Orthogonal targeting of osteoclasts and myeloma cells for radionuclide stimulated dynamic therapy induces multidimensional cell death pathways

Abstract

Rationale: Multiple myeloma (MM) is a multifocal malignancy of bone marrow plasma cells, characterized by vicious cycles of remission and relapse that eventually culminate in death. The disease remains mostly incurable largely due to the complex interactions between the bone microenvironment (BME) and MM cells (MMC). In the "vicious cycle" of bone disease, abnormal activation of osteoclasts (OCs) by MMC causes severe osteolysis, promotes immune evasion, and stimulates the growth of MMC. Disrupting these cancer-stroma interactions would enhance treatment response. Methods: To disrupt this cycle, we orthogonally targeted nanomicelles (NM) loaded with non-therapeutic doses of a photosensitizer, titanocene (TC), to VLA-4 (α4ß1, CD49d/CD29) expressing MMC (MM1.S) and αvß3 (CD51/CD61) expressing OC. Concurrently, a non-lethal dose of a radiopharmaceutical, ¹⁸F-fluorodeoxyglucose ([¹⁸F]FDG) administered systemically interacted with TC (radionuclide stimulated therapy, RaST) to generate cytotoxic reactive oxygen species (ROS). The in vitro and in vivo effects of RaST were characterized in MM1.S cell line, as well as in xenograft and isograft MM animal models. Results: Our data revealed that RaST induced non-enzymatic hydroperoxidation of cellular lipids culminating in mitochondrial dysfunction, DNA fragmentation, and caspase-dependent apoptosis of MMC using VLA-4 avid TC-NMs. RaST upregulated the expression of BAX, Bcl-2, and p53, highlighting the induction of apoptosis via the BAK-independent pathway. The enhancement of multicopper oxidase enzyme F5 expression, which inhibits lipid hydroperoxidation and Fenton reaction, was not sufficient to overcome RaST-induced increase in the accumulation of irreversible function-perturbing α,ß-aldehydes that exerted significant and long-lasting damage to both DNA and proteins. In vivo, either VLA-4-TC-NM or αvß3-TC-NMs RaST induced a significant therapeutic effect on immunocompromised but not immunocompetent MM-bearing mouse models. Combined treatment with both VLA-4-TC-NM and αvß3-TC-NMs synergistically inhibited osteolysis, reduced tumor burden, and prevented rapid relapse in both in vivo models of MM. Conclusions: By targeting MM and bone cells simultaneously, combination RaST suppressed MM disease progression through a multi-prong action on the vicious cycle of bone cancer. Instead of using the standard multidrug approach, our work reveals a unique photophysical treatment paradigm that uses nontoxic doses of a single light-sensitive drug directed orthogonally to cancer and bone cells, followed by radionuclide-stimulated generation of ROS to inhibit tumor progression and minimize osteolysis in both immunocompetent murine and immunocompromised human MM models.

Keywords: Cerenkov radiation; bone marrow; multiple myeloma; nanomicelles; orthogonal drug delivery; photosensitizer; tumor microenvironment.

Figure 1

Schematic illustration of orthogonal RaST. NMs are loaded with a light-sensitive drug, TC, but directed to different cells that express VLA-4 (MM) and αvß3 (OCs) in the BM. The expression of GLUT1 on both MM and OCs provides additional selectivity for eradicating only cells that express GLUT1 and either VLA-4 or αvß3.

Figure 2

Non-lethal amounts of TC and [¹⁸F]FDG converge in target cells to induce cell death. (A) VLA-4 targeted NM loaded with 0.1 µg TC did not lead to cell death; Quantitative data are shown as percent of live cells compared to the untreated cells. (B) [¹⁸F]FDG administered at 3.7 MBq-27.75 MBq (0.1 mCi-0.75 mCi) induce moderate degree of cell death; (C) VLA-4-TC-NM RaST induced about 40% MM1.S cell death 72 h after treatment; statistical significance was determined using unpaired two-tailed t-test and a 95% confidence level. The difference between 48 h and 72 h VLA-4-TC-NM RaST was significant (P = 0.0228). (D) VLA-4-TC-NM RaST-induced cell death requires the expression of VLA-4 by MM1.S cells; MM1.S cells expressing VLA-4 (WT) were compared to VLA-4-TC-NM RaST-resistant MM1.S cells (NR) expressing low levels of VLA-4 and VLA-4 knockout cells (CD49dKO). The data were analyzed with one-way ANOVA and Dunnett's multiple comparisons tests. Untreated versus RaST + WT: P < 0.0001; untreated versus VLA-4-TC-NM RaST + RaST-resistant (RaST^NR) cells: P = 0.0473.

Figure 3

Mechanism of VLA-4-TC-NM RaST. (A) VLA-4-TC-NM RaST generates significantly more ROS (P=0.0039) compared to either [¹⁸F]FDG (NS) or VLA-4-TC-NM alone (NS) when compared to no treatment; 4 x 10⁵ MM1.S cells were treated with [¹⁸F]FDG, VLA-4-TC-NM, VLA-4-TC RaST, or left untreated. ROS measurements were performed 72 h later using H₂DCFDA. (B) Hydroperoxydation of PUFA increased in cells treated with VLA-4-TC-NM RaST. MM1.S cells were plated and treated as in (A). The level of PUFA hydroperoxidation was measured with a lipid peroxidation assay 72 h later. VLA-4-TC-NM and VLA-4-TC-NM RaST produced significantly more lipid hydroperoxidation (P = 0.001 and P < 0.0001, respectively) than no treatment control. (C) VLA-4-TC-NM RaST-induced PUFA hydroperoxidation generated reactive aldehydes 72 h after administration. The cells were treated with VLA-4-TC-NM RaST or left untreated. 3ß,5α,6ß-THC-triol, 7-ketocholesterol, and malondialdehyde levels were significantly higher (P = 0.02, P = 0.02, P = 0.009, respectively) in treated cells; multiple t-tests were used for the statistical analysis. (D) Caspase-3 level was significantly higher in cells treated with VLA-4-TC-NM RaST compared to [¹⁸F]FDG and VLA-4-TC-NM (P < 0.0001 and P = 0.003, respectively) after 72 h of treatment; one-way ANOVA and Tukey's multiple comparisons tests were used for the statistical analysis. (E) Treatment of MM1.S cells with VLA-4-TC-NM RaST significantly increased the cell death (VLA-4-TC-NM RaST versus untreated: P = 0.0003; VLA-4-TC-NM RaST versus [¹⁸F]FDG: P = 0.005; VLA-4-TC-NM RaST versus VLA-4-TC-NM: NS). The cells were treated as in (A). After 72 h, the cells were treated with 0.1 µg PI, and the amount of cell-associated PI was determined with Flow Cytometry; one-way ANOVA and Tukey's multiple comparisons tests were used for the statistical analysis. (F) dsDNA breaks were significantly more abundant 72 h after VLA-4-TC-NM RaST. The cells were treated as in (A). After 72 h, the cells were stained with anti-γH2AX and DAPI. Flow Cytometry analysis showed an overall increase in dsDNA breaks during the S-phase. Specifically, dsDNA breaks were significantly more frequent during VLA-4-TC-NM RaST treatment compared to VLA-4-TC-NM (P = 0.003), [¹⁸F]FDG (P = 0.0002), and the untreated cells (P < 0.0001); 2way ANOVA and Tukey's multiple comparisons tests were used for the statistical analysis. (G) Western blotting showed increased levels of apoptosis-related proteins 72 h after VLA-4-TC-NM RaST. The cells were treated as in (A). (H) Ferroptosis inhibitor Ferrostatin-1 significantly inhibited VLA-4-TC RaST mediated cell death (P = 0.015). MM1.S cells were plated as in (C). Ferrostatin-1 was administered concomitantly with VLA-4-TC RaST and VLA-4 control at the final concentration of 10 µM. Three-way ANOVA and Tukey's multiple comparisons tests were used for statistical analyses.

Figure 4

RaST induces multiple cell death pathways. (A) Proposed mechanisms of RaST-induced multidimensional cell death pathways. Red arrows are based on data presented in this manuscript. (B) Ratios of relative abundances of proteins from either VLA4-TC-NM RaST treated cells or the untreated cells (RaST/NT). Red filled circles: RaST/NT ≥ 1.5; green filled circles: RaST/NT ≤ 0.5. Triplicates of each experimental condition were analyzed.

Figure 5

(A) CD49d and CD61 expression by MM1.S cells in culture; MM1.S cells were grown in culture with CM for 1-3 weeks prior to flow cytometry analysis. (B) VLA-4-TC-NM RaST and VLA-4-TC-NM + αvß3-TC-NM RaST comparison in MM.S1 cells in vitro. (C) αvß3 expression by MM1.S cells in vivo compared to a control. Animals with disseminated MM were euthanized 2-4 weeks after the inoculation and the bone marrow (BM) was examined with flow cytometry for αvß3 expression. MDA-MB-231 human mammary adenocarcinoma cells grown in culture were used as a positive control. (D) αvß3 and GLUT1 expression by murine OCs. Bone marrow from femurs and tibias of 5 weeks old C57BL/6-KaLwRij male mice was harvested and differentiated to OCs as described in Shioi, et al. . Proteins of interest detected with GLUT1 (clone D3J3A Rabbit mAb) and αvß3 (clone SJ19-09) antibodies. (E), αvß3-TC-NM RaST in OC in vitro. OCs were differentiated from BMM for 7-14 days prior to the experiment. αvß3-TC-NM RaST was performed as in (B). Averages of each dataset were compared to the untreated control using a one-way ANOVA and Tukey's multiple comparisons test.

Figure 6

Dual cancer-osteoclast RaST in mice. (A) Fox Chase SCID beige MM1.S xenograft animals untreated or treated once weekly for five weeks with αvß3-TC-NM RaST, VLA-4-TC-NM RaST, and αvβ3-TC-NM + VLA-4-TC-NM RaST; MM1.S/CBR/GFP cells (1 x 10⁶ cells per animal) were implanted IV and the therapy was initiated after the whole body radiance reached 1 x 10⁶ photons/sec/cm²/steradian (p/s/cm²/sr). Tumor progression was monitored with weekly BLI. αvβ3-TC-NM + VLA-4-TC-NM RaST significantly reduced tumor progression compared to no treatment (P = 0.0007). (B) Representative bioluminescence images of Fox Chase SCID beige MM1.S xenograft animals treated with αvβ3-TC-NM + VLA-4-TC-NM RaST and the untreated controls. (C) C57BL/6-KaLwRij 5TGM isograft animals were implanted with 5TGM/CBR/GFP cells (1 x 10⁶ cells per animal) and monitored as in (A). The test animals were untreated or treated once weekly for five weeks with αvβ3-TC-NM RaST, VLA-4-TC-NM RaST, and αvβ3-TC-NM + VLA-4-TC-NM RaST. αvβ3-TC-NM + VLA-4-TC-NM RaST significantly inhibited tumor progression compared to no treatment (P = 0.0317). (D) Representative bioluminescence images of C57BL/6-KaLwRij 5TGM isograft animals treated with αvβ3-TC-NM + VLA-4-TC-NM RaST and the untreated controls. Two-way ANOVA and Tukey's multiple comparisons were used for the statistical analysis.

Figure 7

αvß3-TC-NM RaST leads to OC depletion. (A) OCs in femurs of Fox Chase SCID beige MM1.S xenograft and B, C57BL/6-KaLwRij 5TGM isograft treated with indicated RaST and visualized with TRAP immunohistochemical stain. OC content in (C), Fox Chase SCID beige MM1.S xenograft and (D), C57BL/6-KaLwRij 5TGM isograft was quantified as the total number of TRAP-positive cells per section using Zeiss ZEN 2.3 software. In both models, αvß3-TC-NM RaST, either alone or in combination with VLA-4-TC-NM, significantly reduced the OC population (Fox Chase SCID beige MM1.S xenograft: P = 0.0005 and P = 0.0006, respectively; C57BL/6-KaLwRij 5TGM isograft: P < 0.0001 and P < 0.0001, respectively). Two-way ANOVA and Tukey's multiple comparisons were used for statistical analysis.