Revising the Genomic Structure of the Human GM-CSF Receptor Alpha (GMRα) Subunit: Characterization of a Novel Membrane-Spanning GMRα Isoform Which Contains a Previously Undescribed Exon.

The cytokine granulocyte-macrophage colony-stimulating factor (GM-CSF) is a key player in the processes of hematopoiesis and inflammation. The functional effects of GM-CSF are mediated by binding of the cytokine to the cell surface GM-CSF receptor, which is comprised of a ligand-binding alpha subunit (GMRα) and a signal-transducing beta subunit (GMRβ). In addition, there are at least 2 soluble GMRα isoforms (sGMRα)-one generated by alternative splicing, and a second produced by proteolytic release of the GMRα extracellular domain from the cell surface.

While examining expression of the spliced soluble and transmembrane (tmGMRα) isoforms of GMRα in neutrophils by RT-PCR, we detected a third, higher molecular weight, GMRα transcript. This was surprising since the tmGMRα transcript is comprised of all 13 exons of the published GMRα gene structure (Nakagawa et al., 1994). We therefore hypothesized that we had discovered a previously undescribed exon of the GMRα gene.

The novel GMRα (nGMRα) transcript was subsequently cloned from human neutrophils. Sequencing of nGMRα indicated inclusion of a 102 nucleotide sequence between exons 10 and 11 of the GMRα gene. Interestingly, the novel exon and surrounding intronic sequence appear to represent an Alu-repeat element, indicating that this exon is likely primate-specific. Segments of these repetitive DNA elements become inserted into mature mRNAs by way of splicing in a process termed “exonization”. In fact, it now seems that more than 5% of the alternatively spliced exons in the human genome are Alu-derived. At the protein level, this novel transcript is predicted to contain all 400 amino acids of the tmGMRα protein as well as an additional 34 amino acids within the membrane-proximal region of the extracellular domain. We have expressed the novel GMRα clone, as well as a soluble version of novel GMRα (sol-nGMRα), in the murine factor-dependent cell line Ba/F3. The sol-nGMRα protein is secreted from these cells into conditioned media, as expected. We have purified sol-nGMRα protein and used it to determine the affinity of nGMRα for GM-CSF. By flow cytometry, we were unable to detect full-length nGMRα on the surface of Ba/F3 cells, but we were able to detect nGMRα in conditioned media from nGMRα-expressing Ba/F3 cells. In addition, while tmGMRα-expressing Ba/F3 cells are able to proliferate in response to human GM-CSF in the absence of murine growth factors, the nGMRα-expressing Ba/F3 cells do not respond to GM-CSF. Our results indicate that the full-length nGMRα protein is proteolytically cleaved from the cell surface or alternatively, that its subcellular localization is otherwise disrupted.

As the nGMRα transcript was first detected in neutrophils isolated from a human donor, we sought to determine whether nGMRα would be expressed by other donors and in other hematopoietic cell types. Our results indicate that all donors tested express nGMRα, and that nGMRα is present in all hematopoietic cell types expressing the tmGMRα transcript, although tmGMRα nonetheless represents the predominant transcript.

It is becoming increasingly clear that GM-CSF signaling is more complex than was previously thought. In order to understand the actions of GM-CSF in the clinical setting, it is critical that we first define all components of the GM-CSF signaling axis.