Platelet and myeloid lineage biases of transplanted single perinatal mouse hematopoietic stem cells

Dear Editor, Hematopoietic stem cells (HSCs) residing in adult mouse bone marrow (BM) are highly heterogenous, 1 ranging from HSCs executing their full multilineage potential to those with platelet (P)-biased/restricted output. 2 – 4 These distinct lineage-biased propensities are sustained upon secondary transplantation, and thus largely intrinsic. 1 While adult HSCs derive from de ﬁ nitive fetal HSCs, 5 it is unclear to what degree the lineage biases/restrictions observed in adult BM HSCs, or any other distinct lineage biases/ restrictions, have been established before birth. De ﬁ nitive HSCs expand in the mid-gestation fetal liver, and at embryonic day (ED) 17.5, they start migrating to the BM, the primary hematopoietic site throughout life. 5 Previous transplantation studies of single fetal HSCs, assessing only white blood cell lineages, identi ﬁ ed myeloid (M)-biased and lymphoid (L)-biased HSCs in ED14.5 and ED18.5 liver. 6 Whereas previous studies identi ﬁ ed P-biased/restricted adult HSCs, 2 – 4 the propensity of fetal and perinatal HSCs (pnHSCs) to replenish the critical platelet and erythroid (E) lineages has yet to be investigated at a clonal level in mice or humans. Herein, we systematically analyzed post-transplantation ﬁ ve-lineage blood replenishment by single pnHSCs at ED19.5/postnatal day 0 (PD0). Utilizing a Vwf - tdTomato tg/ + ; Gata1 -eGFP tg/ + mouse transplantation model, 2 we assessed clonal contribution over time to blood P, E, M, B and T cells by single Lineage − Sca1 + Kit + (LSK)


Dear Editor,
3][4] These distinct lineage-biased propensities are sustained upon secondary transplantation, and thus largely intrinsic. 1While adult HSCs derive from definitive fetal HSCs, 5 it is unclear to what degree the lineage biases/restrictions observed in adult BM HSCs, or any other distinct lineage biases/ restrictions, have been established before birth.
Definitive HSCs expand in the mid-gestation fetal liver, and at embryonic day (ED) 17.5, they start migrating to the BM, the primary hematopoietic site throughout life. 5Previous transplantation studies of single fetal HSCs, assessing only white blood cell lineages, identified myeloid (M)-biased and lymphoid (L)-biased HSCs in ED14.5 and ED18.5 liver. 6Whereas previous studies identified P-biased/restricted adult HSCs, [2][3][4] the propensity of fetal and perinatal HSCs (pnHSCs) to replenish the critical platelet and erythroid (E) lineages has yet to be investigated at a clonal level in mice or humans.Herein, we systematically analyzed posttransplantation five-lineage blood replenishment by single pnHSCs at ED19.5/postnatal day 0 (PD0).Utilizing a Vwf-tdTomato tg/+ ;Gata1-eGFP tg/+ mouse transplantation model, 2 we assessed clonal contribution over time to blood P, E, M, B and T cells by single Lineage − Sca1 + Kit + (LSK)CD150 + CD48 − HSCs from liver and BM (Fig. 1a; Supplementary information, Fig. S1 and Table S1).22% of mice transplanted with a single liver or BM pnHSC showed ≥ 0.1% long-term (LT) contribution to at least one blood cell lineage after 25-26 weeks (Supplementary information, Fig. S1a).Over 90% of reconstituting single pnHSCs (from liver or BM, Supplementary information, Fig. S1) replenished all blood lineages (Fig. 1a), in contrast to < 50% of single adult BM LT-HSCs. 2 Whereas a third of single adult LT-HSCs contribute exclusively to PEM lineages, 2 no pnHSCs were P-or PEM-restricted (Fig. 1a, b) in liver or BM (Supplementary information, Fig. S1b).Likewise, whereas half of adult LT-HSCs possess P/PEM-bias, 2 only one pnHSC from BM and none from liver displayed P-bias, and < 15% displayed PEM-bias (Fig. 1c; Supplementary information, Fig. S1c).A small fraction of pnHSCs were PEMB-restricted at 25-26 weeks (Fig. 1a, b), representing the only lineage restriction of pnHSCs at this time.However, while most P-and PEM-restricted adult HSCs remain restricted in primary recipients, 2 for most pnHSCs showing PEMB-restricted reconstitution at 25-26 weeks, T cell output became apparent with time (Supplementary information, Fig. S2a,  b).Thus, overall, transplanted pnHSCs show less P-and PEM-bias/ restriction and more consistent lymphoid output than adult HSCs.
Many multilineage pnHSCs showed no prominent lineage bias (Fig. 1c), sustaining LT production of all blood lineages, and replenishing hematopoietic stem and progenitor cells (HSPCs) in primary and secondary recipients (Fig. 1d, e; Supplementary information, Fig. S2c-g), confirming their extensive self-renewal and stable lineage bias.
In agreement with published ED18.5 data, 6 many pnHSCs showed L-biased and L-restricted LT contribution to blood (Fig. 1a-c; Supplementary information, Fig. S1b, c).However, L-bias of multilineage-reconstituting LSKCD150 + CD48 − cells only became apparent weeks or months post-transplantation, and all L-restricted cases contributed to PEM lineages at earlier time points (Fig. 1a).The L-restricted contribution was typically preceded by L-biased reconstitution (Fig. 1a; Supplementary information, Fig. S3a, b), and primary L-biased reconstitution patterns became L-restricted upon secondary transplantation, with drastically reduced PEM reconstitution (Fig. 1f-i).HSPC replenishment was low and inconsistent in primary recipients of Lbiased/restricted HSCs (Supplementary information, Fig. S3c, d), and absent in secondary recipients (Fig. 1g, i), demonstrating that L-biased or L-restricted blood contribution reflects the prolonged longevity of lymphocytes compared to short-lived myeloid lineages, 7 rather than a lymphopoietic bias of LT-HSCs.Of relevance to our findings, recent clonal tracking of native hematopoiesis suggested that myeloid and lymphoid cells are mainly derived from embryonic multipotent progenitors (eMPPs) independently of HSCs, at least in the first months after birth, and that eMPPs remain the principal source of lifelong lymphoid contribution. 8In our studies, many pnHSCs, enriched among Vwf-tdTomato − LSKCD150 + CD48 − cells, gave LT robust reconstitution primarily of lymphocytes, and further analysis of reconstituted BM demonstrated that these L-biased/restricted cells were not true LT-HSCs, and might therefore be more closely related to shortterm (ST) HSCs or eMPPs.
Our single HSC transplantations were performed under hematopoietic stress conditions using myelo-ablated recipients, therefore mimicking clinical BM transplantation rather than steady-state hematopoiesis, although, as in corresponding studies of adult HSCs, 2 lineage-bias remained quite stable many months after transplantation when hematopoiesis had become more normalized.Regardless, the purpose of single HSC transplantations is to perform definitive clonal tracking of blood lineage contribution over time by a single cell with extensive LT self-renewal capacity, and our kinetic fate-mapping of post-transplantation outputs from single pnHSCs provides new insights into the blood replenishment potentials of definitive HSCs at birth, as compared to mice transplanted with adult single HSCs. 2 Whereas a large fraction of adult HSCs replenishes exclusively PEM lineages and frequently are P-biased, these lineage restrictions and biases are rare or absent in pnHSCs.Unlike previous studies, 6 we did not find that M-bias increased in HSCs residing in BM at late gestation, as no significant differences were observed in lineage-biases in liver and BM.Our functional data are congruent with published molecular data 14 demonstrating that liver and BM pnHSCs are transcriptionally similar, and rather cluster by developmental stage.
Our study does not address the earliest time point of P-bias/ restriction emergence, although the low frequency of P-and PEMbias/restriction observed in pnHSCs shows that this property, shared by a large fraction of adult HSCs, emerges primarily after birth.While our single-cell functional and transcriptional data show that the extent of P-and PEM-bias/restriction is very limited in pnHSCs when compared to adult HSCs, in agreement with previous molecular analysis of Mk priming, 15 we nevertheless observed that this process had initiated at birth in a pattern reminiscent of adult HSCs.Indeed, the observed lineage biases (P, PEM, and PEMB) were identical to those of adult HSCs, and P-and PEM-biases were exclusively a property of pnHSCs expressing Vwf-tdTomato, a hallmark of P-and PEM-restricted HSCs in adult BM. 2 To what degree the distinct lineage-biases of HSCs, including Pand PEM-bias, are intrinsically and/or extrinsically determined remains to be explored further.However, the fact that both adult [2][3][4] and perinatal LT-HSCs show similar lineage-biases upon transplantation into primary and secondary recipients, as well as transcriptional lineage priming corresponding to their lineage-bias upon transplantation, suggests that this is at least in part an intrinsically programmed and stable HSC property.The involvement of epigenetic programming in the P-bias of HSCs 16 is also compatible with an interplay between intrinsic and extrinsic cues.