Explore other scientists feedback or go to the Molecular Data page and select a cell label to provide your feedback.
I agree this cluster needs to be split between the left side, which is CD4 Naive (SELL, CCR7+) and CD4 Effector (right side) that also expresses some Trm markers (ITGA1, ITGAE). Would suggest subclustering the right side with the Th17 to help with pulling out the various effector populations.
There is a mixture of naive T cells (left cluster) expressing CCR7, LSEL, and not naive ( cluster on the right lacks these two canonical markers of naive lymphocytes) expressing cytokines (IFNG, TNF, tnfsf13B, tnfrsf25, il2, cd40lg etc). This will fit with an effector phenotype.
This cluster contains a mix of CD4 subsets including Naive. e.g. Cells expressing IL12RB2, TBX21, CCL5, IFNG, TNF suggest some Th1. In our data CD45-RO negative and CD45-RA positive CD4 T cells at the CITE-seq level can be distinguished at the RNA level by a lack of expression of genes such as KLRB1, CCR6, CCL5. These cells are also SELL, LEF1, KLF2, CCR7, TCF7 high.
This needs to be subclustered - the cluster near the CD4 Th17 is definitely not CD8. The other portion of the cluster (where the majority is) likely is CD8 TRM, as it expresses ITGAE, ITGA1, SPRY1. However, to be sure, the CD8 clusters should all be grouped together for reannotation (naive/circulating/effector/memory).
Appears to be a subset expressing CD4 as well
Based on the gene expression, these cells might not be CD8 IEL, based on the literature review on those. The expression of ZBTB7, RUNX3D, HNF4G, etc. are too low, although their variants are highly expressed. Li, Lanasa and Park, Pathways and mechanisms of CD4+CD8αα+ intraepithelial T cell development, CellPress, Trends in Immunology, 2024
I agree there are CD4 Th17 in this cluster, as evidenced by IL17A, CCR6, and CCL20 expression.
Agreed with the cell annotation (IL17A, IL23R)
Agree IL23R, IL17A, RORC, IL4I1, KIT
I would expect there to be some plasmablasts in the plasma cell cluster. I think they can be subclustered out of the general "plasma cells", and will be marked by higher expression of MZB1 and XBP1, as well as some CD27. Could also be in what is being called "Plasma IGM," it's a bit hard to tell from this embedding.
I agree this is Plasma IgA, evidenced by IGHA1 and IGHA2 expression. Hard to tell from this embedding, but could consider separating into IgA1 and IgA2 subpopulations (https://www.nature.com/articles/s41586-024-07571-1, Supp Fig 1).
can be subclustered to identify naïve CD8⁺ T cells in the left region (above CD4 naïve cells); these show high CCR7 and SELL expression.
Split into CD8+ naive (CD8A, CCR7, SELL) and the other cluster (negative for CCR7 and SELL) which appears to be jejunal (and accordingly expressing CCR9).
Also these "CD4+ TH1" cells are negative for CD4, but positive for CD8A/B. When comparing to the CD4TH17 cluster, CD4 is actually among the top downregulated genes in the "TH1" cluster. Merge with CD8+ T cells?
I don't think these are B Marginal Zone. These look more like B naive cells, as they express SELL, TCL1A, IGHD, FCER2. You would also expect B naive cells to cluster with the B memory, which this cluster has.
I think this is likely the naive B cell cluster. FCER2+, CD27-, Foxo1+, CD24hi, TCL1A+, IL4R+. https://www.science.org/doi/10.1126/sciimmunol.abe6291
I think this cluster needs to be sub-clustered because it also contains NK_CD16 cells (the top part of the upper left cluster that is under the cluster annotated as "ILC NK Cells." It doesn't express CD8A or CD8B, and does express FCGR3A, GNLY, GZMB, and NKG7.
can be divided to identify CD8⁺ GZMK⁺ effector T-cells for those highly expressing GZMK.
Leiden Lineage L2 9_2 cluster is CD8 FGFBP2 (FGFBP2, FCGR3A, ...) (Corridoni, D and Garrido-Trigo A)
These do not even look like CD4 cells, as express CD8A and CD8B. Likely some sort of CD8 Trm subset, as express ITGA1, ITGAE, SPRY1. Would suggest reclustering with just the other CD8 circulating/effector/memory subsets, as there is likely heterogeneity here that is hard to tease out without additional leiden clusters.
Also these "CD4+ TH1" cells are negative for CD4, but positive for CD8A/B. When comparing to the CD4TH17 cluster, CD4 is actually among the top downregulated genes in the "TH1" cluster. Merge with CD8+ T cells?
Expressing high levels of CD8A and CD8B and not expressing CD4, likely CD8 T cells instead
I think these are also B memory, rather than GC B Dark Zone. They don't express MK67 or AICDA unlike what is currently annotated as "B cycling".
These cells are CD83high, MYChigh, CXCR5high, but lack AICDA and are MKI67 low-ish. --> I would judget them as light zone GC B cells and think that the "cycling B cell" cluster in fact is the real dark zone GC B cell cluster. Check out details e.g. here. https://www.sciencedirect.com/science/article/pii/S1074761321003010?via%3Dihub
I do think this cluster has Tfh (CXCR5, BCL6, PDCD1, CD40LG), but agree that it needs to further subclustered. For example, I think you could split into at least "mature Tfh" that has higher expression of the Tfh markers, but there is definitely a more "naive" portion with lower expression (https://www.nature.com/articles/s41586-024-07571-1, Supp Figure 1).
This is a heterogeneous cluster, containing at least 3 different subsets. Only 1 of them has the canonical CXCL13, IL21 Tfh. A second subset not expressing these genes but with expression of CXCR5 and CCR6. The third subset has no markers of Tfh, and expresses several heat shock gene markers and CCL20. I would re-annotate this last cluster (number 3_3 in Leiden L2) as Heat shock T cells.
T follicular helper cells (CXCL13 and MAGEH1) (Garrido-Trigo, A)
Agree BCL6, TOX2, CXCR5, CD200
CXCR5, CXCL13, MME indicates TFH
I don't think these are GC B Light Zone. They don't express SUGCT/EBI3 and there are clusters with higher expression of BCL6 (see B naive feedback). I think these are more likely B memory (high in CD27, TNFRSF13B).
I agree these look like B_memory cells. Express CD27 and TNFRSF13B.
For the small cluster closer to CD4 Th17, could this represent IL26-high Th17 cells? We previously observed a PD1⁺ Th17 population with high IL26 expression. Also, there is a subpopulation highly expressing RBMS3, which has been shown to have a suppressive role in the tumor microenvironment.
Contains several subsets, including a GMZA, IL23R and IL12RB2 expressing Th1/Th17-like population
I think these are dark zone B cells, also known as GC I. Evidenced by high cycling markers (MKI67), but also BCL6, AICDA. See Figure 3: https://www.cell.com/trends/immunology/fulltext/S1471-4906(22)00003-5
I think these are dark zone germinal center B cells. Mostly based on AICDA expression. But also MKI67, EZH2, CD9, PCNA, CDK1 https://rupress.org/jem/article/217/10/e20200483/151908/Single-cell-analysis-of-germinal-center-B-cells
I think the part of this cluster with what is currently "B Cycling" is actually light zone B cells (also known as GC II). Evidenced by CXCR5, EBI3, BTK, BLNK, BCL6, SUGCT, CD27 expression. Also makes more sense for light/dark zone germinal center B cells to form a cluster together, and then B naive to actually be part of the cluster with B memory. See Figure 3 for additional marker genes: https://www.cell.com/trends/immunology/fulltext/S1471-4906(22)00003-5
I think there is a couple of problems here with this cluster. I don´t think they are Naive B cells (check out e.g. CD27 expression. Also, the "marginal B cell zone" cluster has a couple of key naive B cell markers FCER2, CD27, Foxo1, CD24hi, TCL1A, IL4R. Lastly, there are a couple of plasmacytoid dendritic cells included in this cluster, the island very far out left of the T cell islands. CLEC4C, IL3RA... And in the myeloid object, pDCs are missing completely. In my opinion, this small pDC island should be removed here and put into the myeloid object. https://www.science.org/doi/10.1126/sciimmunol.abe6291 https://pmc.ncbi.nlm.nih.gov/articles/PMC10470411/
I agree this is Tregs, express FOXP3, CTLA4. Could consider subsetting into pTreg and tTreg based on IL10, IKZF2, GATA3 expression (as described in https://www.nature.com/articles/s41577-023-00890-w or https://www.nature.com/articles/nri.2016.36).
can be subclustered into those highly expressing IL1R1, which appear to be pTregs (relatively right portion); highly expressing IKZF2, which appear to be tTregs (relatively left portion)
Agree with splitting into IKZF2+ (also expressing CCR8 RTKN2) with a highly suppressive phenotype, and IKZF2- (also IL10+).
Agree FOXP3, IL10, CTLA4, IKZF2, TIGIT, CCR4, Can be further split into IKZF2 high and IKZF2 low
This is contamination from the epithelial compartment and should be removed from here (EPCAM, MUC2...)
these cells don't express CD8A nor naive marker genes e.g. SELL - probably not CD8 naïve cells; expression of CCR7, LEF1 not sufficient to claim these are naïve cells
CD4pos, CCR7pos, TCF7pos. Naive CD4+ T cells? (Possibly central memory, maybe a T cell person should double check)
Not expressing CD8A or CD8B but is expressing CD4. So not Naive CD8. The Naive CD8 subset is likely a subset of the current CD8 T cluster, specifically the one above the Naive CD4 T cells on the left of the T cell island. This cluster is Tfh like, ICOS, TOX2, CD200
I think this cluster contains some of the NK cells, specifically NK_CD56 bright, as evidenced by XCL1, IL2RB, NCAM1, and NCR1 expression.
they are the same subset, expressing IL-23R and TNFR25. Actually "ILC2" express more IL22 a marker of ILC3. Merge both into ILC3.
Expressing CD4, not expressing CD8A or CD8B
I agree these appear to be T cycling. They express CD3 genes, tend to form a "loose" cluster and are high in cycling genes (MKI67, TOP2A, etc.)
Annotate as NK cells. they are CD3 negative but expressing NCAM1, KLRD1. Nature Immunology volume 25, pages 1474–1488 (2024)
I agree this is plasma IGG. Expresses general plasma markers of MZB1 and XBP1, SDC1+ (class-switched), and expresses immunoglobulin heavy chains (IGHG3, IGHG1, IGHG2, IGHG4).
Delete. This cluster is noise. Has no B cell marker. Markers include secretory epithelial markers ZG16, CLCA1, MUC2.
I don't think these are ILC2. They don't express PTGDR2, HPGDS, IL9R (https://www.nature.com/articles/s41586-021-03852-1/figures/6). This cluster does express RORC, PCDH9, KIT, and LST1, so more likely ILC3.
they are the same subset, expressing IL-23R and TNFR25. Actually "ILC2" express more IL22 a marker of ILC3. Merge both into ILC3.
Top DEG when comparing to the "ILC3" cluster is IL22 (top up). No signs of increased GATA3, IL4, IL13, AREG... Possibly a mix of ILCs?
This is contamination from the epithelial compartment (MUC2, KRT19...) and should be deleted from here.
I agree this is likely MAIT. In addition to expressing TRAV1-2, also expressed SLC4A10 (https://www.nature.com/articles/s41586-021-03852-1#Sec8).
TRAV, RORC, IL23R Checks out with previous scRNA-seq MAIT T cell profiles. (e.g. here https://www.nature.com/articles/s41590-023-01575-1; this is liver, though)
