Pak A Proteins Are Essential for Hematopoietic Stem/Progenitor Cell (HSC/P) Engraftment Through Regulation of Signaling Pathways Controlling HSC/P Proliferation, Survival, Migration, and the Actin Cytoskeleton

Abstract 917

Rho GTPases, including Rac, integrate multiple extracellular signals and play important regulatory roles in HSC/P functions such as engraftment, retention, migration, adhesion, proliferation, and survival (Gu et al. Science, 2003). Our studies now focus on identifying potential Rac downstream effector proteins important for normal HSC/P function(s). The p21-activated kinases (Pak) are serine/threonine kinases that interact with and are major downstream targets of Rac and Cdc42. There are six human Paks (Pak1–6), which are grouped based on homology into Group A (Pak 1–3) and Group B (4–6) Paks. Paks regulate cytoskeletal organization including stress fiber dissolution, lamellipodia formation and focal adhesion disassembly and mediate activation of MAPK pathways. To identify the possible role(s) of Pak proteins in engraftment, freshly isolated LSK cells from WT (CD45.1+/CD45.2+) BM were transduced with retrovirus containing the Pak Inhibitory Domain (PID), which inhibits Group A Pak protein function or empty vector control (Mieg3); both constructs co-express GFP. 1.0×10⁵ GFP+ LSK+ cells were then isolated and co-transplanted with 5.0×10⁵ BoyJ (CD45.1+) whole bone marrow (WBM) into lethally irradiated C57Bl/6J (CD45.2+) recipients. Percent chimerism was measured at 3- to 24- weeks post BMT. PID transduced LSK+ cells were incapable of contributing to recipient hematopoietic reconstitution (Table 1). To explore the underlying mechanism of this engraftment failure we performed in vivo homing assays and found a 4- and 16- fold decrease, respectively, in BM homing of PID transduced LSK+ vs controls at 12 and 48 hours (p<0.05, for both time points). Altered cell migration of LSK+ cells was confirmed by live imaging microscopy which showed a 4-fold decrease in overall cell displacement in SDF-1-stimulated directed migration in the PID-expressing LSK+ compared to controls and was associated with a two-fold increase in random cell migration of PID-transduced LSK+ cells in transwell migration assays. PID-expressing LSK+ cells also demonstrated abnormal lamellipodia associated with significant increases in both cell surface area and cell perimeter. Because cytoskeletal changes may be linked to alterations in cell growth, we next examined the effect of Pak inhibition on cell survival and proliferation. PID-expressing LSK+ cells had decreased proliferation (17.7% vs 36.8% of cells in S-phase, p<0.05) and increased apoptosis (48.1% vs 16.7% AnnexinV+ cells, p<0.05) when compared to controls, respectively. These phenotypic changes were associated with decreased pERK and pAKT in PID-expressing LSK+. To confirm the importance of Pak activation of these proteins in HSC/P, we performed experiments to rescue the observed engraftment defect by co-transducing PID or Mieg3 with a constitutively active-ERK (ca-MEK1) or ca-AKT. We found ca-MEK1, but not ca-AKT, was able to increase proliferation in vitro (% proliferating cells for PID + empty vector = 6.1% and PID+ ca-MEK1 = 9.5%; p<0.05) and partially but only transiently rescue Pak-deficient HSC/P engraftment (% donor cells for LSK+ transduced with: PID + empty vector =1.5%, PID + ca-MEK1 =15.8%, and PID + ca-AKT =0.5% at 3-weeks post-BMT; p<0.05 for empty vector vs ca-MEK1). Finally, to determine which PakA pathway is critical in HSC engraftment we studied Pak genetic knock-out cells. We found that Pak2^Δ/Δ -but not Pak1^−/− -cells resulted in a profound HSC/P engraftment defect (% Pak2^Δ/Δ vs Pak2^flox/flox and Pak1^−/− vs Pak1^wt/wt: 1.0% vs 26.5% and 35.8% vs 37.4%; p<0.05 and p=ns, respectively at 3-weeks). Taken together, these data suggest that Pak A proteins regulate multiple HSC/P functions and link Rac GTPases to actin cytoskeletal rearrangements, directed cell migration, and proliferation/survival of HSC/P during engraftment.