Longitudinal analysis of invariant natural killer T cell activation reveals a cMAF-associated transcriptional state of NKT10 cells
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
The manuscript by Kane et al. described transcriptional profiles of various subsets of activated iNKT cells using longitudinal scRNA-Seq analysis. The finding that IL-10 producing iNKT cells have a cMAF-associated gene signature similar to Tr1 cells is novel. Overall, the data is well presented, however, functional consequences of some findings require further investigation.
(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.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Immunology and Inflammation (eLife)
Abstract
Innate T cells, including CD1d-restricted invariant natural killer T (iNKT) cells, are characterized by their rapid activation in response to non-peptide antigens, such as lipids. While the transcriptional profiles of naive, effector, and memory adaptive T cells have been well studied, less is known about the transcriptional regulation of different iNKT cell activation states. Here, using single-cell RNA-sequencing, we performed longitudinal profiling of activated murine iNKT cells, generating a transcriptomic atlas of iNKT cell activation states. We found that transcriptional signatures of activation are highly conserved among heterogeneous iNKT cell populations, including NKT1, NKT2, and NKT17 subsets, and human iNKT cells. Strikingly, we found that regulatory iNKT cells, such as adipose iNKT cells, undergo blunted activation and display constitutive enrichment of memory-like cMAF + and KLRG1 + populations. Moreover, we identify a conserved cMAF-associated transcriptional network among NKT10 cells, providing novel insights into the biology of regulatory and antigen-experienced iNKT cells.
Article activity feed
-
-
Author Response
Reviewer #2 (Public Review):
The role of cMAF in the formation of iNKT10 is only suggested ny the transcriptional signatures analyzed here. There is no direct evidence that cMAF is indeed needed to generate iNKT10. This should be investigated.
We thank the reviewer for their comments on the link between IL-10, NKT10 cells, and cMAF. We agree that our study provides evidence that cMAF is a promising candidate regulator of IL-10 production by iNKT cells, and we attempted to address this using gene-specific knockout mice. Since mice lacking expression of cMAF exhibit post-natal mortality and severe developmental defects13 we attempted to breed Maffl/flCd4-cre mice, which have previously been used to study the role of cMAF in T cell function14. However, we were not able to successfully breed enough of these mice to …
Author Response
Reviewer #2 (Public Review):
The role of cMAF in the formation of iNKT10 is only suggested ny the transcriptional signatures analyzed here. There is no direct evidence that cMAF is indeed needed to generate iNKT10. This should be investigated.
We thank the reviewer for their comments on the link between IL-10, NKT10 cells, and cMAF. We agree that our study provides evidence that cMAF is a promising candidate regulator of IL-10 production by iNKT cells, and we attempted to address this using gene-specific knockout mice. Since mice lacking expression of cMAF exhibit post-natal mortality and severe developmental defects13 we attempted to breed Maffl/flCd4-cre mice, which have previously been used to study the role of cMAF in T cell function14. However, we were not able to successfully breed enough of these mice to assay whether or not cMAF is required for the production of IL-10 by iNKT cells. Therefore, our study can only suggest that cMAF is a promising candidate regulator of NKT10 cells based on our transcriptomic data and flow cytomery data showing that production of IL-10 is associated with expression of cMAF. However, we present further correlative or indirect evidence to this effect. It has previously been demonstrated that restimulation of activated iNKT cells at 72 hours post-⍺GalCer results in increased production of IL-10 compared to the stimulation of iNKT cells at steady state15. We found that the frequency of splenic cMAF+ iNKT cells was greatly increased at 72 hours post-⍺GalCer compared to steady state (Figure S3B, Figure S3D) and this increase in expression of cMAF correlated with increased production of IL-10 (Figure S3E-S3F). Therefore, we believe that cMAF is a promising candidate for future work examining the functional landscape of NKT10 cells and we anticipate that our study will be a useful transcriptomic reference for such studies.
The Kronenberg group recently published a similar analysis, using RNAseq and ATACseq. Although I don't believe the cMAF signature was highlighted at the time, one could argue that this previously published study dampens the originality of this manuscript. Although this study (Murray et al.) is clearly acknowledged, the similarities and differences in both the methodology and findings should be clearly discussed.
As the reviewer stated, the excellent study by Murray et al. (2021) did not identify or highlight a population of cMAF+ iNKT cells expressing a regulatory gene signature, as presented in our study, and as the reviewer mentions, we cite and discuss the Murray et al. study in our manuscript. We believe that both studies together provide a comprehensive transcriptomic analysis of iNKT cells after activation, and that ours provides unique insight not found in Murray et al. Our study uses scRNA-Seq rather than bulk RNA-Seq or bulk ATAC-Seq methods, enabling us to study transcriptomic characteristics of activation among heterogeneous iNKT cell subsets without needing to sort pre-identified iNKT cell populations or subsets. It is the use of unbiased scRNA-Seq that allowed us to identify cMAF+ iNKT cells, since this population has not been previously described in the literature. Notably, we also sequenced the largest number of iNKT cells to date, 48,813 cells, to the best of our knowledge, which provides deeper insight. We also performed transcriptomic characterization of activated iNKT cells at different stages of activation to those characterized by Murray et al. Importantly, we profiled the phenotype of iNKT cells at 4 hours post-⍺GalCer and 72 hours post-⍺GalCer, when iNKT cells engage in a rapid cytokine production or undergo proliferation and expansion. This revealed several novel transcriptional insights including rapid metabolic gene reprogramming that occurs twice during this activation timeline. By contrast, Murray et al. focused on analysis of iNKT cells at steady state and 6 days post-⍺GalCer. Finally, we performed transcriptional characterization of adipose iNKT cells in our study, which are known to represent an unusual regulatory population of iNKT cells at steady state16, whereas the study by Murray et al. (2021) did not study adipose iNKT cells. Therefore, we propose that our study complements the excellent work performed by Murray et al. (2021) but provides novel insight in terms of focus, discovery, and scope.
The authors should clearly describe the genes that were used to define iNKT1/2/17 identity in their study. This is important in order to track that identity over time following activation, at it is well known that the expression of some of the markers typically used change following activation. This would bring clarity to the manuscript.
We agree with the reviewer and we had originally removed two clarifying figures for iNKT cell subset identification due to space, but now we have included these two clarifying supplemental figures (Figure S4, Figure S5) to illustrate how we identified NKT1, NKT2, and NKT17 cell subsets in our scRNA-Seq data. We have also added further details to the Methods section (please see the “Downstream scRNA-Seq data analysis” section) and we have changed the title of the activated iNKT cell data in Figure 2A-2D and Figure 4C from “4 hours post-⍺GalCer” to “Activated (4 hours post-⍺GalCer)” to reflect our subset identification protocol as accurately as possible (please see below).
Steady state and activated splenic NKT1, NKT2 and NKT17 cell subset identification was performed as follows: We identified spacial separation and graph-based clustering of five main populations of cells at steady state (Figure S4A). We then used the expression of the published marker genes Tbx21, Zbtb16, Rorc and Mki67 to identify NKT1, NKT2, NKT17 and Cycling cells (Figure S4B). We identified spacial separation and graph-based clustering of three main populations at 4 hours post-αGalCer (Figure S4D). However, we found that Tbx21 and Zbtb16 expression was increased across multiple clusters and did not effectively demarcate NKT1 and NKT2 cells at the RNA level (Figure 2B), and so we instead used the flagship cytokines Ifng, Il4, Il13 and Il17a and Il17f to demarcate NKT1, NKT2 and NKT17 cells. We then combined the identified NKT1, NKT2 and NKT17 cell populations from steady state and 4 hours post-⍺GalCer together (i.e. cells from the same subset at the two different time points were combined together) and performed reclustering of the cells within each subset (Figure S5). It has previously been shown that there can be differences in the activation kinetic of different splenic iNKT cell subsets, for example NKT1 versus NKT2 cells, which may be in part due to physical localization, for example in the red pulp versus the white pulp of the spleen (see Lee et al. 2015)17. We observed a similar phenotype for NKT2 cells in our data, whereby a proportion of NKT2 cells at 4 hour post-⍺GalCer clustered with NKT2 cells from mice that received no ⍺GalCer (Figure S5A). To prevent differences in activation kinetic from biasing our analysis of transcriptional signatures of iNKT cell subset activation, we performed low-level graph-based reclustering within each iNKT cell subset to accurately segregate activated and steady state iNKT cells (Figure S5B). We validated our reclustering using the expression of activation markers and flagship cytokines (Figure S5C-S5D). Finally, these reclustered subset data were recombined and renormalized to generate the final analysis as shown in Figure 2 of the manuscript.
-
Evaluation Summary:
The manuscript by Kane et al. described transcriptional profiles of various subsets of activated iNKT cells using longitudinal scRNA-Seq analysis. The finding that IL-10 producing iNKT cells have a cMAF-associated gene signature similar to Tr1 cells is novel. Overall, the data is well presented, however, functional consequences of some findings require further investigation.
(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.)
-
Reviewer #1 (Public Review):
In this study, the authors performed scRNA-seq analysis of iNKT cells from spleen and adipose tissue at steady state and after short-term, long-term, and repetitive ⍺-GalCer stimulation in vivo. They found iNKT cells undergo rapid and extensive transcriptional remodeling during activation. By reanalyzing published scRNA-Seq data of human iNKT cells, the authors found transcriptional signatures of iNKT cell activation are conserved across species. in addition, they showed, adipose NKT10 cells, had blunted response to ⍺-GalCer and expressed markers associated with Tr1 cells. Furthermore, they demonstrated two memory-like iNKT cell populations, expressing immunoregulatory cytokines and maf (cMAF+ iNKT cells) or cytotoxic markers and klrg1 (KLRG1+ iNKT), were constitutively present in adipose tissue, and were …
Reviewer #1 (Public Review):
In this study, the authors performed scRNA-seq analysis of iNKT cells from spleen and adipose tissue at steady state and after short-term, long-term, and repetitive ⍺-GalCer stimulation in vivo. They found iNKT cells undergo rapid and extensive transcriptional remodeling during activation. By reanalyzing published scRNA-Seq data of human iNKT cells, the authors found transcriptional signatures of iNKT cell activation are conserved across species. in addition, they showed, adipose NKT10 cells, had blunted response to ⍺-GalCer and expressed markers associated with Tr1 cells. Furthermore, they demonstrated two memory-like iNKT cell populations, expressing immunoregulatory cytokines and maf (cMAF+ iNKT cells) or cytotoxic markers and klrg1 (KLRG1+ iNKT), were constitutively present in adipose tissue, and were induced in the spleen following ⍺-GalCer challenge. Overall, this study provides novel insights into the transcriptional program of activated iNKT cells and the phenotype of regulatory iNKT cells. The bioinformatic aspect of this study is well performed, but the immunology and T cell biology aspects could be strengthened.
-
Reviewer #2 (Public Review):
This is an interesting study that provides a "transcriptomic atlas of iNKT cell activation states".
The role of cMAF in the formation of iNKT10 is only suggested ny the transcriptional signatures analyzed here. There is no direct evidence that cMAF is indeed needed to generate iNKT10. This should be investigated.
The Kronenberg group recently published a similar analysis, using RNAseq and ATACseq. Although I don't believe the cMAF signature was highlighted at the time, one could argue that this previously published study dampens the originality of this manuscript. Although this study (Murray et al.) is clearly acknowledged, the similarities and differences in both the methodology and findings should be clearly discussed.
The authors should clearly describe the genes that were used to define iNKT1/2/17 …
Reviewer #2 (Public Review):
This is an interesting study that provides a "transcriptomic atlas of iNKT cell activation states".
The role of cMAF in the formation of iNKT10 is only suggested ny the transcriptional signatures analyzed here. There is no direct evidence that cMAF is indeed needed to generate iNKT10. This should be investigated.
The Kronenberg group recently published a similar analysis, using RNAseq and ATACseq. Although I don't believe the cMAF signature was highlighted at the time, one could argue that this previously published study dampens the originality of this manuscript. Although this study (Murray et al.) is clearly acknowledged, the similarities and differences in both the methodology and findings should be clearly discussed.
The authors should clearly describe the genes that were used to define iNKT1/2/17 identity in their study. This is important in order to track that identity over time following activation, at it is well known that the expression of some of the markers typically used change following activation. This would bring clarity to the manuscript.
-
Reviewer #3 (Public Review):
This paper is of interest to scientists within the field of iNKT cells. The authors conducted scRNA-seq to longitudinally profile activated iNKT cells and generated a transcriptomic atlas of iNKT cells at the activation states. The study suggests that transcriptional signatures of activation are highly conserved among heterogeneous iNKT cell populations and that the adipose iNKT cells undergo blunted activation and display constitutive enrichment of memory like population, plus identifying a conserved cMAF- associated network in NKT10 cells. This study provides some new insights into the NKT biology.
-
