Alternative splicing of the platelet-derived growth factor A-chain transcript occurs in normal as well as tumor cells and is conserved among mammalian species

Genetic variants of PDGF signaling pathway genes predict cutaneous melanoma survival

To investigate whether genetic variants of platelet-derived growth factor (PDGF) signaling pathway genes are associated with survival of cutaneous melanoma (CM) patients, we assessed associations of single-nucleotide polymorphisms in PDGF pathway with melanoma-specific survival in 858 CM patients of M.D. Anderson Cancer Center (MDACC). Additional data of 409 cases from Harvard University were also included for further analysis. We identified 13 SNPs in four genes (COL6A3, NCK2, COL5A1 and PRKCD) with a nominal P < 0.05 and false discovery rate (FDR) < 0.2 in MDACC dataset. Based on linkage disequilibrium, functional prediction and minor allele frequency, a representative SNP in each gene was selected. In the meta-analysis using MDACC and Harvard datasets, there were two SNPs associated with poor survival of CM patients: rs6707820 C>T in NCK2 (HR = 1.87, 95% CI = 1.35-2.59, P_meta= 1.53E-5); and rs2306574 T>C in PRKCD (HR = 1.73, 95% CI = 1.33-2.24, P_meta= 4.56E-6). Moreover, CM patients in MDACC with combined risk genotypes of these two loci had markedly poorer survival (HR = 2.47, 95% CI = 1.58-3.84, P < 0.001). Genetic variants of rs6707820 C>T in NCK2 and rs2306574 T>C in PRKCD of the PDGF signaling pathway may be biomarkers for melanoma survival.

Characterization of platelet-derived growth factor-A expression in mouse tissues using a lacZ knock-in approach

Expression of the platelet-derived growth factor A-chain gene (Pdgfa) occurs widely in the developing mouse, where it is mainly localized to various epithelial and neuronal structures. Until now, in situ mRNA hybridization (ISH) has been the only reliable method to identify Pdgfa expression in tissue sections or whole mount preparations. Validated protocols for in situ detection of PDGF-A protein by immunohistochemistry is lacking. In particular, this has hampered understanding of Pdgfa expression pattern in adult tissues, where ISH is technically challenging. Here, we report a gene targeted mouse Pdgfa allele, Pdgfa^ex4COIN, which is a combined conditional knockout and reporter allele. Cre-mediated inversion of the COIN cassette inactivates Pdgfa coding while simultaneously activating a beta-galactosidase (lacZ) reporter under endogenous Pdgfa transcription control. The generated Pdgfa^{ex4COIN-INV-lacZ} allele can next be used to identify cells carrying a Pdgfa null allele, as well as to map endogenous Pdgfa expression. We evaluated the Pdgfa^{ex4COIN-INV-lacZ} allele as a reporter for endogenous Pdgfa expression patterns in mouse embryos and adults. We conclude that the expression pattern of Pdgfa^{ex4COIN-INV-lacZ} recapitulates known expression patterns of Pdgfa. We also report on novel embryonic and adult Pdgfa expression patterns in the mouse and discuss their implications for Pdgfa physiology.

Analysis of mice lacking the heparin-binding splice isoform of platelet-derived growth factor A

Platelet-derived growth factor A-chain (PDGF-A) exists in two evolutionarily conserved isoforms, PDGF-Along and PDGF-Ashort, generated by alternative RNA splicing. They differ by the presence (in PDGF-Along) or absence (in PDGF-Ashort) of a carboxy-terminal heparin/heparan sulfate proteoglycan-binding motif. In mice, similar motifs present in other members of the PDGF and vascular endothelial growth factor (VEGF) families have been functionally analyzed in vivo, but the specific physiological importance of PDGF-Along has not been explored previously. Here, we analyzed the absolute and relative expression of the two PDGF-A splice isoforms during early postnatal organ development in the mouse and report on the generation of a Pdgfa allele (Pdgfa(Δex6)) incapable of producing PDGF-Along due to a deletion of the exon 6 splice acceptor site. In situations of limiting PDGF-A signaling through PDGF receptor alpha (PDGFRα), or in mice lacking PDGF-C, homozygous carriers of Pdgfa(Δex6) showed abnormal development of the lung, intestine, and vertebral column, pinpointing developmental processes where PDGF-Along may play a physiological role.

Brain abnormalities and glioma-like lesions in mice overexpressing the long isoform of PDGF-A in astrocytic cells

Background

Deregulation of platelet-derived growth factor (PDGF) signaling is a hallmark of malignant glioma. Two alternatively spliced PDGF-A mRNAs have been described, corresponding to a long (L) and a short (S) isoform of PDGF-A. In contrast to PDGF-A(S), the PDGF-A(L) isoform has a lysine and arginine rich carboxy-terminal extension that acts as an extracellular matrix retention motif. However, the exact role of PDGF-A(L) and how it functionally differs from the shorter isoform is not well understood.

Methodology/Principal Findings

We overexpressed PDGF-A(L) as a transgene under control of the glial fibrillary acidic protein (GFAP) promoter in the mouse brain. This directs expression of the transgene to astrocytic cells and GFAP expressing neural stem cells throughout the developing and adult central nervous system. Transgenic mice exhibited a phenotype with enlarged skull at approximately 6-16 weeks of age and they died between 1.5 months and 2 years of age. We detected an increased number of undifferentiated cells in all areas of transgene expression, such as in the subependymal zone around the lateral ventricle and in the cerebellar medulla. The cells stained positive for Pdgfr-α, Olig2 and NG2 but this population did only partially overlap with cells positive for Gfap and the transgene reporter. Interestingly, a few mice presented with overt neoplastic glioma-like lesions composed of both Olig2 and Gfap positive cell populations and with microvascular proliferation, in a wild-type p53 background.

Conclusions

Our findings show that PDGF-A(L) can induce accumulation of immature cells in the mouse brain. The strong expression of NG2, Pdgfr-α and Olig2 in PDGF-A(L) brains suggests that a fraction of these cells are oligodendrocyte progenitors. In addition, accumulation of fluid in the subarachnoid space and skull enlargement indicate that an increased intracranial pressure contributed to the observed lethality.

Differential binding of platelet-derived growth factor isoforms to glycosaminoglycans

The platelet-derived growth factor (PDGF) family comprises disulfide-bonded dimeric isoforms and plays a key role in the proliferation and migration of mesenchymal cells. Traditionally, it consists of homo- and heterodimers of A and B polypeptide chains that occur as long (A(L) and B(L)) or short (A(S) and B(S)) isoforms. Short isoforms lack the basic C-terminal extension that mediates binding to heparin. In the present study, we show that certain PDGF isoforms bind in a specific manner to glycosaminoglycans (GAGs). Experiments performed with wild-type and mutant Chinese hamster ovary cells deficient in the synthesis of GAGs revealed that PDGF long isoforms bind to heparan sulfate and chondroitin sulfate, while PDGF short isoforms only bind to heparan sulfate. This was confirmed by digestion of cell surface GAGs with heparitinase and chondroitinase ABC and by incubation with sodium chloride to prevent GAG sulfation. Furthermore, exogenous GAGs inhibited the binding of long isoforms to the cell membrane more efficiently than that of short isoforms. Additionally, we performed surface plasmon resonance experiments to study the inhibition of PDGF isoforms binding to low molecular weight heparin by GAGs. These experiments showed that PDGF-AA(L) and PDGF-BB(S) isoforms bound to GAGs with the highest affinity. In conclusion, PDGF activity at the cell surface may depend on the expression of various cellular GAG species.

Detection of platelet-derived growth factor (PDGF)-AA in actively healing human wounds treated with recombinant PDGF-BB and absence of PDGF in chronic nonhealing wounds

Some human chronic dermal wounds treated with recombinant platelet-derived growth factor-BB (rPDGF-BB) show increased healing coupled with fibroblast activation and granulation tissue formation. To determine whether endogenous PDGF is associated with healing and nonhealing dermal ulcer phenotypes, we developed monoclonal antibodies capable of recognizing the three isoforms of PDGF, AA, AB, and BB dimers, and capable of discriminating between two alternatively spliced A chain transcripts. We detected little PDGF isoform expression in normal skin and in nonhealing dermal ulcers. In contrast, in surgically created acute wounds and chronic ulcers treated with rPDGF-BB, markedly upregulated levels of PDGF-AA (long form) were found. In both types of wounds, increased PDGF-AA was detected primarily in capillaries and fibroblasts, although in rPDGF-BB-treated chronic wounds, widespread expression of PDGF-AA was somewhat delayed. With continued treatment, the long form of PDGF-AA, which can preferentially bind extracellular matrix, was expressed only in capillaries, while fibroblasts began synthesizing the short form of PDGF-AA. Within capillaries, all endothelial cells and varying numbers of pericytes and smooth muscle cells contained PDGF-AA. In all wounds, macrophages and keratinocytes were not a major contributor. While PDGF-BB and PDGF-AB were present in a minority of healing wounds, they were usually present at lower levels than PDGF-AA. PDGF-beta receptors, which bind only PDGF-BB and not other isoforms, were found in normal skin and granulation tissue, providing a molecular basis for treating human chronic wounds with exogenous rPDGF-BB.

Accumulation of PDGF B and cell-binding forms of PDGF A in the extracellular matrix

PDGF is a powerful mitogen initially identified within platelets, but also shown to be produced by a wide variety of cell types. PDGF is encoded on two separate genes. These give rise to three polypeptides, PDGF B and two forms of PDGF A (SA and LA), resulting from alternative splicing of the PDGF A gene primary transcript. We report that in CHO cells transfected with PDGF gene constructs and producing moderate levels of PDGF homodimers, much of the PDGF LA and B produced, but little if any SA, is found in the matrix laid down beneath the cells. Immunoreactive PDGF in cells, and in matrix below expressing cells, was visualized by laser confocal microscopy. Western blotting of protein in matrix extracts, cell extracts, and secreted into the growth medium was used to demonstrate that the range of PDGF A polypeptides seen in the matrix was overlapping with those reported previously to be cell associated in cell types such as NIH3T3 and COS 7. This suggests that attachment to matrix or cell surface may be alternative fates for these polypeptides, with fate dependent on the characteristics of the producing cells. Immunoreactive PDGF A and B could be partially released by incubation of matrix material with heparin but not with other glycosaminoglycans. Digestion of matrix with chondroitin ABC lyase but not heparitinase or collagenase displaced some PDGF from its attachment sites. The results indicate attachment of PDGF to matrix proteoglycans, at least partly through the glycosaminoglycan moieties, and perhaps to additional components. The significance of matrix deposition for PDGF action is discussed.

Heparin-like glycosaminoglycans influence growth and phenotype of human arterial smooth muscle cells in vitro. II. The platelet-derived growth factor A-chain contains a sequence that specifically binds heparin

Synthetic oligopeptides were used to study the specificity of the interaction between heparin and platelet-derived growth factor (PDGF) in competition experiments. DNA synthesis in PDGF-dependent human arterial smooth muscle cell (hASMC) cultures was used as a biological tracer of PDGF activity. Oligo-108-124 (corresponding to amino acid residues 108-124 of the long PDGF A-chain isoform) had no effect on DNA synthesis in itself but competed at 10(-10) M concentration effectively with PDGF for binding to heparin and released the block on thymidine incorporation induced by heparin. Poly-lysine-serine (lysine:serine ratio 3:1) was also effective but at a considerably higher concentration (10(-6) M). Poly-arginine-serine did not compete with PDGF for heparin as deduced from the cell assay. This suggested that among basic amino acids, lysine was more important than arginine for heparin binding. Deletion of lysine residues 115 and 116 in Oligo-108-124 abolished its effect on the interaction between PDGF and heparin in the cell assay. Likewise, Oligo-69-84 (corresponding to the PDGF A-chain residues 69-84), with three lysine residues interrupted by a proline, was ineffective. In Oligo-108-124, the lysine residues are interrupted by an arginine. Our results suggested that the binding between PDGF and heparin is specific and that the amino acid sequence [-Lys115-Lys116-Arg117-Lys118-Arg119-] is of major importance. They do not however, exclude other domains of the PDGF A or B chains as additional binding sites for heparin nor do they exclude the possibility that heparin and the PDGF receptor share a common binding site.

Compartmentalization of PDGF on extracellular binding sites dependent on exon-6-encoded sequences

The PDGFs are a family of molecules assembled as disulfide-bonded homo- and heterodimers from two distinct but highly homologous polypeptide chains (PDGF-A and PDGF-B). Two PDGF A-chain transcripts, which arise from alternative usage of the 69-bp exon 6 and exon 7, give rise to two forms of PDGF-A. In spite of the conservation of two PDGF A-chain forms over at least 350 million years, no differences in their biological activities have been identified. We have investigated the activity of the sequence encoded by the alternatively spliced exon 6 of the PDGF A-chain (peptide AL). Addition of peptide AL at 10(-5)-10(-9) M to cultured endothelium and smooth muscle induced a dose-dependent, 3-20-fold increase in PDGF in conditioned media within 30 min. Peptide AL had no detectable effect on A- or B-chain transcript levels, and decrease in culture temperature did not prevent rapid release of PDGF. In human umbilical vein endothelial cells treated with peptide AL, the PDGF release was principally PDGF-BB, while in smooth muscle cells it was primarily PDGF-AA. The capacity to induce release of PDGF is shared by the homologous peptide encoded by exon 6 of the B-chain of PDGF. Binding studies and cross-linking analysis are consistent with a charge-based association of exon 6 sequences with membrane- and matrix-associated heparan-sulfate proteoglycans. We hypothesize that translation of exon 6 of the A- or B-chain of PDGF results in compartmentalization of these forms of PDGF with HS-PG, whereas forms lacking this sequence would be soluble and diffuse.

Identification of a cell retention signal in the B-chain of platelet-derived growth factor and in the long splice version of the A-chain

The B-chain homodimer of platelet-derived growth factor (PDGF) is only very inefficiently secreted and remains largely associated with the producer cell; in contrast, the dimer of the short, and most common, splice variant of the A-chain is secreted. To identify the structural background to the differences in the secretory pattern between the different isoforms of PDGF, a set of chimeric PDGF A/B cDNAs was generated and expressed in COS cells. Analyses of the biosynthesis and processing of the corresponding products led to the identification of a determinant for cell association in the carboxy-terminal third of the PDGF B-chain precursor. Introduction of stop codons at various positions in the carboxy-terminal prosequence of the PDGF B-chain localized this determinant to an 11-amino-acid-long region (amino acids 219-229). This region contains an 8-amino-acid-long basic sequence that is homologous to a sequence present in an alternatively spliced longer version of the PDGF A-chain. In contrast to the short splice variant, the long splice A-chain version, like the B-chain, was found to remain predominantly cell associated. Thus, we have identified a conserved sequence that inhibits the secretion of some of the PDGF isoforms. Our data also suggest that switching of splicing patterns can be a mechanism to regulate the formation of secreted or cell-associated forms of PDGF-AA and possibly other growth factors.

Alternatively spliced platelet-derived growth factor A-chain transcripts are not tumor specific but encode normal cellular proteins

Two platelet-derived growth factor A-chain proteins, termed short and long A chains, are generated as a result of alternative mRNA splicing of exon 6 of the A-chain gene. S1 nuclease mapping and polymerase chain reaction analyses demonstrate that both short and long A-chain transcripts are expressed in a variety of normal tissues. In addition, immunohistochemical localization of long A-chain protein reveals a cellular distribution identical to that observed with platelet-derived growth factor heteroserum.

Alternatively spliced platelet-derived growth factor A-chain transcripts are not tumor specific but encode normal cellular proteins

Two platelet-derived growth factor A-chain proteins, termed short and long A chains, are generated as a result of alternative mRNA splicing of exon 6 of the A-chain gene. S1 nuclease mapping and polymerase chain reaction analyses demonstrate that both short and long A-chain transcripts are expressed in a variety of normal tissues. In addition, immunohistochemical localization of long A-chain protein reveals a cellular distribution identical to that observed with platelet-derived growth factor heteroserum.