Despite much effort, the development of mouse models for human MM remains a challenge [35
]. A major advance in this area of research has been provided recently by the generation of a transgenic mouse line with spontaneous MYC activation driven by Activation-Induced Deaminase (AID) [36
]. These transgenic mice developed bone marrow plasma cell tumors that recapitulate many features of human MM. However, rearrangements of the MYC gene are present in only 15% of MM patients [37
], calling for the development of additional mouse models that target distinct signaling pathways important in MM pathogenesis. Chromosomal translocations and rearrangements at the NF-κB2 locus have been shown to occur in primary human MM [8
]. The human MM cell lines JK6L and CAG also carry genetic mutations in the NF-κB2 gene [6
]. These mutations lead to the generation of C-terminally truncated NF-κB2 proteins similar to the tumor-derived NF-κB2 mutant p80HT [11
]. Currently, no mouse models are available for human MM with aberrant activation of NF-κB2 signaling. We have recently reported that transgenic mice with targeted expression of p80HT in lymphocytes developed predominantly B-lineage lymphomas with the tumor incidence of 79% by 70
]. In this investigation, we conducted detailed histological and immunohistochemistry examination of 12 tumor samples from the previous study [13
], which revealed that half of them (6/12) were plasma cell tumors. To corroborate this finding, we generated additional p80HT mice with a focus on their development of plasma cell tumors. Approximately 40% of the newly generated p80HT mice produced M-protein by 1
year of age. Most of these M-protein-positive p80HT mice developed plasma cell tumors with diffuse osteoporosis. Some of them also had osteolytic bone lesions and/or significant accumulation of plasma cells in the bone marrow. These findings provide the first direct evidence for a causal role of NF-κB2 mutation in the pathogenesis of plasma cell tumors that share some key histopathological and clinical features of human MM.
The transgenic mice express p80HT in both T and B cells, and our previous study suggests that p80HT promotes tumor development primarily by enhancing the survival of T and B cells [13
]. Also, the lymphoma development in p80HT mice is characterized by a prolonged latent period [13
], suggesting that additional genetic and/or epigenetic alterations are required for the malignant transformation of p80HT lymphocytes and their clonal expansion. We speculate that this might be the major reason why p80HT mice develop a wide spectrum of B cell lymphomas including plasma cell tumors, as well as T cell lymphomas, which depend on the type and developmental stage of the lymphocytes that have acquired secondary genetic and/or epigenetic alterations.
We have previously identified TRAF1 as a target gene critical for the oncogenic activity of p80HT [13
]. TRAF1 deficiency reestablished B cell homeostasis and significantly delayed the tumor development in p80HT mice [13
] (unpublished data). However, constitutive overexpression of TRAF1 in lymphocytes is not tumorigenic in mice [38
] suggesting that additional target genes must be critical for the development of plasma cell tumors in p80HT mice. We performed gene expression profiling of B cells from M-protein-positive p80HT mice for the reason that genes activated in these plasma cell precursors are anticipated to drive the development of plasma cell tumors. It is important to note that the three M-protein-positive p80HT mice used for the microarray assay might not develop plasma cell tumors in the end. Nonetheless, the gene expression profiling revealed the activation of many genes, in addition to TRAF1, in p80HT B cells that are known to promote the proliferation and survival of MM cells, as well as the differentiation of B cells to plasma cells.
IL-15 is the second most upregulated gene in p80HT B cells. IL-15 is important to the proliferation of MM cells and their ability to evade apoptosis [23
]. An autocrine loop between IL-15 and its receptor has been identified as a mechanism for tumor cell expansion in MM [39
]. The upregulation of both IL-15 and its receptor IL-15Ra (+1.7 fold, P
0.045) in the B cells of p80HT mice suggest that this pathway is important in the pathogenesis of plasma cell tumors in our mouse model.
Another top upregulated gene is IL-10. Using the transcription factor database search tool DECODE (SaBiosciences) and data from the University of California Santa Cruz Genome Browser, we identified potential two κB-binding sites in the IL-10 promoter (unpublished data), suggesting that p80HT may directly upregulate IL-10 expression. IL-10 exerts its biological functions primarily through Stat3 [30
]. It has long been recognized that IL-10 promotes the differentiation of B cells to plasma cells [22
]. This action of IL-10 is likely mediated, at least in part, by Blimp1, a known target gene of Stat3 [33
]. Consistent with the notion, Blimp1 expression was significantly upregulated (+1.7 fold, P
0.007) in p80HT B cells. Blimp1 is a transcriptional repressor required for the formation and maintenance of mature plasma cells [26
]. Blimp1 is also required for the formation of plasmacytoma in a mouse model [44
]. Other Stat3 target genes include cyclin D1, cyclin D2, and survivin [32
], and they were all markedly upregulated in p80HT B cells.
Survivin is a member of the inhibitor of apoptosis protein family with dual roles in regulation of cell cycle progression and apoptosis [45
]. Survivin expression is increased during MM progression, and knockdown of survivin induces cell death in human MM cells [29
]. It has been shown previously that the human survivin gene promoter region contains κB-binding sites [46
], and our sequence examination revealed 4 potential κB-binding sites within the mouse survivin promoter region (unpublished data). These observations suggest that p80HT may transcriptionally upregulate survivin expression either directly or indirectly through the IL-10-Stat3 signaling pathway.
Cyclins D1, D2 and D3 (encoded by CCND1, D2 and D3) interact with and activate cyclin-dependent kinase 4 (Cdk4) or Cdk6 to facilitate the G1/S cell-cycle transition [47
]. Upregulation of cyclin D expression occurs in the vast majority of MM tumors and has been considered a crucial and early oncogenic event in MM pathogenesis [27
]. Approximately 20% of MM tumors show elevated levels of cyclin D1 or D3 as the result of chromosomal translocations that juxtapose potent immunoglobulin (Ig) gene enhancers next to CCND1 (11q13) or CCND3 (6p21) [27
], and ~7% of MM tumors have Ig translocations involving c-MAF (16q23) or MAFB (20q11), which encode transcription factors that target CCND2 [27
]. In the rest of MM tumors, the transcription factors responsible for upregulating the expression of cyclin D genes remain to be identified. Cyclin D1 is a known target gene of the NF-κB1 signaling pathway [49
]. However, to the best of our knowledge, a role for NF-κB2 in regulation of cyclin D expression has not been previously described. Thus, our findings that cyclin D1 and cyclin D2 were significantly upregulated in p80HT B cells suggest a novel mechanism for early activation of cyclin D genes in the development of MM.
On the basis of above discussion, we suggest a model for p80HT activation of the IL-10-Stat3 signaling pathway to promote plasma cell expansion and tumor development by regulating the generation, proliferation and survival of plasma cells (Figure ). Please note that p80HT may also directly activate the transcription of cyclin D1, cyclin D2 and survivin genes. In addition, other p80HT-regulated genes, including IL-15, CD27, CD28, CD30, CD30L, CD80, and TRAF1 may contribute significantly to the tumorigenic process.
A simplified model for p80HT targeting the IL-10-Stat3 signaling pathway to promote plasma cell expansion and tumor development.
To determine the clinical relevance of our findings, we assessed the response of human MM cell lines to the inhibition of Stat3, a key downstream component of the IL-10 signaling pathway. Our investigation revealed a correlation between the activity of Stat3 and the sensitivity to Stat3 inhibition in MM cells. Also consistent with our model, Stat3 signaling was found to be essential for high-level expression of cyclin D1 and survivin in MM cells, providing a molecular mechanism for the critical role of Stat3 signaling in the proliferation and survival of MM cells. Thus, targeting the Stat3 signaling pathway may represent a therapeutic strategy for human MM.