We read with interest the recent article by Stein et al studying the monocytoid B cells in comparison to other B-cell subsets such as mantle cells, germinal center B cells, and splenic marginal zone B cells.1 In that study, 11 cases of Piringer Kuchinka lymphadenitis were selected for the study of monocytoid B cells. The phenotype of the monocytoid B cells was investigated using immunohistochemistry and in situ hybridization on formalin-fixed paraffin-embedded tissues. Single-cell polymerase chain reaction (PCR) was used to study the mutation pattern and immunoglobulin heavy chain variable gene usage (VH) in monocytoid B cells from 5 cases.
The study confirmed our findings that monocytoid B cells are heterogeneous and comprise both naive and memory B cells, as well as the fact that an important fraction of the monocytoid B cells but not the splenic marginal zone B cells are proliferating.2 But differences are reported by Stein et al with regard to the presence of B-cell clones, as well as the level of somatic hypermutation in the monocytoid B-cell and the splenic marginal zone. We previously reported that the monocytoid B-cell zones and, to a lesser extent, the splenic marginal zone may harbor B-cell clones.2 In contrast, no B-cell clones could be identified by Stein et al.1 We believe that the different results obtained are likely due to different methods that have been used to study the clonal relationship of the monocytoid B cells. We performed a denaturing gradient gel electrophoresis (DGGE) analysis of the amplified immunoglobulin heavy chain (IgH) complementarity determining region 3 (CDR3) products from microdissected clusters of Ki-67 positive and negative monocytoid and splenic marginal zone B cells.2 By using DGGE, small clones of B cells were demonstrated in the monocytoid B-cell and splenic marginal zones. In this study we analyzed about 500 cells in each of the microdissected marginal zones. In contrast, Stein et al investigated the rearranged immunoglobulin genes of single monocytoid B cells by directly sequencing the amplified IgH products; 14 to 25 rearranged immunoglobulin heavy chain gene sequences were obtained per case. The latter numbers of sequences may not be enough to demonstrate the presence of B-cell clones. In addition, we observed that clones were more frequent in zones with a higher number of proliferating cells, a finding that we have recently confirmed by other methods (manuscript in preparation). Indeed, the proliferating cell fraction in monocytoid B-cell zones varies, a finding that is also reported by Stein et al (between 10 to 35 percent of the monocytoid B cells, in their study). Thus the selection of cells or zones for study can to a great extent determine the outcome of the results on clonality and may additionally explain the differences between Stein et al's and our findings.
Stein et al also reports differences between their study and ours regarding the percentage of cells with somatic mutations to be found between monocytoid B cells and (splenic) marginal zone B cells. To explain this, the authors claim that germinal center cells might have been mistaken for monocytoid B cells in our study and therefore has yielded a higher number of cells with somatic hypermutations. Even without being experienced morphologists, this event would have been unlikely in view of the combined staining for CD23 and IgD prior to microdissection. This combined staining highlights both follicle centers and their mantles. We argue that Stein et al's and our findings are basically similar if one pools the cells with no or only low-level somatic hypermutations and compares those with the number of cells with a high level of somatic hypermutations. This seems a relevant distinction in terms of the affinity maturation of the immune response. As such, it is clear from both our studies, as well as that of others, that the monocytoid B cells, as well as the splenic marginal zone B cells, are heterogeneous with respect to the number of somatic hypermutations.1-4
Stein et al further argues that the lower number of cells with high-level somatic hypermutations among monocytoid B cells warrants their distinction from (splenic) marginal-zone B cells. But analysis of their data shows that the range of monocytoid B cells with somatic hypermutations varies from 13.3 to 50 percent and the number of accumulated mutations per rearranged immunoglobulin gene can reach up to 14 depending on the case and the cells analyzed. This variability is also seen in marginal-zone B cells in the spleen. Thus the relative percentage of cells with low-level and high-level somatic hypermutations differs between various monocytoid B cell zones, as well as between monocytoid B cells and splenic marginal zone B cells. This can be explained by different phases of the immune response at which tissues are analyzed and by obvious differences in antigenic challenge provoking the immune response. In that respect, it is relevant to note that, both in Stein's study and ours, lymph nodes suspect for toxoplasma infection are compared with normal spleen. The even higher number of B cells with somatic mutations in the marginal zone of Peyer patch, as well as the higher number of mutations in the rearranged immunoglobulin genes of these cells as compared to the splenic marginal zone B cells, can be explained similarly. Therefore, heterogeneity with respect to the occurrence and number of somatic hypermutations in the immunoglobulin genes of the B cells is a hallmark of the marginal zone, including the monocytoid B-cell zone. Scoring this heterogeneity is not appropriate to recognize different B-cell subsets.
The authors further claim that weak signals for gamma transcripts are another distinctive feature of monocytoid B-cells. But the authors have not studied the presence of these transcripts in other marginal zone B cells whereby the uniqueness of this feature for monocytoid B cells is uncertain. In addition, in situ hybridization does not discriminate beween germ-line gamma transcripts.5 Therefore, the importance of this finding to define monocytoid B cells as constituting a distinct B-cell subset is not established.
In conclusion, we believe that Stein et al's data do not essentially differ from our own and do not provide convincing evidence that monocytoid B cells are different from marginal zone B cells at other sites.
Monocytoid B cells are different from marginal zone B cells
The assignment of monocytoid B cells to known B-cell subsets has been a source of constant debate for many years now. More recent molecular data provided both by Tierens et al1-1 and by our group1-2 has further intensified this debate. In their letter and in their recent publication, Tierens et al stated that monocytoid B cells are identical to nodal marginal zone B cells and belong to the same B-cell compartment as marginal zone B cells of the spleen. As a result, they have consistently used the term “nodal marginal B cells” for monocytoid B cells. The authors have, however, overlooked several important facts that would bring their concept into doubt.
Tierens et al's statement that monocytoid B cells and marginal B cells belong to the same B-cell compartment is based on the claim that the mentioned cells have “similar” immunophenotypic features and a “similar” immunoglobulin (Ig) rearrangement pattern. But our much more detailed phenotypical analysis provides several pieces of evidence to suggest that monocytoid B cells are distinct from marginal zone B cells. The Ki-B3 epitope of CD45RA is consistently expressed by monocytoid B cells, and DBA44 is detectable in approximately 20% of these cells. In contrast, both molecules are completely absent from splenic and nodal marginal zone B cells. Conversely, Bcl-2 and IgM, which are consistently detectable in marginal zone B cells, are completely (or predominantly) absent from monocytoid B cells. Further striking differences between both types of B cells concern the admixture of other cells. In contrast to marginal zones, areas of monocytoid B cells are nearly completely devoid of T cells but contain neutrophils in significant numbers. These clear-cut differences prompted us to conclude that real monocytoid B cells represent a B-cell subset unrelated to marginal zone B cells of the spleen and of the lymph nodes.
In harmony with this conclusion are our data concerning rearranged Ig genes. Monocytoid B cells in most instances (75%) harbor unmutated Ig genes, indicating their derivation from pre–germinal center B cells. In contrast, marginal zone B cells originate, in the vast majority, from mutated post–germinal center B cells as evidenced by our and others' investigations.1-2,1-3 Tierens et al's findings that the majority of nodal marginal zone B cells carry mutated IgH genes and are thus derived from mutated memory B cells clearly supports our conclusion that monocytoid B cells represent a different B cell subset. Furthermore, Tierens et al found a relatively high number of nonfunctional Ig rearrangements (37.5%) in their nodal marginal zone B cells.1-1 Because there is no subset of B cells other than germinal center B cells that carry nonfunctional IgH rearrangements at that high frequency, it is reasonable to assume that these cells most likely represent germinal center B cells. Because Tierens et al also did not provide an alternative explanation in their letter, a germinal center derivation of these nonfunctional B cells is still a matter of high probability.
Tierens et al's discussion concerning the “clonality” of marginal zone B cells and monocytoid B cells is also unfortunately misleading. Generally speaking, proliferating cells always produce cell clones. The extension of these cell clones depends on various factors such as the number of cells in cycle, duration of proliferation, preferred (biased) mitotic division of certain cells, and so forth. We estimated the growth fraction by determining the proliferation index (Ki-67 staining) and found more monocytoid B cells in cell cycle than marginal zone B cells. In line with this observation, by single-cell analysis we found some identically rearranged monocytoid B cells, whereas marginal zone B cells were unrelated in all instances. In germinal centers many clonally related B cells have been repeatedly demonstrated by many investigations, including our own.1-2,1-4 Tierens et al performed denaturing gradient gel electrophoresis (DGGE) analysis for the determination of clonality, but we have several concerns about the reliability of their results: (1) Germinal centers are known to consistently contain huge numbers of proliferating B cells and large B-cell clones. Unexpectedly, in Tierens et al's study they proved to be clearly polyclonal by DGGE analysis in many instances (see lanes 5, 7, 15, and 17 of their figure 5).1-1 (2) As stated in their paper, “clonal … rearrangements could also be observed when analyzing marginal zones containing only few or no cycling cells.”1(p230) This indicates that DGGE can create artificial clonal patterns. (3) As stated in their letter, “small clones of B cells were demonstrated in the monocytoid B-cell and splenic marginal zones.” But upon comparison of the results provided in their paper (figures 4 and 5), the degree of clonality of the marginal zones clearly exceeded that of the germinal centers in several instances. To conclude, although the number of cells investigated is smaller in a single-cell approach, the results obtained with this technique are obviously more representative and more reliable than those obtained by microdissection and DGGE.
Finally, we are of course aware of the possibility that Igγ germ-line transcripts can occur in B cells. But these germ-line transcripts are mainly found in germinal centers prior to class-switch recombination and plasmacellular differentiation.1-5 Because class switching and plasmacellular differentiation have not been observed in monocytoid B cells, the meaning of Igγ transcripts in these cells requires further investigation.
In summary, our data provide convincing arguments that monocytoid B cells are distinct from marginal zone B cells and represent a unique B-cell subset. It would appear most likely that differences in our results and in Tierens et al's results are due to technical problems, analysis of different cells, and/or the selection of cases unsuitable for the investigation of monocytoid B cells.
