Immunocompetent mouse models of multiple myeloma

Abstract

Immunocompetent murine models of multiple myeloma are critical for understanding the pathogenesis of multiple myeloma and for the development of novel immunotherapeutics. Different models are available in Balb/c and C57Bl strains, each with different advantages and disadvantages. The availability of many transplantable cell lines allows for the conduct of experiments with large cohorts of mice bearing identical tumors, while cell lines that grow in vitro can be used for genetic manipulations. The introduction of human CRBN into these models allows for the study of IMiDs and cereblon based PROTACs in mice. New genetically engineered models based on germinal center cell activation of Nsd2 or Ccnd1 together with constitutive NFkB are being developed to model some of the important genetic subtypes of human multiple myeloma.

Introduction

Multiple myeloma (MM) is a clonal malignancy of long-lived plasma cells with somatically mutated, isotype switched immunoglobulin genes that reside in the bone marrow and is characterized by complex genetic aberrations and structural variations, including immunoglobulin heavy chain translocations and hyperdiploidy. Despite the significant advancements in understanding the molecular pathogenesis of MM, translating these insights into effective and durable treatments has been challenging due to the complexity of the disease and the heterogeneous nature of tumor progression.[1] To address these challenges, the development of robust preclinical models is critical. Syngeneic mouse models of MM have emerged as important tools for dissecting the molecular mechanisms driving the disease and for testing novel therapeutic strategies. These models allow for the controlled study of genetic lesions that are pivotal in myeloma pathogenesis, such as dysregulation of MYC, NFKB, NSD2, CCND1, TP53, and RAS pathways, within a fully immunocompetent environment. Genetically engineered models can recapitulate the stepwise progression of MM from its precursor conditions, including monoclonal gammopathy of undetermined significance (MGUS) and SMM, through to full-blown symptomatic disease. This mirrors the temporal and genetic evolution observed in human MM, allowing researchers to track the impact of specific mutations on disease progression. These models also enable the investigation of interactions between myeloma cells and the bone marrow microenvironment, which is critical for understanding drug resistance and disease relapse. Moreover, mouse models are proving essential for testing immunotherapies that leverage the host’s immune system to combat the disease. With the emergence of T-cell-based therapies, such as chimeric antigen receptor T cells (CAR-T) and bispecific antibodies, immunocompetent mouse models provide the ideal platform to study the dynamics of immune-mediated clearance of myeloma cells and the development of resistance mechanisms. These models also facilitate the exploration of novel therapeutic combinations and their effects on both the tumor and the microenvironment, and the optimization of immunotherapies to minimize toxicity.

In this review, we will discuss how mouse models have advanced our understanding of the molecular and immunological aspects of MM. We will highlight the strengths and limitations of each model and explore their role in preclinical testing of novel therapies that aim to improve outcomes for patients.

Important Characteristics of Human Myeloma to Be Modeled in Mice

To faithfully replicate the pathogenesis of human MM in mouse models, it is essential to capture the key biological and genetic features that define the disease. Human myeloma is characterized by a complex interplay of genetic mutations, epigenetic alterations, and interactions with the bone marrow microenvironment. Several critical aspects must be integrated into any effective mouse model to ensure relevance to the human condition.

1. Cell of origin

As noted above, human MM is a tumor of post-germinal center isotype class-switched PC. One of the major stumbling blocks in the development of mouse models is the activation of genetic lesions earlier in B-cell development prior to passage through the germinal center, which can lead to pre-B cell leukemia, B cell lymphoma, or pre-germinal center PC tumors that are mostly IgM. The hallmark of a post-germinal center cell is the presence of extensive (>2%) somatic hypermutation of the immunoglobulin genes. Off target mutational activity in the germinal center contributes to the development of oncogenic mutations, chromosome translocations, and the expression of neo-antigens that directly affect tumor growth but also its interaction with the microenvironment. The presence of all of these elements is important in creating a faithful mouse model of MM. In addition, there are important functional characteristics of human MM cells to be considered. Human MM cells are primarily localized in the bone marrow, whereas murine models may primarily involve the peritoneum, lymph nodes or spleen. Normal PC, and clonal cells from patients with MG secrete very high levels of immunoglobulin and devote more than half of their transcriptome to the transcription of immunoglobulin genes. With malignant progression human MM cells express less immunoglobulin as a percent of transcriptome: ~40% in newly diagnosed MM, ~15% in relapsed refractory MM, and ~1% in MM cell lines in vitro. At the same time, there is an increase in proliferation associated with progression. Less than 2% of MM cells in newly diagnosed patients are in S-phase, this number increase to 3-4% in aggressive, relapsed refractory MM, and more than 10-15% in MM cell lines growing in vitro.[2]

2. Genetic Lesions and Clonal Evolution

One of the hallmarks of MM is the presence of early genetic lesions such as immunoglobulin heavy chain translocations and hyperdiploidy, which drive the initial clonal expansion of malignant plasma cells. These primary genetic events, almost always involve the direct or indirect dysregulation of a cyclin D family gene, are present in 100% of cells from the earliest stages of the disease, and persist through its progression from MGUS to symptomatic MM[3]. Mouse models that incorporate these early genetic abnormalities promise to more accurately reflect the clonal evolution of the disease. It is likely that the timing of the primary oncogene activation to the germinal center is key to avoid development of pre-germinal center malignancies, and currently is best achieved using IgG1-CRE for conditional oncogene activation[4, 5].

Ideally, secondary genetic events such as activation of MYC, RAS, NFKB would be timed to occur after the clonal expansion induced by the primary event. Currently, this is best achieved using somatic hypermutation to sporadically activate translation of MYC in germinal center B cells in the Vk*MYC model (see below)[6]. Finally, if desired, terminal progression events, such as inactivation of CDKN2C, RB1, or TP53, associated with high proliferation, could be layered in. Currently no mouse model can adequately sequence all these genetic events. However, models that start from the activation of one or two oncogenes can faithfully mimic the clonal heterogeneity seen in MM, with multiple subclones harboring distinct spontaneous secondary mutations in pathways such as MYC or NFkB or p53 that are eventually selected for in vivo growth and contribute to drug resistance and disease relapse. A robust model must also mimic this intratumor genetic diversity, allowing for the study of how different subclones compete with each other, emerge and evolve under selective pressures such as therapy.

3. Interaction with the Bone Marrow Microenvironment

The bone marrow microenvironment plays a crucial role in the pathogenesis of MM, providing a niche that supports tumor growth, survival, dormancy and drug resistance. Reciprocal interactions between myeloma cells and various components of the bone marrow, including stromal cells, osteoclasts, osteoblasts, and immune cells, are mediated by cytokines, chemokines, and adhesion molecules[1]. These interactions must be adequately modeled to understand their contribution to disease progression and therapeutic resistance. For instance, the suppression of osteoblast activity and activation of osteoclasts leads to the characteristic lytic bone lesions in MM, which must be reflected in mouse models to study bone disease and potential therapeutic interventions[7].

4. Immune Evasion and Tumor Immunology

As immunotherapies such as CAR-T cells and bispecific antibodies become increasingly central to MM treatment, understanding the mechanisms of immune evasion and resistance in MM is ever more important. In the premalignant phase there is evidence of MGUS-specific T-cells[8]. Recent studies using mass cytometry and single-cell transcriptomics, have revealed that the bone marrow in MGUS is characterized by extensive transcriptional and immunophenotypic shifts across a variety of immune cells, encompassing T, B, NK, and myeloid lineages[9-11]. These findings collectively indicate that as early as the MGUS stage, there is evidence of altered or dysfunctional profiles within T, B, and NK cells, accompanied by pro-inflammatory changes within myeloid cells. Furthermore, these inflammatory alterations are not confined to the myeloid compartment but extend into the stromal cells, suggesting a broader disruption of the bone marrow microenvironment early in the disease course[12]. Human MM cells have developed various strategies to escape immune surveillance, including upregulations of checkpoint inhibitors such as PD-L1[13], secretion of immunosuppressive cytokines such as IL6, IL10, TGF-beta, IL18[14-16], and alteration of the tumor microenvironment to exclude immune effector cells.

Mouse models of MM must incorporate these immunological aspects to accurately reflect the human disease. This is particularly important in the context of studying T-cell-based therapies, where the interaction between the immune system and the tumor must be faithfully modeled to optimize therapeutic efficacy and manage immune-related toxicities.

5. Therapeutic Fidelity, Resistance and Relapse

Many classes of drugs have been evaluated in phase 1/2 clinical trials for the treatment of MM based on promising pre-clinical studies in vitro and in xenograft mouse studies. However, only a relatively few classes of drugs have been approved to treat patients (alkylators, glucocorticoids, proteasome inhibitors, IMiDs, Selinexor, monoclonal antibodies, bispecific antibodies and CAR-T). A faithful mouse model that more accurately reflects the therapeutic responses in patients can help prioritize novel therapeutics for evaluation in clinical trials[17]. One of the defining characteristics of MM is its capacity to develop resistance to therapy, often leading to relapse after an initial response. This resistance is driven by both intrinsic factors such as transcriptional plasticity[18, 19] and less frequently mutations in the proteasome or immunomodulatory pathways[20, 21], and extrinsic factors related to the bone marrow microenvironment. Mouse models should simulate these mechanisms of resistance to allow for the preclinical testing of therapies designed to overcome them. Models that replicate the gradual accumulation of genetic alterations and the protective effects of the microenvironment are critical for evaluating new drugs and combination strategies aimed at preventing or delaying relapse.

6. Temporal Disease Progression

MM typically follows a prolonged course, with progression from asymptomatic MGUS through SMM and ultimately to symptomatic disease. This temporal progression is punctuated by periods of stability and rapid progression, often in response to environmental or genetic triggers. A successful mouse model should replicate this temporal aspect, allowing researchers to study the triggers of disease progression and identify potential early intervention strategies. This is particularly important in understanding how MM evolves over time and how early therapeutic intervention might alter the natural history of the disease.

Pristane-induced and spontaneous plasmacytomas in Balb/c mice

The serendipitous discovery that Balb/c mice with plastic objects embedded intraperitoneally develop plasmacytomas (PCT) in the late 1950s[22], laid the foundation for subsequent studies of PCT pathogenesis. Building on this work, Mike Potter extensively characterized mineral oil-induced PCT in Balb/c mice, demonstrating that the chronic inflammation and reactive tissue, called oil granuloma, composed of macrophages, eosinophils, neutrophils, lymphocytes and plasma cells on the outer layer, creates conditions favorable for the expansion of IgA-producing plasma cells. Notably, PCT generally fail to proliferate in vitro or successfully engraft in Balb/c mice that have not been pre-conditioned with pristane, a finding that guided efforts to identify factors supporting PCT growth[23].

The search for these factors led to the simultaneous identification of IL-6 by several groups[24], highlighting its critical role in PCT expansion. Consistently, IL6-null mice are fully resistant to pristane-induced PCT development. Interestingly, treatment of mice with anti-inflammatory agents, such as NSAIDs and corticosteroids, can inhibit PCT development, although growth inhibition of established tumors is not usually observed. More rapid PCT formation is observed following ectopic oncogene expression (viral or transgenic) of Myc, Bcl2, BclxL, Raf, HRas. Additionally, the over-expression of viral abl, IL-6, or a constitutively active IL-6 receptor (Gp130) induces spontaneous PCT formation in the absence of mineral oil[23, 25].

Induction of mineral-oil induced PCT is highly strain specific, with Balb/c and to a lesser extent, NZB, being by the most susceptible, with this susceptibility linked to a hypoactive mTOR allele[26] and in Balb/c to two missense germline polymorphisms in Cdkn2a[27] and a promoter base substitution in p16INK4a[28]. Notably, PCTs are characterized by a chromosome translocation of MYC to an immunoglobulin loci, and the cloning of these translocation breakpoints led to the identification of the MYC oncogene[29]. Balb/c PCT cell lines (e.g., MPC11, MOPC315, XRPC24), many of which have acquired IL-6 independence and can be grown in vitro, laid the foundation for the development of hybridomas and the production of monoclonal antibodies. While PCTs are occasionally associated with lytic bone lesions in Balb/c recipients[30], these tumors more commonly engraft at extramedullary sites. A notable exception is the MOPC315.BM cell line, derived through repeated intravenous transplantation of bone marrow-localized MOPC315 cells. This line homes in vivo to the spleen and bone marrow, where it causes bone lesions and can be cultured in vitro, facilitating genetic manipulation[31]. One limitation of this line is the low level of immunoglobulin transcription (~1% of transcriptome) making it hard to follow disease burden using serum protein electrophoresis and limiting the modeling of the effects of high protein secretion that is a hallmark of myeloma (Figure 1).

Figure 1. Gene expression characteristics of mouse myeloma model.

RNAseq on plasma cells harvested from mice with normal plasma cells (NPC, gray), monoclonal gammopathy (MG, blue), multiple myeloma (MM, green), MM transplant recipients (R, yellow), or myeloma cell lines grown in vitro (V, pale yellow). Shown is the percent of transcriptome devoted to immunoglobulins, a gene proliferation index (GPi), NFkB index (NFkBi), percent of SNV in the rearranged immunoglobulin variable region gene, and the immunoglobulin heavy chain isotype distribution, calculated using the most abundant isotype identified in each sample.

Of the many genetically engineered models of MM in Balb/c summarized in Table 1, the one that appears to have the most relevance today is the IL6Myc model developed by Siegfried Janz[32, 33]. In this model a human IL6 transgene driven by the MHC H2-L promoter is crossed with a mouse in which Myc is inserted adjacent to the heavy chain intronic enhancer (Eμ) both in a Balb/c background. The mice develop a mix of plasmacytic and plasmablastic tumors in mesenteric lymph nodes and spleen with full penetrance and short latency (3 months). The tumors are mostly IgG or IgA, devote about 18% of the transcriptome to immunoglobulins, and have ~2.5% somatic hypermutation of the immunoglobulin genes. Five cell lines that grow in vitro have been established (IL6Myc-1 to 5). These all share the same VDJ rearrangement, devote 7% of the transcriptome to immunoglobulin and have 3.8% somatic hypermutation (Figure 1). The key characteristics of Balb/c PCTs are summarized in Table 1.

Table 1.

Characteristics of syngeneic murine models of multiple myeloma

Model	Strain	Myc activation	Tissue origin	Anti-apoptotic factor	Latency
Pristane PCT [23]	Balb/c	Ig Tx	Oil granuloma	IL-6	Long
IL6-Tg [63]	Balb/c	Ig Tx	LN	IL-6	Long
IL6Myc Tg [32]	Balb/c	Tg	LN/SPL	IL-6	Short
Bcl-2/Bcl-xL Tg [64,65]	Balb/c	Ig Tx	Peritoneum	BCL2/BCL-xL	Long
Bcl-xL/Myc Tg [66]	C57Bl/6.FVBN	Tg	BM/SPL/LV	BCL-xL	Short
Eμ-v-abl Tg [67]	C57Bl/6.SJL	Ig Tx	LN/BM	ABL	Long
ABL/MYC retrovirus [68]	Multiple	Retrovirus	GI/LN/SPL	ABL	Short
5T [69]	C57Bl/KaLwRij	No	BM	?	Long
Vk*MYC Tg [6]	C57Bl/6	Tg	BM	NFkB	Long
Vk*MYC x EuBCL2 Tg [6]	C57Bl/6	Tg	BM/SPL	BCL2	Medium
Vk*MYC x miR15a/16-1 null [34]	C57Bl/6	Tg	BM/SPL	BCL2	Medium
Vk*MYC x NrasQ61R [60]	C57Bl/6	Tg	BM/SPL/LV	NRAS	Medium
Nsd2/Ikk2ca/IgG1-CRE [4]	C57Bl/6	No	BM/SPL	NFkB	Long
Ccnd1/Ikk2ca/IgG1-CRE [4]	C57Bl/6	No	BM/SPL	NFkB	Long
BCL2/Ikk2ca/IgG1-CRE [5]	C57Bl/6.129S	Ig Tx	BM/SPL	BCL2/NFkB	Medium
MYC/Ikk2ca/IgG1-CRE [5]	C57Bl/6.129S	Tg	BM/SPL	NFkB	Medium

PCT = plasmacytoma; Tg = transgene; Ig = immunoglobulin; Tx = translocation; LN = lymph node; SPL = spleen; BM = bone marrow; GI = gastro intestinal; LV = liver.

5T model in C57Bl/KaLwRij mice

Jiri Radl identified the propensity of the C57Bl/Ka mouse strain to spontaneously develop a MG with advancing age, a phenomenon not observed in other strains such as Balb/c, CH3, or CBA. Subsequent studies determined that the more commonly used C57Bl/6 mouse strain develops a similar level of MG with age[34]. This gammopathy closely resembles the human condition of MG of undetermined significance (MGUS), persisting at a stable low level. An aggressive bone marrow localized MM was noted in 0.5% of two-year old C57Bl/KaLwRij, a subline of C57Bl used in the Netherlands, although a relationship to the preceding gammopathy was not established. Radl capitalized on this rare event by passaging in vivo the MM cells, with the fifth passage giving rise to stable MM cell lines (called 5T as a result). This is a remarkable feat, as we have not been successful at engrafting any of the rare progressive gammopathies from wildtype C57Bl/6 mice into wildtype, non-irradiated C57Bl/6 recipients (Bergsagel & Chesi, personal communication). Over the years, several distinct MM lines, were established by Radl, with 5T2 and 5T33 the most thoroughly characterized. Following transplantation 5T2 is much more indolent, with MM cells restricted to the BM and a serum M-spike detectable starting at 8 weeks. In contrast 5T33 MM cells are detected in the BM, spleen, liver with a serum M-spike at 3 weeks[35]. 5T2, but not 5T33, reliably produces bone lesions that can be detected and followed by standard mammography-type radiography[35-37]. Two sublines of 5T33 adapted for growth in vitro, 5T33vt and 5TGM1 are the most commonly used 5T MM lines[38-40].

All 5T MM lines secrete IgG, exhibit homing to the bone marrow, and cause osteolytic lesions in mice. Despite these similarities to human MM, the 5T MM lines have not undergone somatic hypermutation of immunoglobulin genes, a notable divergence from typical human disease. Between 4-10% of the transcriptome is devoted to immunoglobulins. A unique feature compared to other mouse models of MM is the lack of activation of Myc, suggesting alternative pathways of progression in this model. A genomic analysis identified some shared progression events with human MM, including Rb1 mutation in 5T2, and Trp53 and Kdm6a mutation in 5TGM1[39].

Because Radl primarily conducted his research using the C57Bl/KaLwRij strain, the broader predisposition of C57Bl/6 mice to develop MG was initially underappreciated. Genomic studies comparing C57Bl/KaLwRij to its related C57Bl/6 strain revealed a biallelic deletion of the Samsn1 gene in KaLwRij mice. Early work suggested that Samsn1 deficiency contributed to the susceptibility of KaLwRij mice to MM, as reintroducing Samsn1 into 5TGM1 cells suppressed their growth in vivo[41, 42]. However, subsequent findings revealed that Samsn1-expressing 5TGM1 cells were still able to engraft in Samsn1 wild-type C57Bl/6 mice, implying that the failure to engraft in KaLwRij mice was caused by immune rejection rather than a direct role of Samsn1 in tumor cell biology[43]. Despite these complexities, mice from C57Bl/6 and C57Bl/KaLwRij strains both share the propensity to spontaneously develop MG, and lytic bone lesions upon engraftment with MM cells. The specific characteristics of the tumor cells themselves, as opposed to host factors, presumably dictate the development of bone disease.

Over the years, the 5T MM model has been a valuable tool for studying plasma cell homing to the bone marrow, understanding MM cell dormancy, elucidating the mechanisms underlying bone disease, and investigating novel therapeutic approaches[39, 44-48]. Nevertheless, the model's reliance on the specific KaLwRij strain limits its utility, as tumor engraftment is not feasible in other genetically modified, immunocompetent mouse models. Furthermore, the great majority of the work has been done in a single cell line (5T33/5TGM1), limiting the ability to confirm the observations in an independent cell line.

Vk*MYC

We developed the Vk*MYC mouse model by introducing a mutated, inactive human MYC transgene under the control of the kappa light chain promoter/enhancer into a pure C57BL/6 background. This transgene is sporadically activated by somatic mutations driven by activation-induced cytidine deaminase (AID) within germinal center B cells that revert an in frame stop codon in MYC. The Vk*MYC mice exhibit a slow yet progressive clonal expansion of isotype-switched, somatically hypermutated plasma cells confined to the bone marrow, closely mimicking the pathophysiology of MM. This clonal proliferation is quantifiable over time through serum protein electrophoresis identification of a MG. Spread outside of the bone marrow, primarily in the spleen, is observed in approximately 30% of the mice. Furthermore, about 15% of the mice also develops Burkitt’s lymphoma with age, always associated with spontaneously acquired Trp53 inactivation[6]. While the myeloma cells from this model are generally challenging to propagate in vitro, in 20% of cases they can engraft into non-irradiated wild type C57BL/6 mice, where they lead to a more aggressive, extramedullary form of the disease. More recently, cell lines that grow in vitro have also been generated[2]. Interestingly the development of MM is strain specific as it is not seen when the Vk*MYC transgene is backcrossed onto a Balb/c background[49]. As the Balb/c, unlike C57Bl, strain does not spontaneously develop MG it is tempting to speculate that the germinal center activation of MYC in Vk*MYC acts as a secondary genetic event causing the progression of MG to MM, as in human MM.

Interestingly, immunizing Vk*MYC mice with NP-CGG, a T cell-dependent antigen, accelerates the expansion of NP-specific plasma cells, leading to shorter survival[6]. However, it remains unclear whether this is due to chronic antigen stimulation, or the heightened inflammation associated with the use of complete Freund’s adjuvant. The long median survival (two years) of Vk*MYC mice provides an accurate model for studying the progression of smoldering MM (SMM) to symptomatic disease, including myeloma defining features such as anemia, osteolytic lesions, and kidney disease. The indolent disease course, which may be considered a limitation, does allow for the sporadic acquisition and selection of secondary genetic events necessary for disease progression. Genomic profiling of many Vk*MYC myelomas has revealed significant tumor heterogeneity between individual tumors, associated with an APOBEC mutational signature, and a wide variety of mutations leading to activation of the NFkB and RAS/mTOR signaling pathways. A relatively frequent mechanism of mutagenesis is mediated by spontaneous insertion of endogenous Intracisternal A Particle retroviruses activating Map3k14 and Il6 and inactivating Ncor1, providing a relatively unique method to identify pathways critical for MM progression in this model. Nearly sixty Vk*MYC lines that can be passaged in vivo by transplantation and over twenty that grow in vitro have been thoroughly characterized[2]. The de novo tumors devote about 30% of the transcriptome to immunoglobulin, transplantable tumors about 20%, and in vitro cell lines about 5-10% (Figure 1).

The Vk*MYC model has demonstrated a high level of biological fidelity, making it particularly valuable for evaluating anti-myeloma therapies. The model's predictive accuracy for clinical drug activity has been rigorously validated: of the twelve drug classes of drugs active in this system, nine have demonstrated clinical efficacy, yielding a positive predictive value of 75%. Furthermore, all twelve drug classes that were ineffective in Vk*MYC MM are inactive in clinical settings, providing a negative predictive value of 100%[17, 50]. It is important to note that immunomodulatory drugs (IMiDs) are ineffective in rodent models due to species-specific amino acid variation in murine Crbn that impairs drug binding. To overcome this, a transgenic Vk*MYC line expressing human CRBN was created (Vk*MYC^hCRBN), rendering these mice sensitive to IMiDs. By transplanting Vk*MYC MM with or without hCRBN into recipient mice with or without hCRBN, it has been possible to disentangle the tumor-intrinsic versus extrinsic activity of IMiDs on MM growth[19].

The Vk*MYC model has significantly advanced our understanding of MM biology, including the role of IL-17-producing microbiota in disease progression[51], the role of MM-niche derived IL-18 in myeloid derived suppressor cell immune suppression[14], the potential of TIGIT inhibition for controlling MM, particularly in the context of autologous stem cell transplantation[52, 53], the identification of factors that determine bispecific antibody efficacy[54-57]. Moreover, the immunostimulatory effects of SMAC mimetics have been explored using this model[58]. A line that expresses GFP (Vk14451) has been used to study in vivo interaction of MM cells with the microenvironment[15, 58, 59]. The Vk*MYC transgene contains LoxP sites flanking the kappa 3’-enhancer, enabling conditional knockout of MYC expression via Cre mediated recombination, causing rapid tumor response. A subsequent transgene, Vk*MYCDLox, without LoxP sites, has been developed that can be used with Cre expressing mice, with both models phenotypically indistinguishable and available through the MMRCC repository.

Efforts to accelerate MM development by introducing additional oncogenes or deleting tumor suppressor genes in Vk*MYC mice often result in the emergence of Burkitt’s lymphoma, consistent with the notion that MYC activation in germinal center B cells, when tolerated, drives Burkitt’s lymphoma, though a few exceptions have been noted. One such exception is the cross between Vk*MYC and the EμBCL2 transgenic mouse, where the resulting MM cells become independent of the bone marrow microenvironment, proliferating in extramedullary sites and significantly shortening survival[6]. The Mir15a/Mir16-1 gene cluster, frequently deleted in MM patients with chromosome 13 monosomy, plays a critical role in disease progression. Mice heterozygous or null for Mir15a/Mir16-1, when crossed with Vk*MYC mice, exhibit accelerated disease, with a decrease in median survival from 97 to 90 or 68 weeks in the heterozygous or homozygous mice respectively[34]. Finally, the high frequency of RAS pathway mutations in MM has been modeled by crossing Vk*MYC mice with NRasQ61R stopF, conditionally activating Nras in germinal center B cells by crossing with Cγ1-Cre mice. These mice develop an aggressive form of MM with widespread plasmacytosis within the bone marrow, spleen, lymph nodes, liver and kidneys. Unfortunately, the Vk*MYC transgene used contains loxP sites and is deleted following activation of Cre, complicating the interpretation of the results. (Table 1)[60, 61].

Nsd2/Ikk2ca and Ccnd1/Ikk2ca x Cγ1-Cre

The two most common chromosome translocations in MM are the t(4;14) and t(11;14). To model these, conditional alleles of Nsd2 stopF or Ccnd1 stopF were inserted into the Rosa26 locus. To further mimic the molecular characteristics of human MM, the mice were crossed with Rosa26 Ikk2ca stopF mice, which harbor a constitutively active mutant of Ikk2, designed to drive constitutive NFkB activation. Through the introduction of Cγ1-Cre, transgene expression was selectively induced in germinal center B cells, with germinal center formation stimulated by immunization with NP-CGG. All the studies were performed in pure C57Bl/6 mice. Interestingly, only double transgenic Nsd2/Ikk2ca mice and Ccnd1/Ikk2ca mice, but not the single transgenic counterparts, progressed to disease endpoints between 72-97 and 62-91 weeks of age respectively (overall survival was not reported). These mice exhibited a progressive expansion of oligoclonal CD138+TACI+ plasma cells (PCs) in both the spleen and bone marrow, accompanied by the characteristic hallmarks of MM, including the detection of M-spikes via serum protein electrophoresis and organ damage consistent with disease.

Tumor propagation was achieved in both strains by serial transplantation of splenocytes into immunodeficient hosts, although, on occasion, transgenic B cells outcompeted plasma cells for engraftment. The ability of these plasma cell tumors to engraft in immunocompetent recipients remains an open question. Despite the histological similarities between the tumors of the Nsd2/Ikk2ca and Ccnd1/Ikk2ca mice, gene expression analysis revealed clear segregation of these tumors by principal component analysis, underscoring distinct transcriptional profiles. Notably, these profiles showed significant similarity to human myeloma, validating the biological relevance of the model. However, the absence of somatic hypermutation in immunoglobulin genes in these tumors suggests that the early constitutive activation of NFkB drives the premature exit of B cells from the germinal center, bypassing the typical maturation processes, with their associated off target effects on oncogenes, and neoantigens that are likely critical for MM pathogenesis.

MYC/Ikk2ca and BCL2/Ikk2ca x Cγ1-Cre

Using the same Cγ1-Cre and Ikk2ca mice, two additional models using either a MYC stopF (MIcγ1) or an EμBCL2 (BIcγ1) transgene have been developed. Upon immunization with sheep red blood cells to induce germinal center formation, both crosses rapidly developed widespread plasmacytosis and the clinical hallmarks of myeloma. This resulted in a significantly shortened overall survival, with MIcγ1 mice surviving an average of 208 days and BIcγ1 mice surviving 296 days. In addition to MYC and NFkB, RAS and TP53 are common pathways involved in MM progression. Crosses to mice strains with either conditional activation of KrasG12D or heterozygous Trp53 deletion markedly decreased the overall survival in the MIcγ1 but not the BIcγ1 mice. Unfortunately, the phenotype, serum protein electrophoresis and gene expression profiles of the crosses to MIcγ1 are not presented. The authors report 6 in vitro MM cell lines derived from Trp53-BIcγ1 mice, one of which, MM5080, is able to engraft in C57Bl/6 mice after intravenous injection. MM5080 has undergone deletion of the wildtype Trp53 allele resulting in Trp53 inactivation and has acquired an IgL-Myc translocation. Others have noted that the cross of KrasG12D with Cγ1-Cre often results in T-cell lymphoma and other non-plasma cell tumors[62], which has also been our unpublished experience. The addition of EμCcnd1, EμMaf alleles did not significantly alter disease progression, suggesting that while these genes may play a critical role in MM initiation, they are not as influential in driving disease progression. However, crossing Ikk2ca mice with human NSD2 StopF mice did result in MM development, consistent with findings reported by Winkler and colleagues, indicating that this genetic combination may act synergistically in promoting myeloma progression. Interestingly, the two transgenic lines exhibited distinct disease courses. The BIcγ1 mice displayed a gradual disease progression over 4-6 months that mirrored the transition from MGUS to SMM and ultimately to MM associated with the acquisition of Ig-Myc translocation, reflecting the typical human disease trajectory. In contrast, the MIcγ1 mice progressed rapidly over 2 months from an MGUS-like phase to full-blown MM. The genetic event(s) associated with this latter progression remain to be determined. In both models, disease progression was associated with increased infiltration of natural killer (NK) cells and effector CD8+ T cells, alongside a rise in CD4 Tregs. Notably, the ratio between these two cell types emerged as a key determinant of the mice's response to immune checkpoint inhibition, suggesting that immune dynamics within the tumor microenvironment may significantly influence therapeutic outcomes. Like the Nsd2 and Ccnd1 crosses to Ikk2ca-Cγ1 cre, the MYC and BCL2 crosses do not generate MM with somatic mutation of the immunoglobulin genes limiting their ability to model the role of the germinal center reaction in MM pathogenesis. Despite the robust disease phenotypes observed in these models, both MIcγ1 and BIcγ1 mice have been maintained on a mixed C57BL/6-129Sv genetic background, which has precluded transplantation experiments and limits the utility of the cell lines that have been established to further investigate the tumor's behavior in different host settings.

Conclusion

Immunocompetent mouse models have been pivotal in advancing research in cancer biology and immunotherapy. These models are commonly employed to evaluate the activity, safety, toxicity, and effectiveness of new therapeutic candidates, serving as essential tools for generating preclinical data that inform future clinical studies. Despite these models' significant contributions, no single mouse model has fully recapitulated the full spectrum of MM due to the complex heterogeneity of the disease and its dependence on the bone marrow microenvironment, making it challenging to replicate in vivo.

There are a variety of immunocompetent models of MM with their own advantages and disadvantages. One of the primary considerations when selecting a model is the choice between de novo and transplantable models. De novo models are more accurate representations of the tumor-microenvironment interactions over time, closely mimicking the spontaneous cell transformation and genetic changes seen in human cancers. The Vk*MYC model is particularly notable in this regard, as it captures the sporadic mutations occurring in a small number of cells. These models are especially useful for studying genetic predispositions, environmental influences (such as diet and microbiome), and the effects of therapies in an intact immune system that has adapted over time to spontaneous tumor development. However, de novo models come with challenges, including higher costs and the lengthy time required for breeding, genotyping, and aging mice for experimental use, often extending beyond a year. While additional genetic modifications can speed up tumor development, they may also affect the natural progression of the disease, potentially altering experimental outcomes.

In contrast, transplantable models offer several practical advantages, including more rapid and predictable MM development and the ability to develop cohorts of mice with the same tumor for controlled experiments. One key difference among these models lies in the mouse strain from which they are derived. For instance, the C57Bl/6 strain is known for its stronger Th1 immune response, while the Balb/c strain is more inclined toward a Th2 response. These immunological differences can significantly influence experimental results. Among the transplantable models, several Balb/c plasmacytoma lines exist, with the MOPC315.BM line standing out as one of the most reliable models for studying human disease. The 5TGM1 line is widely employed to investigate interactions between MM and bone. Furthermore, the Vk*MYC model has given rise to over 50 transplantable lines, each with varying levels of aggressiveness and unique genetic features. Many of these lines express human cereblon (CRBN), making them particularly valuable for research on immunomodulatory drugs (IMiDs) and their mechanisms of action.

There are four new genetically engineered models based on germinal center activation of NFkB together with Nsd2, Ccnd1, Myc or Bcl2. The first two have the advantage that they model the t(4;14) and t(11;14) chromosome translocations important in human MM, while the latter two have the advantage that they induce much more rapid tumor development. Unfortunately, all these models lack somatic mutation of the immunoglobulin genes, likely because early activation of NFkB stimulates premature exit from the germinal center. This is problematic because ectopic mutagenesis generated during the germinal center reaction is fundamental to oncogene activation, tumor suppressor gene inactivation, and neo-antigen development in MM. Future studies should avoid cγ1-cre activation of Ikk2ca, perhaps seeking methods to induce activation of NFkB after clonal expansion selected in the germinal center or using other oncogenes/tumor suppressor genes activated by cγ1-cre to accelerate tumorigenesis.

Unfortunately, no mouse strain is currently available to model hyperdiploid (HRD) human MM, which occurs in approximately 50% of patients and is characterized by trisomies of odd number chromosomes. As the molecular mechanisms driving HRD MM remain largely unknown, the creation of a HRD MM strain remains a challenge for the future.

Table 2.

Selected murine myeloma cell lines

Name	Strain	In vitro growth	Ig transcription (%)	SHM (%)	Engraftment (%)	OS (Days)	Known drivers
Vk12598 [2]	C57Bl/6	No	22	3.6	70%	44	MYC, Ncor1, Kdm6a, Egr1, Il6st
Vk12653 [2]	C57Bl/6	No	26	1.3	100%	69	MYC, Kdm6a, Map3k14, Cdkn2a, Rb1
Vk14451GFP [56,58,59]	C57Bl/6	No	22	1.2	73%	160	MYC, Cyld, Dusp2, Tent5c
Vk27181 [54]	C57Bl/6	No	35	0	68%	75	MYC, Kdm6a, Pou2af1, Pten, Map3k14
Vk29790hCRBN [19,54]	C57Bl/6	No	30	2.4	75%	188	MYC, NrasQ61K
Vk32908hCRBN [54]	C57Bl/6	No	8	2.7	100%	50	MYC, Ncor1, Dusp2, Mcl1
Vk22284CRE-ERT2 [2]	C57Bl/6	No	20	7.1	70%	100	MYC, Pten, Ltbr
Vk12598VITRO [2]	C57Bl/6	Yes	7	3.6	40%	84	MYC, Ncor1, Kdm6a, Egr1, Il6st, Dusp2
Vk12653VITRO [2]	C57Bl/6	Yes	16	1.4	50%	130	MYC, Kdm6a, Map3k14, Cdkn2a, Rb1
VQ-D1 [60,61,70]	C57Bl/6	No	17	1.4	100%	80	MYC, NrasQ61R, Sp140
VQ-D2-4938 [61,70]	C57Bl/6	Yes		2.7			MYC, NrasQ61R
MM5080 [5]	C57Bl/6.129S	Yes		0		31	Trp53, Bcl2, Ikk2, IgL-Myc
MOPC315.BM [31]	Balb/c	Yes	1.4	2.9	100%	30	IgH-MYC
5T2 [37,39]	C57Bl/KaLwRij	No		0	100%	90	Rb1
5T3vv [39,40]	C57Bl/KaLwRij	No		0	100%	45	Trp53, Nf1
5T3vt [40]	C57Bl/KaLwRij	Yes		0	100%	45
5TGM1 [39,71]	C57Bl/KaLwRij	Yes	4-10	0	100%	45	Trp53, Kdm6a

GFP – express green fluorescent protein

hCRBN – express human CRBN and respond to IMiDs

CRE-ERT2 – express a tamoxifen-inducible CRE resulting in deletion of Vk*MYC transgene

TRE-OmoMyc – express a tetracycline-inducible OmoMyc that inhibits Myc dimerization