Novel therapeutic strategies for multiple myeloma

Abstract

Multiple Myeloma (MM) is a plasma cell malignancy which remains incurable despite of the recent emergence of multiple novel agents. Importantly, recent genetic and molecular analyses have revealed the complexity and heterogeneity of this disease, highlighting the need for therapeutic strategies to eliminate all the clones. Moreover, the bone marrow microenvironment, including stromal cells and immune cells, plays a central role in MM pathogenesis, promoting tumor cell growth, survival, and drug resistance. New classes of agents including proteasome inhibitors, immunomodulatory drugs, monoclonal antibodies, and histone deacetylase inhibitors have shown remarkable efficacy; however, novel therapeutic approaches are still urgently needed to further improve patient outcome. In this review, we discuss the recent advances and future strategies to ultimately develop MM therapies with curative potential.

Introduction

Multiple myeloma (MM) is characterized by the clonal proliferation of malignant plasma cells in the bone marrow (BM), lytic bone lesions, and immunodeficiency, associated with monoclonal protein in the blood and/or urine. It accounts for 1% of all cancers and more than 10% of all hematological malignancies. In spite of recent advances in treatment including high-dose therapy and novel agents such as bortezomib, thalidomide, and lenalidomide, MM remains fatal due to development of drug resistance in the context of BM microenvironment [1-4]. To overcome this drug resistance, a number of therapeutic approaches have been developed in recent years [5]. For example, new-generation proteasome inhibitors including carfilzomib, ixazomib, and marizomib are active even in the setting of bortezomib-resistant MM. Pomalidomide, a next-generation immunomodulatory drug, has shown activity even in 17p (p53) deleted MM [6]. Excitingly, monoclonal antibodies such as elotuzumab (anti-SLAMF7, also known as CS1) and daratumumab (anti-CD38) show promising clinical efficacy, especially in combination with lenalidomide. In this review, we focus on new therapeutic approaches to increase endoplasmic reticulum stress, target signal transduction, trigger epigenetic modulation, as well as induce anti-MM immune responses in the BM niche. The overview of novel therapeutic approaches is shown in Figure 1.

Figure 1. The overview of novel therapeutic approaches for multiple myeloma (MM).

The scheme of novel therapeutic targets and treatment options (#1–13) discussed in this review article are highlighted. The specific treatment options and representative drugs are also shown below.

1: IRE1α inhibitors (MKC-3946 [42], STF-083010 [45]), 2: HSP90 inhibitors (17-AAG, TAS-116 [30]), 3: PI3K inhibitors (CAL-101 [76]), 4: Akt inhibitors (perifosine, afuresertib [74], TAS-117 [23], MK-2206 [72]), 5: mTOR inhibitors (rapamycin, everolimus, temsirolimus), 6: MEK inhibitors (selumetinib), 7: NF-κB inhibitors (PBS-1086 [87]), 8: HDAC inhibitors (vorinostat, panobinostat, ricolinostat [107], BG45 [112]), 9: EZH2 inhibitors (UNC1999 [120]), 10: synthetic miRNAs (miR-29b [123], miR-34a [124]), 11: Bromodomain inhibitors (JQ1 [128]), 12: PD-1/PD-L1 antibodies (CT-011 [143]), 13: PDE5 inhibitors PIs (proteasome inhibitors): bortezomib, carfilzomib, ixazomib, marizomib IMiDs (immunomodulatory drugs): thalidomide, lenalidomide, pomalidomide Anti-SLAMF7 antibody: elotuzumab Anti-CD38 antibody: daratumumab

1. Targeting the unfolded protein response induced by endoplasmic reticulum stress

The endoplasmic reticulum (ER) is a cellular organelle responsible for gluconeogenesis, lipid synthesis, and Ca²⁺ storage. In the ER, secretory or membrane proteins are folded properly to form their functional structure. However, extracellular insults/stress such as low nutrients, hypoxia, or drugs can disrupt protein synthesis and folding, thereby inducing accumulation of misfolded proteins in the ER and resulting in increased ER stress. The unfolded protein response (UPR) is an adaptive response to ER stress condition by increasing biosynthetic capacity and decreasing the biosynthetic burden on the ER in order to maintain cellular homeostasis and cell survival [7, 8]. However, when the stress cannot be compensated by the UPR, apoptosis is then triggered as a terminal cellular response [9]. In general, activation of the UPR is initiated through three different ER transmembrane proteins and their downstream pathways: inositol-requiring enzyme 1α (IRE1α), protein kinase RNA (PKR)-like ER kinase (PERK), and activating transcription factor 6 (ATF6). During unstressed conditions, these proteins are inactivated by interacting with molecular chaperone immunoglobulin-heavy-chain-binding protein (BiP)/GRP78. However, when unfolded proteins accumulate in the ER, then BiP/GRP78 dissociates from these sensor proteins to prevent aberrant aggregation of the proteins, thereby triggering downstream UPR signaling [10].

During the UPR, IRE1α is oligomerized and autophosphorylated, followed by activation of its endoribonuclease domain and triggering of splicing of X-box binding protein 1 (XBP1) mRNA. More specifically, activated IRE1α endoribonuclease cleaves a 26 nucleotide intron from XBP1 mRNA, resulting in a translational frame-shift to turn unspliced XBP1 (XBP1u: inactive) into spliced XBP1 (XBP1s: active) [11]. XBP1 acts as a crucial transcription factor in the UPR, regulating genes responsible for protein folding and ER associated degradation (ERAD) to process misfolded proteins [12]. PERK is a serine/threonine kinase which phosphorylates eukaryotic translation-initiation factor 2α (eIF2α), leading to inhibition of the translation of new protein synthesis and thereby reducing protein overload in the ER [13]. In the UPR, ATF6 is transported to the Golgi apparatus and cleaved into active transcription factor regulating ER chaperones, including XBP1 [14]. Importantly, under prolonged and uncompensated stress conditions, the UPR triggers cellular apoptosis, also called terminal UPR. In this process, a pro-apoptotic transcription factor C/EBP homologous protein (CHOP), also known as GADD153, is induced via PERK and downstream ATF4 pathways, with downregulation of BCL2 followed by caspase-dependent apoptosis [15, 16].

Myeloma cells produce excess M proteins that cause high basal levels of ER stress and require strict ER quality control for protein synthesis. Therefore, targeting the UPR induced by ER stress represents a promising novel therapeutic strategy in MM [17, 18].

1.1 Ubiquitin-proteasome system

Bortezomib, a dipeptide boronic acid analogue mediating selective and reversible inhibition of the 26S proteasome, has dramatically changed the outcome of MM patients [19, 20]. Since a number of proteins are substrates of the proteasome, one of the major cytotoxic mechanisms induced by bortezomib is increased ER stress [21]. More specifically, the ubiquitin-proteasome pathway facilitates the removal of misfolded proteins accumulated in the ER by induction of the ERAD; conversely, bortezomib blocks this pathway and induces the terminal UPR, thereby leading to apoptosis via upregulation of CHOP. Following the remarkable success of bortezomib, different classes of proteasome inhibitors with reversible and/or irreversible inhibition of chymotrypsin-like, trypsin-like, and/or caspase-like activities, have been developed. All these agents inhibit activity of the 20S proteasome and have similar biologic impact in MM cells [22]. Indeed, we have confirmed that carfilzomib, a 20S proteasome inhibitor with irreversibly chymotrypsin-like activity , also triggers the ER stress response in MM cells [23].

1.2 Heat shock proteins

Heat shock proteins (HSP) were initially defined as a family of proteins induced by heat shock, that act as molecular chaperones to stabilize and/or correctly fold their client proteins [24]. Among these HSPs, HSP90 is the most well-investigated in many cancers, including MM. HSP90 is a highly conserved molecular chaperone which interacts with numerous number of client proteins with crucial roles in cancer pathogenesis. Therefore, HSP90 is an attractive therapeutic target in cancer [25]. In MM, HSP90 is abundantly expressed and mediates the anti-apoptotic effects conferred by the BM microenvironment by maintaining client proteins involved in critical signaling pathways including NF-κB, phosphatidylinositol 3-kinase (PI3K)/Akt, and RAS/ mitogen-activated protein kinase (MAPK) which mediate MM cell survival, drug resistance, and proliferation, respectively [26]. A geldanamycin analog 17-allylamino-17-demethoxy-geldanamycin (17-AAG) has been extensively studied in preclinical models of MM: it increases UPR components including CHOP protein, triggering terminal UPR followed by cell death via apoptosis [27]. Inhibition of HSP90 triggers ER stress both by preventing HSP90 from folding proteins properly via GRP94 in the ER and by destabilizing the ER-stress sensor IRE1α [28, 29]. Although HSP90 inhibitors demonstrate promising anti-tumor effects in preclinical studies, some clinical trials have been discontinued due to adverse effects [25]. Among these side effects, ocular toxicity is the dose-limiting factor, which should be avoided by next-generation HSP90 inhibitors. Indeed, Suzuki et al. have recently reported that a novel HSP90α/β selective inhibitor TAS-116 exhibits anti-MM activities in preclinical settings without ocular toxicity [30].

HSP70 inhibitors also show significant anti-MM activities [31, 32]. Since HSP70 is increased as a compensatory response to HSP90 inhibition, dual inhibition of HSP90 and 70 may be a promising therapeutic option [33]. Moreover, targeting upstream of HSP is also promising in MM therapy: inhibition of heat shock transcription factor 1 (HSF1), which regulates multiple HSPs including HSP90, 70, and 27, induces apoptosis in MM cells [34]. Interestingly, the inhibition of PI3K/Akt pathway downregulates HSF1 and downstream HSP70, sensitizing MM cells to HSP90 inhibition [35].

1.3 IRE1α-XBP pathway

Among the three cascades of the UPR, IRE1α-XBP pathway has been most extensively studied in cancers, including MM [36]. XBP1 is required for differentiation to normal plasma cells [37] and is overexpressed in human MM cells [38]. Moreover, knockdown of XBP1 sensitizes MM cells to cell stress-induced apoptosis [39], whereas overexpression of its spliced variant (XBP1s) has been used to generate a murine model of MM, further supporting its critical role in MM pathogenesis [40]. Based on these reports, the therapeutic potential of targeting XBP1 has been investigated in MM. For example, targeting XBP1 splicing by inhibiting IRE1α has triggered anti-tumor activities in preclinical studies [41-45]. We have demonstrated that an IRE1α RNase domain selective inhibitor MKC-3946 blocks XBP1 splicing, which is associated with modest MM cell growth inhibition in vitro. Importantly, however, it significantly inhibits tumor growth in a mouse xenograft model of human MM [42], suggesting that basal levels of ER stress may be higher in vivo in the tumor microenvironment due to hypoxia or low levels of nutrients. Moreover, MKC-3946 significantly enhances ER stress and cytotoxicity induced by bortezomib or 17-AAG in vitro; it also enhances tumor growth inhibition in combination with bortezomib in vivo in a xenograft murine model, suggesting potential clinical utility of combination treatment.

A recent study has demonstrated that BM cellular components such as BM stromal cells (BMSCs) and osteoclasts are dependent of XBP1 splicing. Since the BM microenvironment plays a crucial role in MM cell pathogenesis, modulation of XBP1s represents a novel therapeutic approach in MM in the context of the BM milieu [46]. Moreover, XBP1 may also be a therapeutic target in immunotherapy, since vaccination of smoldering MM patients with human leukocyte antigen (HLA)-A2(+) immunogenic peptides derived from XBP1 antigens can induce a tumor-specific cytotoxic T-cell response [47-49].

Contribution of basal level of XBP1 to the treatment response is still controversial. For example, lower XBP1s/u ratio is correlated with better response to thalidomide treatment [50]; in contrast, bortezomib is more effective in patients with high XBP1 levels [51, 52]. More recently, Leung-Hagesteijn et al. have reported that XBP1s is required for bortezomib-induced cytotoxicity in MM; and conversely, that the XBP1s-negative MM subset with arrested secretory maturation becomes resistant to bortezomib [53]. Although MM cells with high XBP1s are more sensitive to proteasome inhibitors due to high basal levels of ER stress, inhibition of XBP1 splicing can enhance response to bortezomib in this setting.

Regarding other ER stress sensor proteins, ATF6 and PERK, a previous study has reported that downregulation of either ATF6 or PERK triggers MM cell death, due to induction of autophagy especially by PERK silencing [54]. These results further confirm the UPR as a potential therapeutic target in MM.

2. Targeting signal transduction

The Interaction of MM cells with BMSCs stimulates secretion of IL-6, IGF-1, VEGF, SDF1α, TGFβ, HGF, and TNFα from both BMSCs and MM cells. These cytokines activate multiple signaling pathways in MM cells, including the phosphatidylinositol 3-kinase (PI3K)/Akt pathway, JAK/STAT3 pathway, RAS/RAF/MEK/ERK pathway, and NF-κB pathway, resulting in cell growth, anti-apoptosis, and drug resistance in MM cells [55, 56]. Therefore, the molecules in these signaling cascades represent therapeutic targets in MM.

2.1 PI3K/Akt pathway

PI3K/Akt pathway has been extensively studied in various types of cancers, including MM, and mediates tumor cell survival, growth, and drug resistance [57-60]. Among the three classes of PI3K, class I PI3K is composed of a catalytic subunit p110 and an adaptor/regulatory subunit p85, and may be the most relevant in MM pathogenesis. It is activated via upstream receptor tyrosine kinases by various growth factors and/or cytokines at the inner membrane, followed by phosphorylation and activation of downstream serine-threonine protein kinase Akt (also known as protein kinase B). Akt in turn phosphorylates and activates many substrates, including GSK3α/β and FKHR proteins, to facilitate cell proliferation, survival, and protection against apoptosis [57, 61, 62]. Akt also mediates activation of mammalian target of rapamycin (mTOR), a key regulator of cell metabolism and autophagy [63, 64]. Since the PI3K/Akt pathway is activated by many cytokines and cell adhesion to BMSCs, it plays a crucial role in MM cell survival and drug resistance in the context of the BM microenvironment [55, 56, 65, 66]. Therefore, targeting PI3K/Akt pathway is a promising therapeutic option in MM [59, 67].

Akt, a master regulator in this pathway, has been extensively investigated in both preclinical and clinical studies in MM. We have previously shown that an alkylphospholipid perifosine, which inhibits Akt activity [68, 69], shows anti-MM activities and enhances MM cytotoxicity induced by bortezomib in preclinical setting [70]; however, it has off-target effects [70, 71]. Clinical trials of perifosine in MM were discontinued due to limited clinical response. A recent study has shown the in-vitro efficacy of a selective Akt inhibitor MK-2206 [72]. Most recently, we have demonstrated that a selective and potent allosteric Akt inhibitor TAS-117 induces significant growth inhibition in MM cells in vitro and in vivo [23]. TAS-117 triggers both apoptosis and autophagy, as well as ER stress response. Importantly, TAS-117 enhances cytotoxicity induced by proteasome inhibitors (bortezomib and carfilzomib) by inhibition of proteasome inhibitor-induced Akt activation and induction of fatal ER stress, associated with increased CHOP expression. These results suggest cross talk between Akt pathway and unfolded protein response in MM cells. Indeed, previous studies have shown that Akt inhibition activates PERK pathway in the UPR [73]. Moreover, an ATP-competitive Akt inhibitor afuresertib exhibits safety and clinical activity in a phase I clinical trial [74], suggesting the clinical potential of selective Akt inhibitors in MM. In addition to direct effects on MM cells, Akt inhibition also affects BM cellular components. For example, we found that TAS-117 inhibits IL-6 secretion from BMSCs by modulating NF-κB activity [23]. Moreover, Akt inhibition blocks osteoclastogenesis and bone osteolysis, further supporting clinical utility of Akt inhibitors in MM patients with bone disease [75].

Small molecule PI3K inhibitors have also been developed [61]. Ikeda et al. have reported that the blockade of p110δ, a component of class I PI3K, triggers apoptosis and autophagy in vitro, as well as inhibits tumor growth in preclinical in-vivo models [76]. mTOR inhibitors, such as rapamycin and its analogues everolimus and temsirolimus, have also been investigated in the treatment of cancers, including MM. While these drugs inhibit only mTORC1 and lead to activation of Akt due to negative feedback of mTORC2 [77], dual inhibition of mTORC1 and mTORC2 in MM cells demonstrates significant growth inhibition [78]. Moreover, dual inhibition of PI3K/mTOR or insulin-like growth factor 1 receptor (IGF1R)/mTOR also triggers significant anti-tumor effects in MM [79, 80].

2.2 Other signaling pathways

RAS/RAF/MEK/ERK pathway plays a pivotal role in the pathogenesis of MM, and high frequency of mutations in NRAS, KRAS and BRAF genes have been reported [81]. Although small molecule inhibitors targeting this pathway (e.g. MEK inhibitor selumetinib) have been developed and show remarkable anti-tumor effects in preclinical settings, their clinical activities to date are limited [82]. Importantly, this pathway is upregulated by Akt inhibition, and dual inhibition of both pathways using Akt inhibitors and MEK inhibitors shows synergistic anti-tumor effects [70, 72, 83]. More recently, 3-phosphoinositide–dependent protein kinase 1 (PDPK1) inhibitor has been shown to induce growth inhibition and apoptosis in MM cells through blockade of both RSK2, a mediator of MAPK/ERK pathway, and Akt [84], further indicating potential clinical utility.

NF-κB signaling pathway also plays a key role in the pathogenesis of MM, by transcriptional regulation of genes involved in proliferation, survival, and drug resistance [85]. NF-κB activity in MM cells is mediated via both canonical and non-canonical pathways. Bortezomib was considered to acts as a canonical NF-κB inhibitor by blocking proteasomal degradation of IκBα; however, bortezomib activates constitutive canonical NF-κB activity in MM cells, and the combination of bortezomib with IKKβ inhibitor triggers synergistic MM cell growth inhibition [86]. Importantly, there is cross-talk signaling between canonical and non-canonical pathways: inhibition of non-canonical pathway activates canonical pathway and vice versa. Indeed, Fabre at al. reported that PBS-1086, a small molecule inhibitor which simultaneously blocks both canonical and non-canonical NF-κB pathways, induces significant MM cell growth inhibition, even in bortezomib-resistant cells [87].

3. Targeting epigenetic modulation

In the development of myeloma, stepwise gene mutations in cells are accumulated, resulting in genetic complexity and molecular heterogeneity. Importantly, chromosomal deletions or gains, translocations, as well as point mutations, are correlated with clinical features and prognosis in MM [81, 88]. Moreover “epigenetic modification”, defined as alteration of gene expression without alteration of DNA sequences, is also important in the pathogenesis of cancers, including MM [89, 90]. Specifically, epigenetic modifications are categorized into two major types: DNA methylation and histone modification. Indeed, many oncogenes and tumor-suppressor genes are affected by these epigenetic modifications. Therefore, epigenetic therapies are under evaluation in many cancers. In MM, aberrant DNA methylation with affected expression of specific genes is observed during disease development and progression [91, 92]. However, the mechanisms of aberrant DNA methylation in MM remain unclear; therefore, targeting DNA methylation has not yet been fully developed. In contrast, histone modifications have been characterized, and histone deacetylase inhibitors have already shown promising results in MM. Histones are major protein components of chromatin, where they complex with DNA to form the nucleosome. The lysine residues of histone tails can undergo various chemical modifications including acetylation, methylation, ubiquitination, phosphorylation, and sumoylation, which alter gene transcription [93].

3.1 Histone acetylation

Gene transcription is actively regulated by histone acetylation, which is strictly mediated by both histone acetyltransferases (HATs) and histone deacetylases (HDACs). Therefore, HDACs are an attractive therapeutic target in cancers including MM. HDACs consist of 4 classes according to their structure, localization, and function: class I (HDAC 1,2,3, and 8), class IIa (HDAC 4,5,7, and 9), class IIb (HDAC 6 and 10), class III (also known as sirtuins), and class IV (HDAC11) [94, 95]. Importantly, HDACs target not only histones, but also non-histone proteins in cancer cells [96]. These proteins relevant in MM pathogenesis include HSP90, p53, STAT3, β-catenin, Sp1, c-Myc, Rel-A, phosphatase and tensin homolog (PTEN), murine double minute 2 (MDM2), and tubulin [97]. In preclinical studies, non-selective HDAC inhibitors show anti-tumor activities. Class I and II HDAC inhibitors such as vorinostat (also known as SAHA) [98], MVP-LAQ824 (LAQ824) [99], panobinostat (LBH589) [100, 101], and belinostat (PXD101) [102] show anti-tumor effects in MM. A class I HDAC inhibitor romidepsin (FR901228, FK228) [103] also induces apoptosis in MM cells. Although non-selective HDAC inhibitors such as vorinostat combined with bortezomib have shown efficacy, the progression-free survival (PFS) advantage of combination versus bortezomib with placebo was less than one month due to attendant toxicities including fatigue, diarrhea, and thrombocytopenia [104]. Of note, panobinostat (LBH589) has been FDA approved for treatment of relapsed MM due to 4-month PFS advantage of panobinostat with bortezomib/dexamethasone versus bortezomib/dexamethasone alone [105]. .

To avoid the side effects due to broad inhibition by non-selective HDAC inhibitors, class- and/or isoform-selective HDAC inhibitors have recently been developed. Specifically, HDAC6 plays a crucial role in aggresome formation, an alternative system for degradation of ubiquitinated proteins in the lysosome, binding to misfolded ubiquitinated proteins on the one hand and to dynein motility complexes on the other to shuttle proteins to the aggresome. Therefore, HDAC6 inhibitors coupled with bortezomib can block ubiquitinated protein degradation in both proteasomes and lysosomes. Indeed, first-in-class HDAC6-selective inhibitor tubacin shows synergistic MM cytotoxicity with bortezomib, associated with significant accumulation of polyubiquitinated proteins and cell stress [106]. Based on this encouraging result, a selective HDAC6 inhibitor ricolinostat (ACY-1215) has subsequently been developed for clinical use. It shows synergistic anti-MM activities with both bortezomib and lenalidomide in preclinical setting [107, 108], and derived phase I/II clinical trials in MM are ongoing [109, 110]. The combination of HDAC6 inhibitors with the second-generation proteasome inhibitor carfilzomib shows even more potent in-vitro synergistic cytotoxicity than bortezomib, providing the rationale for clinical trials of this combination [111]. Finally, Minami et al. have shown that an HDAC3-selective inhibitor BG45, alone and in combination with bortezomib, triggers significant anti-MM activities both in vitro and in vivo [112].

3.2 Histone methylation

MM set domain (MMSET, also known as WHSC1/NSD2) is a histone methyltransferase overexpressed by t(4;14) subset (15%) of MM and associated with poor prognosis. MMSET mainly induces dimethylation of lysine 36 at histone H3 (H3K36me2), leading to active target gene transcription [113]. MMSET plays a crucial role in pathogenesis of t(4;14) MM by promoting cell proliferation through repression of miR-126* and subsequent enhancement of c-MYC protein levels [114, 115]. Therefore, MMSET is a therapeutic target in t(4;14) MM, and selective inhibitors are under development.

Enhancer of zeste homolog 2 (EZH2) is a histone-lysine N-methyltransferase, which also represents a novel therapeutic target in MM. EZH2 is the catalytic component of polycomb repressive complex 2 (PRC2), which triggers H3K27me3 to repress the transcription of target genes, thereby promoting growth in cancers, including lymphoma [116]. Moreover, EZH2 is overexpressed with progression from MGUS to MM [117]. Importantly, EZH2 knockdown leads to inhibition of MM cell growth [118]. In contrast, ubiquitously transcribed tetratricopeptide repeat X chromosome (UTX, also known as KDM6A), a histone demethylase which removes methyl groups at H3K27, is inactive due to mutation in a subset of MM cases [119], suggesting that increased H3K27me3 by UTX inactivation contributes to MM pathogenesis. Indeed, Rizq et al. have reported that an EZH2 inhibitor UNC1999 demonstrates both in-vitro and in-vivo anti-tumor activities in MM [120].

3.3 Other epigenetic targets

MicroRNAs (miRNAs) are a class of noncoding RNAs which regulate gene expression and are associated with oncogenic or tumor-suppressive functions [121]. Indeed, abnormal expression of various miRNAs contributes to pathogenesis of MM [122]. Among these miRNAs, miR-29b reduces global DNA methylation and inhibits cell cycle progression in MM, and synthetic miR-29b mimics exert in-vivo anti-MM activity in preclinical models [123]. miR-34a also acts as a negative regulator of MM cell growth, and systemic delivery of formulated miR-34a inhibits growth of MM xenografts in SCID mice [124]. Ongoing investigations are evaluating therapeutic systems to assure optimal delivery of miRNAs to MM cells [125].

Another possible epigenetic target in MM are bromodomains, which act as a reader of epigenetic marks to recognize acetylated lysine residues at histone tails, thereby leading to modulation of chromatin structure and target gene expression [126]. A selective small molecule inhibitor of bromodomain protein BRD4, JQ1 [127], has demonstrated potent anti-MM activities both in vitro and in vivo, associated with downregulation of c-Myc [128], and clinical trials of BRD4 inhibitors in MM are ongoing.

4. Targeting immune system in the bone marrow microenvironment

Immune surveillance system is impaired in MM patients, providing the impetus to develop strategies to restore host anti-MM immunity [129]. Allogeneic stem cell transplantation (AlloSCT) can provide a potent donor graft-versus-MM effect and lead to long-term disease free survival. However, due to treatment-related morbidity and mortality, alloSCT is utilized only in the context of a clinical trial to avoid attendant toxicity while exploiting graft-versus-MM effect [130]. To date, the most successful immune-based approach in MM is immunomodulatory drugs (IMiDs) thalidomide, lenalidomide, and pomalidomide. A recent study has shown that the direct anti-MM activity of IMiDs is mediated via binding to E3 ubiquitin ligase cereblon, triggering proteasomal degradation of Ikaros family zinc finger proteins 1 and 3 (IKZF1 and IKZF3) [131]. The immunomodulatory effects by IMiDs include activation of natural killer (NK) or NKT cells, stimulation of both CD4+ and CD8+ T cells, and inhibition of regulatory T cells [132, 133]. Interestingly, IMiDs induce epigenetic modification of suppressor of cytokine signaling (SOCS) 1 gene in MM cells and modulate SOCS1-mediated cytokine signaling in immune effector cells, thereby integrating immune responses against MM cells [134].

4.1 Immunological checkpoint PD-1/PD-L1

Programmed cell death 1 (PD-1) is a type I transmembrane protein expressed on the surface of activated T cells, interacting with its two ligands PD-L1 and PD-L2. The association between PD-L1 on target cells and PD-1 on T and effector cells acts as an immunological checkpoint to suppress anti-tumor immunity [135]. Indeed, PD-1 is largely expressed on tumor-infiltrating T cells, and PD-L1 is upregulated on multiple human solid tumors, allowing tumor cells to escape from host immune response [136]. Recently, a series of studies have demonstrated the remarkable clinical efficacy of PD-1/PD-L1 blockade in cancers including melanoma, lung cancer, and lymphoma [137-140]. Importantly, MM cells express PD-L1, which is further upregulated in the BM microenvironment [141, 142]. Furthermore, PD-1 expression is upregulated on NK or T cells in MM patients [143, 144]. These results indicate that the PD-1/PD-L1 axis plays crucial roles in host immune surveillance in MM. Indeed, growth of MM cells is inhibited in PD-1-deficient mice [145], and an anti-PD-1 antibody CT-011 both enhances NK-cell cytotoxicity against MM cells [143] and also enhances activated T-cell responses to vaccination with autologous dendritic/MM cell fusion [144]. PD-L1 blockade also contributes to inhibition of MM-tumor growth [145], and anti-PD-L1 therapy in combination with lymphodepletive irradiation facilitates T cell-mediated anti-MM activity [146]. Although autologous stem cell transplantation in combination with cell-base vaccine administration does not prolong survival in the 5T33 murine model of MM, addition of PD-L1 blockade to this combination treatment prolongs host survival [147]. These reports suggest that targeting PD-1/PD-L1 pathway holds great promise in the treatment of MM patients [148], alone or in combination with other immune-based therapies including monoclonal antibodies, immunomodulatory drugs, and/or vaccines.

4.2 Myeloid-derived suppressor cells

Myeloid-derived suppressor cells (MDSCs) are a population of early myeloid cells which are expanded in the tumor microenvironment and suppress T-cell proliferation and cytokine production, while promoting tumor cell growth [149, 150]. In MM, the number of MDSCs is increased in both peripheral blood and BM [151, 152]. Moreover, MDSCs facilitate the growth of MM cells through suppression of T cell-mediated immune responses [151]. Interestingly, phosphodiestrerase-5 (PDE5) inhibitors have been reported to reduce the function of MDSCs [153], and treatment with a PDE5 inhibitor resulted in a dramatic clinical response in a patient with end-stage relapsed/refractory MM, associated with decreased MDSCs [154].

5. Future directions and closing remarks

In this article, we have discussed the potential of novel targeted therapeutic approaches in MM, which still remains incurable with current therapeutic options. Since MM is heterogeneous, associated with complex gene abnormalities and multiple signaling aberrations, the strategy of targeting single genes, gene products, or single signaling pathways may not suppress MM cell growth. Therefore, combination strategies are still key for MM treatment, targeting not only MM cells, but also the tumor microenvironment, including host immunity. In order to reach the goal to conquer this intractable cancer, further new treatment options with novel therapeutics should be developed.