Forkhead box P1 (FOXP1) is a member of the forkhead box (FOX) family of transcription factors that are defined by a common DNA-binding domain termed the forkhead box or winged helix domain. The FOXP1 gene maps to a tumour suppressor locus at 3p14.1 and is aberrantly expressed at both the mRNA and protein levels in a variety of carcinomas [1Banham AH, Beasley N, Campo E, et al. The FOXP1 winged helix transcription factor is a novel candidate tumor suppressor gene on chromosome 3p Cancer Res 2001; 61: 8820-9.]. In breast cancer patients the loss of FOXP1 protein expression correlates with a poor prognosis, consistent with a role as a tumour suppressor [2Fox SB, Brown P, Han C, et al. Expression of the forkhead transcription factor FOXP1 is associated with estrogen receptor-a and improved survival in primary human breast carcinomas Clin Cancer Res 2004; 10: 3521-7.]. In contrast, the high level expression of the FOXP1 protein in large B-cell Non Hodgkin’s lymphomas (NHL) correlates with a poor prognosis [3Barrans SL, Fenton JA, Banham A, et al. Strong expression of FOXP1 identifies a distinct subset of diffuse large B-cell lymphoma patients with poor outcome Blood 2004; 104: 2933-5.-5Sagaert X, de Paepe P, Libbrecht L, et al. Forkhead box protein P1 expression in mucosa-associated lymphoid tissue lymphomas predicts poor prognosis and transformation to diffuse large B-cell lymphoma J Clin Oncol 2006; 24(16 ): 2490-7.]. While this might initially appear to be counter intuitive there are plenty of examples in the published literature of transcription factors, such as Maf [6Pouponnot C, Sii-Felice K, Hmitou I, et al. Cell context reveals a dual role for Maf in oncogenesis Oncogene 2006; 25(9 ): 1299-310.] and NF-κB [7Perkins ND. NF-kappaB: tumor promoter or suppressor? Trends Cell Biol 2004; 14(2 ): 64-9.], that can display dual roles as either tumour suppressors or oncogenes, depending on the cellular context. Certainly the alterations in apoptosis and cellular proliferation observed in cardiac tissues from Foxp1 knockout mice have indicated that the affects on these processes are regulated both temporally and spatially during development, suggesting that there is complex context dependent regulation of Foxp1 function in vivo [8Wang B, Weidenfeld J, Lu MM, et al. Foxp1 regulates cardiac outflow tract, endocardial cushion morphogenesis and myocytes proliferation and maturation Development 2004; 131: 4477-87.].

The identification of recurrent chromosome translocations targeting the FOXP1 gene in both mucosal associated lymphoid tissue (MALT) lymphomas and in diffuse large B-cell lymphomas (DLBCL) suggested that FOXP1 de-regulation might have an important role in lymphomagenesis [9Streubel B, Vinatzer U, Lamprecht A, et al. T(3; 14)(p14.1 q32) involving IGH and FOXP1 is a novel recurrent chromosomal aberration in MALT lymphoma Leukemia 2005; 19: 53-8.-11Fenton JA, Schuuring E, Barrans SL, et al. t (3; 114)(p14. q32) results in aberrant expression of FOXP1 in a case of diffuse large B-cell lymphoma Genes Chromosomes Cancer 2006; 45(2 ): 164-8.]. Interestingly these translocations commonly target the coding region of the FOXP1 gene [12Goatly A, Bacon CM, Nakamura S, et al. FOXP1 abnormalities in lymphoma: translocation breakpoint mapping reveals insights into deregulated transcriptional control Mod Pathol 2008; 21(7 ): 902-11.]. This is consistent with our data suggesting that N-terminal truncation of FOXP1 (by alternative splicing) may be a key event in normal B-cell activation that is retained in the activated B-cell-like (ABC) DLBCL subtype [13Brown PJ, Ashe SL, Leich E, et al. Potentially oncogenic B-cell activation-induced smaller isoforms of FOXP1 are highly expressed in the activated B cell-like subtype of DLBCL Blood 2008; 111(5 ): 2816-4.]. Furthermore, N-terminal truncation of foxP1 is an oncogenic event in a retroviral insertion model identifying novel oncogenes that cause avian nephroblastoma [14Pajer P, Pecenka V, Kralova J, et al. Identification of potential human oncogenes by mapping the common viral integration sites in avian nephroblastoma Cancer Res 2006; 66(1 ): 78-86.]. This raises the possibility that constitutive expression of smaller FOXP1 isoforms may have an oncogenic role in ABC-like DLBCL.

The majority of studies have focussed on FOXP1 in malignancy and thus there is still relatively little information concerning the roles of FOXP1 in normal tissues. Studies using knock out mice have identified an essential role for Foxp1 in embryonic development with phenotypes affecting cardiac [8Wang B, Weidenfeld J, Lu MM, et al. Foxp1 regulates cardiac outflow tract, endocardial cushion morphogenesis and myocytes proliferation and maturation Development 2004; 131: 4477-87.], neuronal [15Dasen JS, De Camilli A, Wang B, et al. Hox repertoires for motor neuron diversity and connectivity gated by a single accessory factor FoxP1 Cell 2008; 134(2 ): 304-16., 16Rousso DL, Gaber ZB, Wellik D, et al. Coordinated actions of the forkhead protein Foxp1 and Hox proteins in the columnar organization of spinal motor neurons Neuron 2008; 59(2 ): 226-40.], lung [17Shu W, Lu MM, Zhang Y, et al. Foxp2 and Foxp1 cooperatively regulate lung and esophagus development Development 2007; 134(10 ): 1991-2000.] and B-cell development [18Hu H, Wang B, Borde M, et al. Foxp1 is an essential transcriptional regulator of B cell development Nat Immunol 2006; 7(8 ): 819-26.]. Foxp1 is also required for monocyte differentiation and macrophage function [19Shi C, Zhang X, Chen Z, et al. Integrin engagement regulates monocyte differentiation through the forkhead transcription factor Foxp1 J Clin Invest 2004; 114(3 ): 408-18., 20Shi C, Sakuma M, Mooroka T, et al. Down-regulation of the forkhead transcription factor Foxp1 is required for monocyte differentiation and macrophage function Blood 2008; 112(12 ): 4699-711.], adipocyte differentiation and is implicated in human heart failure. Recently FOXP1 silencing in ex vivo expanded bone marrow mesenchymal stem cells has been reported to suppress their self-renewal capacity and their adipogenic differentiation [21Kubo H, Shimizu M, Taya Y, et al. Identification of mesenchymal stem cell (MSC)-transcription factors by microarray and knockdown analyses, and signature molecule-marked MSC in bone marrow by immunohistochemistry Genes Cells 2009; 14(3 ): 407-24.].

A brief description of FOXP1 expression in B cells in reactive tonsil has already reported that this protein is expressed in the majority of mantle zone B cells, a variable proportion of germinal centre (GC) cells and in many cells in the interfollicular area [1Banham AH, Beasley N, Campo E, et al. The FOXP1 winged helix transcription factor is a novel candidate tumor suppressor gene on chromosome 3p Cancer Res 2001; 61: 8820-9.]. In addition to staining haematopoietic cells, FOXP1 expression was also observed in both tonsillar epithelium and endothelium. In the current study we have employed double immunoenzymatic labelling techniques to identify the absence of FOXP1 protein expression in normal CD138⁺ tonsillar plasma cells. Despite the presence of FOXP1 transcripts, the vast majority of both normal and abnormal bone marrow plasma cells also lacked FOXP1 protein expression. Silencing of FOXP1 protein expression may have an important biological role during terminal B-cell differentiation to plasma cells and represents a novel mechanism by which FOXP1 expression is regulated.

Using double immunoenzymatic labelling studies we have identified the absence of FOXP1 protein expression in terminally differentiated CD38⁺, CD138⁺ or VS38c⁺ plasma cells from tonsil (Fig. 1). The occasional CD38⁺/FOXP1⁺ and CD138⁺/FOXP1⁺ lymphocytes were noted, indicating that a relatively rare population of terminally differentiated B cells do express FOXP1.

Fig. (1) Double immunoenzymatic labelling of formalin-fixed paraffin embedded tonsil sections. FOXP1 (brown) is rarely co-expressed with markers of terminally differentiated B cells (CD138, CD38, VS38c; blue). The black arrow head in the top righthand panel illustrates a rare plasma cell that is double labelled with CD138 and FOXP1.

While analysing FOXP1 expression in lymphoma cell lines we were surprised to find that the JJN3 MM cell line expressed relatively high levels of the FOXP1 mRNA (Fig. 2). The levels were comparable to those in the HLY1 and DB DLBCL cell lines, which we have previously shown to strongly express the FOXP1 mRNA and protein [13Brown PJ, Ashe SL, Leich E, et al. Potentially oncogenic B-cell activation-induced smaller isoforms of FOXP1 are highly expressed in the activated B cell-like subtype of DLBCL Blood 2008; 111(5 ): 2816-4.]. The NCIH929 and RPMI8226 cell lines expressed low levels of FOXP1 mRNA, while none was detectable in the THIEL cell line. Western blotting studies showed that FOXP1 protein expression was barely detectable in the majority of MM cell lines, in contrast to the highly FOXP1-positive OCI-Ly3 and DB DLBCL cell lines used as controls (Fig. 3). Interestingly the JJN3 cell line weakly expressed both the full-length FOXP1 protein and similar levels of the two smaller isoforms that are also present in the OCI-Ly3 DLBCL cell line. Using immunoenzymatic labelling to double stain the MM cell lines for FOXP1 expression (brown) and CD138 (blue) we observed no FOXP1 protein expression in the NCIH929 and THIEL cell lines, while the occasional FOXP1 positive nucleus was detected in the RPMI8226 and JJN3 cell lines.

Fig. (2) FOXP1 relative mRNA expression in MM and DLBCL-derived cell lines. Highest expression levels are observed in ABC-like DLBCL, although the GC-DLBCL line DB and the MM line JJN3 also express significant levels of FOXP1 transcripts.

Fig. (3) A. Western blotting of nuclear extracts from DLBCL and MM-derived cell lines. Probing with the anti-FOXP1 monoclonal antibody JC12 shows a lack of FOXP1 protein expression in MM as compared with DLBCL samples. Weak expression of three FOXP1 protein isoforms was detectable in the JJN3 cell line. TBP was used to indicate sample loading. B. Double immunoenzymatic labelling of MM-derived cell lines; FOXP1 staining in brown and CD138 in blue. The Presence of occasional FOXP1+CD138+ double labelled cells was detected in RPMI8226 cells, whereas no FOXP1 labelling was detectable in the NCIH929 cell line.

This study was extended to investigate, whether abnormal and malignant plasma cells in bone marrow also lacked FOXP1 expression. In whole bone marrow aspirates from Croatian patients without plasma cell malignancy or those with MGUS or MM there was no significant difference in FOXP1 mRNA levels between patient groups (Fig. 4a). Although, most samples expressed relatively high levels of FOXP1 mRNA there was no correlation with the expression levels of CD138 mRNA, which was used to assess the level of plasma cell infiltration (data not shown; Campbell et al., Br J Haem 2010; in press).

Fig. (4) FOXP1 relative mRNA expression in (A) total bone marrow aspirates and (B) purified CD138+ plasma cells. MM = multiple myeloma at diagnosis; MM-R = MM patients at relapse; MM-T, refers to samples taken from patients post-treatment; SM = smoldering myeloma; P = plasmacytoma; WM = Waldenstroms Macroglobulinaemia; AL = amyloid; R = reactive marrow; NHL= non-Hodgkin's lymphoma; no MM, includes a heterogenous group of patients whose plasma cells should not be malignant or pre-malignant. * These patients did not fall into one of the above categories. Bone marrow trephines from cases 7 and 8 were reported as LPL or myeloma, clinically the diagnosis was LPL; case 9 had a bone plasmacytoma, the bone marrow aspirate was not reported to show increased plasma cells and the trephine was inadequate.

Immunolabelling studies subsequently identified the widespread expression of the FOXP1 protein in CD138-negative bone marrow cells (Fig. 5). Therefore, FOXP1 mRNA expression was also further investigated in CD138⁺ purified bone marrow plasma cells from a cohort of patients from Oxford to exclude other cell populations from the expression analysis (Fig. 4b). FOXP1 was differentially expressed between the CD138⁺ bone marrow samples, but there was no evidence for elevated FOXP1 mRNA expression in malignant plasma cells when compared to those from MGUS patients or patients without plasma cell malignancy. However FOXP1 mRNA was detectable in all the plasma cell samples, a significant proportion of which (26/30, 86.6%) expressed levels comparable to that seen in some strongly positive DLBCL cell lines, such as DB and the ABC-DLBCL lines OCI-Ly3 and OCI-Ly10 [13Brown PJ, Ashe SL, Leich E, et al. Potentially oncogenic B-cell activation-induced smaller isoforms of FOXP1 are highly expressed in the activated B cell-like subtype of DLBCL Blood 2008; 111(5 ): 2816-4.]. Interestingly, the levels of FOXP1 mRNA in both normal and abnormal CD138⁺ bone marrow plasma cells was generally higher than those observed in the MM-derived cell lines. These data suggest that FOXP1 mRNA expression is not commonly transcriptionally silenced in normal or abnormal plasma cells.

Fig. (5) Double immunoenzymatic labelling of bone marrow samples; FOXP1 brown and CD138 blue. As indicated by the top four panels the majority of pre-malignant and malignant CD138+ plasma cells lacked expression of the FOXP1 protein, although there were many CD138-FOXP1+ cells in this tissue. The detection of >30% FOXP1+/CD138+ cells was restricted to two MM cases, one from Oxford (bottom left) and the other from Zagreb (bottom right).

Double immunoenzymatic labelling was used to investigate the expression of FOXP1 at the protein level in CD138⁺ plasma cells. Paraffin embedded bone marrow trephines were available from fifteen Croatian patients (including nine MM patients and two MGUS patients) of which ten (patients numbered; 1, 2, 3, 10, 12, 22, 23, 27, 31 and 32) corresponded to those where the expression levels of FOXP1 mRNA had been determined in whole bone marrow. There was no FOXP1 protein detected in fourteen of the patients’ biopsies. Samples with occasional rare FOXP1⁺ plasma cells were scored as negative throughout this study. The exception was patient number 3, where weak/moderate expression of the FOXP1 protein was observed in the nuclei of the majority of the CD138⁺ population (Fig. 5, bottom right). This 61 year old male patient was diagnosed with multiple myeloma, IgG kappa paraprotein of 63g/l, albumin 22g/l, Beta2microglobulin 7.0, (stage 3 ISS), with >80% plasma cell infiltrate in the aspirate/trephine and had multiple osteolytic lesions. He initially received VAD and thalidomide that had to be discontinued due to the cardiac side effects (bradycardia, arrhythmia). He was continued on Dexamethasone with which he achieved partial remission. The patient received tandem autologous stem cell transplants and is still alive two years after diagnosis. An unusual feature of this patient was that the diagnostic bone marrow sample was taken while he was recovering from sepsis (Streptococcus pneumoniae). We cannot exclude the possibility that this event may have affected the expression of FOXP1; certainly cytokines such as IL-2, IL-10 and TGF-beta have a key role in the development and function of regulatory T-cells characterised by the expression of the related FOXP3 protein. However, further studies will be needed to determine whether cytokines do indeed also have a role in modulating FOXP1 expression and function. Interestingly, this case had one of the lowest levels of FOXP1 mRNA expression suggesting that FOXP1 mRNA levels do not reflect tumoural FOXP1 protein expression in whole bone marrow aspirates.

FOXP1 protein expression was also investigated by double immunoenzymatic labelling in a larger series of patients from Oxford. Including routinely fixed bone marrow trephines from non-malignant reactive marrows (n=10), patients with MGUS (n=11) and a cohort with MM (n=60). The vast majority of normal and malignant bone marrow CD138⁺ plasma cells lacked FOXP1 protein expression. The presence of CD138-negative cells expressing FOXP1 served as an internal control for the immunolabelling technique. FOXP1 protein expression in more than 30% of CD138⁺ cells was only observed in one MM biopsy, taken at time of diagnosis (Fig. 5, bottom left). The patient had fairly typical clinical features including: age 69 years at diagnosis, 90% plasma cell infiltrate on trephine, no bone marrow involvement, normal calcium, anaemic at diagnosis with haemaglobin of 8.7, normal renal function, IgA lambda paraprotein, albumin 37, Beta2microglobulin 10.8 and was still alive after 7 months follow up. A minor population of FOXP1⁺/CD138⁺ plasma cells (<10%) was also observed in two MM cases and one MGUS case. These data suggest that FOXP1 is not frequently expressed at the protein level in MM.

Foxp1 has been shown to function during the pro-B to pre-B cell transition in the early stages of B-cell development (Hu et al., 2006) and regulates Rag gene expression. Our immunolabelling studies have already shown widespread, high level FOXP1 expression in lymphoid tissue, in particular in the naïve B cells of the mantle zone and variable but often strong expression in B cells within the germinal centre [1Banham AH, Beasley N, Campo E, et al. The FOXP1 winged helix transcription factor is a novel candidate tumor suppressor gene on chromosome 3p Cancer Res 2001; 61: 8820-9.]. However, here we show that expression of the FOXP1 protein appears to be tightly regulated in the terminal stages of B-cell development, rarely being expressed in normal plasma cells. However, FOXP1 transcripts were expressed at relatively high levels in CD138⁺ bone marrow plasma cells from patients without MM.

This pattern of FOXP1 protein expression, with occasional nuclear positivity, was found to be similar in MM cell lines and FOXP1 transcripts were fairly abundantly expressed in the JJN3 myeloma cell line. In the two MM cell lines in which FOXP1 protein was detectable, it appeared that this was restricted to moderate levels of nuclear expression in a very minor population of CD138⁺ cells. Similarly, we observed that the FOXP1 protein is rarely expressed in malignant plasma cells, yet the FOXP1 transcript is abundant in samples of purified CD138⁺ cells recovered from MM and MGUS BM aspirates. It was noticeable that both normal and malignant CD138⁺ primary bone marrow cells tended to expressed considerably higher levels of the FOXP1 mRNA than was observed in three of the four MM-derived cell lines. It is possible that factors in the bone marrow microenvironment may upregulate FOXP1 mRNA expression in bone marrow plasma cells. Certainly expression of the related FOXP3 transcription factor in regulatory T cells is significantly affected by cytokines, such as IL-2 and the signalling molecule TGF-β [22Josefowicz SZ, Rudensky A. Control of regulatory T cell lineage commitment and maintenance Immunity 2009; 30(5 ): 616-25.]. These data suggest that the expression of the FOXP1 protein in both normal and abnormal plasma cells is regulated at a post-transcriptional level.

While we were preparing this manuscript, another group published a paper and a case report describing FOXP1 protein expression in MM; FOXP1 mRNA expression was not analysed in either study [23Korac P, Skrtic A, Peran I, et al. Atypical FOXP1 expression in malignant plasma cells that show several simultaneous translocations Histopathology 2009; 54(6 ): 770-1.]. The case report described a FOXP1⁺ MM case with rather atypical clinical features, hyperploidy and multiple IGH translocations in which the FOXP1 gene was amplified in CD138⁺ cells (12-15 copies) [23Korac P, Skrtic A, Peran I, et al. Atypical FOXP1 expression in malignant plasma cells that show several simultaneous translocations Histopathology 2009; 54(6 ): 770-1.]. In a larger series of 13 MGUS and 60 MM patients they identified >30% nuclear FOXP1 protein expression in one MGUS and 12 MM biopsies (20%) [24Korac P, Peran I, Skrtic A, et al. FOXP1 expression in monoclonal gammopathy of undetermined significance and multiple myeloma Pathol Int 2009; 59(5 ): 354-8.]. Korac et al., detected FOXP1 protein expression at a much higher incidence than in our study, as we detected >30% FOXP1 nuclear positivity in only 3% of MM patients (2/68) and none of the MGUS patients (n=13).

The reason(s) for the difference in the findings are unclear. Both groups have used the same monoclonal anti-FOXP1 antibody to detect protein expression in formalin-fixed paraffin-embedded bone marrow trephines by immunohistochemistry (although different labelling kits were used). There was nothing to suggest that FOXP1 expression differed between the cases that we studied from Oxford versus those from Croatia. However, we found that double labelling enabled us to detect weak FOXP1 protein expression levels that would have been difficult to detect in the presence of a nuclear counterstain and to verify that the positive cells were indeed CD138⁺ plasma cells. As illustrated by the data presented in Fig. (5) we observed variable and often significant levels of FOXP1 expression in the CD138-negative population in bone marrow.

Korac and colleagues identified FOXP1 copy number changes in 24/60 MM and 5/13 MGUS [24Korac P, Peran I, Skrtic A, et al. FOXP1 expression in monoclonal gammopathy of undetermined significance and multiple myeloma Pathol Int 2009; 59(5 ): 354-8.]. How this relates to the expression of the FOXP1 protein is unclear as the abnormalities were detected at a much higher frequency than was FOXP1 protein expression. The authors did not comment on whether the MM cases reported to express the FOXP1 protein included those with the highest FOXP1 gene copy number. Further study will be necessary to determine whether these genetic changes genuinely contribute to aberrant FOXP1 protein expression in MM. Certainly our data suggest that the FOXP1 mRNA is commonly expressed in both normal and malignant plasma cells in the absence of FOXP1 protein expression.

The rarity of FOXP1 protein expression in normal plasma cells is consistent with the detection of FOXP1 protein expression in only a minority of MM. Other transcription factors with a role in early B-cell development are also involved in terminal differentiation to plasma cells, including PAX5 which is silenced during plasma cell development and IRF4 which is up-regulated during plasma cell differentiation. Thus it is not without precedent to suggest that the absence of FOXP1 protein expression may also be functionally important during terminal B-cell differentiation to plasma cells. It will be interesting in the future to investigate the affects that expressing FOXP1 has on the terminally differentiated phenotype of plasma cells. In particular whether FOXP1 regulates the expression of those cell surface markers and transcription factors whose silencing or induction plays a crucial role in this phenotype. The post-transcriptional regulation of FOXP1 protein expression in both normal and malignant plasma cells provides an explanation as to why its loss of expression would not have been detected by studies of gene expression changes occuring during plasmacytic differentiation. FOXP1 mRNA levels are indicative of the FOXP1 protein expression patterns in lymphomas derived from mature B cells. Therefore, it will be interesting to discover the mechanism of this novel post-transcriptional regulation of FOXP1 expression in terminally differentiated B cells.