Proteins consisting of two PDGF-D polypeptide chains, polynucleotides encoding the polypeptides, and materials and methods for making the proteins are disclosed. Each of the polypeptide chains consists of, from amino terminus to carboxyl terminus, the following operably linked segments: P1-P2-h-C.sub.H2-C.sub.H3; P1-P2-C.sub.H2-C.sub.H3; h-C.sub.H2-C.sub.H3-P2-P1; or C.sub.H2-C.sub.H3-P2-P1. Within these polypeptide chains, P1 is a first polypeptide segment as shown in SEQ ID NO:2 or SEQ ID NO:4 from amino acid x to amino acid y, wherein x is an integer from 246 to 258, inclusive, and y is an integer from 365 370, inclusive; P2 is a second polypeptide segment consisting of from 4 to 20 amino acid residues; h is an immunoglobulin hinge region or portion thereof; and C.sub.H2 and C.sub.H3 are C.sub.H2 and C.sub.H3 domains of an immunoglobulin heavy chain, respectively. Within the protein, the two polypeptide chains are joined by one or more disulfide bonds, each of the chains is optionally glycosylated, and the protein binds to and activates cell-surface PDGF receptors.