Novel Integrated Autologous Hematopoietic Stem Cell Tracking in Nonhuman Primates Reveals Successive Pattern of Multi-Lineage Reconstitution after Total Body Irradiation

Hematopoietic stem cell transplantation (HSCT) has been used to cure patients for over four decades, yet the kinetics and clonal dynamics of repopulation by multipotent HSCs remains unresolved. Two sequence-based methods have been used to measure HSC contribution in preclinical models, integration site (IS) analysis and DNA barcoding. Both methods are accomplished by retrovirus tagging of HSCs and each confers different methodological constraints. We hypothesized that IS analysis and DNA barcoding together would provide more robust HSCT reconstitution data if applied to the clinically relevant nonhuman primate autologous HSCT model. We developed a high complexity (~1.2 million), DNA-barcoded LV library encoding the chemotherapy resistance gene MGMT(P140K) and enhanced green fluorescent protein (GFP) to measure hematopoietic reconstitution following total body irradiation (1020 cGy) and autologous HSCT in two pigtailed macaques. In both animals, hematopoietic recovery from TBI was achieved within 1-2 months after HSCT. We observed ~1 and 12% GFP+ peripheral blood white blood cells (PB WBCs) following hematopoietic recovery in these animals, respectively. We performed barcode and IS retrieval from PB WBCs collected early (1 month) after HSCT and observed <0.001% of infused clones contributing to hematopoiesis by DNA barcoding. We identified 42% fewer clones by IS analysis of the same sample, but with similar contributions of the most dominant clones identified by each method. This suggests DNA barcoding provides greater retrieval of low abundance clones. Over 1 year following HSCT, a total of 5,089 (0.4%) and 20,799 (1.7%) infused clones contributed to hematopoiesis in each animal, respectively, with polyclonal contribution patterns. At 1 year post-transplant, we analyzed PB WBC subsets and identified a total of 116 and 2,246 clones respectively contributing to each of three lineages (T, B, and granulocyte), suggesting <0.2% of infused, gene-modified CD34⁺ cells displayed multi-lineage potential. Some of these clones emerged in the periphery as early as 1 month after HSCT, indicating a successive pattern of multi-lineage clonal behavior. Adjusting for marking levels, we estimate that 1 in 10,000 transplanted CD34⁺ cells gave rise to trilineage hematopoiesis at 1 year post-HSCT. We then challenged these animals with chemotherapy (O⁶-benzylguanine and BCNU) to provide gene modified cells with a selective advantage in vivo. We observed moderate myelosuppression following chemotherapy with stable increases in GFP⁺ PB WBCs to ~2.5 and 22% in each respective animal. Analysis of the clonal repertoire after chemotherapy identified a significant number of new clones not previously identified in the first year after transplant. This suggests that >1 in 10,000 infused CD34⁺ cells is capable of trilineage hematopoiesis. In sum, the combination of retrovirus integration site analysis with molecular barcode analysis provides synergistic methods in understanding the kinetics and clonal dynamics of hematopoietic reconstitution following autologous HSCT and revealed a successive pattern of multi-lineage clonal behavior.