Figure 3

Figure 3

Phylogenetic tree of thrombospondins. Thrombospondin sequences comprising the EGF domains, TSP type-3 repeats and the C-terminal domains from zebrafish (z) and mouse (m) were aligned using the PILEUP program of the GCG package, using the default parameters. As a full-length sequence for zebrafish thrombospondin 1a was not present in the databases, we amplified the lacking N-terminal sequence by RT-PCR (see Supplementary Figure 4). The tree was constructed using the programs PROTEIN DISTANCE, Fitch-Margoliash and CONSENSE of the PHYLIP package version 3.695. Bootstrap analyses using 100 replicates were performed to show the significance. The numbers indicate the statistical weight of the individual branches. Drosophila (d) thrombospondin was used as outgroup.