Purification, characterization, and cDNA cloning of a galactose-specific C-type lectin from Drosophila melanogaster.

We purified a lectin from a pupal extract of Drosophila melanogaster. This lectin agglutinated trypsinized and glutaraldehyde-fixed bovine red blood cells in the presence of calcium or magnesium. The hapten sugar of this lectin was galactose. The molecular mass of the intact lectin was determined to be 41 kDa, and it comprised 14- and 17-kDa subunits. The 17-kDa subunit was shown to be a glycosylated form of the 14-kDa subunit. Analysis of the cDNA for this lectin revealed that the 14-kDa subunit consists of 163 amino acid residues and contains all residues conserved in various C-type lectins. It was suggested that the Drosophila lectin and Sarcophaga lectin share some properties and function similarly in defense and development, but probably they are not structural homologues.

We purified a lectin from a pupal extract of Drosophila melanogaster. This lectin agglutinated trypsinized and glutaraldehyde-fixed bovine red blood cells in the presence of calcium or magnesium. The hapten sugar of this lectin was galactose. The molecular mass of the intact lectin was determined to be 41 kDa, and it comprised 14-and 17-kDa subunits. The 17-kDa subunit was shown to be a glycosylated form of the 14-kDa subunit. Analysis of the cDNA for this lectin revealed that the 14-kDa subunit consists of 163 amino acid residues and contains all residues conserved in various C-type lectins. It was suggested that the Drosophila lectin and Sarcophaga lectin share some properties and function similarly in defense and development, but probably they are not structural homologues.
We have been studying the humoral lectin of Sarcophaga peregrina (flesh fly), focusing mainly on its biological function (18 -20). The Sarcophaga lectin was found in the hemolymph of immunized third instar larvae of Sarcophaga during a study on antibacterial proteins of this insect. Although naive larvae do not contain Sarcophaga lectin, it is induced promptly when larvae are injected with foreign cells such as bacteria or when their body wall is simply injured. The same lectin was found to appear at the embryonic and pupal stages in naive animals (21). Subsequently, it was suggested that Sarcophaga lectin is essential for the differentiation of imaginal discs at the pupal stage (22). Thus, it is becoming more and more clear that this lectin participates in both the defense and development (ontogeny) of this insect, which indicates that a single lectin can play two independent biological roles (23,24). To determine the biological roles of Sarcophaga lectin more precisely, it would be useful to identify the Drosophila counterpart of the Sarcophaga lectin, if any, because genetical techniques are available for Drosophila.
In this paper, we report the purification, characterization, and cDNA cloning of a Drosophila lectin. So far, no lectin activity has been reported in Drosophila. We found that intact red blood cells are not effective for the detection of Drosophila lectin activity. This seems to be the reason why lectin activity was not detected in Drosophila for a long time. Like the Sarcophaga lectin, the Drosophila lectin is a C-type lectin and specifically recognizes galactose. From the similarity of the expression of the genes for the Drosophila and Sarcophaga lectins, we propose the functional similarity of these two lectins. However, their molecular masses are different, and their amino acid sequence identity is at most 25%.

Animals and Pupal Extract Preparation-Drosophila melanogaster
CS2 pupae collected within 4 -8 h after pupation were used throughout. Packed pupae (30 ml) were homogenized in 30 ml of buffered insect saline (10 mM Tris/HCl, pH 7.4, 130 mM NaCl, 5 mM KCl, and 1 mM CaCl 2 ) and then centrifuged at 35,000 ϫ g for 10 min, and the resulting supernatant (pupal extract) was used as the starting material for purification of the Drosophila lectin.
Assay of Hemagglutinating Activity-Bovine red blood cells (BRBC) 1 trypsinized and fixed with glutaraldehyde were used as target cells. These cells were prepared according to the method of Nowak et al. (25). Briefly, a suspension of 4% (w/v) BRBC in 100 mM phosphate buffer, pH 7.4, containing 40 mM NaCl was treated with 1 mg/ml trypsin for 1 h at 37°C. The cells were washed well with phosphate-buffered saline (PBS; 75 mM phosphate buffer, pH 7.2, containing 75 mM NaCl) and then suspended in PBS at a concentration of 20%. Glutaraldehyde was added to this suspension (final, 1% (v/v)), which was then mixed well and kept for 1 h at room temperature. The cells were successively washed well with PBS containing 0.1 M glycine and PBS, respectively, and then stored as a 10% suspension in PBS (BRBC suspension).
To measure lectin activity, 25 l of a serially diluted sample was mixed with 25 l of a 2.5% BRBC suspension containing 1% (w/v) bovine serum albumin in a well of a V-bottomed microtiter plate and then incubated for 1 h at 37°C. Agglutinated BRBC formed a diffuse mat, whereas unagglutinated BRBC formed a clear dot on the bottom of a well. Lectin activity (titer) was expressed as the reciprocal of the maximum dilution of the test sample causing hemagglutination.
Purification of Drosophila Lectin-About 50 ml of the pupal extract was mixed well with an equal volume of the 10% BRBC suspension and then kept for 1 h at 4°C. The cells were then washed well with buffered insect saline containing 1 mM CaCl 2 by repeated centrifugation until the absorbance at 280 nm dropped below 0.005. Then the cells were suspended in 10 ml of buffered insect saline containing either 2 mM EDTA instead of CaCl 2 (EDTA-eluted lectin) or 0.2 M galactose (galactose-eluted lectin) and incubated for 1 h at 4°C to elute the Drosophila lectin from the BRBC. Then the suspension was centrifuged to remove the cells, and the clear supernatant was collected. The galactose-eluted lectin was dialyzed against buffered insect saline for 12-16 h to remove galactose. CaCl 2 was added to the EDTA-eluted lectin at a final concentration of 4 mM. Homogeneous Drosophila lectin was eluted from the BRBC with both elution procedures.
Determination of Partial Amino Acid Sequences of Drosophila Lec-* This work was supported by a Grant-in-Aid for Specifically Promoted Research from the Ministry of Education, Science, Sports and Culture of Japan and by a Grant-in-Aid BMP (Bio Media Program) 96-V-1-5-8 from the Ministry of Agriculture and Fisheries of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) D83705.
tin-The purified Drosophila lectin (10 -40 g) was subjected to SDSpolyacrylamide gel electrophoresis and then electrophoretically blotted onto a polyvinylidene difluoride filter in 10 mM CAPS buffer, pH 11.0, containing 10% (v/v) methanol. After staining with Ponceau S, the band corresponding to the 14-kDa subunit of Drosophila lectin was excised, destained with 0.2 N NaOH for 1 min, and then incubated in acetic acid containing 0.5% (v/v) polyvinyl pyrrolidone-40 for 30 min at 37°C. Each filter was then cut into pieces and treated with lysyl-endopeptidase in 100 l of 50 mM Tris/HCl buffer, pH 9.5, at an enzyme protein ratio of 1:30 by weight for 20 h at 30°C. Acetonitrile was then added to the reaction mixture at a final concentration of 5% (v/v) followed by sonication for 5 min. The mixture was then centrifuged at 10,000 ϫ g for 1 min to remove the pieces of filter, and then the supernatant was applied to a reverse-phase HPLC column of Synchropak RP-P (C 18 , 250 ϫ 4.1 mm; Synchropak Inc., Leiden) connected to a Water's HPLC system. Peptides were eluted with a linear gradient of 0 -70% solution B (0.09% (v/v) trifluoroacetic acid in 70% acetonitrile) in solution A (0.09% trifluoroacetic acid). The resulting peptides were applied to a Shimadzu PPSQ-10 protein sequencer (Shimadzu, Tokyo, Japan).
cDNA Cloning of Drosophila Lectin-We determined the sequences of five peptides derived from the 14-kDa subunit of the purified Drosophila lectin. Then we synthesized degenerate primers corresponding to RELNSELVTF and DPNSLFK, which were included in two of these peptides, and performed reverse transcription-polymerase chain reaction with these primers using pupal mRNA as a template. The amplified sequence comprising 296 base pairs was cloned and used as a probe for screening Drosophila early pupal cDNA library in gt10. For this, recombinant phages were plated with Escherichia coli C600 hfl(a) at a density of about 20,000 plaques/plate, and the resulting plaques were blotted onto two nylon filters. Prehybridization was performed in 3 ϫ SSC (1 ϫ SSC ϭ 150 mM NaCl, 15 mM sodium citrate), 10 ϫ Denhardt's solution (1 ϫ Denhardt's solution ϭ 0.02% (w/v) each of Ficoll-400, bovine serum albumin, and polyvinylpyrrolidone) containing 25 g/ml of single-stranded salmon sperm DNA at 50°C for 5 h. Hybridization was performed in 3 ϫ SSC, 10 ϫ Denhardt's solution containing 25 g/ml of single-stranded salmon sperm DNA and the 32 P-labeled probe at 50°C for 12 h. Then the filters were washed three times with 2 ϫ SSC containing 0.1% SDS for 20 min each time at room temperature and then at 50°C. The blot was then exposed to x-ray film at Ϫ80°C.
For sequencing, the isolated cDNA clone in gt10 was first subcloned into pGEM-3ZF(ϩ). The nucleotide sequences of both strands were determined using a Taq Dye Deoxy Cycle Sequencing kit (Applied Biosystems).
Northern and Southern Blot Hybridizations-Northern blot hybridization was performed in 50% (v/v) formamide, 5 ϫ SSC, 1 ϫ Denhardt's solution, 50 mM phosphate buffer, pH 6.5, 0.2 mg/ml single-stranded salmon sperm DNA for 16 h at 42°C. Then the filters were washed twice for 10 min each at room temperature and once for 15 min at 42°C with 2 ϫ SSC containing 0.1% SDS. For Southern blot hybridization, prehybridization was performed in 6 ϫ SSC, 10 ϫ Denhardt's solution containing 0.5% SDS, and 0.1 mg/ml single-stranded salmon sperm DNA at 42°C for 16 h. Then hybridization was performed under the same conditions. After hybridization, the filters were washed twice for 10 min each time at room temperature with 2 ϫ SSC containing 0.1% SDS, twice for 10 min each time with 0.2 ϫ SSC containing 0.1% SDS, and finally once for 15 min at 42°C with 0.2 ϫ SSC containing 0.1% SDS. The DNA used as a probe was 32 P-labeled Drosophila lectin cDNA. The probe was labeled with [␣-32 P]dCTP using a multiprime labeling kit.
Other Methods-DNA manipulations, including restriction enzyme digestions, DNA ligation, plasmid isolation, and E. coli transformation, were carried out by standard methods. Chromosomal localization of the Drosophila lectin gene was performed according to the Drosophila laboratory manual (26). RNA was extracted by the guanidine-thiocyanate method (27). Protein was determined by the method of Lowry et al. (28) using bovine serum albumin as a standard.

Detection of Drosophila Lectin in the Pupal
Extract-Sarcophaga lectin is known to be synthesized by the fat body at the pupal stage and to be secreted into the hemolymph (20), so a pupal extract of Drosophila was thought to be an appropriate source for the purification of Drosophila lectin, if any. We examined the hemagglutinating activity of the pupal extract using various red blood cells. No intact or trypsinized and glutaraldehyde-fixed red blood cells of bovine, horse, sheep, guinea pig, and chicken were agglutinated in the presence of the pupal extract, but significant agglutination was detected only when trypsinized and glutaraldehyde-fixed BRBC were used. It is known that the susceptibility of red blood cells to agglutinin generally increases when they are treated with trypsin (29).
We examined the effects of various sugars and divalent cations on the hemagglutinating activity of the pupal extract. As shown in Table I, galactose and lactose caused 50% inhibition of the hemagglutinating activity at concentrations of 0.1 and 1 mM, respectively. Other sugars so far tested had almost no effect, except D-glucose, which inhibited the agglutinating activity at a much higher concentration. We found that the hemagglutinating activity required divalent cations. Essentially no hemagglutinating activity was detected in the absence of a divalent cation, but a saturation level of activity was detected in the presence of 1 mM CaCl 2 . MgCl 2 was also effective, but its effect was a little weaker than that of CaCl 2 (Table II). These results suggested that the hemagglutinating activity in the pupal extract of Drosophila is due to a galactose-binding C-type lectin like the Sarcophaga lectin.
The hemagglutinating activity in the pupal extract was found to significantly decrease when it was kept at room temperature for several hours, suggesting that Drosophila lectin in the extract is unstable.
Purification of Drosophila Lectin-Taking advantage of the galactose specificity and calcium dependence, we were able to purify the Drosophila lectin in the pupal extract to homogeneity in a single step using the affinity to trypsinized and glutaraldehyde-fixed BRBC. The Drosophila lectin bound to BRBC was eluted with either a galactose or EDTA solution. The eluted Drosophila lectin gave 14-and 17-kDa protein bands on SDS-polyacrylamide gel electrophoresis, but the intensity of the 17-kDa band was much fainter than that of the 14-kDa one. As summarized in Table III, the galactose solution was better than the EDTA one for elution of Drosophila lectin from BRBC in terms of its specific activity and recovery. Therefore, we routinely used the former solution.
To determine whether or not the 14-and 17-kDa proteins are both subunits of Drosophila lectin, we further subjected the purified lectin to HPLC on a molecular sieve column (Superose 12). As shown in Fig. 1A, the Drosophila lectin was eluted as a single peak, and its molecular mass was estimated to be 41 kDa. Fractions corresponding to this peak were subjected to SDS-polyacrylamide gel electrophoresis. Again one major 14-kDa and one minor 17-kDa band were detected, as shown in  Fig. 1B. These results suggested the possibility that Drosophila lectin is essentially a homotrimer of the 14-kDa protein, and this subunit is sometimes partly glycosylated, resulting in the 17-kDa protein.
To examine this possibility, we treated Drosophila lectin with N-glycanase to see if its electrophoretic profile changed. As is evident from Fig. 2, the 17-kDa band disappeared with an increase in the dose of N-glycanase, but the 14-kDa band remained unchanged. Moreover, no new band appeared with the disappearance of the 17-kDa band, indicating that the deglycosylated 17-kDa protein comigrated with the 14-kDa protein.
These results strongly suggested that the 17-kDa protein is a glycosylated form of the 14-kDa protein and that it is converted to the 14-kDa protein on treatment with N-glycanase.
We determined the amino acid sequences of the amino-terminal 20 residues of the 17-and 14-kDa proteins and found that they were identical, being REKFSIQVNEGNTFGALVKA, indicating that they are derived from the same gene. Possibly, a minor population of Drosophila lectin contains a glycosylated subunit.
cDNA Cloning of Drosophila Lectin-We isolated cDNA for the Drosophila lectin. For this, we first determined the partial amino acid sequences of peptides derived from the 14-kDa protein. The sequences are: peptide 1, RELNSELVTFET-DQEFDAVT; peptide 2, FELNDRPCSQDPNSLFK; peptide 3, FETESLNWYEAYEA; peptide 4, YICEAPEMETIS; and peptide 5, EHCIHLGYIYK. Then we synthesized two degenerate oligodeoxyribonucleotide primers corresponding to parts of peptides 1 and 2, respectively, and performed reverse transcription-polymerase chain reaction with these primers, using pupal mRNA as a template. The resulting polymerase chain reaction product (296 base pairs) was cloned. Using this clone as a probe, we isolated two hybridization positive clones on the screening of about 100,000 plaques derived from a cDNA library of Drosophila pupae. The complete nucleotide sequences of both clones were determined and found to be identical. This cDNA encoded a protein consisting of 183 amino acid residues, and the sequences of all of peptides 1-5 derived from the 14-kDa protein and the amino-terminal 20 residues were included in this sequence, indicating that this is Drosophila lectin cDNA. The complete nucleotide sequence together with the putative amino acid sequence encoded by this cDNA are shown in Fig. 3A.
The amino-terminal amino acid residue was assigned as Arg at position 21. Therefore, the 14-kDa protein consists of 163 amino acid residues, and the 20 residues from the first Met are thought to be a leader peptide containing a signal sequence, which was supported by the hydropathy profile of this protein shown in Fig. 3B. On comparison of the sequence of Drosophila lectin with that of Sarcophaga lectin, the molecular size of Drosophila lectin was found to be about 2 ⁄3 that of Sarcophaga lectin, and Drosophila lectin lacked a large part corresponding to the carboxyl-terminal region of Sarcophaga lectin. The sequence of about 30 residues at the amino-terminal region was unique to Drosophila lectin, but most of the other sequence corresponded to that of carbohydrate recognition domain. Drosophila lectin contained 9 amino acid residues conserved in the carbohydrate recognition domains of known C-type lectins, as shown in Fig. 4. Moreover, Glu, Asp, and Asn residues, which were recently pointed out to be responsible for sugar and Ca 2ϩ binding (30), were also conserved in Drosophila lectin. Thus, this is clearly a C-type lectin. The Sarcophaga lectin was the lectin with the highest percentage of identity with the Drosophila lectin in the Protein Information Resource, SWISS-PROT, and Protein Research Foundation protein data bases. The overall sequence identity between the Sarcophaga and Drosophila lectins was at most 25%, but the sequence identity between their carbohydrate recognition domains was much higher.
The sequence Asn-Xaa-Ser or Thr is known to be the consensus sequence for the N-glycosylation site (31). There was one   1. Analysis of the purified Drosophila lectin on a molecular sieve column. A, the galactose-eluted lectin was subjected to HPLC on a Superose 12 column, which was calibrated with the following molecular mass markers: bovine serum albumin (66 kDa); ovalbumin (45 kDa); ␣-chymotrypsinogen (26 kDa); and cytochrome c (12 kDa). E, absorbance at 280 nm; q, hemagglutinating activity. B, electrophoretic profiles of fractions containing hemagglutinating activity. Lane numbers correspond to fraction numbers. The positions of the 14and 17-kDa proteins are indicated by arrows. such sequence, Asn-Gly-Ser, in this lectin. Possibly, this Asn residue in the 17-kDa subunit is glycosylated.
Expression of the Drosophila Lectin Gene-To determine the functional similarities of the Sarcophaga and Drosophila lectins, we investigated the expression of the Drosophila lectin gene by Northern blot hybridization. It is known that the Sarcophaga lectin gene is activated when the body wall of third instar larvae is injured (21). In Drosophila, the situation was essentially the same as in Sarcophaga. The Drosophila lectin gene was activated 4 -6 h after the larval body wall had been pricked with a thin needle, as shown in Fig. 5A. This was especially clear in second instar larvae, because no appreciable basal level expression of the Drosophila lectin gene was detected in naive larvae. In third instar larvae, significant basal level expression was detected, which was different from in Sarcophaga larvae, but this expression was clearly enhanced on their body injury.
The Sarcophaga lectin gene is known to be expressed transiently at the embryonic and early pupal stages during the normal development of Sarcophaga (21). We examined the expression of the Drosophila lectin gene at various developmental stages of Drosophila. As shown in Fig. 5B, no significant expression of the Drosophila lectin gene was detected in embryos or first or second instar larvae. However, its expression commenced in third instar larvae and was enhanced in early pupae but then decreased significantly at the late pupal stage. Significant expression of the Drosophila lectin gene was also detected in adults, but a clear difference existed between males and females. The expression in males was much more than that in females, suggesting that this lectin participates in a male-specific function in adults.
Analysis of the Drosophila Lectin Gene-To determine the copy number of the Drosophila lectin gene, we performed Southern blot hybridization with Drosophila DNA digested FIG. 3. Nucleotide sequence of cDNA for the Drosophila lectin and its hydropathy profile. A, nucleotide numbers are shown above the nucleotide sequence, and the deduced amino acid sequence is shown using one-letter symbols below the nucleotide sequence. The numbers of amino acid residues starting from the first Met are given at the right of each line. The determined partial amino acid sequences are underlined, and a possible N-glycosylation site is boxed. The asterisk and double underlining denote the termination codon and poly(A) addition signal, respectively. B, the distribution of hydrophobic and hydrophilic domains was analyzed by the method of Kyte and Doolittle (34). The numbers of amino acid residues are shown at the bottom. The data are plotted as hydrophobic and hydrophilic portions above and below the vertical line, respectively. with various restriction enzymes. The results are shown in Fig.  6. A single band was detected irrespective of the restriction enzyme used, indicating that the Drosophila lectin gene is a single copy gene like the Sarcophaga lectin one (32). On chromosome mapping, the Drosophila lectin gene was found to be located at position 37B on the left arm of the second chromosome, as shown in Fig. 7. DISCUSSION We isolated and characterized a galactose-binding C-type lectin from a pupal extract of Drosophila. We succeeded in detecting lectin activity by using trypsinized and glutaraldehyde-fixed bovine red blood cells. The carbohydrate specificity and divalent cation requirement of the Drosophila lectin were almost the same as those of the Sarcophaga lectin (18). However, analysis of Drosophila lectin cDNA revealed that the sequence identity between the Drosophila and Sarcophaga lectins was at most 25%. There are many proteins with much higher sequence identity in Sarcophaga and Drosophila. For instance, the sequence of Drosophila cecropin A was found to be identical with that of Sarcophaga sarcotoxin IA (33). Considering these facts, the 25% sequence identity of the two lectins is too low to assume that they are structurally related proteins and that the genes for these lectins were derived from a common ancestral gene. Moreover, the molecular masses of these two lectins are very different.
Sarcophaga lectin was originally isolated from the hemolymph of injured larvae, indicating that it is a hemolymph lectin (18). Although Drosophila lectin was soluble in the pupal extract, we cannot exclude the possibility that it is a membrane-bound lectin, because several hydrophobic regions were found in the molecule on hydropathy analysis, whereas no appreciable hydrophobic region was detected in Sarcophaga lectin (32). Therefore, it is difficult to conclude at this stage that the Drosophila lectin is a true structural homologue of the Sarcophaga lectin, if any, in Drosophila. However, as the pupal extract does not seem to contain other lectins, it would be worthwhile determining whether or not this Drosophila lectin is a functional homologue of the Sarcophaga lectin.
Drosophila lectin was suggested to be a trimer of the 14-kDa subunit, and one of these subunits was sometimes glycosylated, resulting in a 17-kDa subunit. However, the molar ratio of the 14-and 17-kDa subunits was estimated to be more than 2:1, judging from the intensities of the electrophoretic bands of these two subunits of Drosophila lectin. Therefore, all the lectin molecules do not necessarily contain a glycosylated subunit, and the biological relevance of glycosylation of the Drosophila lectin is not clear. This situation was essentially the same as in the case of the Sarcophaga lectin.
We found that expression of the Drosophila lectin gene was enhanced significantly by pricking of the larval body wall with a thin needle, suggesting that it is a defense molecule like the Sarcophaga lectin. Expression of the Drosophila lectin gene was also enhanced at an early pupal stage when imaginal discs differentiate. These expression patterns of the Drosophila lectin gene are very similar to those of the Sarcophaga lectin gene (21). As Sarcophaga lectin was shown to be a defense molecule as well as an autocrine regulator of imaginal disc differentiation (24), Drosophila lectin is assumed to have similar functions in the defense and development of Drosophila.
Although the two lectins are assumed to have common roles, they also seem to have their own roles. The Sarcophaga lectin gene was expressed at an embryonic stage (21), whereas no significant expression of the Drosophila lectin gene was detected in embryos. The Drosophila lectin gene was expressed in third instar larvae, whereas no significant expression of the Sarcophaga lectin gene was detected in third instar larvae (21). These differences in gene expression may be due to a functional difference between the two lectins. In adults, expression of the Drosophila lectin gene in males is significantly higher than that in females, irrespective of their age, which is also different from the expression of the Sarcophaga lectin gene.
We think that the Sarcophaga and Drosophila lectins share some properties and function similarly in the defense and development of these insects. Because the Drosophila lectin gene was assigned on the second chromosome, analyses of mutant and/or transgenic flies will provide more information on the function of this lectin.