B-cell lymphomas: getting in the zone!

David Tarlinton

Germinal center (GC) B cells, defined by histology as belonging to dark (DZ) and light zones (LZ), are given unique, corresponding molecular identities in humans by Victora et al in this issue of Blood,1 identities that, surprisingly, appear in many B-cell lymphomas.

Germinal centers are the sites of mutation-driven diversification of immunoglobulin variable (V) region genes and their subsequent selection based on binding affinity.2 Early physical descriptions of these structures identified DZ and LZ that served to partition the activities that were attributed to GCs. In a landmark model of GC function, it was proposed that B cells underwent proliferation and mutation in the DZ before migrating to the LZ to undergo selection for improved binding to the immunizing antigen, an interaction that could result in the differentiation of the B cell into a memory B cell or a plasma cell.3 Alternatively, LZ B cells could be induced to re-enter the DZ to undergo additional rounds of proliferation and mutation, initiating another cycle of selection and differentiation. In this cyclic re-entry model, the molecular machinery driving V gene mutation—now known to be the enzyme activation-induced cytidine deaminase (AICD)—was associated intimately with B cell division and thus separated from the processes associated with differentiation, processes that are entirely dependent on the activity of the GC resident, CD4+ helper T cells, now known as follicular helper T (Tfh) cells.4 An expectation of this model, in addition to that of the B-cell movement between zones, was that there would be molecular differences between the 2 populations of GC B cells, reflecting their different activities.

The mechanics underpinning GC function remained uncertain until the application of multiphoton microscopy to these structures during immune responses in mice (reviewed in Victora and Nussenzweig5). Victora and colleagues used a photo-activatable form of the fluorescent protein GFP to uniquely label and thus isolate DZ and LZ B cells based on their in vivo location.6 Once isolated by flow cytometry, these 2 cell types were interrogated by molecular methods to reveal unique attributes, including the genes up-regulated uniquely in each zone and thus providing molecular signatures for each population. These signatures, not surprisingly, included cell-surface proteins that facilitated the routine isolation and characterization of these B cells. Chief among these markers were CXCR4, already identified as crucial for the correct function of GCs in mice,7 CD83 and CD86. Thus in mice, GC DZ B cells are CXCR4hi, CD86lo, and CD83lo while LZ B cells are the inverse, CXCR4lo, CD83hi, CD86hi. This identification allowed detailed characterization of these cells, which validated many aspects of the original model3 with DZ cells being significantly enriched for proliferation and LZ B cells interacting with GC T cells, the driving force behind selection and differentiation.

Now, Victora and colleagues have applied the mouse fractionation scheme to human GC B cells and found, more or less, complete concordance.1 Human tonsil GC B cells can be partitioned by CXCR4 and CD83 (CD86 is not useful), the populations thus identified correspond to the expected histologic locations, proliferation as measured by DNA content occurs predominantly in the DZ and, in addition, there is significantly increased expression of AICD in DZ B cells. Victora et al clarify other attributes of these 2 B-cell types. For example they note no major difference in cell size or granularity but increased expression of CD27—a marker normally associated with memory—on DZ B cells.1

Now able to identify and purify these 2 types of human GC B cells, Victora et al analyzed gene expression.1 They identified a modest number of differences between DZ and LZ B cells and, perhaps not surprisingly, found that the majority of the differences between the human B-cell subsets were also different between the equivalent mouse GC B-cell subsets. Thus, they defined through this cross-species comparison, gene-expression signatures that identify robustly and across species, DZ and LZ B cells.1 In fact, the authors make the point that the DZ and LZ B cell gene expression patterns are so similar that they should be considered stages of activation within 1 population rather than 2 discrete populations.

The final aspect of the work reported by Victora and colleagues was the examination of human GC-derived B-cell lymphomas for the DZ and LZ gene-expression signatures.1 The authors examined a series of B-cell non-Hodgkin lymphomas, comprising several distinct lymphoma subtypes, and found all types except 1 were assigned reproducibly and with great confidence to the LZ expression group. This included diffuse large B-cell lymphomas (DLBCL) of activated B cell (ABC) and GC types, follicular lymphoma and many cases of Burkitt lymphoma (BL). The remaining fraction of BL, however, segregated reasonably closely with the DZ expression signature. Closer examination of the BL cases revealed their DZ and LZ partition corresponded very closely with a previous division of these lymphomas into molecular and nonmolecular BL.8 It is concluded that the tumors resemble the LZ expression profile predominantly because of their expression of CD40/LMP1 and related signaling response genes, while loss of these signatures confers the DZ match.

In summary, there now is a clear manner to recover GC B cells from 2 compartments that are associated with the key functions of the GC: mutation, proliferation, selection, and differentiation. This will surely lead to greater insights into the initiation, propagation, and cessation of these crucial processes in both healthy and disease states. Similarly, combining knowledge of the migration route of GC B cells and the ability to recover the cells at the beginning and end of the track will also lead to better understanding of the complex but essential process of cell movement. The association of the zonal expression signatures with particular B-cell lymphomas may also lead to further insight into the origin or occurrence of the transformative events that collectively constitute lymphoma development.


  • Conflict-of-interest disclosure: The author declares no conflict of interest. ■