In our journal club this week, we went a little more experimental and selected a BioRxiv paper yet to be published - it can be found at https://www.biorxiv.org/content/early/2017/07/13/163402 and currently bears the title "Heterogeneous Responses of Hematopoietic Stem Cells to Inflammatory Stimuli are Altered with Age". We were inspired by the eLife Labs article on PREReview to give this a go and hopefully the authors and wider biological community benefit from this interaction.As mentioned above, this was our first attempt at discussing a preprint paper from BioRxiv so we had a few opening questions and general discussion about the concept between the 15 or so members of the journal club. Our groups (Laurenti and Kent) are based at the WT / MRC Cambridge Stem Cell Institute in Cambridge, UK (https://www.stemcells.cam.ac.uk/) and the opinions and comments found below are a summary of the group discussion rather than the opinion of any single researcher partaking in the journal club.First question: Was reading this any different than a normal paper?Some people had worried that a pre-publication paper would suffer from major gaps in organisation or presentation and this was not the case here. Overall, there was not a substantial difference from published papers and people agreed that the structure of the paper was good and the content relatively easy to follow. That said, one cynical comment was raised, querying whether the peer review process prohibited authors from telling their version of the story (e.g., one interpretation), that might actually diverge from the interpretation of the broader field (i.e., peer review makes the article "less readable" but also "more accurate"). Second question: Would you continue to explore pre-print servers for new content?Resounding "yes" from majority of researchers (many of whom had not heard of preprint servers prior to the discussion)---We then entered the discussion phase and came up with the following "review" based on the discussion:In their manuscript, "Heterogeneous Responses of Hematopoietic Stem Cells to Inflammatory Stimuli are Altered with Age", Mann et al., identify and transcriptionally profile at the single cell level a novel subset of aged cells primed to respond to inflammatory stimuli. Overall, the manuscript read very well, had a logical progression and was easily considered to be "publication standard" with respect to these aspects (i.e., it was in better shape from an editing perspective than some manuscripts!). The data were presented nicely and generally supported the statements in the abstract, but several areas of concern were identified through our discussion and a number of questions arose that might be considered by the authors: Concerns:1) Unknown LT-HSC frequency of cells isolated from aged animals: LSK SLAM is a well-described phenotype for functional HSCs in young mice (~35-50% functional HSCs), but it remains unknown what the precise frequency of functional HSCs in that phenotype in aged mice would be. Perhaps the best evidence comes from the de Haan group (http://jem.rupress.org/content/208/13/2691.long) where they studied LSK SLAM 34- EPCR+ cells and showed that there was at least 2-fold less functional activity on a per HSC basis and this arguably is a much more stringent sorting criteria. Therefore, the cells being isolated as "aged LT-HSCs" in this current manuscript are likely less than 10% functional HSCs, making interpretation of all downstream data difficult to ascribe to "LT-HSCs". Moreover, the frequency of phenotypic LT-HSCs is highly variable (from 100 to 4300 per million cells), making the number of input mice (n=4) concerning, since they might represent a large proportion of one single mouse (e.g., 100-fold more phenotypic HSCs). Overall, our discussion group agreed that the data presented re: inflammatory signatures were still interesting - it was just that the claims that it was a functional LT-HSC specific (or myeloid-biased LT-HSC specific) signature were less robust.2) Proportional, not absolute, changes in chimerism: Figure 1 appears to display the proportion of cells that were T, B, gran, or myeloid without showing the absolute levels of chimerism. It therefore precludes any reviewer from appropriately determining whether a "myeloid bias"really exists (e.g., an absolute increase in myeloid cell production) or whether it is an absence of lymphoid cell production (lymphoid deficient). Without differences in the absolute myeloid cell number, it changes the interpretation of the data since both cell types would produce the same number of cells and young myeloid progenitors would be present (but not responding) to the stimulus.3) in vitro 2h stimulus vs. in vivo effect: The in vivo effect of a single dose treatment seems to be temporary (Figure 1 B-E), whereas the in vitro effect (from a 2 hr stimulation) is sufficient to change the fate of the aged LT-HSCs that get transplanted. It remains unclear how the progeny of these LT-HSCs continue to self-renew in a way which creates progeny HSC with the same properties for sustaining long term hematopoiesis. A few items came up in discussion: 1) Is there selective killing of balanced/lymphoid biased HSCs? 2)Are secondary transplantations "myeloid-biased" as well (e.g., how strong is the continuation of the pattern?)Additional questions arising in discussion:1) Could the authors speculate on the origin of the HSC expansion? It has been described by several groups that the HSC subtype balance shifts with age but nobody has clearly addressed whether the lymphoid deficient HSCs that accumulate with age are "old" balanced HSCs that have lost lymphoid potential or an accumulation of "lymphoid deficient (or alpha / My-biased)" HSCs. 2) Would young HSCs from malignant mouse models (or patient samples) also bear the aged/inflammatory signature?minor items:the mega-biased HSCs were first described by the Jacobsen and Nakauchi groups, not the de Haan and Graf groups as mentioned in discussion.
