLEY, WILLIAM S. ALLISON, AND NATHAN 0. KAPLAN
The Department of Chemistry, University of California, San Diego, La Jolla, Calif. 92037
Contributed by Nathan 0. Kaplan, March 19, 1973
About 80% of the aminoacid sequence of ABSTRACT dogfish (Squalus acanthius) M4 lactate dehydrogenase (EC 1.1.1.27) has been elucidated. Several sequence homologies with peptides from pig H4 and pig M4 lactate dehydrogenase are identified. Histidine 195 is homologous to the essential histidine residue in pig H4 lactate dehydrogenase. Similarities in the sequence around the "essential" cysteine residue of lactate dehydrogenase, glyceraldehyde-3phosphate dehydrogenase, and yeast and liver alcohol dehydrogenases are delineated.
Lactate dehydrogenase is a tetrameric, NAD-requiring enzyme. The subunits are identical and have a molecular weight of 36,000 (1); no cooperative effects of the subunits have been demonstrated (2, 3). At least two types of lactate dehydrogenase are found in significant amounts in most species;
the M4 isozyme predominates in skeletal muscle and the H4
isozyme in more aerobic tissues such as heart and kidney cortex (4, 5). The two types are distinguishable from one another by electrophoresis, aminoacid composition, immunological methods, and kinetic properties (5-7). They show significant differences in their response to substrate inhibition, and this difference may be the basis of distinct physiological roles (5, 8). Many lactate dehydrogenases have been characterized to varying extents, and it appears that homologous isozymes of the same enzyme type from various species are more similar in their properties to each other than are the two corresponding forms in any one given species. Nevertheless, mixed hybrids of M and H subunits are observed in vivo (4, 5) and can be readily formed in vitro (9) suggesting a considerable degree of structural similarity between the two proteins. The sequence reported here has been elucidated for the M4 isozyme from dogfish (Squalus acanthius) muscle. This same dogfish M4 isozyme has been characterized by x-ray diffraction (10-13), and the correlation of the aminoacid sequence with the crystallographic structure will be discussed in more detail (33). The enzymatic activity of lactate dehydrogenase is inhibited by most sulfhydryl reagents, and there are four functionally significant cysteine residues per tetramer (14). A dodecapeptide containing this cysteine residue has been sequenced from several lactate dehydrogenases (1, 15, 16). The sequences of the N-terminal 18 residues of dogfish M4 lactate dehydrogenase (1), of several small peptides from pig H4 lactate dehydrogenase (17), and of a peptide containing an essential histidine residue have been reported (11, 18). Preliminary evidence for involvement of one, and possibly as many as three, arginine residues (19) and of one tyrosine residue (20, 21) per subunit in enzymatic activity has been presented.

MATERIALS AND METHODS

dehydrogenase was purified by the procedure of Pesce et al. (6). An aminoacid analysis of this protein is given in Table 1. Before proteolytic digestion, the homogeneous protein was carboxymethylated in 10 mM Tris HCl (pH 8.3)-0.5 mM EDTA-8 M urea with a 2-fold excess of [14C]iodoacetic acid for 4 hr at room temperature (240), followed by extensive dialysis. The sequence was determined primarily from peptides isolated after tryptic and chymotryptic digestion. Thirty-five tryptic peptides were purified and sequenced out of an expected 40 pflptides; the remaining peptides were partially sequenced. The carboxymethylated lactate dehydrogenase was also treated with maleic anhydride, which specifically blocks the lysine residues (22). Tryptic digestion of this maleated protein yields 10 peptides, seven of which were completely sequenced. Finally, the peptide fragments resulting from treatment with cyanogen bromide (23) were characterized. Since there are 11 methionine residues per subunit, 12 peptides should be obtained after cyanogen bromide treatment; the sequence of 10 of these peptides was elucidated. All sequencing was done by the dansyl-Edman procedure (24). Dansyl-amino acids were identified by thin-layer chromatography on 4 X 4 inch polyamide sheets (25). Aminoacid compositions were determined on a Beckman model 120C amino acid analyzer. Peptides were purified by numerous methods, including both Sephadex and ion exchange column chromatography, paper electrophoresis at pH 6.5, 1.9, 3.5, and 8.9, and paper chromatography in butanolacetic acid-water-pyridine 15:3:12: 10. Amide identification was based on electrophoretic mobilities at pH 6.5 (26). The Numbering of the Polypeptide Chain is that proposed by Rossmann et al. (11) based on the 2.5-A map of the apoenzyme. Subsequent improvement of the crystallographic resolution (13) as well as sequence information has shown that residues 21, 82, and 104 in the amino terminal region should be deleted and an extra residue should be inserted at position 245. The total number of residues is therefore somewhat less than the 331 originally proposed. Nevertheless, until the entire sequence is complete, this numbering scheme shall be adhered to. Of the tentative total of 329 aminoacid residues per subunit, 263 are reported here in three portions: the amino terminal region (1-115), residues 134-205, and the carboxy terminal region (253-331).

RESULTS AND DISCUSSION

Purification and Peptide Isolation. Dogfish M4 lactate
Amino Terminal Region (1-115). The sequence of this region was established from proteolytic digests (Fig. 1). The 13 tryptic peptides in this region have been sequenced completely except for the region in parentheses [residues 29-33],
Proc. Nat. Acad. Sci. USA 70

(1973)

Lactate Dehydrogenase Sequence
ACETYLTHR- ALA- LEU- LYS- ASP- LYS- LEU- ILE- GLY- HIS- LEU- ALA- THR- SER- GLN- GLU- PRO- ARG- SER- TYR-
ASN- LYS- I LE- THR- VAL- VAL- GLY(Cys,

TM-1 C-330

Ala,Asx,Val,GIy )MET- ALA- ASP- ALA- ILE- SE R- VAL- LEU- MET- LYS- ASP- LEU- ALA- ASP- GLU- VAL- ALA- LEU- VAL- ASP- VAL- MET- GLU- ASP- LYS-
T-6,7 | |CNBr-LEU- LYS- GLY- GLU- MET- MET- ASP- LEU- GLU- HIS- GLY- SER- LEU- PHE- LEU- HIS- THR- ALA- LYS- ILE- VAL- SER- GLYLYS- ASP-TYR- SER- VALI8iT-8 Ii T-9,10 TM- CNBr-5 CNBr-4 I 110 SER- ALA- GLY- SER- LYS- LEU- VAL- VAL- ILE- TH R- ALA- GLY- ALA- ARG- GLN- GLNGLU- GLY- GLU- SER- ARG- LEU- ASN- LEU- VAL- GLN- ARGT-13 T-11 T-12

|CNBr- 2 C-11

FIG. 1. Aminoacid sequence of the N-terminal region of dogfish M4 lactate dehydrogenase. The numbering of the residues is as indicated in the text; the deletions are designated as blank spaces at positions 21, 82, and 104. Type of digest from which the peptide was isolated; T = tryptic; C = chymotryptic; CNBr = cyanogen bromide; TM = tryptic peptide from maleated protein. Positions 16 and 17 have been modified from the original reported sequence (1).
where only the composition has been established. Four CNBr fragments from this region were also characterized, and with these peptides, eight of the 12 overlapping regions of the tryptic peptides were determined. Identification of the acetyl substitution of the amino terminal threonine was confirmed by nuclear magnetic resonance spectroscopy of the peptide acetyl-Thr-Ala-Leu*. This first third of the molecule is involved primarily in coenzyme binding. In addition, the section extending from residue 98-114 forms the loop that differs significantly in Conformation in the crystal structures of the apoenzyme and ternary complex (11). It is interesting to note that two of the tryptic peptides in this portion of the polypeptide chain, T-12 and T-13, are homologous to peptides isolated from pig H4 and M4 lactate dehydrogenase (17) (Fig. 2). Aminoacid Residues 134-205. This region (Fig. 3) appears to contain many of the residues important 'for lactate and pyruvate binding. An unusual feature of the total sequence is the uneven distribution of arginine residues. Each subunit contains only nine arginine residues, and this fragment contains four of these arginines in a cluster of consecutive arginine tryptic peptides. The order of these fragments was established by a single cyanogen bromide peptide. The "essential" cysteine residue is found in this region and from x-ray crystallographic data was identified as residue 165 (11). The tryptic peptide containing this cysteine has been isolated from several different lactate dehydrogenases, including both M4 and H4 isozymes, and in all cases the sequence has been closely conserved (15, 16). If a single deletion is assumed to have occurred in the lactate dehydrogenase sequence, a reasonable comparison can be made of the sequence of this peptide with the sequence sumrounding the "essential" cysteine in yeast alcohol dehydrogenase (27), horse-liver alcohol dehydrogenase (27, 28), and glyceraldehyde-3-phosphate dehydrogenase (29, 30) (Fig. 4). If, in addition, consideration is given to those amino acids whose codons may be related by a single nucleotide base change in the mRNA sequence (16), the similarities in the sequences of these peptides become even more apparent. The correlation of functionally similar amino acids is also reasonably good. In lactate dehydrogenase the "essential" cysteine is not one of the most reactive cysteines in the molecule (16), and, sur*

prisingly, correlation of the sequence with the crystal structure indicated that this cysteine does not participate directly in either substrate or coenzyme binding (33). In contrast, the essential cysteine of glyceraldehyde-3-phosphate dehydrogenase is extremely reactive and forms an S-acyl-enzyme
intermediate (31). Woenckhaus et al. inactivated pig H4 lactate dehydrogenase with 3-(2-bromo-1-[14C]acetyl) pyridine (18). From this inactivated enzyme they isolated a single peptide containing a modified radioactive histidine derivative. Adams et al., on the basis of the sequence of this peptide, subsequently identified this histidine as residue 195 (11). In Fig. 5, the Woenckhaus peptide is compared with the corresponding sequence from dogfish M4 lactate dehydrogenase. This, too, is a region of the molecule where the sequence in these two proteins is highly conserved. Carboxy-Terminal Sequence (253-331). Nine tryptic peptides were isolated and sequenced from the C-terminal region of the protein (Fig. 6). Chymotryptic peptides established the overlapping portion of six of the peptides. Three cyanogen bromide fragments accounted for the final 69 residues of this region, and with these cyanogen bromide peptides the order of the remaining tryptic peptides was determined. This sequence is compared with the known C-terminal sequence of pig H4 lactate dehydrogenase (32) (Fig. 7). Eleven of the final 14 residues are identical. An insertion in
TABLE 1. Aminoacid composition in mol per subunit (36,000 molecular weight) of carboxymethylated dogfish M4 lactate dehydrogenase
LYSINE HISTIDINE ARGININE

CM-CYSTEINE

29.6 ASPARTIC ACID 10.7 THREONINE 8.8 SERINE GLUTAMIC ACID PROLINE GLYCINE ALANINE VALINE METHIONINE 5.7 ISOLEUCINE LEUCINE TYROSINE PHENYLALANINE
33.7 13.0 26.2 26.0 12.0 24.9 19.7 33.1 10.5 19.0 33.1 7.1 6.6
Morelli, R. & Allison, W. S., unpublished results.
This composition is an average of multiple hydrolysates. Duplicate samples each containing 200 ,ug of protein were hydrolyzed under reduced pressure at 105 in 6 N HOC, for 24, 48, and 72 hr.
Biochemistry: Taylor et al.
Proc. Nat. Acad. Sci. USA 70 (1978)

DOGFISH M4

-ARG -GLN -GLN -GLU -GLY -GLU -SER -ARG -LEU -ASN -LEU -VAL -GLN -ARG -ARG GLN -GLN -GLU -GLY -GLX -SER -ARG
LEU -ASN -LEU -VAL -GLN -ARG I GLN -GLN -GLU -GLY -GLX -SER -ARG LEU -ASN -LEU -VAL -GLN -ARG i ---I i

PIG M4

FIG. 2. Possible sequence homologies in the peptide fragment beginning at residue 101. The positioning of the pig heart and pig muscle peptides (17) was established here by alignment with the dogfish M4 sequence.
LYS- GLU- LEU- HIS- PRO- GLU- LEU- GLY-THR-ASP- LYS- ASN- LYS- GLN- ASP-TRP- LYS- LEUIFT
CNBr SER-GLY- LEU-PRO-MET-HIS-ARG- ILE- ILE-GLY- SER- GLY--CYS-ASN- LEU- ASP-SER-ALAT
ARG-PHE-ARG-TYR-LEU-MET-GLY-GLU-ARG-LEU-GLY-VAL-HIS-SER (Cys, Leu, Val, lie.
Gly)TRP-VAL- ILE-GLY-GLN-HIS-GLY-ASP-SER-VAL-PRO-SER-VAL-TRP-MET(Asx.
FIG. 3. Aminoacid sequence of residues 134-205 in dogfish Mt lactate dehydrogenase.

IE3 lIE GLYSERGLY-

CYS ASP CYS AS

AG -PHE A

rvLkIlILE ILVSERG Y| i IE -r [[ ]-[S:]-R1{ I-SER ATH LEE -ILE LLYSJ -LM~iET ,rALI
DOGFISH M4 LACTATE DEHYDROGENASE PIG H4 LACTATE DEHYDROGENASE '17)
GLYCERALDEHYDE-3-P DEHYDROGENASE*

-LEU -ALAU-PRO

-CYS -ARG

-E ASP

-ASP -HIS -VAL -THR
HORSE LIVER ALCOHOL DEHYDROGENASE (28.29)
YEAST ALCOHOL DEHYDROGENASE (28) --VAL -YS n'-THR A --HIS -ALA-TRP THS of several dehydrogenases. A single deletion is shown FIG. 4. Comparison of the sequence around the "essential" cysteine (underlined) in the lactate dehydrogenase sequence between Gly-164 and Cys-165, which tends to maximize any possible homologies. indicates those residues that are, identical; indicates those aminoacid residues whose codon could differ from the corresponding dogfish M4 lactate dehydrogenase by a single nucleotide base change in the mRNA sequence. *Identical sequence for yeast (29), pig muscle (29), and rabbit muscle (30).

--TYR {

-VAL -ILE -GLY -GLN -HIS -GLY -ASP -SER -VAL -PRO -SER -VAL -TRP-

PIG H4

VAL -ILE -GLY -GLU -HIS -GLY -ASP -SER -VAL -PRO -SER -VAL -TRP
FIG. 5. Sequence homologies around histidine 195. *Essential histidine residue in pig 1 lactate dehydrogenase (10, 18).
- SE R- VAL- ALA- ASP- LEU- ALA- GLN- TH R- LE- MET- LYS- ASN- LEU- CYS- ARG- VAL- HIS- PRO- VA.- SERT I - TI:C C
I_________________ CNBr THR- MET- VAL- LYS- ASP- PHE- TYR- GLY- ILE- LYS- ASP- ASN- VAL- PHE- LEU-SER- LEU-PRO-CYS- VALT T -, ,-

LEU- ASN- ASX- GLY- LE- SE R- HIS-C Ys - ASN- LE- VAL- LYS- MET- LYS- LEU- LYS- PRO- ASP- GLU- GLU-
GLN GLN- LEU-GLN- LYS-SER-ALA-THR-THR- LEU-TRP-ASP-ILE-GLN-LYS-ASP-LEU-LYS-PHE T T T II
FIG. 6. Aminoacid sequence of the carboxy-terminal region of dogfish M4 lactate dehydrogenase.

(1978)

-PRO -
300 -SER -HIS -Cys -ASN -ILE -VAL -LYS -MET -LYS

DOGFISH

ARG -LEU -LYS -ASP -ASP -GLU -VAL -ALA -GLN

-GLY -

PIG I-
310 ASP -GLU -GLU -GLN -GLN -LEU LEU -THR -SER -ASN -VAL -ILE

FG-LN Ls-ER

GLN YS-ASN

320 AA-THR -

AL-ASP-

-H ASP -{3 THR GL Y

-GLY -fj

LYSE LEU VS-g

-ASP -LEU
FIG. 7. Comparison of the C-terminal sequence of lactate dehydrogenase from pig H4 (31) and dogfish M4 isozymes. residues that are identical.

indicates those

the pig H4 sequence after residue 317 maximizes the homology; nevertheless, it is clear that the sequence becomes more divergent in the region extending from residue 306-314.

CONCLUSION

Although the sequence in two regions of the molecule remains to be confirmed, it is already possible to align the known sequence of dogfish M4 lactate dehydrogenase with the crystallographic structure. From the crystallographic data and various chemical and kinetic information it has also been possible to identify precisely many of the specific residues important for substrate and coenzyme binding. A preliminary comparison of sequence homologies between dogfish M4 and pig H4 and M4 lactate dehydrogenase suggests that the similarities may be extensive even though they are different types isolated from different species.
This research was supported by Research Grants GM 16979, GM 13901, and CA 11683 from the National Institutes of Health. S.T. was also the recipient of a Public Health Service Postdoctoral Fellowship CA 34758 and a Career Development Award GM 70244 from the National Institutes of Health. 1. Allison, W. S., Admiraal, J. & Kaplan, N. 0. (1969) J. Biol. Chem. 244, 4743-4749. 2. Heck, H. d'A. (1969) J. Biol. Chem. 244, 4375-4381. 3. Schwert, G. W., Miller, B. R. & Peanasky, R. J. (1967) J. Biol. Chem. 242, 3245-3252. 4. Appella, E. & Markert, C. L. (1961) Biochem. Biophys. Res. Commun. 6, 171-176. 5. Cahn, R. D., Kaplan, N. O., Levine, L. & Zwilling, E. (1962) Science 136, 962-969. 6. Pesce, A. J., McKay, R. H., Stolzenbach, F. G., Chan, R. D. & Kaplan, N. 0. (1964) J. Biol. Chem. 239, 1753-1761. 7. Pesce, A. J., Fondy, T. P., Stolzenbach, F. G., Castillo, F. & Kaplan, N. 0. (1967) J. Biol. Chem. 242, 2151-2167. 8. Everse, J., Berger, R. L. & Kaplan, N. 0. (1970) Science 168,
9. Chilson, 0. P., Costello, L. A. & Kaplan, N. 0. (1965) Biochemistry 4, 271-281. 10. Adams, M. J., Ford, G. C., Leikoek, R., Lentz, P., Jr., McPherson, A., Jr., Rossmann, M. G., Smiley, I. E., Schevitz, R. W. & Wonacott, A. J. (1970) Nature 227, 1098-1103. 11. Rossmann, M. G., Adams, M. J., Buehner, M., Ford, G. C., Hackert, M. L., Lentz, P. J., Jr., McPherson, A., Jr., Schevitz, R. W. & Smiley, I. E. (1971) Cold Spring Harbor Symp. Quant. Biol. 36, 179-191. 12. Adams, M. J., Buehner, M., Chandrasekhar, K., Ford, G. C., Hackert, M. L., Liljas, A., Lentz, P., Jr., Rao, S. T., Rossmann, M. G., Smiley, I. E. & White, J. L. (1972) in ProteinProtein Interactions, eds. Jaenicke, R. & Helmreich, E. (Springer-Verlag, New York), p. 139. 13. Adams, M. J., Liljas, A., Rossmann, M. G. & McPherson, A., Jr. (1973) J. Mol. Biol., in press.

14. DiSabato, G. & Kaplan, N. 0. (1963) Biochemistry 2, 776781. 15. Fondy, T. P., Everse, J., Driscoll, G. A., Castillo, F., Stolzenbach, F. G. & Kaplan, N. 0. (1965) J. Biol. Chem. 240, 4219-4234. 16. Holbrook, J. J., Pfleiderer, G., Mella, K., Volz, M., Leskowac, W. & Jeckel, D. (1967) Eur. J. Biochem. 1, 476-481. 17. Pfleiderer, G., Woenckhaus, C. J., Jeckel, D. & Mella, K. (1970) in Pyridine Nucleotide-Dependent Dehydrogenases, ed. Sund, H. (Springer-Verlag, New York, pp. 145-155). 18. Woenckhaus, C., Berghauser, J. & Pfleiderer, G. (1969) Hoppe-Seyler's Z. Physiol. Chem. 350, 473-483. 19. Schwert, G. W. & Yang, P. C. (1972) Biochemistry 11, 22182224. 20. DiSabato, G. (1965) Biochemistry 4, 2288-2296. 21. Jeckel, D., Anders, R. & Pfleiderer, G. (1971) HoppeSeyler's Z. Physiol. Chem. 352, 769-779. 22. Butler, P. J. G., Harris, J. I., Hartley, B. S. & Leberman, R. (1959) Biochem. J. 112, 679-689. 23. Gross, E. & Witkop, B. (1962) J. Biol. Chem. 237, 18561860. 24. Gray, W. R. & Hartley, B. S. (i963) Biochem. J. 89, 379380 and 59P. 25. Hartley, B. S. (1971) Biochem.J. 119, 805-822. 26. Offord, R. (1966) Nature 211, 591-593. 27. Harris, I. (1964) Nature 203, 30-34. 28. Li, T. K. & Vallee, B. L. (1964) Biochemistry 3, 869-873. 29. Perham, R. N. & Harris, J. I. (1963) J. Mol. Biol. 7, 316320. 30. Harris, J. I., Meriwether, B. P. & Park, J. N. (1963) Nature 198, 154-157. 31. Krimsky, I. & Racker, E. (1955) Science 122, 319-321. 32. Mella, K., Torff, J. J., Folsohe, E. Th.J. & Pfleiderer, G. (1969) Hoppe-Seyler's Z. Physiol. Chem. 350, 28-34. 33. Adams, M. J., Buehner, M., Chandrasekhar, K., Ford, G. C., Hackert, M. L., Liljas, A., Rossmann, M. G., Smiley, I. E., Allison, W. S., Everse, J., Kaplan, N. 0. & Taylor, S. S. (1973) Proc. Nat. Acad. Sci. USA 70, in press. APPENDIX Map as an Aid in the

1236-1238.

Electron Density Sequencing of Peptides
MARGARET J. ADAMS*, GEOFFREY C. FORD, PAUL J. LENTZ, JR.t, ANDERS LILJAS, AND MICHAEL G. ROSSMANN Department of Biological Sciences, Purdue University, West Lafayette, Indiana 47907
An electron density distribution, at a nominal resolution of 2.0 X, was used as a guide and check on the positioning of selected
* Present address: Department of Molecular Biophysics, Zoology Department, South Parks Road, Oxford, England. t Present address: Wallenberg laboratoriet, Uppsala Universitet, Dag Hammarskjblds vag 21, Uppsala, Sweden.