Our single cell atlas of human B cells is out in @ImmunityCP! We screened the B cell surface proteome and then followed up with functional analyses. We identify twelve populations across four lymphoid tissues. https://www.cell.com/immunity/fulltext/S1074-7613(20)30268-5 1/16

Using a multiplexed mass cytometry approach, we screened the expression of 351 surface molecules on human B cells, mostly looking at CD markers and other proteins associated with immunological function or signaling. 2/16

We identified 98 surface molecules expressed by human B cells and evaluated their expression on canonical B cell gates. 3/16

We found that naive B cells consistently had lower expression of surface molecules across gene ontology terms and when averaged across all surface markers, confirming that they exist in a more anergic state than their memory counterparts. 4/16

We identified CD45RB as a marker for memory cells. While CD45RB+ CD27+ had the highest IgH mutation rates, CD45RB+ CD27- cells had also undergone somatic hypermutation, to a lesser extant. 5/16

We put together the most informative markers onto a single CyTOF panel and clustered the data to identify new, more granular B cell subsets. We identified ten unique (and gate-able) populations of B cells in peripheral blood. 6/16

These populations had both unique expression profiles and differential isotype usage. 7/16

We found population-level differences in various functional readouts including metabolism, RNA and polypeptide biosynthesis, and immune signaling in response to BCR and CD40 stimulus. 8/16

By phospho-CyTO, we found all B cell populations were somewhat responsive to stimulation, but in particular CD95+ Memory (a mostly class-switched CD27+ CD45RB+ population) and CD19hi CD11c+ Memory (an atypical memory population) had the most potent responses. 9/16

We show that CD19hi CD11c+ Memory is highly enriched for T-Bet and PD-1! As this cell type has been described as prevalent in various autoimmune conditions, this PD-1 expression could in part explain autoimmune effects seen in cancer patients given PD-1 checkpoint blockade. 10/16

We also investigated B cell profiles in human tonsil, lymph node, and bone marrow. While we found that bone marrow and peripheral blood had similar compositions of B cells, tonsil and lymph node were very unique. 11/16

In addition to tissue-specific differences within B cell populations, we identified two populations not found in blood: germinal center (GC) and CD39+ Tonsillar B cells. GC cells were in both lymph node and tonsil, while CD39+ Tonsillar cells were exclusive to the tonsil. 12/16

CD39+ Tonsillar B cells are mostly class-switched, low for surface Ig, and lacking CD38, CD27, and CD21. While similar to CD19hi CD11c+ Memory (which are absent in the tonsil), these cells are distinct due to expression of CD20, CD22, CD72, CD40, CD184, and of course, CD39. 13/16

In summary, we identified 12 B cell populations across four lymphoid tissues and provided comprehensive profiles of each. We hope this work will serve as a foundation for future investigation of human B cells in infection, autoimmunity, and cancer. 14/16

All cytometry data is available at https://flowrepository.org/id/FR-FCM-Z2MA  and https://flowrepository.org/id/FR-FCM-Z2MC . All code is available at https://github.com/davidrglass/atlas. 15/16

Many thanks to my co-authors @sideroblast @jp_oliveria @felixjhartmann @ArielA_Calderon @drmarglass @lisa_wagar @Bendall_Lab and Sam, Luciene, and Mark who are not on twitter. Also thanks to the editors, reviewers, and my in-house readers @leeat_keren and @ErinMcCaffrey14! 16/16