Investigating the Vascular Toxicity Outcomes of the Irreversible Proteasome Inhibitor Carfilzomib

Abstract

Background: Carfilzomib's (Cfz) adverse events in myeloma patients include cardiovascular toxicity. Since carfilzomib's vascular effects are elusive, we investigated the vascular outcomes of carfilzomib and metformin (Met) coadministration.

Methods: Mice received: (i) saline; (ii) Cfz; (iii) Met; (iv) Cfz+Met for two consecutive (acute) or six alternate days (subacute protocol). Leucocyte-derived reactive oxygen species (ROS) and serum NO_x levels were determined and aortas underwent vascular and molecular analyses. Mechanistic experiments were recapitulated in aged mice who received similar treatment to young animals. Primary murine (prmVSMCs) and aged human aortic smooth muscle cells (HAoSMCs) underwent Cfz, Met and Cfz+Met treatment and viability, metabolic flux and p53-LC3-B expression were measured. Experiments were recapitulated in AngII, CoCl₂ and high-glucose stimulated HAoSMCs.

Results: Acutely, carfilzomib alone led to vascular hypo-contraction and increased ROS release. Subacutely, carfilzomib increased ROS release without vascular manifestations. Cfz+Met increased PGF2α-vasoconstriction and LC3-B-dependent autophagy in both young and aged mice. In vitro, Cfz+Met led to cytotoxicity and autophagy, while Met and Cfz+Met shifted cellular metabolism.

Conclusion: Carfilzomib induces a transient vascular impairment and oxidative burst. Cfz+Met increased vascular contractility and synergistically induced autophagy in all settings. Therefore, carfilzomib cannot be accredited for a permanent vascular dysfunction, while Cfz+Met exert vasoprotective potency.

Keywords: autophagy; carfilzomib; endoplasmatic-reticulum stress; vascular smooth muscle cells; vasculature.

Figure 1

Carfilzomib leads to an acute hypo-contractile phenotype in murine vessels and a significant increase in leucocyte-derived reactive oxygen species (ROS) release, independently of metformin coadministration. Relaxation curves of (A) % relaxation to acetylcholine (ACh; n = 5 per group), (B) tension (g) following ACh-relaxation (*** p < 0.005 vs. control; #### p < 0.001 vs. Met; two way ANOVA, Tukey’s multiple comparison test, main column effect; n = 5 per group) and (C) graph of body weight (g; n = 5 per group). Relaxation curves of (D) % relaxation to nitroglycerine (Gtn) (n = 5 per group), (E) tension (g) following Gtn-relaxation (*** p < 0.005 vs. control; two way ANOVA, Tukey’s multiple comparison test, main column effect; n = 5 per group) and (F) PGF2α maximal contraction (g; * p < 0.05, one way ANOVA, Tukey’s multiple comparison test; n = 5 per group). Curves of (G) Zymosan A and (H) PdBu oxidative burst (counts/min) in course of time (min; *** p < 0.005 vs. control; #### p < 0.001 vs. Met; †††† p < 0.001 vs. Cfz; two way ANOVA, Tukey’s multiple comparison test, main column effect; n = 5 per group). (I) Graphs of circulating cell count (×10³) in whole blood samples. Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent the control (Normal Saline; NS 0.9%), red lines/columns represent Cfz (8 mg/kg, ip), green lines/columns Met (140 mg/kg, per os) and purple lines/columns Cfz+Met (8 mg/kg, ip and 140 mg/kg, per os, respectively) in the 2 days protocol (acute protocol). Cfz, carfilzomib; Met, metformin; Neu, Neutrophils; Mon, Monocytes; Lym, Lymphocytes; WBCs, White Blood Cells.

Figure 2

Carfilzomib did not exhibit any vascular deficits in subacute protocol, whilst it induced increased leucocyte derived-ROS release, which was partially inhibited by metformin. Coadministration of Cfz and Met led to an increased vascular reactivity to PGF2α. Relaxation curves of (A) % relaxation to acetylcholine (ACh; n = 5 per group), (B) tension (g) following ACh-relaxation (** p < 0.01 vs. control; ## p <0.01 vs. Met; ††† p < 0.005 vs. Cfz; two way ANOVA, Tukey’s multiple comparison test, main column effect; n = 5 per group) and (C) graph of body weight (g; n = 5 per group). Relaxation curves of (D) % relaxation to nitroglycerine (Gtn; n = 5 per group), (E) tension (g) following Gtn-relaxation (**** p < 0.001 vs. control; #### p <0.001 vs. Met; †††† p < 0.001 vs. Cfz two way ANOVA, Tukey’s multiple comparison test, main column effect; n = 5 per group) and (F) PGF2α maximal contraction (g) (* p < 0.05, ** p < 0.01; one way ANOVA, Tukey’s multiple comparison test; n = 5 per group). Curves of (G) Zymosan A and (H) PdBu oxidative burst (counts/min) in course of time (min; *** p < 0.005 and **** p < 0.001 vs. control; ### p< 0.005 and #### p < 0.001 vs. Met; †† p < 0.01, †††† p < 0.001 vs. Cfz; two way ANOVA, Tukey’s multiple comparison test, main column effect; n = 5 per group). (I) Representative graphs of circulating cell count (×10³) in whole blood samples. Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent control (Normal Saline; NS 0.9%), red lines/columns Cfz (8 mg/kg, ip), green lines/columns Met (140 mg/kg, per os) and purple lines/columns Cfz+Met (8 mg/kg, ip and 140 mg/kg, per os, respectively) in the 6 days protocol (subacute protocol). Cfz, Carfilzomib; Met, Metformin; Neu, Neutrophils; Mon, Monocytes; Lym, Lymphocytes; WBCs, White Blood Cells.

Figure 3

Carfilzomib and metformin monotherapies do not induce any changes in the vascular phenotype, while the combination of Cfz+Met leads to increased collagen deposition on the vessels and a synergistic induction of an ER-stress/AMPKα/LC-3B dependent autophagy. (A). (Upper Panel) Representative images of hematoxylin–eosin (H&E) staining of murine aortas (magnification 40×; scale bar 40 μm; Lower Panel) Representative images of Sirius red staining (magnification 20×; scale bar 40 μm). Graphs of (B) aortic wall thickness (μm) and (C) collagen thickness (μm) (* p < 0.05, ** p < 0.01, one way ANOVA, Tukey’s multiple comparison test; n = 3 per group). (D) Representative Western blot images of the studied protein targets and their phosphorylated forms. (E) Relative densitometry analysis of Western blot analysis, presented as fold change of the control (* p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 3 per group). Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent the control (Normal Saline; NS 0.9%), red lines/columns represent Cfz (8 mg/kg, ip), green lines/columns Met (140 mg/kg, per os) and purple lines/columns Cfz+Met (8 mg/kg, ip and 140 mg/kg, per os, respectively) in the 6 days protocol (subacute protocol). Cfz, Carfilzomib; Met, Metformin.

Figure 4

Carfilzomib did not affect the NO pathway but increased icam-1 and tf expression. Metformin partially abrogated icam-1 mRNA increase. Graphs of (A) serum NOx concentration (μM; n = 5–7 per group) and (B) graph of aortic 3-nitrotyrosine (ng/mg protein; fold change of the control) of treated mice (n = 5 per group). (C) RT-PCR mRNA expression of proinflammatory and regulatory endothelial genes and redox molecules (n = 4–5 per group; * p <0.05, ** p < 0.01; one way ANOVA, Tukey’s multiple comparison test). Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent the control (Normal Saline; NS 0.9%), red lines/columns represent Cfz (8 mg/kg, ip), green lines/columns Met (140 mg/kg, per os) and purple lines/columns Cfz+Met (8 mg/kg, ip and 140 mg/kg, per os, respectively) in the 6 days protocol (subacute protocol). Cfz, Carfilzomib; Met, Metformin.

Figure 5

Carfilzomib and metformin coadministration induced LC3-B-dependent autophagy in the aortas of aged mice, while Cfz increased serum NO_x content without leading to a nitro-oxidative stress in the aortas. Graphs of serum NO_x concentration (μM) of (A) young (3–4 months of age) and aged (15–17 months of age) mice (n = 4 per group). (B) Aged mice treated with carfilzomib (Cfz), metformin (Met) or the combination of the drugs (n = 5 per group). (C) Graph of aortic 3-nitrotyrosine (3-NT; ng/mg protein; fold change of the control) of aged treated mice (n = 5 per group). (D) Representative Western blot images of aged murine aortas and (E) relative densitometry analysis of the selected proteins and their phosphorylated forms (n = 5 per group; * p < 0.05, ** p < 0.01, *** p < 0.005, one way ANOVA, Tukey’s multiple comparison test). Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent the control (Normal Saline; NS 0.9%), red lines/columns represent Cfz (8 mg/kg, ip), green lines/columns Met (140 mg/kg, per os) and purple lines/columns Cfz+Met (8 mg/kg, ip and 140 mg/kg, per os, respectively) in the 6 days protocol (subacute protocol). Cfz, Carfilzomib; Met, Metformin.

Figure 6

Carfilzomib and metformin induce cytotoxicity in primary murine vascular smooth muscle cells (prmVSMCs) in vitro. Graphs of (A) MTT absorbance (540 nm) expressed as the fold change of the control (**** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 4 per group), (B) cellular confluency expressed as the fold change of the control (n = 3 per group), (C) ATP production rate (pmol/min) and % metabolic contribution under conditions of normal glucose (10 mM) originating from oxidative phosphorylation and glycolysis (*** p < 0.005, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 3 per group) and (D) ATP production rate (pmol/min) and % metabolic contribution under conditions of low glucose (1 mM) originating from oxidative phosphorylation and glycolysis (* p < 0.05, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 3 per group). Time-course graphs of the proton efflux rate (PER) and oxygen consumption rate (OCR) under the conditions of (E) normal glucose (10 mM) and (F) low glucose (1 mM), as studied with ATPase inhibitor oligomycin (Oligo) and cytochrome c reductase inhibitor antimycin-a (Anti-A) and complex I inhibitor rotenone (Rot). (G) Merged immunofluorescence confocal images of prmVSMCs stained against F-actin (Phalloidin, green), p53 (red) and DAPI (blue) as well as the negative control for the staining (rabbit isotype control and anti-rabbit IgG H&L secondary antibody). Graphs of (H) integrated fluorescence density of p53/DAPI expressed as the fold change of Normal Saline (NS) 0.9%, serving as the control vehicle and (I) total fluorescence intensity of LC3B/DAPI as measured in the automated microscope per well (mean total fluorescence intensity of individual VSMCs in the well; ** p < 0.01, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 3 per group). Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent Normal Saline (NS) 0.9%, red lines/columns represent Cfz (0.3 μΜ), green lines/columns Met (10 mM) and purple lines/columns Cfz+Met (0.3 μΜ, 10 mM respectively). Cfz, carfilzomib; Met, metformin; prmVSMCs, primary murine vascular smooth muscle cells.

Figure 7

Carfilzomib and Cfz+Met do not induce cytotoxicity in senescent human aorta smooth muscle cells in vitro. Met shifts cellular metabolism to glycolysis. Graphs of (A) MTT absorbance (540 nm) expressed as the fold change of Normal Saline (NS) 0.9% (** p < 0.01; one way ANOVA, Tukey’s multiple comparison test; n = 6 per group), (B) cellular confluency expressed as the fold change of Normal Saline (NS) 0.9% (n = 3 per group), (C) ATP production rate (pmol/min) and (D) % metabolic contribution under conditions of normal glucose (10 mM) originating from oxidative phosphorylation and glycolysis, (E) ATP production rate (pmol/min) and (F) % metabolic contribution under conditions of low glucose (1 mM) originating from oxidative phosphorylation and glycolysis (* p < 0.05,** p < 0.01, *** p < 0.01, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 3 per group). Time-course graphs of the proton efflux rate (PER) and oxygen consumption rate (OCR) under conditions of (G) normal glucose (10 mM) and (H) low glucose (1 mM), as studied with ATPase inhibitor oligomycin and cytochrome c reductase inhibitor antimycin-a and complex I inhibitor rotenone. Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Blue lines/columns represent Normal Saline (NS) 0.9%, red lines/columns represent Cfz (0.3 μΜ), green lines/columns Met (10 mM) and purple lines/columns Cfz+Met (0.3 μΜ, 10 mM respectively). Cfz, carfilzomib; Met, metformin; HAoSMCs, human aorta smooth muscle cells; Rot, rotenone; Oligo, oligomycin; Anti-A, antimycin A.

Figure 8

Cfz+Met synergistically induce autophagy in senescent human aorta smooth muscle cells (HAoSMCs) in the presence of cardiovascular risk stimuli in vitro. Graphs of (A) MTT absorbance (540 nm, n = 3–5 per group) and (B) cellular confluency expressed as the fold change of the control (untreated senescent HAoSMCs; * p < 0.05, ** p < 0.01, *** p < 0.01, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n = 3 per group). (C) Representative immunofluorescence confocal images of HAoSMCs under normal conditions (untreated cells) or in the presence of AngII (100 nM), CoCl2 (150 μΜ) and high-glucose (25 mM). (D) Integrated fluorescence density of p53/DAPI expressed as the fold change of the control (untreated senescent HAoSMCs) and (E) total fluorescence intensity of LC3B/DAPI as measured in the automated microscope per well (mean total fluorescence intensity of individual VSMCs in the well; * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001; one way ANOVA, Tukey’s multiple comparison test; n=3 per group). Data are presented as mean ± SEM and individual values are presented as scatter column graphs. Black columns represent the control (untreated cells), blue columns represent Normal Saline (NS) 0.9%, red columns represent Cfz (0.3 μΜ), green columns Met (10 mM) and purple columns Cfz+Met (0.3 μΜ, 10 mM respectively). Cfz, carfilzomib; Met, metformin; HAoSMCs, human aorta smooth muscle cells.

Figure 9

Schematic representation of the effect of carfilzomib and metformin coadministration on the vessels. White panels represent independent molecular targets, red arrows indicate Cfz-dependent pathways, green arrows represent Met-dependent pathways. Cfz, carfilzomib; Met, metformin.