Cellular origin(s) of chronic lymphocytic leukemia: cautionary notes and additional considerations and possibilities

Abstract

Several cell types have been suggested as giving rise to chronic lymphocytic leukemia (CLL), and these suggestions have reflected the sophistication of technology available at the time. Although there is no consensus as to the normal cellular counterpart(s) in the disease, an antigen-experienced B lymphocyte appears required based on surface membrane phenotypes and gene expression profiles. However, what is still unclear is whether a single or multiple normal precursors were stimulated to evolve into CLL and at what stage(s) this occurred. A unifying, parsimonious theory is that CLL clones with either mutated or unmutated IGHVs derive from marginal zone B cells. However, evidence for remarkably similar B-cell receptor amino acid sequence and striking differences in polyantigen and autoantigen-binding activity, found in some but not all CLL clones, challenge a single-cell derivation for CLL. In this Perspective, we summarize data regarding normal counterparts of CLL cells and suggest that a multistep process of leukemogenesis is important to consider when assigning a cellular origin for this disease. Finally, although available data do not definitively identify the cell(s) of origin, we offer possibilities for single- and multiple-cell origin models as straw men that can be improved on and hopefully lead to final answers to this puzzle.

Figure 1

Maturation of B lymphocytes. This schema is compiled from information derived from murine and human studies of B-cell differentiation. Note that the B-1 cell lineage has not been identified in humans; therefore, the scheme is based on information from mouse studies. The schema is not all-inclusive; it focuses on issues relevant to the discussion of possible precursors of CLL cells. (A) Normal B-lymphocyte maturation occurs in 2 phases: “foreign antigen independent,” occurring primarily in fetal liver or bone marrow (depending on species); and “primarily foreign antigen dependent,” occurring in the periphery in LNs and spleen. Autoantigen interaction could influence B cells expressing complementary BCRs during both phases in positive and negative manners, although the extent to which this occurs remains unresolved. For classic (non–B-1 cell) B-lymphocyte maturation, BCR-mediated signaling is important from at least the pre–B-cell stage of differentiation when a BCR composed of a rearranged and translated IGHV/D/J rearrangement associates with the monomorphic surrogate L chain. BCR-mediated signaling through this receptor is either constitutive or induced by binding of autologous epitopes within the local environment; in the latter case, the diversity of the bound epitopes is limited because not all translated IGHV/D/J rearrangements associate with the translated surrogate L-chain receptor and also because the monomorphic structure of translated surrogate L chain does not provide antigen-binding diversity. Engagement of this receptor leads to positive selection and clonal amplification of B-cell precursors with restricted IGHV/D/J structures that can be autoreactive. At the immature B-cell stage of maturation, when translated IGHV/D/J and IGK/J or IGL/J rearrangements are paired, negative selection for higher-affinity autoreactive BCRs occurs; this process involves clonal deletion, receptor editing of the rearranged κ or λ L chain, or receptor revision of the IGHV gene. It is assumed that the same process occurs for cells of the B-1 lineage. Note that some B lymphocytes with low affinity for autoantigens do exit to the periphery. Thus, in the normal setting, B cells with no or low affinity for autoantigen move to the periphery as transitional B cells. These cells are selected into the mature B-cell repertoire as either MZ B cells or follicular (FO) B cells by T cell-independent or T cell-dependent BCR stimulation, respectively. At this step, smIgD appears. It is assumed that T cell-independent signaling is important for B-1 cells in this regard. (B) Illustration of GC and some of the consequences of antigen experience of B-1 cells, MZ B cells, and FO B cells. MZ B cells (IgM^highIgD^low) take up residence in the MZ of the spleen or MZ-like areas in human tonsillar subepithelial areas, dome regions of Peyer patches, and subcapsular regions of LNs. When they encounter T cell-independent antigens, they can respond without undergoing isotype class switching or IGV gene mutations and give rise to plasma cells secreting polyreactive, unmutated IgMs that serve as protectors from virulent microbes or eliminators of catabolic autologous molecules. MZ B cells can undergo isotype class switching or IGV gene mutation, again giving rise to plasma cells that produce oligoreactive/monoreactive Igs. Switched and nonswitched memory cells can be generated as well. B-1 cells, like MZ B cells, respond to T-independent stimulation and give rise to plasma cells producing primarily polyreactive unmutated IgMs that provide protection against infections and remove cellular debris. See the text for suggestions that B-1 cells and MZ cells, when antigen experienced, could be “memory-like.” FO B cells are the main source of mutated, isotype switch protective Igs. When these cells bind foreign antigen within (follicular mantle) or outside (blood or lymph) solid tissues, they migrate to LNs or spleen and, with the help of nonlymphoid cells and T cells, build GCs and initiate GC reactions. Within the dark zone, activated FO cells (centroblasts) down-regulate BCRs and expand at an extremely rapid rate, acquiring IGV mutations with each round of cell division. Subsequently, cells move into the basal light zone where they re-express BCRs (centrocytes) and have the opportunity to bind the initial stimulatory foreign antigen that is held in its native state on the surface of follicular dendritic cells. Those cells with BCRs that can bind these antigens are provided with a temporary survival signal; access to these antigens is limited and in this way B cells with higher affinity for the antigen are preferentially selected to avoid apoptosis and proceed to the apical light zone. In this area, B cells that have bound the initiating antigen express fragments on their cell surfaces, associated with major histocompatibility complex class II molecules, and receive a permanent survival signal from antigen-specific T cells (“follicular T cells”). The affinity-matured, often isotype-switched B cells then differentiate into plasma cells and produce oligoreactive/monoreactive Igs or become memory cells; some of these memory cells migrate to the MZ and to the bone marrow (latter not shown). Finally, AID, the enzyme required for IGV mutations and isotype class switching, is definitely expressed in FO B cells during GC reactions. It is still unclear whether AID is expressed in human B cells at more immature stages of maturation (eg, immature B cells and transitional B cells; A) or that mature down different pathways (eg, MZ B cells and B-1 cells; B).