ABSTRACT Thermal adaptation at the biochemical level has been demonstrated in previous studies of bony fishes and other vertebrates. One criterion that has been used to assess the possible presence of thermal adaptation is the Michaelis-Menten constant, or KK values have been shown to be conserved among teleosts and some other vertebrates within their physiological temperature ranges. One possible mechanism by which homologous proteins may conserve K at extreme temperatures is by changes in primary structure. The relationship between temperature adaptation and changes in primary structure has been explored in teleost fishes, but has yet to be studied in elasmobranchs. In this study, the variation of pyruvate Kp with temperature and the deduced amino acid sequences of four elasmobranch Aj-LDH orthologs were compared. Both the subtropical and the cold temperate species of sharks were found to display conservation of pyruvate Kp over physiological temperature ranges, while the subtropical and the Antarctic species of rays did not. The deduced amino acid sequences of the shark enzymes exhibited different regions of variability than the LDH's of teleosts previously studied, while the sequences of the ray enzymes exhibited similar regions of variability to those of teleost orthologs. It was concluded that there may a limited number of sites at which temperature adaptation in AA-LDH occurs, and that even closely related organisms do not necessarily use the same sequence changes to adapt to temperature. INTRODUCTION Life occurs at a wide range of temperatures, from below zero to above 100 degrees Celsius. Because life processes including development, mitosis, and membrane function are affected by the thermal environment in which they occur, organisms must adjust to their physiological temperature ranges. This adjustment must occur at all levels, including the function of proteins. For example, because glycolysis is a nearly universal biological process, glycolytic enzymes must function similarly in a broad range of thermal environments. In affecting the chemical environment in which a protein functions, temperature influences the assembly of subunits, global protein stability, substrate binding and conformational changes during catalysis. In order to compensate for the perturbing effects of temperature, proteins may undergo changes in primary structure through evolution. These changes do not have to be dramatic; substitutions are found outside the active site, in regions that do not change the chemistry of catalysis, but affect the movement of catalytic domains (Holland et al. 1997). Moreover, differences in stability and flexibility may only consist of a glycine/alanine substitution (Fields and Somero, 1997). Such changes have been explored in teleost fishes (Fields and Somero, 1997; Fields and Somero in prep.), and the study of elasmobranchs may help us discover whether evolutionarily distinct organisms use similar sequence changes in their enzymes to adapt to their thermal environments. Elasmobranchs provide an interesting study system not only because of their evolutionary distance from bony fishes but also because of their unique biochemistry. Elasmobranchs such as sharks and rays concentrate urea and trimethylamine oxide (TMAÖ) in their body fluids in order to maintain osmotic balance with the surrounding seawater. Although urea is a protein denaturant, elasmobranch enzymes clearly function at physiological urea concentrations. They may even require the presence of urea to exhibit enzyme kinetics comparable to other organisms (Yancey and Somero 1978). However, it has also been demonstrated that TMAÖ may counteract the effects of urea in some organisms (Yancey and Somero 1980). The combination of elasmobranchs’ novel biochemistry and evolutionary distance from teleosts makes the study of their temperature adaptation at the macromolecular level of interest to biochemists, evolutionary biologists, and comparative biologists. Iused lactate dehydrogenase-A (Az-LDH) to explore the mechanism by which homologous proteins might adapt to different thermal environments. Aj-LDH is the enzyme that converts pyruvate and NADH to lactate and NADt during anaerobic glycolysis. In order to describe temperature adaptation, I compared the PYR) sequences and apparent Michaelis-Menten kinetic constants of pyruvate (Kp among four elasmobranch Az-LDH orthologs. Kn values are typically very sensitive to temperature, usually rising as measurement temperature is increased. However, in previous studies comparing various organisms' pyruvate Kp's over a range of temperatures, it has been found that enzymes tend to conserve Kp at the temperatures at which they function (Somero 1995, Holland et al. 1997). In this study, changes in primary structure were compared with the different ways in which PrR varied with temperature among the four species' LDH orthologs. the Km Knowledge of LDH 3-D structure (Abad-Zapatero et al. 1987) and conformational changes during catalysis (Dunn et al. 1991) were used to interpret correlations between kinetics and sequence variation, and to compare these correlations to biochemical temperature adaptation in teleosts. METHODS Study Organisms The four species of elasmobranchs used in this study were the nurse shark Ginglymostoma cirratum (subtropical), the file tail cat shark Parmaturus xaniurus (cold temperate), the cow nosed ray Rhinoptera bonasus (subtropical), and the Antarctic bat skate Bathyraja eatonii.G. cirratum was collected in the Florida Keys, and shipped on dry ice to Hopkins Marine Station by Elizabeth Clarke of the University of Miami. P. xaniurus was collected from Monterey Bay Canyon and donated to this study by the Monterey Bay Aquarium. R. bonasus was collected from Tampa Bay, Florida, and sent on dry ice for the study by Joseph Torres of the University of Southern Florida. B. eatonii was collected near Palmer Station in Antarctica by Peter Fields of Hopkins Marine Station. All species were stored at -80°C until use. Determination of Km-PYR of A4-LDH The Az-LDH of each species was purified using an agarose-oxamate affinity column (Yancey and Somero 1978). A homogenate of the white muscle of each species was allowed to pass down the oxamate affinity column, with NADH present in excess to ensure efficient binding of the LDH. The LDH was eluted with pyruvate, which binds the protein with a greater affinity than oxamate, and collected in fractions of approximately 3ml each. The fractions were tested for LDH activity, and the active fractions were pooled. Äfter dialysis against 50 mmol 11 potassium phosphate buffer, pH 6.8, the purified protein was frozen at -80°C. -PYR) was then determined The apparent Michaelis-Menten constant of pyruvate (Km at 0, 10, 20, and 30 degrees C for each species. The pyruvate Ky's were determined using a Perkin Elmer Lambda 3B spectrophotometer equipped with a temperature- controlled cell attached to a Lauda RM6 recirculating water bath. The temperature of each cuvette was maintained at +/- 0.2°C. The reaction solutions were prepared using an imidazole chloride buffer (160mM imidazole chloride, 300uM NADH; pH 7.0 at 20°C), which has been shown to exhibit similar pH to intracellular conditions over a wide range of temperatures. For each Kp eight concentrations of pyruvate were used: 0.5, 0.4, 0.3, 0.2 0.15 0.1 0.075, and 0.05 mmol 1-1. The spectrophotometer was set to record the absorption of light with wavelength 340 nm, the wavelength at which NADH absorbs most strongly. The decrease of [NADH] which followed the addition of 25 ul of purified Az-LDH to the imidazole/NADH/pyruvate solution in each cuvette was plotted against time. Conversion of absorbance to concentration using Beer's law then yielded a series of velocities in units of change in concentration of NADH (and thus pyruvate, as they decrease together in a 1:1 ratio) per minute. Pyruvate concentration v. velocity data points were entered into the Wilman computer program (1986) to calculate Kp using weighted linear regression as outlined by Wilkinson (1961). Sequencing of LDH-A CDNA Messenger RNA (mRNA) was purified from each elasmobranch species using microscopic magnetized beads (Dynabeads; Dynal, Inc.). 25-mers of poly¬ deoxythymidylate (poly-T) attached to the beads bound to the poly-adenylate tails of the mRNA in a white muscle homogenate from each species. The beads were then washed and the isolated mRNA eluted and stored at -80°C. Complementary DNA (CDNA) was synthesized and the Az-LDH gene amplified using a Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) procedure (Promega). LDH-A¬ specific primers were first designed using regions of agreement in barracuda (genus Sphyraena; Holland et al. 1997) and dogfish (genus Squalus; Abad-Zapatero et al. 1987) LDH-A sequences. The 3' ends of the A4-LDH CDNA sequence were synthesized and amplified using the 3'RACE procedure (GibcoBRL). AII CDNA products were visualized using an ethidium-bromide stained agarose gel, and purified using a gel extraction procedure (Qiagen). The CDNA sequence was obtained by amplification of the purified CDNA products using a dye-labeled terminating nucleotide protocol (Perkin-Elmer Corporation), and the amplification products were visualized on the Applied Biosystems model 373A automated sequencer. The resulting sequences were aligned and translated using the GCG software program (Oxford Molecular Group). RESULTS Apparent Kp-PYR. Upon comparison of the two shark species, it was found that the „PTR of the cold temperate species, P. xaniurus, was higher than that of the Km subtropical species, G. cirratum, at every temperature (Fig 1A). This trend is consistent with earlier studies of temperature effects on Km in organisms adapted to widely differing thermal environments (Yancey and Somero, 1978; Holland et al. 1997; Fields and Somero, 1997; Fields and Somero in prep.). Upon comparison of the two ray species, however, R. bonasus, the warm-adapted species, exhibited higher Knvalues at every temperature above 0°C when compared to the Antarctic species, B. eatonii (Fig 1B). At 0°C, B. eatonii displayed a Km value slightly greater than that of R. bonasus. Amino acid sequences. There were fourteen differences among the 180 residues of amino acid sequence obtained from the two species of rays. Of those fourteen differences, four were non-conservative. (A non-conservative change is defined as one in which the two residues were of different character, one being polar and one hydrophobic, for example.) There were twenty-eight differences among the 220 residues of sequence obtained from both of the shark species, twelve of which were non-conservative. The differences were distributed non-randomly throughout the four A4-LDH sequences; regions of high conservation among all species include residues 96-110, 155-180, 190-210, and 245-254 (Fig. 2). These regions include the catalytic loop (aaf 96-113, Abad-Zapatero et al., 1987) and substrate and cofactor binding residues (e.g., Arg107, Asp167, Arg170, His194, Thr249; Abad-Zapatero et al., 1987; Sakowicz et al., 1993). [Please note that the residue numbers in this paper correspond to the sequences I obtained; they are not identical to the sequence numbering systems used in previous studies. DISCUSSION In analyzing these data on Kp and deduced amino acid sequences, I am attempting to answer three questions: 1) Do the four elasmobranch Az-LDH orthologs exhibit temperature-adaptive changes? 2) When compared to teleosts, are these elasmobranch species using similar primary structure changes in their AA-LDH sequences to adapt to temperature? 3) How might the primary structure changes lead to the observed kinetic differences between the elasmobranch AA-LDH orthologs studied? Temperature adaptation as assessed by kinetic data. At every temperature of measurement, the shark LDH ortholog of P. xaniurus, which experiences habitat temperatures from 4° to 10°C (G. Somero, pers. comm.), exhibited higher Kp values than the G. cirratum ortholog, which is found at temperatures ranging from 20° to 32°C (J. Torres, University of Southern Florida, pers. comm.). This pattern is consistent with earlier studies which show that organisms tend to share comparable Kn values within their physiological temperature ranges (Fields and Somero 1997, Yancey and Somero 1978, Holland et al. 1997, Fields and Somero in prep.). Upon comparison of the shark K patterns, it is apparent that the Kp values for each shark within its temperature range are very similar (Fig. 1A). The rays, in contrast, do not exhibit Kp conservation in this study. For each temperature of measurement above zero, the LDH-A ortholog of B. eatonii, which experiences habitat temperatures from -1.86° to 1°C (Eastman 1993), displays Km values below those of the R. bonasus Az-LDH, which is found at temperatures ranging from 20° to 32°C (. Torres, University of Southern Florida, pers. comm.) (Fig. 1B). The lower Kp values, and corresponding higher pyruvate affinities of the Antarctic species' Az-LDH homologue are puzzling. Because affinity rises (and Km drops) as temperature decreases, one would expect that enzymes functioning at colder temperatures would compensate by exhibiting lower intrinsic affinity (and higher Kp) at any given temperature. One possible explanation for the rays' kinetic data is that the presence of organic solutes such as urea and TMAO within the organisms under physiological conditions alter the kinetics of the enzymes. It is possible that the ray A4-LDH homologues display affinity (Kp) conservation under physiological conditions. Mechanisms of temperature adaptation in elasmobranchs and teleosts. I compared the subtropical elasmobranch Az-LDH sequences to their cold habitat counterparts to explore the primary structure differences that might occur in organisms adapted to different temperatures. In the rays, non-conservative differences occurred with higher frequency in two regions: the aH helix and the BH-alG loop (Fig. 2). The first region of high variability between the two species of rays was the aH helix, at the carboxy end of the monomer (aaf 311-327). Helix aH plays a significant part during catalysis because aH, along with helix alG-a2G and the catalytic loop (aaft 97- 107) form the mouth of the molecule, moving toward one another during formation of the internal vacuole during catalysis (Dunn et al. 1991) (Fig. 3). The most dramatic sequence difference between the rays in the aH region is at the amino end of the helix. A proline residue in the subtropical species is replaced by a serine in the Antarctic species, a pattern that has also been shown to occur in Antarctic teleosts (Fig. 2). Proline is a unique residue because its side chain is bonded not only to the alpha carbon, but also to nitrogen, forming a five-membered ring. This ring prevents proline from assuming a broad range of conformations. Replacing a proline with a serine would therefore increase the flexibility of helix aH relative to the rest of the molecule. The second region of high variability between the rays was the BH-alG loop (aat 207-226). The BH-alG loop links part of the Aj-LDH active site (containing residues His194, Asp167 and Arg172) to one part of the 'mouth' of the monomer, helix alG¬ a2G, which contains Tyr140, a residue required for enzyme activity. The ßH-alG loop, which does not possess secondary structure, possesses a high temperature factor (Abad-Zapatero et al. 1987) and may ease the movement of the alG-a2G helix during conformational changes during catalysis (Fields and Somero in prep.) (Fig. 3). Hence it is a possible source of adjustment for increased flexibility of catalytically important regions in the enzyme. In the ßH-alG loop, there are two non-conservative differences between the subtropical and Antarctic ray species. In the first, a serine at position 211 in the subtropical species (R. bonasus) is replaced by a glutamine in the Antarctic species (B. eatonii). In the second, a glutamine at position 226 in the subtropical species is replaced by an aspartate in the Antarctic species (Fig. 2). In both of these substitutions, the Antarctic residue is more polar than the subtropical residue. This increase in the polarity of the BH-alG loop has also been shown to occur in Antarctic teleosts. Because ßH-alG forms hydrophobic interactions with the neighboring subunit, increased polarity in this region might correspond to increased flexibility of the monomer. This increased flexibility results when the interaction of the loop with the neighboring monomer is reduced while interaction with the surrounding polar solvent is increased. An increase in the flexibility of ßH-alG might enable ease of movement of alG-a2G during catalysis. These changes should favor a higher catalytic rate constant (kcat) for the enzyme of the Antarctic species, a conjecture to be tested in future work with these enzymes. In the sharks, the highest concentration of non-conservative amino acid substitutions occurs from residue 9 to residue 21, near helix aA, in the amino terminal arm of the Ag-LDH monomer (Fig. 3). This region of the monomer forms polar inter-subunit interactions with the neighboring monomer (Abad-Zapatero et al. 1987. In three of the four non-conservative changes I observed, a polar residue in the subtropical shark was replaced by a charged residue in the cold temperate shark. In the fourth non-conservative difference, a charged residue in the subtropical species was replaced by a polar residue in the cold temperate species (Fig. 2). Because I did not obtain the portion of sequence in G. cirratum which interacts with the amino arm in neighboring monomers, it is not possible to analyze how these changes in polarity of amino arm residues might affect the overall protein stability. However, a single substitution at position 8 has been shown to lead to changes in Km and thermal stability in barracuda A4-LDH (Holland et al. 1997). Therefore, it is possible that the high variability within the shark LDH sequences in the amino terminal arm contributes to their differences in kinetic data. While analysis of the shark sequence differences is limited, it is noteworthy that the shark species do not display high variability in the ßH-alG loop, as previously seen in teleosts. It is also interesting that the two species of rays seem to use similar primary sequence changes to the teleosts in thermal adaptation of their Aj-LDH orthologs. Two conclusions may be drawn from these variability patterns: 1)There may be a finite number of areas in the A4-LDH sequence that allow temperature adaptation. 2)Elasmobranchs do not all use similar sequence changes to adapt to different temperatures. Flexibility and thermal adaptation of proteins. My analysis of the relationship between proteins' primary sequence and temperature adaptation can be explained by a theory developed by Fields and Somero (in prep.) as follows. At a given temperature, a protein may exist in a number of different conformations. As the temperature increases, the number of possible conformations and the rate at which the protein changes between them increases. However, the amount of time spent in any one conformation decreases as temperature increases. Thus the amount of time spent in a conformation appropriate for binding decreases, and Kp increases. Alternatively, increased flexibility of a protein due to primary structure changes can also cause the protein to undergo conformational changes more rapidly. Thus changes in the flexibility of a protein and changes in the temperature at which it functions can cause similar (or counteracting) effects on the kinetic function of a protein. Therefore, the perturbing effects of temperature may be compensated for by varying the flexibility of the protein through changes in primary structure. 10 ACKNOWLEDGMENTS Thank-you to George Somero and Peter Fields for their enduring patience and willingness to teach. I would also like to thank Jonathon Stillman, Andy Gracey, Lars Tomanek, and Rachel Ream for their kind tolerance. LITERATURE CITED Abad-Zapatero, C. et al. 1987. Refined Crystal Structure of Dogfish MA Apo-lactate Dehydrogenase. J. Mol. Biol. 198:445-467. Dunn, C. R. et al. 1991. Design and synthesis of new enzymes based on the lactate dehydrogenase framework. Phil. Trans. R. Soc. Lond. 332:177-184. Eastman, J. T. Antarctic Fish Biolo y: Evolution in a Unique Environment. San Diego: Academic Press Inc., 1993. Fields, P. A. and Somero, G. N. 1997. Amino Acid Sequence Differences Cannot Fully Explain Interspecific Variation in Thermal Sensitivities of Gobiid Fish A4 lactate Dehydrogenases (A-LDHs). J. exp. Biol. 200:1839-1850. Holland, L. Z. et al. 1997. Evolution of Lactate Dehydrogenase-A Homologs of Barracuda Fishes (Genus Sphyraena) from Different Thermal Environments: Differences in Kinetic Properties and Thermal Stability Are Due to Amino Acid Substitution Outside the Active Site. Biochemistry 36:3207-3215. Sakowicz, R. 1993. Threonine 246 at the Active Site of the L-Lactate Dehydrogenase of Bacillus stearothermophilus Is Important for Catalysis but Not for Substrate Binding. Biochemistry. 32:12730-12735. Somero, G. N. 1995. Proteins and Temperature. A. Rev. Physiol. 57:43-68. Stock, D. W. and Powers, D. A. 1995. The CDNA sequence of the lactate dehydrogenase-A of the spiny dogfish (Squalus acanthius): Corrections to the amino acid sequence and an analysis of the phylogeny of vertebrate lactate dehydrogenases. PYR) of A4-LDH from the shark Fig. 1. (A) Apparent Michaelis-Menten constants (Kr species G. cirratum and P. xaniurus. (B) Apparent Michaelis-Menten constants (Km¬ PYR) of A4-LDH from the ray species R. bonasus and B. eatonii. Fig. 2. Deduced regions of amino acid sequences of LDH-A for the four elasmobranch species studied, as well as the dogfish (S. acanthius) sequence (Stock and Powers 1995). Residues in bold type indicate non-conservative differences within the ray sequences or shark sequences. The text above the sequences indicates the location of second degree structure (Abad-Zapatero et al. 1987). Fig. 3. Cartoon of dogfish Az-LDH structure after Abad-Zapatero et al. (1987). One monomer is shown with selected secondary structure labeled. 0.40 - 0.35 0.30 0.25 0.20 0.15 0.10 0.05 - 0.00 0.40 0.35 0.30 0.25 - 0.20 0.15 0.10 - 0.05 0.00 Fig. 1 —— Temp. V P. xaniurus A — Temp. v G. cirratum 5 10 15 20 25 30 35 Temperature (°C) —0— Temp. V B. eatonii + Temp. V R. bonasus B I 5 10 15 20 25 30 35 Temperature (°C) Fig. 2 -B—— — — — — — — — — —— BA-- -A¬ -BB SSS -SODKGS HNKXTIVGVG AVXMACAISI LMKDLADELA 50 P. xaniurus S. acanthius MATLKDKLIG HLATSQEPRS YNKITVVGVG AVGMACAISI LMKDLADEVA S--G OLXTSOPGHS NNKITIVGVG AVGMACAISI LLKDLADEVA G. cirratum ssss sss ssssss wewnnnnnwe B. eatonii wwwww R. bonasus —10 — — — — — — — — — — — — ——— 8c--- pD--- P. X. LVDVMEDKLK GEMMDLOHGS LFLOTSKIVS GKDYKVSAGS KLVVITAGAR 100 S. a. LVDVMEDKLK GEMMDLOHGS LFLHTAKIVS GKDYSVSAGS KLVVITAGAR G. c. LVDVMEDKLK GEMMDLOHGS LFLHTRKIVS GKDYSVSAGS KLVVVTAGAR s ossssss sssssss ossssss ssssww B. e. Ph osssss ossssss osssss ssww -(D----------GE------ -1F---- P. x. QOEGESRLNL VORNVNIFKH IIPDIVKHSP NCIILVVSNP VDILTYVAWK 150 QOEGESRLNI S. a. VORNVNIFKF IIPDIVKHSP DCIILVVSNI VDVLTYVAWK QQEGESRLNL VORNVNIFKH IIPDIVKNSP DCIILVVSNP G. c. VDXLTYVAWK s sssss wosssss o TVK B. e. Wsssssss o s wwwuwwwa R. b. -2F----------- -BG--- P. X. FRYLMGERLG IHSSSCHAWV IGEHGDSSVP 200 LSGLPMHRVI GSGCNLDSAR S. a. IGEHGDSSVP LSGLPMHRII GSGCNLDSAR FRYLMGERLG VHSSSCHGWV IGEHGDSSVP G. c LSGLPMHRVI GSGCNLDSAF FRYLMGERLG IHSSSCHAWV FRYLMGERLG IHSSSCHGWV IGEHGDSSVP B. e. LSGLPMHRVI GSGCNLDSAR R. b. LSGFXMHRVI GSGCNLDSAR FRYLMGERLG IHSSSCHGWV IGEHGDSSVP -a1G—----26—- ---- -83- SLKEIHPDIG TEKDKENWKK VHKDVVDSAY EVIKLKGYTS 250 VWSGMNVAGV P. X. TDKDKENWKK LHKDVVDSAY EVIKLKGYTS VWSGMNVAGV SLKELHPELG TDHDKENWKA IHKE G. c VWSGMNVAGV NLKELHPEIC TDKDKEDWKK VHRDVVDSAY EVIKLKGYTS VWSGMNVAGV NLKELHPEIG VWSGMNVAGV SLKELHPEIG TDKDKENWKK VHKDVVDSAY EVIKLKGYTS -BK- -B1- -03G- P. X. WAIGMSVADL FHGIKNDVFL SLPCVLDNHG 300 GETIMKNLCK VHPISTMVKD S. a. WAIGLSVADL AETIMKNLCR VHPVSTMVKD FYGIKNDVFL SLPCVLDNHG sssss sss wonnn G. C. WAIGMSVADM AETLMKNLRR VHPVSTMVKN FYGITNDVFL SLPSVLDNHG B. e. WAIGMSVADL AETIMKNLRR VHPISTMVKT FYGITNEVFL SLPSVLDNNG R. b. -BM-- -H-- ITNVVKMVLK PEEENQLQOS ATTLWDIQKD L va 350 P. X. ISNIVKMKLK PDEEQOLOKS ATTLWDIQKD LKF'TSPL'P SYAAVTSMII s. a. s wos ISNVVKMTLK SEEEAKLOOS ATTLWNIQKD LKFS e R. b. ISNIVKMTLK PDEEAKLOHS AT O L (