By transplanting rhesus macaques with autologous CD34+ hematopoietic stem and progenitor cells (HSPCs) labeled with lentivirally delivered high diversity oligonucleotide barcodes, we are able to track complex clonal contributions to hematopoiesis over time and across lineages via low cycle PCR and high-throughput sequencing of the inserted barcode. We reported on patterns of reconstitution for the first 9 months following engraftment in a recent publication (Wu et al, Cell Stem Cell, 2014). For the first 1 to 3 months short lived unilineage progenitors predominated, followed by increased clonal diversity and the emergence of first shared myeloid (My)/B clones, and then My/B/T clones. A major novel finding was a distinct ontogeny for NK cells, with the major CD16+ blood NK population showing non-overlapping origin with My/B and T lineages through 6-9 months. We now report on hematopoietic clonal patterns up to 23 months post-transplant in four animals.

In all four animals, a group of high contributing NK clones contributed minimally to other lineages until 9 to 12 months, at which point these unilineage NK clones extinguished and a new generation of NK high contributors appeared. In two animals, these emerging clones continued to be either unilineage or highly biased towards the NK lineage, while in two other animals, they began to contribute highly to all lineages, a discrepancy which possibly derives from differences between animals in barcode marking level. Over the 17 and 23 months since transplant tracked in the two highly marked animals, 5.6±1.4 and 9±0.9 out of the top 10 NK contributing clones, respectively, were highly biased towards NK and away from other lineages, as determined by unsupervised hierarchical clustering of correlations between high contributing clones. Though shared My/B/T progenitors emerged after 6 months in both animals, 3.1±1.3 and 6.4±0.9 out of the T cell top 10 contributing clones were highly biased towards T cells and away from all other lineages up to 17 and 23 months. In contrast to the highly biased NK contributors, these T biased clones did not extinguish in the time surveyed so far, and more frequently were detected at low levels in the My/B lineages. The degree of bias of these high contributing clones towards the NK and T lineages was steady over time, suggesting that strong clonal bias is stable. The top 10 clones in NK contributed a significantly greater fraction of hematopoiesis (20±4% and 38±10%) than the top 10 clones in the T lineage (15±2% and 13±2%) in both animals (p=0.01, p<0.0001).

In all lineages at all time points from 1 to 23 months, a small subset (14±3%) of clones detected at each individual time point contributed at least 50% of barcoded hematopoiesis. Polyclonality increased following initial hematopoietic reconstitution but started to fall around 6 months, reflecting the increased contribution of certain high contributing, totipotent clones. These clones did not extinguish after beginning to contribute, and have characteristics of long-term repopulating HSCs.

We observed novel instability in the lineage bias of some individual HSPC contributions. For instance, an early totipotent clone which was a top 10 producer of granulocytes halted B cell contribution almost completely at 4.5 months, before regaining it by 5 months later and becoming the highest contributing clone in the entire animal. The high degree of bias towards and away from certain lineages observed in many long term repopulating HSCs is not necessarily in violation of hypothesized progenitor hierarchies, but it indicates the presence of poorly understood nuances in clonal control. In addition to the long-term dominant unbiased HSCs whose hematopoietic contribution increased over time, we noticed a cohort of clones biased towards a combined myeloid/B cell lineage.

The varied clonal patterns observed between animals could result from individual differences in transplant progenitor composition or dose, or degree of recipient endogenous HSPC depletion. Understanding the lifespan and population dynamics of repopulating clones in macaques after transplantation is relevant for optimizing human hematopoietic stem cell transplants and also provides an approach for identifying progenitor populations, inferring mechanisms of cell fate control, and calculating rates of differentiation.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.