Historically, by detailing immune functions in vivo and in vitro a model of specialized cell types in immunology and hema-topoiesis were mapped primarily on the basis of expressed cell surface antigens—many of which were codified by using single-cell analysis and fluorescence-based cytometry (31
). Because cell-surface proteins represent only a small proportion of the repertoire of gene products governing cell behavior, intracellular proteins (33
) are also critical in defining cell types. Because surface and intracellular molecules work together in concert to support different cellular roles, it might be expected that proteins governing specialized immunological cell functions (T cell receptor, B cell receptor, or cytokine receptors) are modulated in a coordinated manner as cells transit developmental pathways from stem cell precursors to differentiated endpoints.
We monitored 13 surface markers to identify immune cell types and 18 intracellular epitopes in order to interrogate intracellular signaling biology in healthy human bone marrow. We examined the signaling dynamics of these 18 intracellular markers in response to 13 ex vivo stimulation conditions (such as IL-7 or GM-CSF), including those shown to have prognostic value in leukemia, lymphoma, and myeloproliferative disorders [such as granulocyte colony stimulating factor (G-CSF)] (10
). Cell populations were first defined on the basis of conventional surface expression gates, ultimately identifying 24 immunological populations in human bone marrow (fig. S5
). The induced intracellular signaling responses (changes in phosphorylation state) in these populations, as compared with those of an untreated control, were summarized as a heat-map (). Unsupervised, hierarchical clustering of the phosphorylation responses allowed distinction of biologically related cell types (T cell subsets) by their signaling behavior alone, demonstrating that signaling capacities are closely tied to cellular lineage (fig. S6
). Several canonical signaling responses that mapped to manually determined cell types are shown in . These extremely specialized responses, such as the tight restriction of IL-7–mediated pSTAT5 responsiveness in T cells (, arrow 4) (38
) or lipopolysaccharide (LPS)–stimulated phosphorylation of the mitogen-activated protein kinase (MAPK) p38 (p-p38) responsiveness in monocytes (, arrow 5) (39
), suggest the existence of correlations between signaling events and surface marker–defined boundaries, thus presenting an opportunity to establish a unified view of immune signaling during hematopoiesis.
Fig. 3 Signaling functions mark developmental transitions in hematopoietic progression. (A) A heatmap summary, ordered developmentally by cell type and stimulation condition, of the status of 18 intracellular functional markers in cells treated with 1 of 13 (more ...)
signaling observations ( and fig. S10A
) for each replicate bone marrow, it was necessary to filter the data set in order to arrive at the most significant and potentially novel observations. Using a one-sample t
-test, over 500 observations were observed with a Bonferroni-adjusted significance of P
< 0.05 in each replicate bone marrow for a total of 860 unique responses (fig. S7
and table S3
). Of the 248 observations overlapping between patient marrows, 28 belonged exclusively to cells residing in the human hematopoietic progenitor cell compartment [hematopoietic stem cells (HSCs), multipotent progenitors (MPPs), granulocyte/macrophage progenitors (GMPs), and megakaryocyte-erythroid progenitors (MEPs)], including G-CSF induction of pSTAT3 in the most primitive cell types, HSC and MPP (40
). This same signaling behavior correlated with negative prognosis in acute myeloid leukemia (10
), suggesting that, as in the case of other malignancies, there may be a selective advantage for cells to mimic the properties of their most primitive counterparts.
For a more objective and fine-grained view of these cell type–specific responses, free of the biases of conventional 1D and 2D surface marker categorization, we overlaid the signaling behavior of the 18 functional epitopes on the tree structure using a similar approach as described for the immunophenotype staining panel (), allowing the intracellular signaling status to be visualized on the previously annotated tree structure (). Nodes were colored according to the magnitude of the difference in their median responses relative to the untreated control. This effectively eliminated the subjectivity of manual classification and improved the resolution of the heatmap (), separating the 24 manually assigned cell types into 282 logically connected nodes of phenotypically distinct, but locally similar, cell clusters.
The stimuli that corresponded closely with cell types identified manually in the heatmap also exhibited appropriately specific responses when overlaid on the tree structure—specifically, IL-7/pSTAT5 in T cells, B cell receptor (BCR)/phosphorylated B cell linker protein (pBLNK) exclusively in immature and mature B cells, and LPS/p-p38 restricted to the monocyte compartments (), with the latter corresponding to the expression of the LPS co-receptor (CD14) (). A complete set of the effects of 13 stimuli on 18 different functional markers is presented as tree plots (fig. S8
) along with a confirmatory analysis of a second bone marrow (fig. S9
With multiple matching canonical signaling pathways to validate the approach, we examined the data set for previously unidentified or unexpected signaling behaviors. For example, although pSTAT5 activation by IL-3 () was commensurate with IL-3Rα (CD123) expression levels () in myeloid cells, IL-3–mediated pSTAT3 activation was unexpectedly absent in mature B cells in spite of abundant presence of the receptor (, blue arrow). This suggests that mature B cells share some, but not all, IL-3 signaling mechanisms with myeloid cell types.
Other responses, such as phosphorylation of the protein tyrosine kinases Btk and Itk mediated by IFNα or ribosomal protein S6 by G-CSF, were less tightly confined, exhibiting a range of activity that spanned multiple cell types (). Yet other responses showed a signaling “gradient,” as exemplified by pervanadate (PVO4
)–mediated disruption of the kinase/phosphatase balance upstream of the adenosine 3′,5′-monophosphate (cAMP) response element–binding protein (CREB) transcription factor. A gradient of responses, highest in HSCs, decreased gradually along the path of B cell maturation (). A range of NFκB signaling responses, as measured by monitoring total IκBα levels, were observed across monocyte, NK and T cell subsets following TNFα stimulation (, light blue nodes). As in the CREB response to PVO4
described above (), the consistency of responses within the different T cell subsets suggests tightly regulated differences in signaling molecules that underlie the discrete functional roles of these related cell types. Together, these varied signaling responses across algorithmically defined partitions dictated solely by surface marker immunophenotype imply the existence of different classes of developmental transition points: (i) precise transitions, which are characterized by coordinated changes in cell signaling, such as the IL-7/pSTAT5 response in T cells and the LPS/p-p38 response in monocytes (), and (ii) continuous developmental progressions, which are characterized by gradual gain or loss of expression of certain kinases or phosphatases, as highlighted by PVO4
/pCREB () in B cells (28
). The latter is indicative of fine-grained changes in regulatory architecture that track with immuno-phenotype within conventionally defined hema-topoietic compartments and provides an opportunity to explore the mechanisms that define these distinctive regulatory phenomena.