Targeting ubiquitin-specific protease-7 in plasmacytoid dendritic cells triggers anti-myeloma immunity

To the Editor

Proteasome inhibitors (PIs) are a backbone standard of care in the treatment of relapsed/refractory and newly diagnosed multiple myeloma (MM) [1, 2]. Although PIs represent a major advance, not all MM responds, and relapse develops due to the development of drug-resistance [2]. The mechanism(s) that contribute to PI-resistance include mutations in PI-binding site in the β-5 chymotrypsin-like proteasome subunit [3, 4]; increased transcription and synthesis/activity of proteasome catalytic subunits [4]; insufficient blockade of proteasomal activity by PI [5]; induction of XBP1 and heat shock proteins [5]; repression of 19S subunits via DNA methylation [6]; and upregulation of alternative protein degradation pathway (e.g., aggresome/autophagy). In addition to the 26S proteasome, the ubiquitin-proteasome pathway harbors many other potential sites for pharmacological intervention, with the potential to overcome PI-resistance. In this context, our prior studies showed that genetic and biochemical inhibition of deubiquitylating enzyme USP7, upstream of 20S proteasome, can trigger apoptosis in MM cells resistant to PI therapies via p53/p21 signaling axis, without affecting proteasomal activity [7].

Besides tumor-intrinsic mechanism(s), the interaction of MM cells with BM accessory cells and immune effector cells also stimulates tumor growth, drug-resistance, and immune suppression. For example, cell–cell contact between immunologically-defective plasmacytoid dendritic cells (pDCs) with MM cells, T- or NK-effector cells in the BM milieu induces MM cell proliferation, as well as inhibits innate and adaptive immune responses. Using our co-culture models of patient pDCs, T cells, or NK cells with autologous MM cells [8, 9], we have identified mechanisms of immune suppression and validated novel targeted therapies to restore anti-MM immunity. For example, triggering maturation/activation of MM patient pDCs via TLR7/9 agonist or anti-PD-L1 Abs in these models restores their ability to stimulate T-cell proliferation and cytolytic activity against MM cells [8-10]. Our prior study showed that targeting USP7 DUB in MM cells overcomes PI-resistance [7]. In the current study, we utilized our co-culture models of pDCs, T cells, or NK cells with autologous patient MM cells to examine whether USP7 inhibition also stimulates anti-MM immunity.

We first examined whether USP7 inhibition affects the maturation of pDCs. For these studies, we utilized a novel potent and irreversible inhibitor of USP7 XL177A that was developed through our recent medicinal and structure-guided approach [11]. MM patients’ pDCs were treated with XL177A (0.2 μM) and examined for changes in the expression of markers of activation/maturation. XL177A induces significant upregulation of CD83 (p = 0.0115), CD86 (p = 0.0015), and HLA-DR (p = 0.0067) on MM pDCs (Fig. 1A). Our earlier study showed a similar degree of increase in pDC activation markers (e.g., HLA-DR) in CpG-ODNs/TLR9 agonist-treated pDCs [10]. Of note, treatment of pDCs with concentrations higher (1.0 μM) than those at which XL177A (0.2 μM) activates pDCs does not affect the viability of pDCs (Supplementary Fig. 1). As in our prior study, treatment of pDCs with PI bortezomib showed no significant induction of these markers (data not shown).

Fig. 1. Pharmacological inhibition of USP7 activates MM patient pDCs and triggers pDC-induced MM-specific CD8⁺ CTLs.

A MM patient BM-pDCs were treated with XL177A (0.2 μM) for 24 h, and subjected to multicolor staining using fluorophore-conjugated Abs against pDC activation/maturation markers CD83, CD86, and HLA-DR, followed by flow cytometric analysis. The bar graph shows the percentage change in expression of indicated markers in untreated versus XL177A (0.2 μM)-treated pDCs (n = 9) MM patient BM samples; mean ± SD; CD83: p = 0.0115; CD86: p = 0.0015; HLA-DR: p = 0.0067. B MM patient BM T cells were co-cultured with autologous pDCs (n = 10) at 1:10 (pDC:T cell) ratio in the presence or absence of XL177A (0.2 μM) for 3 days. After washing to remove XL177A, cells were cultured with autologous MM cells pre-stained with CellTracker Violet (T/MM; 10:1 ratio) for 24 h, followed by 7-AAD staining and quantification of CTL-mediated MM cell lysis by FACS. Left panel: Representative FACS scatter plot showing the decrease in the number of viable CellTracker-positive MM cells. Right panel: Plot shows quantification of CD8+ CTLs-mediated MM cell lysis, reflected in CD138⁺ MM cell viability (n = 10 MM patient BM samples; mean ± SD; p < 0.05). (Note: All studies using MM patient samples were performed following IRB-approved protocols at Dana-Farber Cancer Institute/Brigham and Women’s Hospital, Boston, USA. Informed consent was obtained from all patients in accordance with Helsinki protocol, and patient samples were de-identified prior to their use in experiments.)

We next assessed whether XL177A induced-pDC activation correlates with IFN-α secretion. MM patient pDCs were treated with DMSO control or XL177A (0.2 μM) for 24 h; then supernatants were collected and subjected to IFN-α secretion assays (Verikine Human IFN-α Elisa kit, PBL Assay Science, USA). A significant increase in IFN-α levels was observed in supernatants from XL177A-versus DMSO-treated pDCs (1.5-fold increase; n = 3; p = 0.002) (Supplementary Fig. 2). Treatment of pDCs with high concentrations of XL177A (1.0 μM) induced a more robust IFN-α production (data not shown). As a control, treatment of pDCs with inactive enantiomer XL177B showed no significant induction of IFN-α secretion (Supplementary Fig. 2). Taken together, these findings suggest that USP7 inhibition via XL177A activates MM patient pDCs.

We next examined whether USP7 blockade-induced-pDC activation restores their ability to induce MM-specific cytotoxic T lymphocyte (CTL). For these studies, we utilized our previously established co-culture models of patient pDCs, T cells, or NK cells with autologous MM cells [8, 9]. Using these models, we have identified mechanisms of immune suppression and validated novel targeted therapies to restore anti-MM immunity. Freshly isolated MM patient BM CD8⁺ T cells were co-cultured with autologous pDCs (n = 10 patients) at 1:10 (pDC:T cell) ratio in the presence or absence of XL177A (0.2 μM) for 3 days. After washing to remove XL177A, these cells were cultured for 24 h with autologous MM cells pre-stained with CellTracker (Life Technologies, USA) Violet (T/MM; 10:1 ratio), followed by 7-AAD staining and quantification of CTL-mediated MM cell lysis using flow cytometry. As shown in Fig. 1B (scatter plot and bar graph), XL177A stimulates a marked increase in MM-specific CD8⁺ CTL activity (p = 0.0001), evidenced by decreased viable patient MM cells. Among the ten patient samples analyzed, four patients had newly diagnosed untreated MM, and six patients had relapsed MM resistant to bortezomib, lenalidomide, and dexamethasone. We compared the extent of MM cell killing in untreated versus resistant patients. Specifically, XL177A-stimulated CTLs triggered a significant killing of tumor cells obtained from untreated newly diagnosed patients (31% MM cell lysis ± 10% std dev; four patients; p = 0.009) and from drug-resistant patients (21% MM cell lysis ± 8% std dev; six patients; p = 0.003) (Supplementary Fig. 3). The variation in CTL-induced tumor cell killing observed in untreated and resistant patients may be due to differential expression of T-cell responsive immunoreceptors or immune checkpoints on tumor cells, as well as genetic heterogeneity and drug-resistant characteristics of MM.

Prior studies showed that NK cells in MM are immunologically dysregulated [12]. We therefore next utilized our autologous pDCs-NK cells-MM cells co-culture models to assess whether USP7 blockade alters the anti-MM activity of NK cells. Purified NK cells from MM patient BM were co-cultured with autologous pDCs in the presence of XL177A (0.2 μM) or DMSO control for 3 days; cells were then washed to remove the drug and cultured for 24 h with autologous MM cells. XL177A significantly increased NK cell-mediated MM cell lysis (Fig. 2A, scatter plot and bar graph). Together, our findings, therefore, show that USP7 inhibition activates pDCs and restores their ability to stimulate MM-specific CTL- and NK cell-mediated cytotoxicity.

Fig. 2. USP7 blockade triggers NK cell-mediated MM cell lysis and decreases T regulatory cells in MM patient BM.

A MM patient BM NK cells were co-cultured with autologous pDCs at 1:10 (pDC:NK cell) ratio in the presence or absence of XL177A (0.2 μM) for 3 days. After washing to remove XL177A, cells were cultured with autologous MM cells pre-stained with CellTrace violet (10:1 NK cell:MM cell ratio) for 24 h, followed by 7-AAD staining and quantification of MM cell lysis by FACS. Left panel: representative FACS scatter plot showing a decrease in viable CellTrace Violet-positive MM cells. Right panel: Bar graph shows quantification of NK-mediated MM cell lysis. Data were obtained from 3MM patient BM samples (mean ± SD; p < 0.05). The student’s t-test was utilized to derive statistical significance, and the minimal level of significance was p < 0.05 (Graph Pad PRISM version 6, La Jolla, California, USA). B MM patient BM-MNCs cells were treated with CD3/CD28 cocktail for 2 days to stimulate T cells. After 2 days, cells were washed, resuspended in fresh medium, and treated with control DMSO, XL177A (0.2 μM), or its enantiomer XL-177B (0.2 μM) for 5 days, followed by quantification of Treg. (CD4⁺/CD25⁺/Foxp3⁺) population using flow cytometry (n = 3 MM patient BM). For these studies, cells were first gated on CD3⁺ cells, followed by selection of CD4⁺/CD25⁺ dual-positive population, and subsequent quantification of the CD4⁺/CD25⁺/Foxp3⁺ Tregs.

As for pDCs and NK cells, T regulatory cells (Tregs) in the MM BM milieu also exhibit abnormal immune function [13]. A prior study showed that USP7 modulates FOXP3, an essential transcriptional factor for Treg functioning [13]. We therefore next examined whether XL177A affects Tregs homeostasis. MM patient BM-MNCs were treated with indicated concentrations of XL177A for 5 days, followed by multicolor staining and flow cytometric quantification of CD4⁺/CD25⁺/Foxp3⁺ Tregs. XL177A significantly decreased Tregs (CD4⁺/CD25⁺/Foxp3⁺) cells (11–12% decrease; p = 0.009; 3MM patients) (Fig. 2B). Prior studies have reported a similar extent of drug-induced reduction in the Treg cell population as significant and biologically relevant [14, 15]. As a negative control, treatment with XL177B, an enantiomer of XL177A, did not impact Tregs (Fig. 2B). These findings suggest that the USP7 blockade represents a therapeutic strategy to decrease dysfunctional Tregs in MM.

In summary, we here show that selective pharmacological inhibition of USP7 DUB enzyme: activates MM patient pDCs; triggers MM-specific CTL- and NK-cell-mediated cytotoxicity against autologous MM cells; as well as decreases abnormal Tregs in MM. Our study, therefore, highlights the therapeutic potential of targeting USP7 to restore both innate (pDCs) and adaptive (T, NK cells, Tregs) immune responses in MM. The current study, taken together with our prior findings showing that inhibition of USP7 overcomes PI-resistance [7] in MM cells, indicate that therapeutic targeting of USP7 represents a promising novel strategy to overcome both PI-resistance and immune suppression in MM.