CD38 promotes hematopoietic stem cell dormancy via c-Fos

Article activity feed

1. 

   Review Commons

   Apr 14, 2023

   Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   The authors would like to thank the reviewers for their valuable comments and suggestions. We have carefully considered all of the points raised and revised our manuscript accordingly. In the rebuttal letter below, we have extensively discussed all the different concerns and adjustments we made to our work. In what follows the reviewers’ comments are in blue and the authors’ responses are in black. The additions and changes to the main and supplementary text of the manuscript are highlighted in yellow.

   *Reviewer #1 (Evidence, reproducibility and clarity (Required)): *

   *In their paper entitled "CD38 promotes hematopoietic stem cell dormancy via c-Fos", …

   Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   The authors would like to thank the reviewers for their valuable comments and suggestions. We have carefully considered all of the points raised and revised our manuscript accordingly. In the rebuttal letter below, we have extensively discussed all the different concerns and adjustments we made to our work. In what follows the reviewers’ comments are in blue and the authors’ responses are in black. The additions and changes to the main and supplementary text of the manuscript are highlighted in yellow.

   *Reviewer #1 (Evidence, reproducibility and clarity (Required)): *

   *In their paper entitled "CD38 promotes hematopoietic stem cell dormancy via c-Fos", Ibneeva et al., present a set of data predominantly from mouse HSCs where they explore the cell cycle kinetics and self-renewal capacity of LT-HSCs expressing (or not) CD38. They perform a series of sophisticated in vitro and in vivo experiments, including transplantations and single cell cultures and arrive at the conclusion that CD38 can fractionate LT-HSCs that are more deeply quiescent. Overall, it is an interesting question and would be of interest to experimental hematologists. That said, I had a number of issues that concerned me throughout the manuscript with regard to the robustness of the conclusions around CD38 and I have tried to detail these below.

   Major concerns: *

   *1) Novelty - It was unclear what the relationship of this CD38+ fraction had with other "segregators" of LT-HSCs - e.g., how does it compare with the Sca1 fractionation of Wilson et al, Cell Stem Cell 2015 or Gprc5c of Cabezas-Wallschied Cell 2017? Even if CD38 fractionated LT-HSCs, it was unclear what it would give beyond these two molecules (especially re: Sca-1 which is also a cell surface marker). *

   Response:

   We agree with the reviewer that further elaboration of this point with additional data would be helpful. We compared the expression of Sca-1 in the population of LT-HSCs (Lin- Kit+ Sca-1+ CD48- CD150+ CD34- CD201+) based on the gating strategy from the paper Wilson et al, Cell Stem Cell 2015. We found that all LT-HSCs (independent of CD38 expression) express Sca-1 at a high level and can be quantified as Sca-1hi (we have added these data in Fig. S2A). Thus, CD38 subfractionates LT-HSCs, and considering that we have shown that CD38+ are more quiescent (Fig. 3) and have higher repopulation capacity compared with CD38- LT-HSCs (Fig. 2E-G), we conclude that CD38 should be used in addition to Sca-1 to define dormant LT-HSCs.

   We found that CD38+ dormant HSCs expressed Gprc5c mRNA at higher levels than CD38- LT-HSCs (Fig. 5D). Therefore, we cannot exclude that CD38+ and Gprc5c+ identify the same population of dormant HSCs. However, Cabezas-Wallscheid Cell 2017 used the reporter Gprc5c-EGFP mouse strain, which is not widely available. In contrast, we propose to use readily available antibodies against CD38 for efficient isolation of dormant HSCs. Moreover, to define CD38+ dormant HSCs, researchers do not need to use the CD38KO mice as a negative control, it would be sufficient to use total bone marrow cells to identify the CD38+ population for gating dHSCs (we have added this information to Fig. S2C and in the text: line 119-121: “We demonstrated that total bone marrow cells can be used to define the CD38+ fraction in the absence of CD38 knock-out mice (CD38KO) (Fig. S2C), providing the possibility of an internal positive control for easy identification of CD38+ cells”.

   *Claims of CD38+ superiority in transplantation - I was surprised with the claim of CD38 negative cells being a less functional HSC when they are clearly still very strong in secondary transplantation assays. Both 38+ and 38- cells strongly repopulate secondary animals and only 5 mice were shown in the Figure. The legend suggests another experiment was undertaken, but these data are not presented. Did they substantially differ in their chimerism in primary and secondary animals? Was the magnitude of difference between the two fractions similar in both experiments? Is there a reason that the data could not be plotted on the same graph?

   We have added the data from the second experiment to the graphs and changed the figure legend accordingly (Fig. 2D-H), now for primary transplantation n=8, for secondary transplantation n=6 vs 7. These data show the same trend of higher repopulation capacity of CD38+ LT-HSCs compared to CD38- LT-HSCs, although with the larger magnitude of difference in primary transplantation. We agree with the reviewer that CD38- LT-HSCs strongly repopulate secondary animals. However, the higher percentage of chimerism in peripheral blood and bone marrow for CD38+ LT-HSC progeny indicates their superior repopulation and self-renewal capacity compared to CD38- counterparts.

   Also, the typical experiment to establish a quantitative difference in HSC production would be a limiting dilution analysis with a much larger number of recipient animals - without such data it is difficult to ascertain how different the two fractions really are.

   While we appreciate the reviewer's suggestion to include additional data on the amount of repopulating HSCs, we respectfully disagree as we believe that this information is beyond the scope of the current study, which only aims to assess the functional superiority of CD38+ LT-HSCs over CD38- LT-HSCs in side-by-side comparisons. Assessment of donor-derived cells’ frequency in peripheral blood and bone marrow relative to the frequency of competitors after transplantation of the same amount of HSCs (so-called chimerism level) is a widely accepted assay in the field to demonstrate the difference in the functionality between two HSC fractions (Sanjuan-Pla et al., Nature 2013; Gekas C and Graf T, Blood 2015; Bernitz J.M et al. Cell 2016; and others, including papers cited by the reviewer: Wilson et al., Cell Stem Cell 2015 and Cabezas-Wallscheid et al., Cell 2017). A limiting dilution experiment will provide more detailed characteristics of two HSC fractions, namely the quantitative difference (how many cells from the sorted population can repopulate). However, this experiment will not significantly change our conclusion that the CD38+ LT-HSC fraction is superior in repopulation and self-renewal capacity compared to the CD38- LT-HSC fraction, as sufficiently demonstrated in Fig. 2E-G.

   Furthermore the claim that CD38- HSCs do not ever produce CD38+ cells is a bit premature with so few mice and confusingly presented data (e.g., Fig 2I is 5 pooled mice in a single histogram plot - were these concatenated flow files? If so, how were they normalised? Did the other experiment look the same? And were all CD38+ HSCs capable of giving rise to both CD38+ and CD38- cells or was it a subfraction of mice/samples?).

   The plot provided in Fig. 2I is a FACS analysis of pooled cells from mice transplanted with CD38+ or CD38- LT-HSCs (we added a detailed explanation in figure legend 2, lines 701-703). We provided data from the second experiment in Fig. S2G. All CD38+ LT-HSCs could give rise to both CD38+ and CD38- HSC; we added data in Fig. S2H.

   Cell Cycle status differences and grades of quiescence - Ki67 and DAPI are really quite tricky for discerning G0 versus G1 and no flow cytometry plots are provided for the reader to assess how this has been done. Could another technique (e.g., Hoechst/Pyronin) be used to confirm the results? Perhaps more concerning is the variability of the assay in the authors own hands. If I am interpreting things correctly, the plots in 3G, 3H and 3I in the platelet depletion, pIpC and 5FU experiments are >10% higher in the CD38- control arm than the data in 3A which make me worried about the robustness of the cell cycle assay to distinguish G0 from G1.

   Ki67 and DAPI staining is a widely accepted technique for distinguishing G0 from G1. We provide flow cytometry plots in Fig. S2F (original figures, S3B - updated figures), which the referee may have overlooked. We added a reference to the Fig. S3B to figure legend 3 to make it more transparent for the readers. We would like to clarify the reviewer’s concern regarding the slightly different frequency of CD38- cells in the G0 phase of the cell cycle at steady state in Fig. 3A (original figures). Fig. 3A compares the cell cycle stages between CD38- and CD38+ HSCs, while Fig. 3B compares the same parameters for CD38- vs CD38+ LT-HSCs, which are enriched for quiescent HSCs by using additional surface markers. Therefore, it is correct to compare the data for LT-HSCs under stress (Fig. 3G-I, original figures) with the data for LT-HSCs at steady state in figure 3B (original figures). To make it less confusing for the reader, since the entire Figure 3 is devoted to LT-HSCs, we have moved Figure 3A to the supplementary Figures (Fig. S3A).

   All experiments for Fig. S3A&3A, 3F, 3G, and 3H (updated figures), were performed separately, and we did not compare mice from different experiments to avoid differences due to technical details. However, the groups of mice for each specific treatment (ctrl vs. treatment at different time points) were analyzed on the same day, using the same amount of cells, the same master mix of antibodies, and the same FACS machine and settings to compare ctrl vs. treated mice (we added this information in the Materials and Methods section, lines 388-391). In addition, we performed a BrdU incorporation assay and label retention assay using H2B-GFP mice, which support our finding that CD38+ LT-HSCs are more quiescent than CD38- cells in the steady state.

   Minor points: Figure 3I was really confusing - it says it is the gating strategy for GFP retaining LT-HSCs, but only shows GFP versus cKit

   We reformulated the figure legend for 3D: “Representative plot defining GFP+ cells in LT-HSCs.”

   Figure 4B suggests that only 40% of CD38+ cells divide in the first 3 days - are there survival differences or are the cells sat there as single cells? It would be important to carry these further to see if cells eventually divide.

   This is a relevant and crucial point addressed by the reviewer. We did not find any significant difference in the survival of cells. We have added this data to the supplementary data - Fig. S4Q-R.

   Reviewer #1 (Significance (Required)):

   I believe the study will be of interest to specialist readers in the HSC field, especially those working on quiescence and G0 exit. At present, I think the conclusion of a true subfractionation is a bit premature, but there are pieces of data that do look exciting and warrant further investigation. It was a little unclear how this would advance beyond Sca-1 or Gprc5c fractionation for finding more primitive HSCs, but having cleaner markers is always a useful advance for the field.

   We thank the reviewer for his/her positive evaluation of our study. In our work, we compared several functional aspects of CD38+ and CD38- LT-HSCs:

   1. We used four techniques (Ki67 and DAPI staining, BrdU incorporation assay, label retention assay, single-cell division tracing assay) and showed that CD38+ LT-HSCs are more quiescent than CD38- cells.
   2. We performed a serial transplantation assay and found that although CD38- LT-HSCs have the long-term repopulation capacity, they repopulate significantly less effectively than CD38+ LT-HSCs.
   3. We used a combination of surface markers (Lin- Kit+ Sca-1+ CD48- CD150+ CD34- CD201+) to define LT-HSCs; all of which belong to the Sca-1hi population according to Wilson et al, 2015. We further separated Sca-1hi LT-HSCs into CD38+ and CD38- cells and found that they differ in the repopulation capacity and quiescence in steady state and upon hematological stress. We conclude that CD38 surface staining should be used on top of Sca-1 to sort dormant LT-HSCs.
   4. We found that CD38+ dormant LT-HSCs differ from CD38- cells in gene expression and response to CD38 and c-Fos inhibitors. CD38+ LT-HSCs are characterized by higher cytoplasmic Ca2+ and cell cycle inhibitor p57 levels than CD38- LT-HSCs. Thus, we demonstrated that CD38 is not only a marker but also has a functional role in mediating HSC dormancy. We discovered that CD38/cADPR/Ca2+/c-Fos/p57 axis regulates CD38+ HSC dormancy. Taken together, our findings demonstrate that CD38+ LT-HSCs have superior properties compared to CD38- LT-HSCs and can be classified as dHSCs, providing a simple approach for their isolation and further study. Moreover, we uncovered the CD38-mediated molecular mechanism regulating HSCs dormancy.

   Regarding my own expertise - I have spent ~20 years in the field undertaking single cell assays of normal and malignant mouse and human HSCs, including many of the core functional assays described in this paper and consider myself very familiar with the topic area.

   Reviewer #2 (Evidence, reproducibility and clarity (Required)):

   Although the experiments were well done and supported their testing hypothesis, but the overall novelty of the whole work is not that strong and this is because:

   -the use of CD38 to identify/select and to test mouse LT-HSCs' function in vivo (although not commonly used nowadays) was demonstrated a few times more than 20 years ago (Randall, et al., 1996: PMID: 8639761 and Tajima et al., 2001; PMID: 11313250); in fact, the authors didn't even reference/acknowledge these papers which they should have done so; hence, most of the results in Fig.2 were already known (despite this current work gave a more detailed/better analysis);

   We agree with the reviewer that the previous findings using CD38 to separate HSPCs should be appreciated; however, we would like to point out that while the studies by Randall, et al., 1996: PMID: 8639761 and Tajima et al., 2001; PMID: 11313250 employ only 3 markers to discriminate HSPC (Lin- Sca-1+ Kit+), in our study, we performed for the first time a very detailed characterization of CD38+ cells using surface markers that were not available 20 years ago. We analyzed not only the HSC compartment but also different populations of multipotent progenitors. Modern surface marker combinations for the LT-HSC isolation allow us to show that both populations: CD38- and CD38+, can be classified as LT-HSCs in contrast to the data of Randall et al, where the authors did not find any long-term repopulating activity in the CD38- KLS compartment. Moreover, we showed the hierarchical relationships between these two populations. We appreciate the previous findings and recommendations of the reviewer, and have added citations (Randall, et al., 1996: PMID: 8639761 and Tajima et al., 2001; PMID: 11313250) and comment in the discussion section, lines 267-271:

   In contrast to previous studies reporting that only CD38+ HSPC compartment from adult mice contains LT-HSCs (42, 43), in our study we demonstrated using modern surface marker combinations for the isolation of LT-HSCs that while both populations: CD38- and CD38+, can be classified as LT-HSCs, only CD38+ LT-HSCs display characteristics of dormant HSCs (4).’’

   -it is known the generic roles of CD38 in producing cADPR, ADPR, etc and these can induce Ca2+ oscillation in cells; despite that, it was nicely demonstrated here that in mouse HSCs cADPR was the main signalling mediator;

   We thank the reviewer for pointing this out; indeed, it has not been shown before how Ca2+ is regulated in HSCs.

   the roles of cADPR in human CD34+ were demonstrated (Podesta et al., 2023; PMID: 12475890: when CD34+ HSPCs were primed in vitro with cADPR it resulted in enhanced short-term while maintaining long-term (secondary transplant) engraftment in NOD/SCID mice, probably (mechanisms were not determined at that time) inducing cycling/expansion of human CD34+CD38+ progenitors while inhibiting cycling (hence, better long-term maintenance) of CD34+CD38- HSPCs); on this note; the data presented in Fig.4 K and S5 should be eliminated as it adds little to their story and it can be quite confusing when comparing to mouse data unless the authors wish to explore in a more detailed way the human part.

   We appreciate the reviewer’s valuable suggestion. However, we respectfully disagree with their interpretation because we do not believe that the technical aspects of the cited paper (Podesta et al., 2003; PMID: 12475890) are robust enough to support their conclusions. Podesta et al. concluded that *in vivo *and *in vitro *treatment with a high dose of cADPR (25-fold higher than the physiological dose, according to the authors' estimation) stimulates the expansion of HSC and progenitor cells. At the same time, they did not use any surface markers to define populations and studied total mononuclear cord blood cells, so no conclusions can be drawn regarding CD34+ CD38+ and CD34+ CD38- dynamics. Unfortunately, we cannot confirm the reliability of the HSC engraftment data presented by Podesta et al. This is because they did not analyze the chimerism of human cells in peripheral blood and bone marrow for sixteen weeks post-transplantation, which is considered a standard time period for assessing long-term engraftment of human HSCs in the field (Brehm M.A. et al., Blood 2012, Cosgun K.N. et al., Cell Stem Cell 2014, Takagi S. et al. Blood 2012). Instead, they counted only some CD34+ cells at three and eleven weeks after transplantation. Therefore, the role of cADPR in the regulation of human HSC quiescence remained unknown.

   In our original study, we showed that blocking the CD38 ecto-enzymatic activity stimulated both human HSC and mouse HSCs to exit from the G0 phase of the cell cycle. The role of CD38 enzymatic activity can be conservative for mice and humans and needs to be further investigated in future studies on human HSCs. For this reason, we decided to keep Fig. 4K and S6 in the paper.

   -Ca2+ induction in cells can induce c-fos expression (as in an early response gene); in many cell types hence, it was not a surprising finding;

   We agree with the reviewer that it has been shown previously that Ca2+ induction in cells could induce c-fos expression (as an early response gene to stress). However, we have shown for the first time that Ca2+ regulates c-Fos expression in LT-HSCs under steady-state conditions.

   -c-fos was demonstrated to suppress cell cycle entry of dormant hematopoietic stem cells (Okada et al., 1999: PMID: 9920830).

   In the cited publication (Okada et al., 1999: PMID: 9920830) the authors have only analyzed the in vitro proliferation and colony formation of Lin- Sca-1+ cells in the IFNα/β inducible c-Fos overexpression model. This population mainly contains progenitor cells and only 0.004% of dormant LT-HSCs (please find below an estimation of LT-HSC frequency). Therefore, the role of c-Fos in the regulation of dormant HSC cell cycle entry remained unexplored.

   It would be useful to do ChIP-seq to determine to confirm that c-fos regulates p57 expression.

   We have shown that inhibition of c-Fos transcriptional activity inhibits p57 expression (Fig. 6G). ChIP–seq with antibody against c-Fos will answer whether c-Fos directly activates the expression of p57. However, we can only isolate 200-300 CD38+ LT-HSCs from all bones of one mouse. Unfortunately, the ChIP-seq with such an amount of cells is technically very difficult, which explains the absence of publications using ChIP-seq for studying transcription factors in LT-HSCs. We added in the Discussion section that we couldn’t exclude indirect regulation of p57 expression by c-Fos, lines 307-308:” In contrast, although we couldn’t exclude indirect regulation of p57kip2 expression by c-Fos, our data clearly reveal that inhibiting the interaction between c-Fos and DNA in dHSCs reduced protein levels of the cell cycle inhibitor p57kip2 and stimulated cell cycle entry.”

   So overall, many of the findings were already out there and the authors gathered many of the pieces of the puzzle and put them together (and demonstrated) in a nice and well-thought manner. This work does add useful information to the scientific community but unfortunately is not ground-breaking. It may contribute to other fields beyond hematopoiesis where CD38 function may play a role.

   Thank you very much for the positive review of our work. As mentioned by the reviewer, CD38 is expressed by other normal (lymphocytes, Kupffer cells (Tarrago M.G. et al., Cell Metabolism 2018)) and cancer cells, e.g. hematological malignancies, lung cancer, prostate cancer (Hogan K.A. et al. Frontiers in Immunology, 2019),) but has not been studied in the context of quiescence regulation. Currently, anti-CD38 monoclonal antibodies are used to treat malignancies (Daratumumab) by mediating cytotoxicity (Lokhorst H.M et al., N. Engl. J. Med, 2015). However, the inhibition of CD38 enzymatic activity has not been used broadly. Therefore, our study can be groundbreaking and open new directions in anti-cancer therapy.

   Reviewer #2 (Significance (Required)):

   In this manuscript, the authors investigated the potential roles of CD38 (mainly) in mouse HSCs quiescent; the authors dissected the potential molecular mechanism by which this occurred, and it was via CD38/cADPR/Ca2+/cFos/p57Kip2. The authors used a combination of transplantation assays to test the importance of CD38 in vivo, followed by a series of simple in vitro experiments (mainly using pharmacological means) to dissect the molecular mechanisms. The manuscript is well-written/explained and the data presented is solid. There are no major issues in terms of reproducibility and clarity in this work.

   We would like to thank the reviewer again for the detailed positive feedback.

   Read the original source
2. 

   Review Commons

   Apr 14, 2023

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   Although the experiments were well done and supported their testing hypothesis, but the overall novelty of the whole work is not that strong and this is because:

   • the use of CD38 to identify/select and to test mouse LT-HSCs' function in vivo (although not commonly used nowadays) was demonstrated a few times more than 20 years ago (Randall, et al., 1996: PMID: 8639761 and Tajima et al., 2001; PMID: 11313250); in fact, the authors didn't even reference/acknowledge these papers which they should have done so; hence, most of the results in Fig.2 were already known (despite this current work gave a more detailed/better analysis);
   • it is …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   Although the experiments were well done and supported their testing hypothesis, but the overall novelty of the whole work is not that strong and this is because:

   • the use of CD38 to identify/select and to test mouse LT-HSCs' function in vivo (although not commonly used nowadays) was demonstrated a few times more than 20 years ago (Randall, et al., 1996: PMID: 8639761 and Tajima et al., 2001; PMID: 11313250); in fact, the authors didn't even reference/acknowledge these papers which they should have done so; hence, most of the results in Fig.2 were already known (despite this current work gave a more detailed/better analysis);
   • it is known the generic roles of CD38 in producing cADPR, ADPR, etc and these can induce Ca2+ oscillation in cells; despite that, it was nicely demonstrated here that in mouse HSCs cADPR was the main signalling mediator;
   • the roles of cADPR in human CD34+ were demonstrated (Podesta et al., 2023; PMID: 12475890: when CD34+ HSPCs were primed in vitro with cADPR it resulted in enhanced short-term while maintaining long-term (secondary transplant) engraftment in NOD/SCID mice, probably (mechanisms were not determined at that time) inducing cycling/expansion of human CD34+CD38+ progenitors while inhibiting cycling (hence, better long-term maintenance) of CD34+CD38- HSPCs); on this note; the data presented in Fig.4 K and S5 should be eliminated as it adds little to their story and it can be quite confusing when comparing to mouse data unless the authors wish to explore in a more detailed way the human part.
   • Ca2+ induction in cells can induce c-fos expression (as in an early response gene); in many cell types hence, it was not a surprising finding;
   • c-fos was demonstrated to suppress cell cycle entry of dormant hematopoietic stem cells (Okada et al., 1999: PMID: 9920830).

   It would be useful to do ChIP-seq to determine to confirm that c-fos regulates p57 expression.

   So overall, many of the findings were already out there and the authors gathered many of the pieces of the puzzle and put them together (and demonstrated) in a nice and well-thought manner. This work does add useful information to the scientific community but unfortunately is not ground-breaking. It may contribute to other fields beyond hematopoiesis where CD38 function may play a role.

   Significance

   In this manuscript, the authors investigated the potential roles of CD38 (mainly) in mouse HSCs quiescent; the authors dissected the potential molecular mechanism by which this occurred, and it was via CD38/cADPR/Ca2+/cFos/p57Kip2. The authors used a combination of transplantation assays to test the importance of CD38 in vivo, followed by a series of simple in vitro experiments (mainly using pharmacological means) to dissect the molecular mechanisms. The manuscript is well-written/explained and the data presented is solid. There are no major issues in terms of reproducibility and clarity in this work.

   Read the original source
3. 

   Review Commons

   Apr 14, 2023

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   In their paper entitled "CD38 promotes hematopoietic stem cell dormancy via c-Fos", Ibneeva et al., present a set of data predominantly from mouse HSCs where they explore the cell cycle kinetics and self-renewal capacity of LT-HSCs expressing (or not) CD38. They perform a series of sophisticated in vitro and in vivo experiments, including transplantations and single cell cultures and arrive at the conclusion that CD38 can fractionate LT-HSCs that are more deeply quiescent. Overall, it is an interesting question and would be of interest to experimental hematologists. That said, I had a number of issues that concerned me throughout …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   In their paper entitled "CD38 promotes hematopoietic stem cell dormancy via c-Fos", Ibneeva et al., present a set of data predominantly from mouse HSCs where they explore the cell cycle kinetics and self-renewal capacity of LT-HSCs expressing (or not) CD38. They perform a series of sophisticated in vitro and in vivo experiments, including transplantations and single cell cultures and arrive at the conclusion that CD38 can fractionate LT-HSCs that are more deeply quiescent. Overall, it is an interesting question and would be of interest to experimental hematologists. That said, I had a number of issues that concerned me throughout the manuscript with regard to the robustness of the conclusions around CD38 and I have tried to detail these below.

   Major concerns:

   1. Novelty - It was unclear what the relationship of this CD38+ fraction had with other "segregators" of LT-HSCs - e.g., how does it compare with the Sca1 fractionation of Wilson et al, Cell Stem Cell 2015 or Gprc5c of Cabezas-Wallschied Cell 2017? Even if CD38 fractionated LT-HSCs, it was unclear what it would give beyond these two molecules (especially re: Sca-1 which is also a cell surface marker)
   2. Claims of CD38+ superiority in transplantation - I was surprised with the claim of CD38 negative cells being a less functional HSC when they are clearly still very strong in secondary transplantation assays. Both 38+ and 38- cells strongly repopulate secondary animals and only 5 mice were shown in the Figure. The legend suggests another experiment was undertaken, but these data are not presented. Did they substantially differ in their chimerism in primary and secondary animals? Was the magnitude of difference between the two fractions similar in both experiments? Is there a reason that the data could not be plotted on the same graph? Also, the typical experiment to establish a quantitative difference in HSC production would be a limiting dilution analysis with a much larger number of recipient animals - without such data it is difficult to ascertain how different the two fractions really are.

   Furthermore the claim that CD38- HSCs do not ever produce CD38+ cells is a bit premature with so few mice and confusingly presented data (e.g., Fig 2I is 5 pooled mice in a single histogram plot - were these concatenated flow files? If so, how were they normalised? Did the other experiment look the same? And were all CD38+ HSCs capable of giving rise to both CD38+ and CD38- cells or was it a subfraction of mice/samples?).

   1. Cell Cycle status differences and grades of quiescence - Ki67 and DAPI are really quite tricky for discerning G0 versus G1 and no flow cytometry plots are provided for the reader to assess how this has been done. Could another technique (e.g., Hoechst/Pyronin) be used to confirm the results? Perhaps more concerning is the variability of the assay in the authors own hands. If I am interpreting things correctly, the plots in 3G, 3H and 3I in the platelet depletion, pIpC and 5FU experiments are >10% higher in the CD38- control arm than the data in 3A which make me worried about the robustness of the cell cycle assay to distinguish G0 from G1.

   Minor points:

   Figure 3I was really confusing - it says it is the gating strategy for GFP retaining LT-HSCs, but only shows GFP versus cKit

   Figure 4B suggests that only 40% of CD38+ cells divide in the first 3 days - are there survival differences or are the cells sat there as single cells? It would be important to carry these further to see if cells eventually divide.

   Significance

   I believe the study will be of interest to specialist readers in the HSC field, especially those working on quiescence and G0 exit. At present, I think the conclusion of a true subfractionation is a bit premature, but there are pieces of data that do look exciting and warrant further investigation. It was a little unclear how this would advance beyond Sca-1 or Gprc5c fractionation for finding more primitive HSCs, but having cleaner markers is always a useful advance for the field.

   Regarding my own expertise - I have spent ~20 years in the field undertaking single cell assays of normal and malignant mouse and human HSCs, including many of the core functional assays described in this paper and consider myself very familiar with the topic area.

   Read the original source
4. 

   Version published to 10.1101/2023.02.08.527614v1 on bioRxiv

   Feb 8, 2023