1. Evaluation Summary:

    This paper is of broad interest to those in the field of immunometabolism, a field which has largely used the mouse as an experimental system. The descriptive work is the first of its kind to demonstrate important aspects of biological variability and hidden aspects of mitochondrial function in human immune cells.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

    Read the original source
    Was this evaluation helpful?
  2. Reviewer #1 (Public Review):

    In this paper, the authors attempted to fill some major knowledge gaps with regards to mitochondrial function in circulating immune cells in humans. The majority of the immunometabolism literature focuses on the mouse as a model system, so this contribution is a welcome addition to the field. Major strengths of the results of this contribution include the unmasking of cell type specific mitochondrial parameters that are hidden when measuring PBMCs in aggregate. In addition, the contributions of platelet contamination serve as a warning to prospective investigators conducting similar studies. The study methods are quite extensive and well thought out. Some minor weaknesses to take into consideration are interpretations of some of the data. Oftentimes distinctions between clear statistically significant findings and trends do not seem to be made, leaving the reader to scour the text and figures to make sure something was not missed. There are a lot of data to take in and understand, certainly there are probably some figures which could be excluded. Overall, these are fixable things, and the authors have done a fine job in achieving their goal to fill the knowledge gap about mitochondrial parameters in immune cell subtypes in humans. This contribution is the first of its kind in this field and will serve as a reference for those conducting immunometabolism studies on human immune cell populations.

    Read the original source
    Was this evaluation helpful?
  3. Reviewer #2 (Public Review):

    In this manuscript, Rausser et al., have assessed age-, sex- and time-driven differences in mitochondrial phenotypes with human PBMCs and their subsets.

    A major finding of this manuscript was that metabolic analysis of bulk PBMC masks a lot of differences that occur due to other factors, especially changes in leukocyte composition. From my perspective in immuno-metabolism, it is already very clear that i) different leukocytes have very different metabolic profiles and ii) the metabolic profile of a defined leukocyte subset can also change due to age-, infection-related changes in both composition and cell activation. Accordingly, analysis of bulk PBMCs will have limited value for defining biological mechanisms due to the complexity of cell subsets within that tissue sample. Perhaps researchers outside immunology are not aware of the complexity inherent in analysis of PBMCs, so there may be value in highlighting these differences, but that finding was of limited novelty.

    The broader observations in the manuscript are often consistent with what is known in the field of immuno-metabolism. However, the mitotype approach is a niche method of assessment of mitochondrial activity. It is unclear how the measures of mitochondrial activity used overlap with other more common measures of immune cell metabolism such as metabolic dyes, mitochondrial imaging, MetFlow, or Seahorse analyses. I remain unclear of how MHI relates to other more common measures of immune cell metabolism and mitochondrial capacity/health- for example, a correlative analysis of the MHI metric and Seahorse assays to validate whether changes in MHI track with a functional measure of mitochondrial flux across ages, sexes or individuals.

    One uncertainty I have is that the group numbers in each age/sex group seemed small (n=2-3) and there are a large number of comparisons. The authors did identify some robust changes in leukocyte composition that are commonly found with ageing, such as a decline in CD8 T cell % and naïve CD8 T cells, in particular, so this group size may be enough to identify robust differences. I do not have the expertise to assess the robustness of the statistical tests performed- other reviewers may be able to comment better there and I defer to those- but I wanted to note my uncertainty on this point.

    One major variable that has not been assessed is variability in processing of bloods. The authors highlight that there is marked variation in samples collected from the same individual week to week but also highlight that platelet contamination can have a major impact on the readouts and that the range of variation in the age/sex cohort is similar to an individual's variation. So does the variability in the mitochondrial assays reflect variation in processing? It would be instructive to sample from the same individual sequentially over 5 days or even the same individual on the same day, process separately, and then perform these assays.

    More broadly, I suggest that, rather than the value of this manuscript being in its biological findings, its value may be more validation of the metric that the authors have developed- the MHI measure. Re-framing to stress this validation with limitations and considerations, rather than the biological findings, could be helpful.

    While this publication has a distinct focus on metabolic phenotypes, the authors would need to include a discussion of their work with regard to a recent publication that has looked at sexual dimorphism and age-related changes in PBMC composition: https://www.nature.com/articles/s41467-020-14396-9

    Fig3 b-i: I'm unclear why CD8 memory phenotype cells are missing from g-i, although they are present in the graphs b-e.

    Supp Fig 2: On the CD4 vs CD8 gate, the gates are mislabelled as CCD4 and CCD8

    I'm unclear on how PBMCs were processed for all of the workflows. In particular, it is unclear whether the cells were frozen and then defrosted at any point. Some PBMC subsets will withstand this process better and PBMCs may require resting time after thawing to return to normal metabolic activity.

    Read the original source
    Was this evaluation helpful?
  4. Reviewer #3 (Public Review):

    The authors characterize the mitochondrial profile of immune cell subsets vs. bulk PBMCs in a small cohort. They successfully point out the heterogeneity between the subjects with regards to the composition of PBMCs. The composition of innate and adaptive immune cell subsets indeed affect the mitochondrial parameters obtained from the bulk PBMCs, therefore, the cell subsets should be handled separately for such analyses. The leads obtained from this study would be useful for further research.

    Read the original source
    Was this evaluation helpful?