Comparative study of promoter activity of three anther-specific genes encoding lipid transfer protein, xyloglucan endotransglucosylase/hydrolase and polygalacturonase in transgenic Arabidopsis thaliana

Promoter sequences of three anther-specific genes, each of which shows sequence identity to lipid transfer protein (LTP12), xyloglucan endotransglucosylase/hydrolase (XTH3), and polygalacturonase (PGA4), were obtained from Arabidopsis thaliana, fused to the #-glucuronidase (GUS) gene, and then introduced into A. thaliana. Histochemical GUS assay showed that the PGA4 promoter was active in the tapetum at the bicellular pollen stage and in tricellular pollen. The promoter of LTP12 and XTH3 directed GUS expression exclusively in the tapetum. The LTP12 promoter was activated from the uninucleate microspore stage, while the XTH3 promoter was activated from the bicellular pollen stage. This type of activation pattern at the late developmental stage of the tapetum has not been reported previously. The promoter sequences employed in this study will be useful for the characterization of genes differentially expressed in anthers.     

Identification of <Emphasis Type="Italic">S</Emphasis>-haplotype-specific F-box gene in Japanese plum (<Emphasis Type="Italic">Prunus salicina</Emphasis> Lindl.)

Self-incompatibility has been studied extensively at the molecular level in Solanaceae, Rosaceae and Scrophulariaceae, all of which exhibit gametophytic self-incompatibility controlled by a single polymorphic locus containing at least two linked genes, i.e., the S-RNase gene and the pollen-expressed SFB/SLF (S-haplotype-specific F-box/S-locus F-box) gene. However, the SFB gene in Japanese plum (Prunus salicina Lindl.) has not yet been identified. We determined eight novel sequences homologous to the SFB genes of other Prunus species and named these sequences PsSFB. The gene structure of the SFB genes and the characteristic domains in deduced amino acid sequences were conserved. Three sequences from 410 to 2,800 bp of the intergenic region between the PsSFB sequences and the S-RNase alleles were obtained. The eight identified PsSFB sequences showed S-haplotype-specific polymorphism, with 74–83% amino acid identity. These alleles were exclusively expressed in the pollen. These results suggest that the PsSFB alleles are the pollen S-determinants of GSI in Japanese plum. Nucleotide sequence data reported are available in the NCBI database under the accession numbers DQ849084–DQ849090 and DQ849118.     

Self-compatibility and incompatibility in tetraploid sour cherry (Prunus cerasus L.)

. Gametophytic self-incompatibility (GSI) typically "breaks down" due to polyploidy in many Solanaceous species, resulting in self-compatible (SC) tetraploid individuals. However, sour cherry (Prunus cerasus L.), a tetraploid species resulting from hybridization of the diploid sweet cherry (P. avium L.) and the tetraploid ground cherry (P. fruticosa Pall.), is an exception, consisting of both self-incompatible (SI) and SC individuals. Since sweet cherry exhibits GSI with 13 S-ribonucleases (S-RNases) identified as the stylar S-locus product, the objectives were to compare sweet and sour cherry S-allele function, S-RNase sequences and linkage map location as initial steps towards understanding the genetic basis of SI and SC in sour cherry. S-RNases from two sour cherry cultivars that were the parents of a linkage mapping population were cloned and sequenced. The sequences of two S-RNases were identical to those of sweet cherry S-RNases, whereas three other S-RNases had unique sequences. One of the S-RNases mapped to the Prunus linkage group 6, similar to its location in sweet cherry and almond, whereas two other S-RNases were linked to each other but were unlinked to any other markers. Interspecific crosses between sweet and sour cherry demonstrated that GSI exists in sour cherry and that the recognition of common S-alleles has been maintained in spite of polyploidization. It is hypothesized that self-compatibility in sour cherry is caused by the existence of non-functional S-RNases and pollen S-genes that may have arisen from natural mutations.     

The rice tapetum-specific gene RA39 encodes a type I ribosome-inactivating protein

A tapetum-specific cDNA encoded by a rice gene, RA39, was isolated by cDNA subtractive hybridization, differential screening and rapid amplification of cDNA ends. RA39 is a single-copy gene in the rice genome. mRNA in situ hybridization indicates that this gene is a tapetum-specific gene, and highly expressed in the tapetal cells at the meiosis and tetrad stages. The RA39 cDNA is 1,013 bp in length with an open reading frame encoding 298 amino acid residues. This cDNA sequence does not show significant homology to any known sequences in GenBank databases, but its deduced amino acid sequence (RA39) has between 19 and 34% sequence identity to ribosome-inactivating proteins (RIPs). Optimal alignment reveals that the five amino acid residues constituting the active site of the ricin A-chain (Tyr⁸⁰, Tyr¹²³, Glu¹⁷⁷, Arg¹⁸⁰ and Trp²¹¹), which are invariant among all RIPs published to date, are conserved in RA39. Recombinant RA39 protein expressed in Escherichia coli was purified to homogeneity. The purified protein exhibits the RNA N-glycosidase activity of RIPs. This demonstrates that RIPs occur in the reproductive organs of rice. The possible function of RA39 in anther development is discussed.     

Production and Isolation of Levan by Use of Levansucrase Immobilized on the Ceramic Support SM-10

The complete amino acid sequence of rye seed chitinase-a (RSC-a) has been analyzed. RSC-a was cleaved with cyanogen bromide and the resulting three fragments, CB1, CB2, and CB3, were separated by gel filtration. The amino acids of the N-terminal fragment CB1 were sequenced by analyzing the peptides produced by digestion with trypsin, lysylendopeptidase, or pepsin of reduced S-carboxymethyl ated or S-aminoethylated CB1. The sequences of fragments CB2 and CB3 were established by sequencing the tryptic peptides from reduced S-carboxymethylated CB2 and CB3, and by aligning them with the sequence of rye seed chitinase-c (RSC-c) to maximize sequence homology. The complete amino acid sequence of RSC-a was established by connecting these three fragments.

RSC-a consists of 302 amino acid residues including hydroxyproline residues, and has a molecular mass of 31,722 Da. RSC-a is basic protein with a cysteine-rich amino terminal domain, indicating that this enzyme belongs to class I chitinases. The amino acid sequence of RSC-a showed that the sequence from Gly60 to C-terminal Ala302 in this enzyme corresponds to that of RSC-c belonging to class II chitinases with 92% identity, and that RSC-a has high similarity to other plant class I chitinases but a longer hinge region and an extra disulfide bond.     

cDNA and protein sequence of a major pea seed albumin (PA 2 : Mr≈26 000)

Summary Pea albumin 2 (PA2:M_r26000) is a major component of the albumin fraction derived from aqueous salt extracts of pea seed. Sodium dodecylsulfate-polyacrylamide gel electrophoresis and chromatography on DEAE-Sephacel resolve PA2 into two closely related components (PA2a and PA2b). A cDNA clone coding for one of these components has been sequenced and the deduced amino acid sequence compared with partial, chemically-determined sequences for cyanogen bromide peptides from both PA2 components. Complete amino acid sequences were obtained for the C-terminal peptides. The PA2 molecule of 230 amino acids contains four imperfect repeat sequences each of approximately 57 amino acids in length.The combined sequence data, together with a comparison of PA2-related polypeptides produced in vitro and in vivo, indicate that PA2 is synthesized without a signal sequence and does not undergo significant post-translational modification. Although both forms of PA2 contain Asn-X-Thr consensus sequences, neither form is glycosylated. Accumulation of PA2 contributes approximately 11% of the sulfur-amino acids in pea seeds (cysteine plus methionine equals 2.6 residues percent). Suppression of levels of PA2 polypeptides and their mRNAs in developing seeds of sulfur-deficient plants is less marked than that for legumin, in spite of the lower content of sulfur-amino acids in legumin.     

Recognition specificity of self-incompatibility in <Emphasis Type="Italic">Pyrus</Emphasis> and <Emphasis Type="Italic">Malus</Emphasis>

Pyrus displays gametophytic self-incompatibility controlled by a single highly polymorphic gene complex termed S locus, which comprises a stylar-expressed gene (S-RNase) tighlty linked with a pollen expressed gene, that determines the specificity of the self-incompatibility locus. Deduced amino acid sequence of ‘Meigetsu’ S ₈ -RNase in Pyrus pyrifolia and ‘Kuerlexiangli’ S ₂₈ -RNase in P. sinkiangensis showed 100% identity. S ₃ -RNase in Malus spectabilis was also found to be similar to S ₈ -RNase in P. pyrifolia with 96.9% identity in the deduced amino acid sequence. The intron, which is generally highly polymorphic between alleles, was also remarkably well conserved within these allele pairs. The intron of PpS ₈ -RNase showed 95.3 and 91.9% identity with PsS ₂₈ -RNase and MsS ₃ -RNase, respectively. Pollen tube growth in styles, pollen tube length in artificial media containing different S-RNases and segregation of S haplotypes in F₁ plants revealed commonality of the recognition specificity between PpS ₈ -RNase and PsS ₂₈ -RNase and between PpS ₈ -RNase and MsS ₃ -RNase. Results suggested that PpS ₈ -RNase, PsS ₂₈ -RNase and MsS ₃ -RNase have maintained the same recognition specificity after the divergence of the two species and that amino acid substitutions found between PpS ₈ -RNase and MsS ₃ -RNase do not alter the recognition specificity.     

The Evolution of Vp1 Gene in Enterovirus C Species Sub-Group That Contains Types CVA-21, CVA-24, EV-C95, EV-C96 and EV-C99

Genus Enterovirus (Family Picornaviridae,) consists of twelve species divided into genetically diverse types by their capsid protein VP1 coding sequences. Each enterovirus type can further be divided into intra-typic sub-clusters (genotypes). The aim of this study was to elucidate what leads to the emergence of novel enterovirus clades (types and genotypes). An evolutionary analysis was conducted for a sub-group of Enterovirus C species that contains types Coxsackievirus A21 (CVA-21), CVA-24, Enterovirus C95 (EV-C95), EV-C96 and EV-C99. VP1 gene datasets were collected and analysed to infer the phylogeny, rate of evolution, nucleotide and amino acid substitution patterns and signs of selection. In VP1 coding gene, high intra-typic sequence diversities and robust grouping into distinct genotypes within each type were detected. Within each type the majority of nucleotide substitutions were synonymous and the non-synonymous substitutions tended to cluster in distinct highly polymorphic sites. Signs of positive selection were detected in some of these highly polymorphic sites, while strong negative selection was indicated in most of the codons. Despite robust clustering to intra-typic genotypes, only few genotype-specific ‘signature’ amino acids were detected. In contrast, when different enterovirus types were compared, there was a clear tendency towards fixation of type-specific ‘signature’ amino acids. The results suggest that permanent fixation of type-specific amino acids is a hallmark associated with evolution of different enterovirus types, whereas neutral evolution and/or (frequency-dependent) positive selection in few highly polymorphic amino acid sites are the dominant forms of evolution when strains within an enterovirus type are compared.     

Polymorphism of <Emphasis Type="Italic">SFBB</Emphasis><Superscript>−γ</Superscript> and its use for <Emphasis Type="Italic">S</Emphasis> genotyping in Japanese pear (<Emphasis Type="Italic">Pyrus pyrifolia</Emphasis>)

Japanese pear (Pyrus pyrifolia) exhibits the S-RNase-based gametophytic self-incompatibility where the pollen-part determinant, pollen S, had long remained elusive. Recent identification of S locus F-box brothers (SFBB) in Japanese pear and apple suggested that the multiple F-box genes are the pollen S candidates as they exhibited pollen specific expression, S haplotype-specific polymorphisms and linkage to the S locus. In Japanese pear, three SFBBs were identified from a single S haplotype, and they were more homologous to other haplotype genes of the same group (i.e., α-, β- and γ-groups). In this study, we isolated new seven PpSFBB ^−γ genes from different S genotypes of Japanese pear. These genes showed S haplotype-specific polymorphisms, however, sequence similarities among them were very high. Based on the sequence polymorphisms of the PpSFBB ^−γ genes, we developed a CAPS/dCAPS system for S genotyping of the Japanese pear cultivars. This new S genotyping system was found to not only be able to discriminate the S ¹–S ⁹, but also be suitable for identification of the mutant S ^4sm haplotype for the breeding of self-compatible cultivars, and detection of new S haplotypes such as S ^k.     

Identification and cloning of related self-incompatibility S-genes in Papaver rhoeas and Papaver nudicaule

 The primary goal of this study was to identify, clone and analyse new S-gene sequences in order to provide a basis for identifying amino acid residues that confer S-allele specificity. Three new putative S-alleles from Papaver rhoeas and Papaver nudicaule were identified using immunological and PCR methods. cDNAs encoding full-length open reading frames of the P. rhoeas S ₈ and P. nudicaule Sn ₁ genes were isolated. Nucleotide sequencing of these cDNAs, together with the partial S ₇ sequence obtained by PCR, was used to derive the corresponding amino acid sequences. It is of interest that the P. nudicaule Sn₁ sequence, which is the first S-allele isolated from another species of Papaver, shares a closer sequence identity to the P. rhoeas S₃ amino acid sequence than S₃ does to S₁ from P. rhoeas. The identity of the S₈ allele was confirmed by expressing the coding region in Escherichia coli and demonstrating that the recombinant protein, designated S₈e, specifically inhibited S ₈ pollen in an in vitro bioassay. Information from sequence analysis of the S₈, Sn₁ and partial S₇ amino acid sequences revealed important information about Papaver S-proteins. It confirmed previous observations based on only two S-alleles, that whilst exhibiting a high degree of amino acid sequence polymorphism ranging from 51.3% to 63.7%, these molecules probably share very similar secondary structures. These studies also revealed that, in contrast to the S-proteins from the Solanaceae and Brassica, amino acid sequence variation is not found in hypervariable blocks, but instead, is found throughout the S-proteins, interspersed with numerous short strictly conserved segments. Received: 16 March 1998 / Revision accepted: 19 May 1998     

Determination of partial genomic sequences and development of a CAPS system of the S-RNase gene for the identification of 22 S haplotypes of apple (Malus × domestica Borkh.)

Information about self-incompatibility (S) genotypes of apple cultivars is important for the selection of pollen donors for fruit production and breeding. Although S genotyping systems using S haplotype-specific PCR of S-RNase, the pistil S gene, are useful, they are sometimes associated with false-positive/negative problems and are unable to identify new S haplotypes. The CAPS (cleaved amplified polymorphic sequences) system is expected to overcome these problems, however, the genomic sequences needed to establish this system are not available for many S-RNases. Here, we determined partial genomic sequences of eight S-RNases, and used the information to design new primer and to select 17 restriction enzymes for the discrimination of 22 S-RNases by CAPS. Using the system, the S genotypes of three cultivars were determined. The genomic sequence-based CAPS system would be useful for S genotyping and analyzing new S haplotypes of apple.     

Expression of a fusion protein of Pyrus pyrifolia S-RNase with glutathione-S-transferase in E. coli

S-RNase is a style-specific ribonuclease which is associated with gametophytic self-incompatibility. An expression vector of a fusion protein of Pyrus pyrifolia(Japanese pear) S3-RNase with glutathione-S-transferase (GST) was constructed and transformed into E. coli. Using this system, the fusion protein, GST-S3-RNase, was expressed as an active form and can be used for screening pollen S-gene product(s).     

Protein Sequences of Two Keto Ester Reductases: Possible Identity as Hypothetical Proteins

We determined the amino acid sequences of two keto ester reductases (YKER-V and -VI) purified from a cell-free extract of Saccharomyces cerevisiae. The N-terminal and internal amino acid sequences of YKER-VI (AcKR) were in agreement with the sequence of hypothetical 36.4-kDa protein (S. cerevisiae chromosome X reading frame ORF YJR105w) in yeast. The N-terminal amino acid sequence of YKER-V was also identical with that of the hypothetical protein coded by yeast chromosome XIV or II. These results suggested that two hypothetical proteins were expressed as keto ester reductases in yeast cells.     

Pollen S–locus F–box proteins of Petunia involved in S–RNase‐based self‐incompatibility are themselves subject to ubiquitin‐mediated degradation

Many flowering plants show self‐incompatibility, an intra‐specific reproductive barrier by which pistils reject self‐pollen to prevent inbreeding and accept non‐self pollen to promote out‐crossing. In Petunia, the polymorphic S–locus determines self/non‐self recognition. The locus contains a gene encoding an S–RNase, which controls pistil specificity, and multiple S‐locus F‐box (SLF) genes that collectively control pollen specificity. Each SLF is a component of an SCF (Skp1/Cullin/F‐box) complex that is responsible for mediating degradation of non‐self S‐RNase(s), with which the SLF interacts, via the ubiquitin–26S proteasome pathway. A complete set of SLFs is required to detoxify all non‐self S‐RNases to allow cross‐compatible pollination. Here, we show that SLF1 of Petunia inflata is itself subject to degradation via the ubiquitin–26S proteasome pathway, and identify an 18 amino acid sequence in the C‐terminal region of S₂‐SLF1 (SLF1 of S₂ haplotype) that contains a degradation motif. Seven of the 18 amino acids are conserved among all 17 SLF proteins of S₂ haplotype and S₃ haplotype involved in pollen specificity, suggesting that all SLF proteins are probably subject to similar degradation. Deleting the 18 amino acid sequence from S₂‐SLF1 stabilized the protein but abolished its function in self‐incompatibility, suggesting that dynamic cycling of SLF proteins is an integral part of their function in self‐incompatibility.     

Two rice <Emphasis Type="Italic">DMC1</Emphasis> genes are differentially expressed during meiosis and during haploid and diploid mitosis

We have cloned two rice homologues of yeast DMC1, a meiosis-specific gene required for recombination between homologous chromosomes. We show that rice DMC1A and DMC1B were produced by a gene duplication event that occurred after rice separated from the common ancestor of the cereals. The predicted proteins contain 344 amino acids, of which all but 7 are conserved between the two homologues. Between bases -1 and -245, the two promoters share six invariant blocks of sequence of 10-28 bp, interspersed in variable sequences. Both DMC1A and DMC1B are expressed in pollen mother cells coincident with meiosis, and in diploid non-meiotic tissues such as calli and root tips. DMC1B is also expressed in haploid male gametophytes during pollen maturation and in diploid zygotic embryos and endosperm after pollination. These data suggest that DMC1B, either alone or in combination with DMC1A, contributes to recombination during meiosis and during haploid and diploid mitosis.     

Isolation and characterization of five novel high molecular weight subunit of glutenin genes from Triticum timopheevi and Aegilops cylindrica

Analysis by SDS-PAGE of total protein fractions from single seeds of Aegilops cylindrica (genomes C and D) and Triticum timopheevi (genomes A and G) showed the presence of three bands corresponding to high molecular weight subunits of glutenin (HMW subunits) in the former and two major bands and a minor band corresponding to HMW subunits in the latter. Three Ae. cylindrica and two T. timopheevi HMW subunit gene sequences, each comprising the entire coding region, were amplified by polymerase chain reaction (PCR) and their complete nucleotide sequences determined. A combination of N-terminal amino acid sequencing of the proteins identified by SDS-PAGE and alignments of the derived amino acid sequences of the proteins encoded by the PCR products identified the Ae. cylindrica HMW subunits as 1Cx, 1Cy and 1Dy, and the T. timopheevi HMW subunits as 1Gx, 1Ax and 1Ay. It was not clear whether or not a 1Gy HMW subunit was present in T. timopheevi. The PCR products from Ae. cyclindrica were derived from 1Cy and 1Dy genes and a silent 1Dx gene containing an in-frame internal stop codon, while those from T. timopheevi were derived from 1Ax and 1Ay genes. The 1Cx, 1Gx and 1Gy sequences were not amplified successfully. The proteins encoded by the five novel genes had similar structures to previously characterized HMW subunits of bread wheat (Triticum aestivum). Differences and similarities in sequence and structure, and in the distribution of cysteine residues (relevant to the ability of HMW subunits to form high M_r polymers) distinguished the HMW subunits of x- and y-type and of each genome rather than those of the different species. There was no evidence of a change in HMW subunit expression or structure resulting from selective breeding of bread wheat. The novel 1Ax, 1Ay, 1Cy and 1Dy HMW subunits were expressed in Escherichia coli, and the expressed proteins were shown to have very similar mobilities to the endogenous HMW subunits on SDS-PAGE. The truncated 1Dx gene from Ae. cylindrica failed to express in E. coli, and no HMW subunit-related protein of the size predicted for the truncated 1Dx subunit could be identified by immunodetection in seed extracts.     

Structural and molecular analysis of self-incompatibility in almond (<Emphasis Type="Italic">Prunus dulcis</Emphasis>)

. A multi-approach was used to study different aspects of self-incompatibility (SI) in almond (Prunus dulcis). First, a population of almond cultivars was characterised as to their individual S-allele combination using separation of stylar protein extracts (non-equilibrium pH gradient electrofocusing) followed by staining for RNase activity, which led to the identification of one putative new allele and several new S-allele combinations. Second, a field pollination scheme was designed to study pollen tube progression and to obtain a spatial and temporal characterisation of this reproductive stage in both incompatible and compatible crosses. In addition, an anti-serum was raised against a synthetic peptide designed from an almond S-protein (S₈) and used for immunological in situ detection in pistil cryosections. S-RNases were found to accumulate intercellularly in the stylar transmitting tissue as previously reported for other rosaceous species. The results are discussed in view of the evolution of the gametophytic SI system and the models proposed for its mechanism. Gametophyte selection is also proposed as an important intraspecific barrier to fertilisation in this species.     

Primary structure and chemical modification of some amino acid residues of bifunctional alginate lyase from a marine bacterium Pseudoalteromonas sp. strain no. 272

The entire amino acid sequence of bifunctional alginate lyase from Pseudoalteromonas sp. strain No. 272 were determined by two approaches, Edman degradation of the peptides obtained from protease digestion of the enzyme protein and analysis of PCR products of the structural gene. The former resulted in incomplete amino acid sequence in the entire sequence, due to lacking of the proper peptides from the protease digestion. To compensate for this lack of sequences we applied the method of PCR of the structural gene that was initially elucidated from the primers designed from N- and C-terminal amino acid sequences of the enzyme. The results of the amino acid sequences from these two approaches showed good agreement. The enzyme consisted of 233 amino acid residues with a molecular mass of 25,549.5, including the sole W and cystine residue. The sequence homology search among the other alginate lyases from different origins indicated that they were very weakly homologous, with the exception of the sequence homology (80.3%) of Pseudoalteromonas elyakovii alginate lyase. The consensus sequence, YFKhG + Y-Q (Wong, T. Y., Preston, L. A., and Schiller, N. L. 2000. Annu. Rev. Microbiol. 54: 289–340) in the C-terminal regions was conserved. The kinetic analyses of chemical modification of some amino acid residues of the enzyme showed that W, K, and Y appeared to be important in the enzyme function.     

Molecular characterization of the mitochondrial elongation factor EF-Tu gene in rice (<Emphasis Type="Italic">Oryza sativa</Emphasis> L.)

The mitochondrial elongation factor EF-Tu (tufM) in rice (Oryza sativa L.) was isolated and characterized. The rice tufM cDNA clone contained 1,726 nucleotides and coded for a 453 amino acid protein including a putative mitochondrial transit peptide of 64 amino acid residues. This coding region was composed of 12 exons and 11 introns. The deduced amino acid sequence showed 62% and 88% identities with rice chloroplast EF-Tu (tufA) and Arabidopsis mitochondrial EF-Tu, respectively. As previously observed for the rice tufA gene, the tufM gene is likely present as one copy in rice. The mitochondrial EF-Tu gene was differentially expressed during flower development, and the other translational EF-Tu genes (chloroplast EF-Tu and cytosolic EF-1 alpha) were also distinctly expressed in a temporal manner. Phylogenetic analysis of the rice tufM gene showed that the mitochondrial tufA homologue of Reclinomonas was more closely related to the mitochondrial tufM genes of flowering plants than fungal and other mitochondrial tuf genes. In addition, the tufM encoded an N-terminal extension showing significant similarity to that of rps14 (or sdhB), which is also a nuclear-encoded rice mitochondrial gene.     

Lactate dehydrogenase from the Antarctic eelpout, Lycodichthys dearborni

As part of our studies to examine the molecular basis of cold-adaptation, we have determined the kinetic properties, thermal stability and deduced amino acid sequence of the enzyme lactate dehydrogenase (LDH) from an Antarctic zoarcid fish, Lycodichthys dearborni. Unlike Antarctic notothenioid fish which are endemic to the Southern Ocean, zoarcid fish are cosmopolitan and have a substantially longer evolutionary history as a sub-order. The A₄-LDH isoform was isolated and purified from the white muscle of L. dearborni. The kinetic parameters K_m^PYR and k_cat were determined at temperatures from 0 to 25°C. K_m^PYR was substantially higher at low temperatures than those from Antarctic and temperate notothenioid fish, whereas k_cat at these temperatures was essentially the same as those of the other fish LDH in this study. The sequence of L. dearborni A₄-LDH was determined from cDNA derived from white muscle RNA and found to be similar to, but distinct from, the A₄-LDH sequences of Antarctic notothenioid fish. Molecular modelling based on the structure of the A₄-LDH from Pagothenia borchgrevinki suggested that three conservative amino acid changes within the core of the protein that are not directly part of the active site but which might nonetheless influence the active site, may be important in cold-adaptation in L. dearborni A₄-LDH, and that several other changes on the surface of the protein might also play a role in cold-adaptation.