Hematopoietic stem and progenitor cell (HSPC) expansion remains an important unmet goal for ex vivo gene therapy based on gene addition and editing to compensate for the negative impact of the gene transfer procedure enabling faster engraftment and less complications. Additionally, ex vivo expansion of corrected cells may improve efficacy at more sustainable manufacturing costs by downscaling transduction. To date, our knowledge of precise mechanisms of action of expansion compounds is limited, and it remains unclear whether cord blood expansion protocols also maintain stemness of mobilized peripheral blood CD34+ cells (mPB), the preferred HSPC source for gene therapy.

We performed serial (day 0,4,8) droplet-based single cell RNA sequencing (scRNAseq) on lentivirally transduced mPB expanded with UM171 to dissect cellular heterogeneity, monitor population dynamics over time and identify a transcriptional profile of primitive cells in culture.

By associating published HSPC gene expression profiles to our scRNAseq dataset from uncultured mPB, we found that 45% of cells harbored a myelo-lymphoid signature. Smaller cell clusters expressed a shared erythroid (ERY) and megakaryocytic (MK) signature (20%), or a more primitive multipotent HSC-like signature (15%) characterized by enhanced JAK/STAT signaling and expression of HSC associated genes (AVP, HOPX, ID3). Unsupervised ordering of cells within pseudotime separated emerging MK/ERYpoiesis (FCER1A, HBD) from lympho-myelopoiesis (CD52, JUN), with intermediate states of more primitive progenitors located in between.

After 4 days in culture, we noted a general increase in nuclear and mitochondrial gene transcription with activation of oxidative metabolism, paralleled by cell cycle activation, as expected from cytokine stimulation. By d8 of culture these changes leveled off but remained higher than uncultured cells. Of note, cells at d8 revealed an activation of cellular stress response pathways (e.g. TNFa, IFNg) hinting towards a compromised culture that may eventually exhaust HSC.

Unsupervised clustering of cultured mPB highlighted a dramatic expansion (70-80%) of MK/ERY progenitor cells with high cycling activity with only 20-30% cells showing myelo-lymphoid transcriptional features. In line, pseudotime analysis highlighted a main ERY and MK trajectory separated from that of cells characterized by the expression of HSPC genes (HOPX, SPINK2) and of an emerging myeloid trajectory (MPO).

To profile HSC in culture, we sorted and sequenced CD34+90+201+ cells from d4 expansion culture (3% of total cells), which we show to contain >70% of SCID repopulating potential. ScRNAseq revealed transcriptional similarity with the myelo-lymphoid progenitor cluster identified in the unsorted d4 culture. Unsupervised clustering of the CD34+90+201+ population revealed cell cycle dependent heterogeneity, identifying a highly quiescent cluster with expression of HSC-like signatures. This cluster was also characterized by relatively low gene expression, possibly reflecting a non-activated cell state consistent with primitive HSPC. Pseudotime analysis produced a four-branched minimum spanning tree, which retained a clear cell cycle and metabolic effect. Top variable genes included cell cycle, glycolytic, mitochondrial and ribosomal genes, identifying different metabolic modules along the branched trajectory. These results highlight that cell heterogeneity within a purified, HSC-enriched population is driven mainly by metabolic activation and cell cycle status.

As a complementary approach, we purified LT-HSC from uncultured mPB (CD34+38-90+45RA-49f+), marked them with CFSE and expanded them in UM171 culture. LT-HSCs expanded on average 3.5 fold in 7 days, with the following distribution: 0 divisions: 3%; 1: 26%; 2: 47%; 3: 21%; 4: 3%. We performed scRNAseq on LT-HSC pre culture and after 7d separating a highly proliferative (≥2 divisions) and quiescent (0 - 1 division) fraction, allowing us to obtain unprecedented insight into the response of engrafting cells to ex vivo culture and set a framework to dissect self-renewal (HSC expansion), HSC maintenance and loss through differentiation as potential culture outcomes. Our combined functional/transcriptomic approach will define new HSC markers in culture and greatly facilitate side-by-side comparison of different expansion protocols towards rapid clinical translation.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution