Hexokinase like protein from Drosophila melanogaster.

Kulkarni, Gauri V., Dileep N. Deobagkar, and Deepti D. Deobagkar.  Molecular Biology Research Laboratory, Department of Zoology, University of Pune, Ganeshkhind, Pune-411007. Maharashtra,  INDIA.

Summary

     A genomic clone from D. melanogaster, when completely sequenced and analyzed, showed the presence of an uninterrupted coding sequence of 1398 bp coding for a 465 amino acid ORF. Although this protein showed overall 31% identity and 56% similarity to the mammalian HK1, critical glucose binding residues are altered in this protein. In mice, a moonlighting role has been assigned to HK1 testis specific isoform in the sperm egg interaction suggesting that the glucose binding ability of hexokinases can be co-opted for carbohydrate binding in the extracellular matrix of the egg. Conservation of such homologue in Drosophila raises very interesting possibilities about moonlighting roles of hexokinases during evolution.

     Hexokinases ({ATP:d-hexose6-phosphotransferase}, EC2.7.1.1) are ubiquitous housekeeping enzymes catalyzing the  phosphorylation of glucose by ATP, which occurs in all eukaryotic and prokaryotic cells as the first step in the utilization of glucose (Grossbard and Schimke, 1966; Knutsen et al., 1969; Faulkner and Jones, 1976; Peters and Neet, 1977).  Mammals show four hexokinase isoforms (Type I-IV) (Grossbard and Schimke, 1966).  Apart from acting as a hexokinase, multiple moonlighting roles have been assigned to hexokinases. Rat brain HK-1 was shown to have autophosphorylating and protein kinase activity (Adams et al., 1994), the yeast hexokinase pII was shown to involved in repression of SUC-II gene, which encodes for an enzyme involved in disaccharide metabolism (Piller et al., 1998). Immunolocalization studies indicated the presence of testes specific HK-1 isoform on the mouse sperm surface suggesting its probable role in sperm and egg zona pellucida (ZP3) interaction (Kalab et al., 1994; Travis et al., 1998). Based on biochemical and genetic analysis, four isoforms of hexokinase from D. melanogaster have been reported (Murray and Ball, 1967; Cavener, 1980; Madhavan et al., 1972; Voelker et al., 1978). Using native gel electrophoresis, Murray and Ball (1967) have described a testes specific isoform Hex T. Hex A and Hex B isozymes are products of a single X-linked locus (8D/ X), which are largely expressed in adult thorax and larval muscle tissue, respectively, while Hex C (52 E/ 2R) enzyme is expressed in adult and larval fat body.  Here, we report the sequence analysis of a 465 amino acid coding ORF, which, although it has homology to hexokinases, shows remarkable alteration in typical substrate binding domains.

     A genomic clone from D. melanogaster genomic library EMBL3 was completely sequenced. Analysis of a 4509 bp region revealed presence of two genes separated by a short variable-length intergenic sequence of 41 bp (Figure 1). The first gene exhibits an uninterrupted coding sequence of 1398 bp, while the second gene contains a coding sequence of 1359 bp that is interrupted by a single intron of 88 bp.  The sequence is deposited in EMBL database (Accession No. AJ 271350). This sequence was localized to 97B on the third polytene chromosome (Data not shown), which is consistent with its localization reported in Drosophila Genome Project (Gadfly Gene CG54430) (Adams et al., 2000)  and is different from that of  HexA (8D/X) and HexC (51B/2R).

    The 465 aa long protein (DHK-465) encoded by the first gene (ORF-1) has a molecular mass of 50.2 kDa  and theoretical pI of 5.99. BLAST analysis of this ORF showed homology to known hexokinases.  It showed 31% identity and 56% similarity with human HK-I isoform. It also showed 32% identity and 55% similarity with Hex C, 36% identity and 62% similarity with Hex A from Drosophila.  The two other ORF’s (ORF-2, ORF-3) together form a complete protein of 453 amino acids (DHK-453) with a calculated molecular mass of 49 kDa and theoretical pI of 8.56. This protein also showed homology to known hexokinases.  It showed 41% identity and 59% similarity with human HK-1; 56% identity and 64% similarity with Hex A; 46% identity and 73% similarity with Hex C from Drosophila. Remarkably, DHK-465 and DHK-453 shared only 36% identity and 60% similarity with each other.

     Further, glucose and ATP binding domains of DHK-465 and DHK-453 were compared with Drosophila hexokinases (Hex A and Hex C), mammalian and yeast hexokinase sequences (Figure 2A and B).  DHK-465 sequence shows the presence of glucose binding domain at 138-165 aa and ATP binding domain at 68-100 aa residues, respectively.  In DHK-453, glucose binding domain resides at amino acid position 133-161 and ATP binding domain resides at position 63-83, respectively.  8 major residues in the glucose binding domain are altered in DHK-465 protein (Figure 2A). Site directed mutagenesis of yeast hexokinase (Y-HKA) demonstrated the importance of Ser158 and Phe160, which are critical for glucose binding (Kraakman et al., 1999). These residues are altered in DHK-465 with alanine and tyrosine, respectively (Figure 2A). Whereas, DHK-453 sequence shows the conservation in these residues.  In addition to the involvement of the amino acid residues of glucose binding domain, site directed mutagenesis of human HK-1, demonstrated that Asp657, Glu708 and Glu742 residues were involved in glucose binding and catalysis (Arora et al., 1991).  Mutation in Glu708 of HK-1 resulted in 50-fold increase in the km for glucose. Asp657 and Glu708 of HK-1 are altered to Val and Asp respectively in DHK-465. Whereas, Glu742 is conserved at 286 position in DHK-465. All these residues are conserved in DHK-453 sequence (Table 1).

     Out of 13 residues important in the ATP binding domain, DHK-465 shows alteration in 5 residues (Figure 2B).  Arg539 of human HK-1 has been shown to be essential for catalysis as it stabilizes the transition state product ADP-HK (Zeng et al., 1998). It is conserved in DHK465 at Arg83.  Mutation of Gly534  affected ATP binding of HK-1 (Zeng et al., 1998).  This residue is altered in DHK-465 with a methionine. Earlier studies from the site directed mutagenesis of human brain HK-1 have shown that Asp532 in the ATP binding domain interacts with Mg2+ ion and Thr680 stabilize the g–phosphoryl group of ATP, and their interactions are important for the stabilization of the transition state (Zeng et al., 1996). Both Asp532 and Thr680 are replaced by glutamate and serine in DHK-465, respectively, (Figure 2B and Table 1). Mutations in Gly862 and Gly679 in human HK-1 has a significant influence on the binding affinity of ATP (Zeng et al., 1998).  These residues are conserved in DHK-465.  Most of the ATP binding residues are well conserved in DHK-453 (Figure 2B and Table 1) as well as other Drosophila hexokinases, Hex C (AF2347469 and AJ309864) and Hex A (AAF46507, AAG23047) (Data not shown).

     Thus, although all these observations suggest that the DHK-465 shows overall similarity to hexokinases, the amino acid residues crucial for substrate binding and catalysis are altered in this protein suggesting that this protein has much less efficiency to act as hexokinase. DHK-453 on the other hand shows conservation of all residues significant for substrate binding and catalysis representing a typical hexokinase. Recent findings suggest that somatic hexokinases may have additional locations and different functions within cells (Adams et al., 1994; Piller et al., 1998; Kalab et al., 1994; Travis et al., 1998). A sperm specific HK-1 isoform in mouse and humans has been attributed a role in sperm and zona pellucida (ZP) binding (Kalab et al., 1994; Travis et al., 1998).  It is therefore likely that the glucose-binding ability of hexokinases is co-opted for carbohydrate binding in the extracellular matrix of the egg (Travis et al., 1998).  The alterations in substrate binding and catalytic residues critical for a hexokinase in DHK-465 raises further interests about the probable functions associated with this protein in Drosophila.