DPP4 (CD26) is a dipeptidyl peptidase that functions by enzymatically cleaving the penultimate proline, alanine or select other amino acids such as serine of proteins, resulting in functional alterations of the protein. We recently published that many cytokines, chemokines and growth factors have putative DPP4 truncations sites and that DPP4 specifically was able to truncate some colony stimulating factors such as GM-CSF and IL-3 with resultant blunting of their activity. However, the mechanism of action of the truncated factors is still unknown and requires further investigation. The expression, and activity, of DPP4 is relevant in normal and malignant hematopoiesis as we have data showing that CD34+ umbilical cord blood cells (UCB) as well as Acute Myelogenous Leukemia (AML) patient samples express active DPP4. Further, specific inhibition of DPP4 increases homing and engraftment of both human UCB and mouse bone marrow cells after transplantation in mice indicating the therapeutic potential of DPP4 activity altering compounds. Due to its potential importance in disease states, and their subsequent treatment, it is relevant to study how the activity of DPP4 alters the functions of the molecules it cleaves, and subsequently their interactions with each other.

DPP4 can cleave the penultimate proline of GM-CSF and IL-3 resulting in truncated forms which have blunted colony stimulating factor activity for hematopoietic progenitor cells (HPC). Since GM-CSF and IL-3 receptors share a common receptor beta chain, we investigated if DPP4 truncation of GM-CSF (TGM) or IL-3 (T3) could inhibit the receptor binding and functional activity of the full length (FL) alternate compound (i.e TGM inhibition of FL3 activity or T3 inhibition of FLGM activity) in the factor dependent TF-1 cell line, UCB cells and in in vivo mouse studies. We determined using TF-1 and UCB that both T3 and TGM bound to the receptors with higher affinity than their FL forms and could blunt the receptor binding of the FLGM and FL3. Additionally, TGM and T3 decreased colony formation induced by either FLGM or FL3 in both TF-1, UCB, and primary AML patient cell samples. Strikingly, this inhibition of colony formation did not require a 1:1 ratio of the full length to truncated forms of these cytokines. Rather, approximately 4-10 fold less truncated molecules could be used to efficiently inhibit the colony formation activity of the full length form, even across molecules. In vivo injection of FL, T, or a mixture of FL/T or T/T factors into DPP4 activity knockout mice followed by colony assays showed the TGM and T3 suppresed the effect of FLGM or FL3 on progenitor cell numbers per femur and diminished cycling of hematopoietic progenitor cells as detected by high specificity tritiated thymidine kill assay.

Proteomic analysis of the effects of full length and truncated factors (FLGM, FL3, TGM, T3) were performed with TF-1 cells where we detected differential protein regulation by the full length vs truncated factors. After 24 hour treatment with 10ng/ml of FLGM or TGM, TF-1 cells displayed statistically significant (p < .05) differences in 26 proteins of which 17 were higher in the FL vs the T, and 9 higher in the T vs FL treated groups. These proteins included, but were not limited to, cell cycle proteins such as CDK6, HDAC6, as well as signal transduction proteins and redox control proteins such as STAM1 and Glutaredoxin. Additionally, alterations in protein phosphphorylation were detected for TF-1 cells treated for 15 or 30 min with the full length vs truncated IL-3 and GM-CSF proteins. Interestingly, the protein expression or phosphorylation levels were not always decreased by the truncated protein compared to the full length. In some cases, the truncated molecules induced an increase in the protein expression or phosphorylation.

These data suggest interesting roles for full length and truncated GM-CSF and IL-3 in both normal and malignant hematopoiesis. Further investigation into the regulation of DPP4, and the roles that full length and truncated factors play during normal and malignant hematopoiesis, are important and will allow for a better understanding of the signficance of DPP4 activity during steady state, stressed, and disease hematopoiesis.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.