- Email: [email protected]
A thrombospondin homologue in Drosophila melanogaster: cDNA and protein structure
Gene 269 (2001) 177±184
www.elsevier.com/locate/gene
A thrombospondin homologue in Drosophila melanogaster: cDNA and protein structure Kenneth W. Adolph* University of Minnesota, Department of Biochemistry, Molecular Biology and Biophysics, 6-155 Jackson Hall, 321 Church St. S.E., Minneapolis, MN 55455, USA Received 12 January 2001; received in revised form 27 February 2001; accepted 14 March 2001 Received by A.J. van Wijnen
Abstract The cDNA of a thrombospondin homologue in Drosophila melanogaster (DTSP) has been sequenced, and structural features of the translated protein analyzed. Thrombospondin proteins had previously been identi®ed only in vertebrates, including human and mouse. Comparison with the genomic sequence revealed that the DTSP gene is divided into 13 exons; The translation initiation site is in exon 2. The transcription start site was analyzed using a PCR procedure, and the longest transcripts were found to initiate about 300 bp 5 0 of the predicted start site. The open reading frame of the DTSP cDNA encodes a protein that has 1060 amino acid residues. The polypeptide is composed of domains or repeats characteristic of the TSP3/TSP4/COMP subfamily of thrombospondin proteins: Amino-terminal domain, four Type II repeats, seven Type III repeats, carboxyl-terminal domain. The protein is highly acidic, particularly in the region of Type III repeats, with an Asp 1 Glu content of 15.8%. A signal peptide was detected at the N-terminus, which indicates that DTSP, like other TSPs, functions as an extracellular protein. Ten Asn residues were identi®ed as potential glycosylation sites. Alignment of the amino acid sequences of the Drosophila TSP with human TSP1±TSP4 and COMP demonstrated a high degree of homology between the four Type II repeats, seven Type III repeats, and C-terminal domain. q 2001 Elsevier Science B.V. All rights reserved. Keywords: Nucleotide sequence; Exon/intron organization; Transcription start site; Open reading frame; Polypeptide domains; TSP3/TSP4/COMP subfamily
1. Introduction The thrombospondin family of proteins consists of ®ve members, in humans and other vertebrates, which can be divided into two subgroups based upon their polypeptide domain structures (Bornstein and Sage, 1994; Adams et al., 1995). TSP1 (Baenziger et al., 1971; Hennessy et al., 1989) and TSP2 (Bornstein et al., 1991; LaBell and Byers, 1993; Adolph et al., 1997) constitute the ®rst subgroup and contain (a) an amino-terminal domain, (b) a procollagen homology region, (c) three Type I (TSP) repeats, (d) three Type II (EGF-like) repeats, (e) seven Type III (Ca 21-binding) repeats, (f) a carboxyl-terminal domain. The mature TSP1 and TSP2 molecules are disul®de-bonded trimers Abbreviations: BLAST, Basic Local Alignment Search Tool; COMP, cartilage oligomeric matrix protein; DTSP, Drosophila melanogaster thrombospondin homologue; EGF, epidermal growth factor; HTSP, human thrombospondin; NCBI, National Center for Biotechnology Information; ORF, open reading frame; PCR, polymerase chain reaction; TSP, thrombospondin; UTR, untranslated region * Tel.: 11-612-625-8467; fax: 612-625-2163. E-mail address: [email protected] (K.W. Adolph).
that are secreted and interact with a variety of cells, extracellular matrix proteins, and plasma proteins. These interactions account for the involvement of TSP1 and TSP2 in processes that include angiogenesis, tumor growth, embryonic development, tissue differentiation, and blood coagulation. For example, interactions involving the Type I repeats are primarily responsible for the function of TSP1 and TSP2 as potent inhibitors of angiogenesis (Iruela-Arispe et al., 1999; Jimenez et al., 2000). TSP3 (Vos et al., 1992; Adolph et al., 1995; Adolph and Bornstein, 1999), TSP4 (Lawler et al., 1993, 1995), and COMP (cartilage oligomeric matrix protein) (Oldberg et al., 1992) make up the second thrombospondin subgroup. These polypeptides have a simpler structure consisting of (a) an amino-terminal domain, (b) four Type II repeats, (c) seven Type III repeats, and (d) a carboxyl-terminal domain. The proteins exist as pentamers of the modular polypeptides and are extracellular. The biological roles of TSP3 and TSP4 are not well understood. COMP is better characterized, since mutations in the COMP gene have been demonstrated to be the cause of human skeletal disorders (Hecht et al., 1995; Briggs et al., 1995).
0378-1119/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S 0378-111 9(01)00441-3
178
K.W. Adolph / Gene 269 (2001) 177±184
The genes for the thrombospondin proteins are at different chromosomal locations. The TSP1 gene, for example, is at 15q15 in humans, TSP2 is at 6q27, and TSP3 is at 1q2124. The TSP1 gene is composed of 22 exons, while the TSP2 and TSP3 genes possess 23 exons (Adolph et al., 1995). The exon sequences and exon/intron organization of each TSP are highly conserved between human and mouse. The nucleotide sequences of the TSP promoter/5 0 ¯anking regions in humans show no signi®cant homologies, even between subgroup members such as TSP1 and TSP2 (Adolph et al., 1997). This strengthens the conclusion that thrombospondin proteins have distinct functions. Besides human and mouse, TSP cDNAs and genes have been characterized for other vertebrate sources, including Xenopus (Lawler et al., 1993), bovine (Ueno et al., 1998), chicken (Lawler et al., 1991), and rat (Oldberg et al., 1992). The results presented here describe a thrombospondin homologue (DTSP) from a non-vertebrate source, the fruit ¯y Drosophila melanogaster. The sequence of the Drosophila TSP cDNA has been determined, and the deduced amino acid sequence has been analyzed and compared to the human thrombospondin amino acid sequences. 2. Materials and methods 2.1. Isolation of cDNA for a thrombospondin homologue in D. melanogaster cDNA corresponding to the DTSP coding exons was prepared by PCR using nested, antisense primers X and Y that included the translation stop codon, with primer Z that included the translation start codon. The sequences of the primers (obtained from Integrated DNA Technologies, Inc.) were: primer X, 5 0 -GTTAATGGCTCTGAGGATTGCG-3 0 ; primer Y, 5 0 -TCAGTCCTGCAACTCCACCTTC-3 0 ; primer Z, 5 0 -ATGATCGACTTCCCGATGCTGA3 0 . The primers were designed from the D. melanogaster predicted gene sequences (Adams et al., 2000). The template in the PCR reactions was Drosophila melanogaster embryo cDNA (QUICK-Clone cDNA; Clontech). The poly A 1 source for the cDNA was given as 18-hr Canton S. embryo. The thermostable DNA polymerase included in the PCR reactions was Expand High Fidelity polymerase mixture (Roche Molecular Biochemicals). The ampli®cation conditions, using a Perkin±Elmer Model 480 thermal cycler, were typically 948C, 45 s; 658C, 45 s; 728C, 2 min; 30 cycles. The products of the PCR reactions were analyzed on 2% agarose/ TAE gels. 2.2. Isolation of cDNAs containing the 5 0 UTR and 3 0 UTR and identi®cation of transcripts with the longest 5 0 UTR To prepare cDNA fragments containing the 5 0 UTR, PCR was carried out using nested, antisense primers A (in exon 6) and B (in exon 4) with each of a series of primers C1, C2, C3 near the 5 0 end. C1 was located at the predicted 5 0 end of
the DTSP transcript, while C2 and C3 were, respectively, 76 and 294 bp 5 0 of C1. Primer A had the sequence 5 0 GCACCTCCTGGTGTCTGTGATC-3 0 ; primer B was 5 0 CGTTCGCACTAATGTCCAACGT-3 0 . C3, which identi®ed the longest DTSP transcript, was 5 0 -TGCGTGTCCTAGGTGCTTCGAT-3 0 . PCR with these primers was carried out under the conditions described above. cDNA fragments containing the 3 0 UTR were prepared with a similar procedure. In this case, PCR used antisense primer W (5 0 -GGCCAACTGGATTGGCAACTCA-3 0 ) placed immediately 3 0 of the potential AATAAA poly A 1 addition site. 2.3. DNA sequencing procedures and sequence analysis DTSP cDNA fragments representing the coding exons, 5 0 UTR, and 3 0 UTR were puri®ed for sequencing from preparative agarose gels. cDNA bands of the correct size were excised from 2% agarose/TAE gels and the cDNAs were isolated with the QIAEX II Gel Extraction Kit (Qiagen). Sequences were determined by automated DNA sequencing using the Applied Biosystems Model 377 sequencer. Sequencing reactions were performed with the ABI Big Dye Terminator Cycle Sequencing kit or the Amersham Thermosequenase II Dye Terminator Cycle Sequencing kit. Both strands of each cDNA fragment were completely sequenced using gene-speci®c primers. GENEPRO (Riverside Scienti®c Enterprises) was employed for alignment and analysis of sequences. The BLAST e-mail server (NCBI) was used to determine homologies with thrombospondin sequences in GenBank. The coding region was identi®ed using the ORF Finder program (NCBI). Transcription factor binding sites were determined with the SIGNAL SCAN program (Prestridge, 1996). The presence of signal peptides was investigated with the SignalP server (Nielsen et al., 1997) (www.cbs.dtu.dk/services). 3. Results and discussion 3.1. cDNA sequence of the Drosophila thrombospondin homologue PCR analysis and DNA sequencing revealed that transcripts representing a thrombospondin homologue are present in D. melanogaster. The complete cDNA sequence, including the 5 0 UTR and 3 0 UTR, was therefore determined: The GenBank accession number for this sequence, and the deduced amino acid sequence, is AF305875. The exon/intron organization of the DTSP gene is shown schematically in Fig. 1 and described in Table 1. The DTSP gene consists of 13 exons. This number is one more than the predicted number in Adams et al. (2000) due to the presence of an intron of 84 bp which divides exon 5 into exons 5 and 6. The intron sequence was not predicted to be an intron in Adams et al. (2000), but was transcribed into a sequence of 28 amino acid residues. Analysis of the cDNA sequence using the NCBI ORF
K.W. Adolph / Gene 269 (2001) 177±184
179
Fig. 1. Diagram representing the genomic region and exon/intron organization of the gene for the Drosophila thrombospondin homologue. The upper part of the diagram shows the genes that are adjacent to the DTSP gene on chromosome 2 of Drosophila melanogaster. The proteins encoded by these genes are: GRHR, gonadotropin-releasing hormone receptor; CG11327, a structural protein involved in the cell cycle; Nhe3, a sodium/hydrogen transporter. The lengths of the arrows and spaces between the arrows are proportional to the sizes of the genes and inter-genic distances. The heads of the arrows indicate the direction of transcription of the genes. The genes are transcribed from the DNA strand de®ned as the reverse-complement or (2) strand (Adams et al., 2000). The 13 exons of the DTSP gene are shown below the genomic region. The horizontal dimensions are proportional to the sizes of the exons and introns given in Table 1. Solid rectangles, or portions of rectangles, represent the translated exons, while open rectangles correspond to the 5 0 and 3 0 untranslated regions. The translation start site is in exon 2. The scale at the bottom of the diagram refers to the exon/intron structure of the DTSP gene. The diagram is based on the results of this research, with the map of the genomic region for the DTSP gene (CG11326) adapted from the FlyBase/GadFly (NCBI) database.
The coding regions for the amino-terminal domain, Type II repeats, Type III repeats, and carboxyl-terminal domain are divided between exons 2±13. Exon 11 is by far the largest exon and codes for amino acids of the Type II and Type III repeats, and the carboxyl-terminal region. The
Finder program revealed an open reading frame of 3183 bp coding for a protein of 1060 amino acids. The translation start site is in exon 2. This is 192 bp 5 0 of the predicted translation start site in exon 4 that is given in Adams et al. (2000), and adds 64 amino acids to the N-terminus. Table 1 Exon/intron organization of the Drosophila thrombospondin homologue a Exon or intron number
Domain or repeat
Exon size (bp) Intron size (bp) Phase class 5 0 splice
3 0 splice
1 2 3 4 5 6 7 8 9 10 11
5 0 UTR Translation start Amino-terminal Amino-terminal Amino-terminal Amino-terminal Amino-terminal Amino-terminal Type II (EGF-like) repeat #1 Type II (EGF-like) repeats #1±#3 Type II (EGF-like) repeats #3±#4, Type III (Ca-binding) repeats #1±#7, Carboxyl-terminal Carboxyl-terminal Carboxyl-terminal, 3 0 UTR
764 195 67 316 130 138 195 118 122 323 1484
6342 1130 56 64 84 2307 135 5911 1858 58 56
-1 1-2 2-0 0-1 1-1 1-1 1-2 2-1 1-0 0-2
GTgagt GTaagt GTaatt GTgagt GTgagt GTaggt GTaagt GTaagt GTaagt GTgagt GTgagt
tttcAG ttgcAG aattAG tgacAG ttgcAG aaacAG ccccAG ttgcAG acgaAG tcgcAG ccacAG
134 160
68 ±
2-1 1-
GTaagg aaacAG ± ±
12 13
a The exon and intron sizes were determined by comparing the experimental cDNA sequence (this work) with the genomic DNA sequence (Adams et al., 2000). The phase class refers to the positions of insertion of introns in codons (Patthy, 1987). The 5 0 and 3 0 splice sequences shown are intronic sequences, with the invariant GT and AG bases in upper case.
180
K.W. Adolph / Gene 269 (2001) 177±184
Table 2 Codon differences between DTSP cDNA and genomic sequences Amino acid position
Exon number
cDNA codon
Genomic codon
Amino acid(s)
20 90 125 128 236 448 581 584 591 617 620 655
2 4 4 4 6 10 11 11 11 11 11 11
GAG ATC ACA GTT GAC CCT CAC CGG CCC CAA CAG CTG
GAT ATT ACG GTG GAT CCG CGC CGA CCT CAG CAA TTG
Glu, Asp Ile Thr Val Asp Pro His, Arg Arg Pro Gln Gln Leu
DTSP domain or repeat structure is characteristic of thrombospondin proteins, in particular the TSP3/TSP4/COMP subfamily. However, the exon/intron organization of the
Drosophila TSP homologue differs from that of human or mouse in having fewer exons: The human TSP3 gene, for example, consists of 23 exons (Adolph et al., 1995).
Fig. 2. Identi®cation of the transcription initiation site of the DTSP gene, and nucleotide sequence of the 5 0 end of the DTSP cDNA. (A) The agarose gel (2%) shows the cDNA fragments generated by PCR (2nd round) using primer B in exon 4 with primer C1 (lane 1), C2 (lane 2), and C3 (lane 3). C1 is located at the 5 0 end of the predicted DTSP cDNA sequence (Adams et al., 2000), C2 is 76 bp 5 0 of C1, and C3 is 294 bp 5 0 of C1. The M lanes contain molecular weight size standards (phiX174/HaeIII digest): 1353, 1078, 872, 603, 310, 281/271, 234, 194 bp. (B) The band seen in lane 3 was excised from preparative agarose gels and puri®ed, and the DNA sequence determined. The nucleotide sequence 5 0 of the predicted DTSP cDNA sequence is shown.
K.W. Adolph / Gene 269 (2001) 177±184
Table 2 lists the sequence differences in the coding region for the cDNA sequence, described in this report, and the genomic sequence (Adams et al., 2000). Twelve base differences were found. With two exceptions, the changes are conservative and result in codons that specify the same amino acid. The exceptions substitute glutamate for aspartate, both acidic amino acids, and histidine for arginine, both basic amino acids. The base changes are therefore likely to be due to Drosophila strain differences. The positions of the base differences are clustered, in particular in a region of exon 11 and also in exon 4. In the 5 0 UTR, differences between the cDNA and genomic sequences were more frequent. The 874 bp of the 5 0 UTR showed 15 sequence differences including base insertions and deletions, or 17 differences/kb. In contrast, the 12 sequence differences in the coding region of 3180 bp correspond to 3.8/kb. 3.2. Identi®cation of DTSP transcripts with the longest 5 0 ends The procedure to identify DTSP transcripts that initiate the furthest upstream employed nested, antisense primers A and B with primers C1, C2, and C3 placed increasingly 5 0 of
181
the predicted transcription start site. The results are shown in Fig. 2A. Lanes 1±3 contain the 2nd round PCR products using primer B with C1 (lane 1), C2 (lane 2), and C3 (lane 3). The band in lane 3 demonstrates that DTSP transcripts are initiated at least 294 bp 5 0 of the predicted start site. The decreased intensity of the band in lane 3, however, indicates that the start site is heterogeneous with fewer of the longest transcripts being initiated. The cDNA band in lane 3 was isolated from preparative agarose gels and the DNA sequence determined. Fig. 2B includes this sequence up to the predicted transcription start site. 3.3. Amino acid sequence of the Drosophila thrombospondin homologue A polypeptide of 1060 amino acids with a molecular weight of 118,247 is encoded by the open reading frame of the DTSP cDNA. This compares to 956 residues for human TSP3 (Adolph et al., 1995) and 1172 residues for human TSP2 (LaBell and Byers, 1993). Aspartic acid is the most abundant amino acid in the DTSP protein: The 123 Asp residues constitute 11.6% of the polypeptide. The Asp 1 Glu content of 15.8% is similar to the value of 15.1% for
Fig. 3. Alignment of amino acids in the region of the Type II (EGF-like) repeats. The existence of four repeats is characteristic of the TSP3/TSP4/COMP subfamily of thrombospondin proteins. The amino acid sequences are for the Drosophila thrombospondin homologue (DTSP), human thrombospondin 3 (TSP3), human thrombospondin 4 (TSP4), and human cartilage oligomeric matrix protein (COMP). Between the DTSP and TSP3 alignments are shown the identities and positives (amino acids with codons differing by a single base change) for these two proteins. The conserved cysteine residues are boxed in each of the repeats. For DTSP, the four Type II repeats extend from amino acids 380 to 561 of the 1060 aa protein. For TSP3, the repeats extend from amino acids 275 to 455 of the 956 aa protein, and for TSP4, they are from amino acids 290 to 461 of the 961 aa polypeptide. For COMP, the repeats are from amino acids 88 to 266 of the 757 aa polypeptide.
182
K.W. Adolph / Gene 269 (2001) 177±184
HTSP3 (Adolph et al., 1995). The Arg 1 Lys 1 His content of the Drosophila protein is 12.9%. The most highly acidic region of the molecule is between amino acids 564±894, and corresponds to the Type III (Ca 21-binding) repeats and part
of the C-terminal domain. The basic residues are concentrated in the amino-terminal (amino acids 44±208) and carboxyl-terminal (amino acids 906±1042) regions of the protein.
Fig. 4. Alignment of amino acids of the seven Type III (Ca 21-binding) repeats. DTSP is the Drosophila thrombospondin homologue, TSP3 is human thrombospondin 3, TSP4 is human thrombospondin 4, and COMP is human cartilage oligomeric matrix protein. The identities and positives of the DTSP vs. TSP3 alignment are also included. Boxes enclose the conserved cysteine residues. For DTSP the seven Type III repeats extend from amino acids 592 to 819, for TSP3 from 488 to 719, for TSP4 from 492 to 723, and for COMP from 297 to 524.
K.W. Adolph / Gene 269 (2001) 177±184
183
Fig. 5. The carboxyl-terminal domain of the Drosophila thrombospondin homologue aligned with the C-terminal domains of human TSP3, human TSP4, and human COMP. The line below the DTSP amino acid sequence shows the identities and positives with the TSP3 amino acid sequence. The C-terminal domains extend from the end of the Type III repeats to the ®nal amino acid of the protein. This includes amino acids 820±1060 for DTSP, 720±956 for TSP3, 724±961 for TSP4, and 525±757 for COMP.
Analysis of the amino acid sequence at the N-terminus using the SignalP server (Nielsen et al., 1997) revealed the presence of a signal peptide most likely cleaved after residue 32: MNWTRVLLIGLTALALTFVEVASLSLDPVASA. The DTSP signal peptide corresponds to those of other TSPs such as HTSP2 (MVWRLVLLALWVWPSTQA) and HTSP3 (METQELRGALALLLLCFFTSAS). The presence of a signal peptide implies that the DTSP protein has extracellular functions, perhaps related to Drosophila development. Ten asparagine residues were identi®ed as potential glycosylation sites. These are at amino acid positions 2, 107, 415, 439, 512, 672, 687, 817, 851, and 936. This compares to ®ve potential sites for HTSP3 (Adolph et al., 1995). The large number of sites for DTSP suggests that glycosylation is likely to be functionally signi®cant. 3.4. Analysis of the DTSP polypeptide domain structure The existence of the amino-terminal domain, Type II (EGF-like) repeats, Type III (Ca 21-binding) repeats, and carboxyl-terminal domain are the major structural features of the DTSP polypeptide. Analysis using the NCBI BLAST
server did not reveal the procollagen homology region and Type I repeats characteristic of TSP1 and TSP2. This fact and the presence of four, not three, Type II repeats indicates that the Drosophila thrombospondin homologue is most similar to the TSP3/TSP4/COMP subgroup. The BLAST homology scores, generated by the BLAST server in aligning amino acid sequences, bear this out. The Drosophila thrombospondin homologue gives BLAST scores, with the human TSPs, of 634 for TSP3, 631 for COMP, 613 for TSP4, 556 for TSP2, and 524 for TSP1. Because the numbers are close, it is not possible to conclude which vertebrate TSP has the greatest homology to DTSP. These BLAST scores are similar to the values in comparing TSPs of the two subfamilies in the same species: For example, the BLAST score for human TSP2 vs. human TSP3 is 605. The alignment of the amino acids of the Type II repeats is included in Fig. 3 for the Drosophila TSP and human TSP3, TSP4, and COMP. For DTSP, the coding sequence for the Type II repeats begins within exon 9 and continues into exon 11. The amino acids of the four repeats comprise 17% of the DTSP polypeptide. In Fig. 3, the conserved cysteine residues characteristic of the Type II repeats are
184
K.W. Adolph / Gene 269 (2001) 177±184
boxed. The amino acid identities and positives in comparing the DTSP and TSP3 sequences are indicated on the second line of the alignments. Overall, 40% of the amino acids are identities and 52% identities plus positives. Fig. 4 shows the alignment of amino acids for the seven Type III repeats. The coding region for the DTSP Type III repeats is all within exon 11. The seven repeats constitute 22% of the DTSP protein, so that the Type II and Type III repeats together make up 39% of the protein. Conserved cysteine residues are indicated by boxes in Fig. 4. The thrombospondins display a gradient of amino acid sequence identity, in comparing the two subfamilies, from about 10% at the N-terminus to 60% at the C-terminus. This gradient of identity is also observed for the Drosophila thrombospondin homologue: The DTSP and TSP3 sequences show greater homology in the region of Type III repeats (58% identities; 68% identities plus positives) than Type II repeats (40%; 52%). The high degree of homology between the DTSP protein and the TSP3/TSP4/COMP subfamily continues through the C-terminal region. This is illustrated by Fig. 5, which contains the alignment of the DTSP carboxyl-terminal domain with those of human TSP3, TSP4, and COMP. The amino acid identities and positives in comparing the DTSP protein with TSP3 are indicated on the second line. The DTSP C-terminal domain has a coding region that begins in exon 11 and continues to the TGA translation stop codon in exon 13. The amino acid residues of the carboxyl-terminal domain comprise 23% of the protein. A high level of sequence conservation between the TSP Cterminal domains is evident in Fig. 5: In comparing DTSP with TSP3, 57% identities and 75% identities plus positives are observed. Acknowledgements This research was supported by a grant from the Minnesota Medical Foundation. References Adams, J.C., Tucker, R.P., Lawler, J., 1995. The Thrombospondin Gene Family. Springer-Verlag, Berlin. Adams, M.D., et al., 2000. The genome sequence of Drosophila melanogaster. Science 287, 2185±2195. Adolph, K.W., Bornstein, P., 1999. The human thrombospondin 3 gene: analysis of transcription initiation and an alternatively spliced transcript. Mol. Cell Biol. Res. Commun. 2, 47±52. Adolph, K.W., Long, G.L., Win®eld, S., Ginns, E.I., Bornstein, P., 1995. Structure and organization of the human thrombospondin 3 gene (THBS3). Genomics 27, 329±336.
Adolph, K.W., Liska, D.J., Bornstein, P., 1997. Analysis of the promoter and transcription start sites of the human thrombospondin 2 gene (THBS2). Gene 193, 5±11. Baenziger, N.L., Brodie, G.N., Majerus, P.W., 1971. A thrombin-sensitive protein of human platelet membranes. Proc. Natl. Acad. Sci. USA 68, 240±243. Bornstein, P., Sage, E.H., 1994. Thrombospondins. Methods Enzymol. 245, 62±85. Bornstein, P., O'Rourke, K., Wikstrom, K., Wolf, F.W., Katz, R., Li, P., Dixit, V.M., 1991. A second, expressed thrombospondin gene (Thbs2) exists in the mouse genome. J. Biol. Chem. 266, 12821±12824. Briggs, M.D., Hoffman, S.M.G., King, L.M., Olsen, A.S., Mohrenweiser, H., Leroy, J.G., Mortier, G.R., Rimoin, D.L., Lachman, R.S., Gaines, E.S., Cekleniak, J.A., Knowlton, R.G., Cohn, D.H., 1995. Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the cartilage oligomeric matrix protein gene. Nat. Genet. 10, 330±336. Hecht, J.T., Nelson, L.D., Crowder, E., Wang, Y., Elder, F.F.B., Harrison, W.R., Francomano, C.A., Prange, C.K., Lennon, G.G., Deere, M., Lawler, J., 1995. Mutations in exon 17B of cartilage oligomeric matrix protein (COMP) cause pseudoachondroplasia. Nat. Genet. 10, 325± 329. Hennessy, S.W., Frazier, B.A., Kim, D.D., Deckwerth, T.L., Baumgartel, D.M., Rotwein, P., Frazier, W.A., 1989. Complete thrombospondin mRNA sequence includes potential regulatory sites in the 3 0 untranslated region. J. Cell Biol. 108, 729±736. Iruela-Arispe, M.L., Lombardo, M., Krutzsch, H.C., Lawler, J., Roberts, D.D., 1999. Inhibition of angiogenesis by thrombospondin-1 is mediated by 2 independent regions within the type 1 repeats. Circulation 100, 1423±1431. Jimenez, B., Volpert, O.V., Crawford, S.E., Febbraio, M., Silverstein, R.L., Bouck, N., 2000. Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1. Nat. Med. 6, 41±48. LaBell, T.L., Byers, P.H., 1993. Sequence and characterization of the complete human thrombospondin 2 cDNA: potential regulatory role for the 3 0 untranslated region. Genomics 17, 225±229. Lawler, J., Duquette, M., Ferro, P., 1991. Cloning and sequencing of chicken thrombospondin. J. Biol. Chem. 266, 8039±8043. Lawler, J., Duquette, M., Whittaker, C.A., Adams, J.C., McHenry, K., DeSimone, D.W., 1993. Identi®cation and characterization of thrombospondin-4, a new member of the thrombospondin gene family. J. Cell Biol. 120, 1059±1067. Lawler, J., McHenry, K., Duquette, M., Derick, L., 1995. Characterization of human thrombospondin-4. J. Biol. Chem. 270, 2809±2814. Nielsen, H., Engelbrecht, J., Brunak, S., von Heijne, G., 1997. Identi®cation of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Prot. Eng. 10, 1±6. Oldberg, A., Antonsson, P., Lindblom, K., Heinegard, D., 1992. COMP (cartilage oligomeric matrix protein) is structurally related to the thrombospondins. J. Biol. Chem. 267, 22346±22350. Patthy, L., 1987. Intron-dependent evolution: preferred types of exons and introns. FEBS Lett. 214, 1±7. Prestridge, D.S., 1996. SIGNAL SCAN 4.0: Additional databases and sequence formats. Comput. Appl. Biosci. 12, 157±160. Ueno, A., Yamashita, K., Nagata, T., Tsurumi, C., Miwa, Y., Kitamura, S., Inoue, H., 1998. cDNA cloning of bovine thrombospondin 1 and its expression in odontoblasts and predentin. Biochim. Biophys. Acta 1382, 17±22. Vos, H.L., Devarayalu, S., de Vries, Y., Bornstein, P., 1992. Thrombospondin 3 (Thbs3), a new member of the thrombospondin gene family. J. Biol. Chem. 267, 12192±12196.
www.elsevier.com/locate/gene
A thrombospondin homologue in Drosophila melanogaster: cDNA and protein structure Kenneth W. Adolph* University of Minnesota, Department of Biochemistry, Molecular Biology and Biophysics, 6-155 Jackson Hall, 321 Church St. S.E., Minneapolis, MN 55455, USA Received 12 January 2001; received in revised form 27 February 2001; accepted 14 March 2001 Received by A.J. van Wijnen
Abstract The cDNA of a thrombospondin homologue in Drosophila melanogaster (DTSP) has been sequenced, and structural features of the translated protein analyzed. Thrombospondin proteins had previously been identi®ed only in vertebrates, including human and mouse. Comparison with the genomic sequence revealed that the DTSP gene is divided into 13 exons; The translation initiation site is in exon 2. The transcription start site was analyzed using a PCR procedure, and the longest transcripts were found to initiate about 300 bp 5 0 of the predicted start site. The open reading frame of the DTSP cDNA encodes a protein that has 1060 amino acid residues. The polypeptide is composed of domains or repeats characteristic of the TSP3/TSP4/COMP subfamily of thrombospondin proteins: Amino-terminal domain, four Type II repeats, seven Type III repeats, carboxyl-terminal domain. The protein is highly acidic, particularly in the region of Type III repeats, with an Asp 1 Glu content of 15.8%. A signal peptide was detected at the N-terminus, which indicates that DTSP, like other TSPs, functions as an extracellular protein. Ten Asn residues were identi®ed as potential glycosylation sites. Alignment of the amino acid sequences of the Drosophila TSP with human TSP1±TSP4 and COMP demonstrated a high degree of homology between the four Type II repeats, seven Type III repeats, and C-terminal domain. q 2001 Elsevier Science B.V. All rights reserved. Keywords: Nucleotide sequence; Exon/intron organization; Transcription start site; Open reading frame; Polypeptide domains; TSP3/TSP4/COMP subfamily
1. Introduction The thrombospondin family of proteins consists of ®ve members, in humans and other vertebrates, which can be divided into two subgroups based upon their polypeptide domain structures (Bornstein and Sage, 1994; Adams et al., 1995). TSP1 (Baenziger et al., 1971; Hennessy et al., 1989) and TSP2 (Bornstein et al., 1991; LaBell and Byers, 1993; Adolph et al., 1997) constitute the ®rst subgroup and contain (a) an amino-terminal domain, (b) a procollagen homology region, (c) three Type I (TSP) repeats, (d) three Type II (EGF-like) repeats, (e) seven Type III (Ca 21-binding) repeats, (f) a carboxyl-terminal domain. The mature TSP1 and TSP2 molecules are disul®de-bonded trimers Abbreviations: BLAST, Basic Local Alignment Search Tool; COMP, cartilage oligomeric matrix protein; DTSP, Drosophila melanogaster thrombospondin homologue; EGF, epidermal growth factor; HTSP, human thrombospondin; NCBI, National Center for Biotechnology Information; ORF, open reading frame; PCR, polymerase chain reaction; TSP, thrombospondin; UTR, untranslated region * Tel.: 11-612-625-8467; fax: 612-625-2163. E-mail address: [email protected] (K.W. Adolph).
that are secreted and interact with a variety of cells, extracellular matrix proteins, and plasma proteins. These interactions account for the involvement of TSP1 and TSP2 in processes that include angiogenesis, tumor growth, embryonic development, tissue differentiation, and blood coagulation. For example, interactions involving the Type I repeats are primarily responsible for the function of TSP1 and TSP2 as potent inhibitors of angiogenesis (Iruela-Arispe et al., 1999; Jimenez et al., 2000). TSP3 (Vos et al., 1992; Adolph et al., 1995; Adolph and Bornstein, 1999), TSP4 (Lawler et al., 1993, 1995), and COMP (cartilage oligomeric matrix protein) (Oldberg et al., 1992) make up the second thrombospondin subgroup. These polypeptides have a simpler structure consisting of (a) an amino-terminal domain, (b) four Type II repeats, (c) seven Type III repeats, and (d) a carboxyl-terminal domain. The proteins exist as pentamers of the modular polypeptides and are extracellular. The biological roles of TSP3 and TSP4 are not well understood. COMP is better characterized, since mutations in the COMP gene have been demonstrated to be the cause of human skeletal disorders (Hecht et al., 1995; Briggs et al., 1995).
0378-1119/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S 0378-111 9(01)00441-3
178
K.W. Adolph / Gene 269 (2001) 177±184
The genes for the thrombospondin proteins are at different chromosomal locations. The TSP1 gene, for example, is at 15q15 in humans, TSP2 is at 6q27, and TSP3 is at 1q2124. The TSP1 gene is composed of 22 exons, while the TSP2 and TSP3 genes possess 23 exons (Adolph et al., 1995). The exon sequences and exon/intron organization of each TSP are highly conserved between human and mouse. The nucleotide sequences of the TSP promoter/5 0 ¯anking regions in humans show no signi®cant homologies, even between subgroup members such as TSP1 and TSP2 (Adolph et al., 1997). This strengthens the conclusion that thrombospondin proteins have distinct functions. Besides human and mouse, TSP cDNAs and genes have been characterized for other vertebrate sources, including Xenopus (Lawler et al., 1993), bovine (Ueno et al., 1998), chicken (Lawler et al., 1991), and rat (Oldberg et al., 1992). The results presented here describe a thrombospondin homologue (DTSP) from a non-vertebrate source, the fruit ¯y Drosophila melanogaster. The sequence of the Drosophila TSP cDNA has been determined, and the deduced amino acid sequence has been analyzed and compared to the human thrombospondin amino acid sequences. 2. Materials and methods 2.1. Isolation of cDNA for a thrombospondin homologue in D. melanogaster cDNA corresponding to the DTSP coding exons was prepared by PCR using nested, antisense primers X and Y that included the translation stop codon, with primer Z that included the translation start codon. The sequences of the primers (obtained from Integrated DNA Technologies, Inc.) were: primer X, 5 0 -GTTAATGGCTCTGAGGATTGCG-3 0 ; primer Y, 5 0 -TCAGTCCTGCAACTCCACCTTC-3 0 ; primer Z, 5 0 -ATGATCGACTTCCCGATGCTGA3 0 . The primers were designed from the D. melanogaster predicted gene sequences (Adams et al., 2000). The template in the PCR reactions was Drosophila melanogaster embryo cDNA (QUICK-Clone cDNA; Clontech). The poly A 1 source for the cDNA was given as 18-hr Canton S. embryo. The thermostable DNA polymerase included in the PCR reactions was Expand High Fidelity polymerase mixture (Roche Molecular Biochemicals). The ampli®cation conditions, using a Perkin±Elmer Model 480 thermal cycler, were typically 948C, 45 s; 658C, 45 s; 728C, 2 min; 30 cycles. The products of the PCR reactions were analyzed on 2% agarose/ TAE gels. 2.2. Isolation of cDNAs containing the 5 0 UTR and 3 0 UTR and identi®cation of transcripts with the longest 5 0 UTR To prepare cDNA fragments containing the 5 0 UTR, PCR was carried out using nested, antisense primers A (in exon 6) and B (in exon 4) with each of a series of primers C1, C2, C3 near the 5 0 end. C1 was located at the predicted 5 0 end of
the DTSP transcript, while C2 and C3 were, respectively, 76 and 294 bp 5 0 of C1. Primer A had the sequence 5 0 GCACCTCCTGGTGTCTGTGATC-3 0 ; primer B was 5 0 CGTTCGCACTAATGTCCAACGT-3 0 . C3, which identi®ed the longest DTSP transcript, was 5 0 -TGCGTGTCCTAGGTGCTTCGAT-3 0 . PCR with these primers was carried out under the conditions described above. cDNA fragments containing the 3 0 UTR were prepared with a similar procedure. In this case, PCR used antisense primer W (5 0 -GGCCAACTGGATTGGCAACTCA-3 0 ) placed immediately 3 0 of the potential AATAAA poly A 1 addition site. 2.3. DNA sequencing procedures and sequence analysis DTSP cDNA fragments representing the coding exons, 5 0 UTR, and 3 0 UTR were puri®ed for sequencing from preparative agarose gels. cDNA bands of the correct size were excised from 2% agarose/TAE gels and the cDNAs were isolated with the QIAEX II Gel Extraction Kit (Qiagen). Sequences were determined by automated DNA sequencing using the Applied Biosystems Model 377 sequencer. Sequencing reactions were performed with the ABI Big Dye Terminator Cycle Sequencing kit or the Amersham Thermosequenase II Dye Terminator Cycle Sequencing kit. Both strands of each cDNA fragment were completely sequenced using gene-speci®c primers. GENEPRO (Riverside Scienti®c Enterprises) was employed for alignment and analysis of sequences. The BLAST e-mail server (NCBI) was used to determine homologies with thrombospondin sequences in GenBank. The coding region was identi®ed using the ORF Finder program (NCBI). Transcription factor binding sites were determined with the SIGNAL SCAN program (Prestridge, 1996). The presence of signal peptides was investigated with the SignalP server (Nielsen et al., 1997) (www.cbs.dtu.dk/services). 3. Results and discussion 3.1. cDNA sequence of the Drosophila thrombospondin homologue PCR analysis and DNA sequencing revealed that transcripts representing a thrombospondin homologue are present in D. melanogaster. The complete cDNA sequence, including the 5 0 UTR and 3 0 UTR, was therefore determined: The GenBank accession number for this sequence, and the deduced amino acid sequence, is AF305875. The exon/intron organization of the DTSP gene is shown schematically in Fig. 1 and described in Table 1. The DTSP gene consists of 13 exons. This number is one more than the predicted number in Adams et al. (2000) due to the presence of an intron of 84 bp which divides exon 5 into exons 5 and 6. The intron sequence was not predicted to be an intron in Adams et al. (2000), but was transcribed into a sequence of 28 amino acid residues. Analysis of the cDNA sequence using the NCBI ORF
K.W. Adolph / Gene 269 (2001) 177±184
179
Fig. 1. Diagram representing the genomic region and exon/intron organization of the gene for the Drosophila thrombospondin homologue. The upper part of the diagram shows the genes that are adjacent to the DTSP gene on chromosome 2 of Drosophila melanogaster. The proteins encoded by these genes are: GRHR, gonadotropin-releasing hormone receptor; CG11327, a structural protein involved in the cell cycle; Nhe3, a sodium/hydrogen transporter. The lengths of the arrows and spaces between the arrows are proportional to the sizes of the genes and inter-genic distances. The heads of the arrows indicate the direction of transcription of the genes. The genes are transcribed from the DNA strand de®ned as the reverse-complement or (2) strand (Adams et al., 2000). The 13 exons of the DTSP gene are shown below the genomic region. The horizontal dimensions are proportional to the sizes of the exons and introns given in Table 1. Solid rectangles, or portions of rectangles, represent the translated exons, while open rectangles correspond to the 5 0 and 3 0 untranslated regions. The translation start site is in exon 2. The scale at the bottom of the diagram refers to the exon/intron structure of the DTSP gene. The diagram is based on the results of this research, with the map of the genomic region for the DTSP gene (CG11326) adapted from the FlyBase/GadFly (NCBI) database.
The coding regions for the amino-terminal domain, Type II repeats, Type III repeats, and carboxyl-terminal domain are divided between exons 2±13. Exon 11 is by far the largest exon and codes for amino acids of the Type II and Type III repeats, and the carboxyl-terminal region. The
Finder program revealed an open reading frame of 3183 bp coding for a protein of 1060 amino acids. The translation start site is in exon 2. This is 192 bp 5 0 of the predicted translation start site in exon 4 that is given in Adams et al. (2000), and adds 64 amino acids to the N-terminus. Table 1 Exon/intron organization of the Drosophila thrombospondin homologue a Exon or intron number
Domain or repeat
Exon size (bp) Intron size (bp) Phase class 5 0 splice
3 0 splice
1 2 3 4 5 6 7 8 9 10 11
5 0 UTR Translation start Amino-terminal Amino-terminal Amino-terminal Amino-terminal Amino-terminal Amino-terminal Type II (EGF-like) repeat #1 Type II (EGF-like) repeats #1±#3 Type II (EGF-like) repeats #3±#4, Type III (Ca-binding) repeats #1±#7, Carboxyl-terminal Carboxyl-terminal Carboxyl-terminal, 3 0 UTR
764 195 67 316 130 138 195 118 122 323 1484
6342 1130 56 64 84 2307 135 5911 1858 58 56
-1 1-2 2-0 0-1 1-1 1-1 1-2 2-1 1-0 0-2
GTgagt GTaagt GTaatt GTgagt GTgagt GTaggt GTaagt GTaagt GTaagt GTgagt GTgagt
tttcAG ttgcAG aattAG tgacAG ttgcAG aaacAG ccccAG ttgcAG acgaAG tcgcAG ccacAG
134 160
68 ±
2-1 1-
GTaagg aaacAG ± ±
12 13
a The exon and intron sizes were determined by comparing the experimental cDNA sequence (this work) with the genomic DNA sequence (Adams et al., 2000). The phase class refers to the positions of insertion of introns in codons (Patthy, 1987). The 5 0 and 3 0 splice sequences shown are intronic sequences, with the invariant GT and AG bases in upper case.
180
K.W. Adolph / Gene 269 (2001) 177±184
Table 2 Codon differences between DTSP cDNA and genomic sequences Amino acid position
Exon number
cDNA codon
Genomic codon
Amino acid(s)
20 90 125 128 236 448 581 584 591 617 620 655
2 4 4 4 6 10 11 11 11 11 11 11
GAG ATC ACA GTT GAC CCT CAC CGG CCC CAA CAG CTG
GAT ATT ACG GTG GAT CCG CGC CGA CCT CAG CAA TTG
Glu, Asp Ile Thr Val Asp Pro His, Arg Arg Pro Gln Gln Leu
DTSP domain or repeat structure is characteristic of thrombospondin proteins, in particular the TSP3/TSP4/COMP subfamily. However, the exon/intron organization of the
Drosophila TSP homologue differs from that of human or mouse in having fewer exons: The human TSP3 gene, for example, consists of 23 exons (Adolph et al., 1995).
Fig. 2. Identi®cation of the transcription initiation site of the DTSP gene, and nucleotide sequence of the 5 0 end of the DTSP cDNA. (A) The agarose gel (2%) shows the cDNA fragments generated by PCR (2nd round) using primer B in exon 4 with primer C1 (lane 1), C2 (lane 2), and C3 (lane 3). C1 is located at the 5 0 end of the predicted DTSP cDNA sequence (Adams et al., 2000), C2 is 76 bp 5 0 of C1, and C3 is 294 bp 5 0 of C1. The M lanes contain molecular weight size standards (phiX174/HaeIII digest): 1353, 1078, 872, 603, 310, 281/271, 234, 194 bp. (B) The band seen in lane 3 was excised from preparative agarose gels and puri®ed, and the DNA sequence determined. The nucleotide sequence 5 0 of the predicted DTSP cDNA sequence is shown.
K.W. Adolph / Gene 269 (2001) 177±184
Table 2 lists the sequence differences in the coding region for the cDNA sequence, described in this report, and the genomic sequence (Adams et al., 2000). Twelve base differences were found. With two exceptions, the changes are conservative and result in codons that specify the same amino acid. The exceptions substitute glutamate for aspartate, both acidic amino acids, and histidine for arginine, both basic amino acids. The base changes are therefore likely to be due to Drosophila strain differences. The positions of the base differences are clustered, in particular in a region of exon 11 and also in exon 4. In the 5 0 UTR, differences between the cDNA and genomic sequences were more frequent. The 874 bp of the 5 0 UTR showed 15 sequence differences including base insertions and deletions, or 17 differences/kb. In contrast, the 12 sequence differences in the coding region of 3180 bp correspond to 3.8/kb. 3.2. Identi®cation of DTSP transcripts with the longest 5 0 ends The procedure to identify DTSP transcripts that initiate the furthest upstream employed nested, antisense primers A and B with primers C1, C2, and C3 placed increasingly 5 0 of
181
the predicted transcription start site. The results are shown in Fig. 2A. Lanes 1±3 contain the 2nd round PCR products using primer B with C1 (lane 1), C2 (lane 2), and C3 (lane 3). The band in lane 3 demonstrates that DTSP transcripts are initiated at least 294 bp 5 0 of the predicted start site. The decreased intensity of the band in lane 3, however, indicates that the start site is heterogeneous with fewer of the longest transcripts being initiated. The cDNA band in lane 3 was isolated from preparative agarose gels and the DNA sequence determined. Fig. 2B includes this sequence up to the predicted transcription start site. 3.3. Amino acid sequence of the Drosophila thrombospondin homologue A polypeptide of 1060 amino acids with a molecular weight of 118,247 is encoded by the open reading frame of the DTSP cDNA. This compares to 956 residues for human TSP3 (Adolph et al., 1995) and 1172 residues for human TSP2 (LaBell and Byers, 1993). Aspartic acid is the most abundant amino acid in the DTSP protein: The 123 Asp residues constitute 11.6% of the polypeptide. The Asp 1 Glu content of 15.8% is similar to the value of 15.1% for
Fig. 3. Alignment of amino acids in the region of the Type II (EGF-like) repeats. The existence of four repeats is characteristic of the TSP3/TSP4/COMP subfamily of thrombospondin proteins. The amino acid sequences are for the Drosophila thrombospondin homologue (DTSP), human thrombospondin 3 (TSP3), human thrombospondin 4 (TSP4), and human cartilage oligomeric matrix protein (COMP). Between the DTSP and TSP3 alignments are shown the identities and positives (amino acids with codons differing by a single base change) for these two proteins. The conserved cysteine residues are boxed in each of the repeats. For DTSP, the four Type II repeats extend from amino acids 380 to 561 of the 1060 aa protein. For TSP3, the repeats extend from amino acids 275 to 455 of the 956 aa protein, and for TSP4, they are from amino acids 290 to 461 of the 961 aa polypeptide. For COMP, the repeats are from amino acids 88 to 266 of the 757 aa polypeptide.
182
K.W. Adolph / Gene 269 (2001) 177±184
HTSP3 (Adolph et al., 1995). The Arg 1 Lys 1 His content of the Drosophila protein is 12.9%. The most highly acidic region of the molecule is between amino acids 564±894, and corresponds to the Type III (Ca 21-binding) repeats and part
of the C-terminal domain. The basic residues are concentrated in the amino-terminal (amino acids 44±208) and carboxyl-terminal (amino acids 906±1042) regions of the protein.
Fig. 4. Alignment of amino acids of the seven Type III (Ca 21-binding) repeats. DTSP is the Drosophila thrombospondin homologue, TSP3 is human thrombospondin 3, TSP4 is human thrombospondin 4, and COMP is human cartilage oligomeric matrix protein. The identities and positives of the DTSP vs. TSP3 alignment are also included. Boxes enclose the conserved cysteine residues. For DTSP the seven Type III repeats extend from amino acids 592 to 819, for TSP3 from 488 to 719, for TSP4 from 492 to 723, and for COMP from 297 to 524.
K.W. Adolph / Gene 269 (2001) 177±184
183
Fig. 5. The carboxyl-terminal domain of the Drosophila thrombospondin homologue aligned with the C-terminal domains of human TSP3, human TSP4, and human COMP. The line below the DTSP amino acid sequence shows the identities and positives with the TSP3 amino acid sequence. The C-terminal domains extend from the end of the Type III repeats to the ®nal amino acid of the protein. This includes amino acids 820±1060 for DTSP, 720±956 for TSP3, 724±961 for TSP4, and 525±757 for COMP.
Analysis of the amino acid sequence at the N-terminus using the SignalP server (Nielsen et al., 1997) revealed the presence of a signal peptide most likely cleaved after residue 32: MNWTRVLLIGLTALALTFVEVASLSLDPVASA. The DTSP signal peptide corresponds to those of other TSPs such as HTSP2 (MVWRLVLLALWVWPSTQA) and HTSP3 (METQELRGALALLLLCFFTSAS). The presence of a signal peptide implies that the DTSP protein has extracellular functions, perhaps related to Drosophila development. Ten asparagine residues were identi®ed as potential glycosylation sites. These are at amino acid positions 2, 107, 415, 439, 512, 672, 687, 817, 851, and 936. This compares to ®ve potential sites for HTSP3 (Adolph et al., 1995). The large number of sites for DTSP suggests that glycosylation is likely to be functionally signi®cant. 3.4. Analysis of the DTSP polypeptide domain structure The existence of the amino-terminal domain, Type II (EGF-like) repeats, Type III (Ca 21-binding) repeats, and carboxyl-terminal domain are the major structural features of the DTSP polypeptide. Analysis using the NCBI BLAST
server did not reveal the procollagen homology region and Type I repeats characteristic of TSP1 and TSP2. This fact and the presence of four, not three, Type II repeats indicates that the Drosophila thrombospondin homologue is most similar to the TSP3/TSP4/COMP subgroup. The BLAST homology scores, generated by the BLAST server in aligning amino acid sequences, bear this out. The Drosophila thrombospondin homologue gives BLAST scores, with the human TSPs, of 634 for TSP3, 631 for COMP, 613 for TSP4, 556 for TSP2, and 524 for TSP1. Because the numbers are close, it is not possible to conclude which vertebrate TSP has the greatest homology to DTSP. These BLAST scores are similar to the values in comparing TSPs of the two subfamilies in the same species: For example, the BLAST score for human TSP2 vs. human TSP3 is 605. The alignment of the amino acids of the Type II repeats is included in Fig. 3 for the Drosophila TSP and human TSP3, TSP4, and COMP. For DTSP, the coding sequence for the Type II repeats begins within exon 9 and continues into exon 11. The amino acids of the four repeats comprise 17% of the DTSP polypeptide. In Fig. 3, the conserved cysteine residues characteristic of the Type II repeats are
184
K.W. Adolph / Gene 269 (2001) 177±184
boxed. The amino acid identities and positives in comparing the DTSP and TSP3 sequences are indicated on the second line of the alignments. Overall, 40% of the amino acids are identities and 52% identities plus positives. Fig. 4 shows the alignment of amino acids for the seven Type III repeats. The coding region for the DTSP Type III repeats is all within exon 11. The seven repeats constitute 22% of the DTSP protein, so that the Type II and Type III repeats together make up 39% of the protein. Conserved cysteine residues are indicated by boxes in Fig. 4. The thrombospondins display a gradient of amino acid sequence identity, in comparing the two subfamilies, from about 10% at the N-terminus to 60% at the C-terminus. This gradient of identity is also observed for the Drosophila thrombospondin homologue: The DTSP and TSP3 sequences show greater homology in the region of Type III repeats (58% identities; 68% identities plus positives) than Type II repeats (40%; 52%). The high degree of homology between the DTSP protein and the TSP3/TSP4/COMP subfamily continues through the C-terminal region. This is illustrated by Fig. 5, which contains the alignment of the DTSP carboxyl-terminal domain with those of human TSP3, TSP4, and COMP. The amino acid identities and positives in comparing the DTSP protein with TSP3 are indicated on the second line. The DTSP C-terminal domain has a coding region that begins in exon 11 and continues to the TGA translation stop codon in exon 13. The amino acid residues of the carboxyl-terminal domain comprise 23% of the protein. A high level of sequence conservation between the TSP Cterminal domains is evident in Fig. 5: In comparing DTSP with TSP3, 57% identities and 75% identities plus positives are observed. Acknowledgements This research was supported by a grant from the Minnesota Medical Foundation. References Adams, J.C., Tucker, R.P., Lawler, J., 1995. The Thrombospondin Gene Family. Springer-Verlag, Berlin. Adams, M.D., et al., 2000. The genome sequence of Drosophila melanogaster. Science 287, 2185±2195. Adolph, K.W., Bornstein, P., 1999. The human thrombospondin 3 gene: analysis of transcription initiation and an alternatively spliced transcript. Mol. Cell Biol. Res. Commun. 2, 47±52. Adolph, K.W., Long, G.L., Win®eld, S., Ginns, E.I., Bornstein, P., 1995. Structure and organization of the human thrombospondin 3 gene (THBS3). Genomics 27, 329±336.
Adolph, K.W., Liska, D.J., Bornstein, P., 1997. Analysis of the promoter and transcription start sites of the human thrombospondin 2 gene (THBS2). Gene 193, 5±11. Baenziger, N.L., Brodie, G.N., Majerus, P.W., 1971. A thrombin-sensitive protein of human platelet membranes. Proc. Natl. Acad. Sci. USA 68, 240±243. Bornstein, P., Sage, E.H., 1994. Thrombospondins. Methods Enzymol. 245, 62±85. Bornstein, P., O'Rourke, K., Wikstrom, K., Wolf, F.W., Katz, R., Li, P., Dixit, V.M., 1991. A second, expressed thrombospondin gene (Thbs2) exists in the mouse genome. J. Biol. Chem. 266, 12821±12824. Briggs, M.D., Hoffman, S.M.G., King, L.M., Olsen, A.S., Mohrenweiser, H., Leroy, J.G., Mortier, G.R., Rimoin, D.L., Lachman, R.S., Gaines, E.S., Cekleniak, J.A., Knowlton, R.G., Cohn, D.H., 1995. Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the cartilage oligomeric matrix protein gene. Nat. Genet. 10, 330±336. Hecht, J.T., Nelson, L.D., Crowder, E., Wang, Y., Elder, F.F.B., Harrison, W.R., Francomano, C.A., Prange, C.K., Lennon, G.G., Deere, M., Lawler, J., 1995. Mutations in exon 17B of cartilage oligomeric matrix protein (COMP) cause pseudoachondroplasia. Nat. Genet. 10, 325± 329. Hennessy, S.W., Frazier, B.A., Kim, D.D., Deckwerth, T.L., Baumgartel, D.M., Rotwein, P., Frazier, W.A., 1989. Complete thrombospondin mRNA sequence includes potential regulatory sites in the 3 0 untranslated region. J. Cell Biol. 108, 729±736. Iruela-Arispe, M.L., Lombardo, M., Krutzsch, H.C., Lawler, J., Roberts, D.D., 1999. Inhibition of angiogenesis by thrombospondin-1 is mediated by 2 independent regions within the type 1 repeats. Circulation 100, 1423±1431. Jimenez, B., Volpert, O.V., Crawford, S.E., Febbraio, M., Silverstein, R.L., Bouck, N., 2000. Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1. Nat. Med. 6, 41±48. LaBell, T.L., Byers, P.H., 1993. Sequence and characterization of the complete human thrombospondin 2 cDNA: potential regulatory role for the 3 0 untranslated region. Genomics 17, 225±229. Lawler, J., Duquette, M., Ferro, P., 1991. Cloning and sequencing of chicken thrombospondin. J. Biol. Chem. 266, 8039±8043. Lawler, J., Duquette, M., Whittaker, C.A., Adams, J.C., McHenry, K., DeSimone, D.W., 1993. Identi®cation and characterization of thrombospondin-4, a new member of the thrombospondin gene family. J. Cell Biol. 120, 1059±1067. Lawler, J., McHenry, K., Duquette, M., Derick, L., 1995. Characterization of human thrombospondin-4. J. Biol. Chem. 270, 2809±2814. Nielsen, H., Engelbrecht, J., Brunak, S., von Heijne, G., 1997. Identi®cation of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Prot. Eng. 10, 1±6. Oldberg, A., Antonsson, P., Lindblom, K., Heinegard, D., 1992. COMP (cartilage oligomeric matrix protein) is structurally related to the thrombospondins. J. Biol. Chem. 267, 22346±22350. Patthy, L., 1987. Intron-dependent evolution: preferred types of exons and introns. FEBS Lett. 214, 1±7. Prestridge, D.S., 1996. SIGNAL SCAN 4.0: Additional databases and sequence formats. Comput. Appl. Biosci. 12, 157±160. Ueno, A., Yamashita, K., Nagata, T., Tsurumi, C., Miwa, Y., Kitamura, S., Inoue, H., 1998. cDNA cloning of bovine thrombospondin 1 and its expression in odontoblasts and predentin. Biochim. Biophys. Acta 1382, 17±22. Vos, H.L., Devarayalu, S., de Vries, Y., Bornstein, P., 1992. Thrombospondin 3 (Thbs3), a new member of the thrombospondin gene family. J. Biol. Chem. 267, 12192±12196.