Dual Targeting of Protein Degradation Pathways with the Selective HDAC6 Inhibitor ACY-1215 and Bortezomib Is Synergistic in Lymphoma

Abstract

Purpose: Pan-class histone deacetylase (HDAC) inhibitors are effective treatments for select lymphomas. Isoform-selective HDAC inhibitors are emerging as potentially more targeted agents. HDAC6 is a class IIb deacetylase that facilitates misfolded protein transport to the aggresome for degradation. We investigated the mechanism and therapeutic impact of the selective HDAC6 inhibitor ACY-1215 alone and in combination with bortezomib in preclinical models of lymphoma.

Experimental design: Concentration-effect relationships were defined for ACY-1215 across 16 lymphoma cell lines and for synergy with bortezomib. Mechanism was interrogated by immunoblot and flow cytometry. An in vivo xenograft model of DLBCL was used to confirm in vitro findings. A collection of primary lymphoma samples were surveyed for markers of the unfolded protein response (UPR).

Results: Concentration-effect relationships defined maximal cytotoxicity at 48 hours with IC50 values ranging from 0.9 to 4.7 μmol/L. Strong synergy was observed in combination with bortezomib. Treatment with ACY-1215 led to inhibition of the aggresome evidenced by acetylated α-tubulin and accumulated polyubiquitinated proteins and upregulation of the UPR. All pharmacodynamic effects were enhanced with the addition of bortezomib. Findings were validated in vivo where mice treated with the combination demonstrated significant tumor growth delay and prolonged overall survival. Evaluation of a collection of primary lymphoma samples for markers of the UPR revealed increased HDAC6, GRP78, and XBP-1 expression as compared with reactive lymphoid tissue.

Conclusions: These data are the first results to demonstrate that dual targeting of protein degradation pathways represents an innovative and rational approach for the treatment of lymphoma.

Figure 1. Selective HDAC6 inhibition with ACY-1215 has activity in lymphoma

(A) Concentration : effect relationship of treatment with increasing concentrations of ACY-1215 on acetylation of α-tubulin after 24 hours in 4 lymphoma cell lines (OCI-Ly10, OCI-Ly7, HBL-2, H9). ACY-1215 selectively targets α-tubulin and not histone 3 as compared to romidepsin after 24 hour exposure in OCI-Ly10 and OCI-Ly7. (B) ACY-1215 has single agent cytotoxic effect across a large panel of lymphoma cell lines. Luminometric assay growth inhibition curves of ACY-1215 on 16 lymphoma cell lines at 48 hours. Growth inhibition IC₅₀ mean values (uM) in 16 cell lines at 48rs is reported in parenthesis. (C) PARP and caspase-3, -8, and -9 cleavage was demonstrated for cells treated with ACY-1215 at 48 hours as measured by Western blot analysis. (D) Basal levels GRP78 expression is higher across 8 cell lines as compared to peripheral blood mononuclear cells (PBMCs). Basal HDAC6 level does not correlate with response to treatment with ACY-1215. Densitometry analysis was performed on scanned immunoblot images using the ImageJ software (NIH). Relative levels of protein expression were calculated and plotted.

Figure 2. Dual targeting of protein degradation pathways with ACY-1215 and bortezomib is synergistic

(A) Four lymphoma cell lines representing ABC-DLBCL, GC-DLBCL, MCL, and TCL (OCI-Ly10, OCI-Ly7, HBL-2, H9 respectively) were treated with increasing concentrations of bortezomib 0nM – 5 nM and ACY-1215 0nM – 1500 nM. Cell viability was measured by luminometric assays at 48 hours. Values represent means expressed as percentages compared with the untreated control; error bars represent SD. Synergy was calculated by Excess over Bliss (EOB) Independence where values greater than 10 represent synergy. Synergy was achieved across all cell lines with EOB as high as 41 in OCI-Ly7. (B) Assessment of apoptosis by Annexin V and propidium iodine in 4 cell lines mentioned in (A). Cells were incubated with ACY-1215 (1 or 1.5 μM) bortezomib (2-5 nM) at their approximate IC_10-20 or the combination. Compared to the untreated control, the combination resulted increased apoptosis across all cell lines at 48 hours. (C) PARP and caspase 3 cleavage was demonstrated for cells treated with the combination of ACY-1215 and bortezomib at 48 hours across cell lines.

Figure 3. Targeting with ACY-1215 and bortezomib leads to marked activation of the Unfolded Protein Response

(A) Treatment of OCI-Ly10 cells with ACY-1215 2μM for 6 and 24 hours. At 6hours acetylation of GRP78 was demonstrated by immunoprecipitation with GRP and Immunoblot with acetyl-lysine. By 24 hours, PERK dissociated from GRP78 as measured by co-immunoprecipitation. Densitometry analysis was performed on scanned immunoblot images using the ImageJ software (NIH). Relative levels of protein expression were calculated and plotted. Treatment of OCI-Ly10 and OCI-Ly7 with ACY-1215 2 μM for 24 hours led to up-regulation of the UPR demonstrated by increased expression of GRP78, PERK, p-eIF2α, and CHOP as measured by Immunoblot. (B) Treatment with the combination of ACY-1215 1.0 μM and bortezomib 2 nM led to increased accumulation of poly-ubiquinated protein compared to treatment with either drug alone in OCU-Ly10 and OCI-Ly7 cell lines at 48 hours. (C) The combination of ACY-1215 and bortezomib led to synergistic induction of the UPR demonstrated by increased expression of GRP78, PERK, p-eIF2α, p-IRE1, and spliced XBP-1 as measured by Immunoblot at 36 hours. (D) Schema of hypothesized mechanism of action of ACY-1215 on modulating the aggresome and the UPR. Treatment with ACY-1215 leads to inhibition of HDAC6 leading to acetylation of α-tubulin and GRP78. Acetylation of α-tubulin leads to disruption of the tubulin-HDAC6-motor dynein complex preventing misfolded proteins from aggresome mediated degradation. Accumulated misfolded proteins then activate the UPR, a quality control mechanism in cells to shuttle proteins back to the endoplasmic reticulum for refolding, or to trigger apoptosis if proteostasis cannot be maintained. ACY-1215 leads to acetylation of GRP78 which releases key regulators of the UPR: IRE1, ATF6, and PERK.

Figure 4. ACY-1215 plus bortezomib led to significant tumor growth delay and prolonged overall survival in a xenograft model of DLBCL (OCI-Ly10)

(A) Weight was recorded every 72 hours. All treatment groups had decreased weight as compared to untreated mice. Mice that had >10% weight loss within the first week regained their weight be day 10. There was only 1 toxic death among all mice treated with ACY-1215 in both the single agent and combination cohorts (N=19). (B) After 1 cycle of therapy, the combination treatment with ACY-1215 and bortezomib led to a significant tumor growth delay compared to either drug alone and untreated control mice (p=0.006). Mean doubling time was significantly prolonged in the combination after 1 cycle of treatment (16.71 days) compared to ACY-1215 (10.68) and bortezomib (10.42) or control (9.4). (C) Overall survival was prolonged after 1 cycle with the combination with median overall survival of 42 days compared to 29.5, 31, and 33 days for control, ACY-1215 and bortezomib cohorts (p<0.05).

Figure 5. Pharmacokinetic and pharmacodynamic effects of ACY-1215 in mice

(A) Serum and tumor tissue was collected from mice at sequential time points and analyzed for concentration of ACY-1215 and bortezomib by LC-MS/MS. Mice were treated with ACY-1215 at 0.5 or 50 mg/kg alone and ACY-1215 50 mg/kg with bortezomib 0.5 mg/kg. Drug concentrations are represented as mean values with standard deviation where more than one mouse sample was available. Cmax in serum was determined at 0.5 hours after injection at 10,888 nM. (B) Graphical representation of ACY-1215 concentration over time. Mice received one dose of ACY-1215 0.5 mg/kg, ACY-1215 50 mg/kg or ACY-1215 50 mg/kg plus bortezomib 0.5 mg/kg for analysis of serum concentration of ACY-1215. For analysis of drug concentration in tumor tissue, ACY-1215 was administered at 50 mg/kg with or without bortezomib. (C) Graphical representation of bortezomib concentration over time analyzed in serum and tumor tissue. Mice received one dose of bortezomib 0.5 mg/kg via i.p. route. (D) Immunohistochemistry staining of acetylated α-tubulin, GRP-78, XBP-1 and TUNEL assay for apoptosis at 6 hours after a single injection of ACY-1215 50 mg/kg with or without bortezomib 0.5 mg/kg. Treated tumor tissue was compared to the spleen of the same mouse and to untreated control mice tumor tissue. Immunoblot analysis of GRP78 and acetylated α-tubulin from whole cell lysates of mouse tumor tissue treated with ACY-1215 alone or in combination with bortezomib. Mice were treated with a single i.p. injection of ACY-1215 at 0.5 mg/kg or 50 mg/kg and analyzed at 4 and 6 hours. Densitometry analysis was performed on scanned immunoblot images using the ImageJ software (NIH). Relative levels of acetylated α-tubulin protein expression were calculated and plotted.

Figure 6. Patient lymphoma samples express high levels of GRP78 and XBP-1 compared to reactive lymph nodes and higher GRP78 staining correlates with prolonged overall survival

(A) Immunohistochemistry staining of HDAC6, GRP78 and XBP-1 in lymphoma patient samples and reactive lymph nodes. Representative cases are shown here. (B) Table of the samples with positive staining of HDAC6, GRP78, XBP-1 and CHOP. Samples were analyzed by lymphoma subtype (DLBCL=diffuse large B-cell lymphoma, FL= follicular lymphoma, MZL= marginal zone lymphoma, MCL= mantle cell lymphoma, TCL= T-cell lymphoma), together as a group and compared to reactive lymph nodes. Lymphoma samples had statistically stronger staining than reactive lymph nodes for HDAC6, GRP78, XBP-1 (p<0.001) but not CHOP. (C) Kaplan Meier curve representing survival (days) in DLBCL patients with +2 verses +1 staining of GRP78. More intense staining correlated with a significantly prolonged survival (p=0.0195). (D) Kaplan Meier curve representing survival (days) in all lymphoma patients with +2 verses +1 staining of GRP78. More intense staining correlated with a significantly prolonged survival (p=0.0315).