HDAC6 inhibitors sensitize non-mesenchymal triple-negative breast cancer cells to cysteine deprivation

Triple-negative breast cancer (TNBC) is a highly malignant type of breast cancer and lacks effective therapy. Targeting cysteine-dependence is an emerging strategy to treat the mesenchymal TNBC. However, many TNBC cells are non-mesenchymal and unresponsive to cysteine deprivation. To overcome such resistance, three selective HDAC6 inhibitors (Tubacin, CAY10603, and Tubastatin A), identified by epigenetic compound library screening, can synergize with cysteine deprivation to induce cell death in the non-mesenchymal TNBC. Despite the efficacy of HDAC6 inhibitor, knockout of HDAC6 did not mimic the synthetic lethality induced by its inhibitors, indicating that HDAC6 is not the actual target of HDAC6 inhibitor in this context. Instead, transcriptomic profiling showed that tubacin triggers an extensive gene transcriptional program in combination with erastin, a cysteine transport blocker. Notably, the zinc-related gene response along with an increase of labile zinc was induced in cells by the combination treatment. The disturbance of zinc homeostasis was driven by PKCγ activation, which revealed that the PKCγ signaling pathway is required for HDAC6 inhibitor-mediated synthetic lethality. Overall, our study identifies a novel function of HDAC6 inhibitors that function as potent sensitizers of cysteine deprivation and are capable of abolishing cysteine-independence in non-mesenchymal TNBC.


Results
HDAC6 inhibitors sensitize non-mesenchymal TNBC to cysteine deprivation. Cysteinedependence is a novel feature in mesenchymal TNBC cells. However, we found that ~ 50% of TNBCs are nonmesenchymal, and enriched expression of epithelial genes ( Fig. S1A-C). These non-mesenchymal TNBC cells, such as HCC70 and HCC38, were cysteine-independent and resistant to erastin, a blocker of cysteine transport (Fig. S1D), distinct from mesenchymal HBL100 and MDA-MB-231 cells. To overcome such resistances in non-mesenchymal tumor cells and identify potential sensitizers, three inhibitors of the histone deacetylase 6 (HDAC6)-Tubacin, CAY10603, and Tubastatin A were identified by the epigenetic compound library screening, which dramatically induce synthetic-lethal death under the cysteine-depleted condition (Fig. 1A,B). We further confirmed that tubacin induced synthetic-lethal cell death in non-mesenchymal MDA-MB-436 and HCC70 cells when cotreated with either cysteine deprivation or erastin; Either tubucin or erastin alone had no significant cytotoxic effect (Fig. 1C,D and Fig. S2A,B). The effectiveness of combined application of tubacin   S2D). Colony formation assay further confirmed that tubacin plus erastin siginificantly suppressed long-term tumor cell growth in comparison with individual drug treatmnet (Fig. S2E). However, the synergistic effect of tubacin and erastin was not evident in MCF10A, a non-cancerous cells (Fig. S2F). Taken together, our epigenetic drug screening identified HDAC6 inhibitors as potent sensitizers capable of promoting synthetic lethality of cysteine depletion in non-mesenchymal tumor cells.

HDAC6 inhibitors synergize with erastin to induce a mixed cell death.
To characterize the death pathway induced by the combination application of erastin and HDAC6 inhibitor, we examined various death and signaling markers, and evaluated the protective effect of different types of cell death inhibitors. Erastin plus tubacin treatment caused significant activation of p38 death signaling, increases of DNA double-strand breaks as indicated by phosphorylation of H2AX, and partial cleavages of PARP1 and caspase-3 ( Fig. 2A,B), while either erastin or tubacin individually failed to activate such death signaling and markers. As the substrate of HDAC6, α-tubulin was strongly acetylated by tubacin. Both necroptosis inhibitor Necrostatin-1 and ferroptosis inhibitor Ferrostatin-1 fully rescued cells from death induced by erastin plus tubacin, while the pan-caspase inhibitor Q-Vad had partial protective effects ( Fig. 2C-E). Similarly, the lethal effects promoted by CAY10603 were prevented by different types of cell death inhibitors (Fig. 2F). This data suggested that HDAC6 inhibitors synergize with erastin to activate a mixed cell death program in non-mesenchymal TNBC.
HDAC6 is not required for the tubacin-mediated synthetic lethality.. Next, we examined whether repression of HDAC6 expression mimics the lethal-promoting effect of tubacin. Unexpectedly, silencing of HDAC6 gene expression by different shRNAs did not promote cell death in the presence of erastin, although the protein level of HDAC6 was reduced and the acetylation level of tubulin was significantly increased (Fig. S3A-C). We hypothesized that shRNA might be unable to sufficiently suppress endogenous HDAC6 to mimic the potency of tubacin. To that end, we employed CRISPR/Cas9 gene editing to knockout HDAC6 in TNBC HCC38 Tubacin synergizes with erastin to activate a lethal gene transcriptional program. Microarray transcriptional profiling analysis was employed to identify the molecular mechanism of tubacin. By the supervised cluster analysis, either erastin or tubucin individually induced mild or few changes in gene expression. However, tubacin plus erastin dramatically induced or repressed expression of large groups of genes ( Fig. 4A), such as apoptotic genes (Bim, BNIP3, and Puma) and genes involved in endoplasmic reticulum and oxidative stress (ATF3, CHOP, and HOMX1). GSEA revealed that the gene responses activated by erastin plus tubacin are similar to gene changes induced by photodynamic therapy (PDT) (Fig. 4B), which is a well-established cancer www.nature.com/scientificreports/ therapy to cause tumor ablation 39 . RT-qPCR and immunoblotting data confirmed that apoptotic genes were strongly induced by the combination treatment, but not by either erastin or tubacin alone (Fig. 4C,D). Gene changes induced by erastin plus tubacin were similar in HDAC6-null cells in comparison with HDAC6 wild type cells (Fig. 4E,F). Taken together, the lethal gene response of tubacin plus erastin reiterated the potential of tubacin to confer cysteine-dependence in non-mesenchymal TNBC via an HDAC6-independent manner.
Tubacin synergizes with erastin to induce cellular zinc response. Ferroptosis is an iron-dependent cell death that is able to be induced by erastin. In our gene expression profiling, gene changes involved in iron metabolism were not observed. Interestingly, the genes involved in zinc transport, ZnT1 and ZnT2, as well as the genes related to zinc storage, metallothioneins MT1G/1H/1 M, were highly induced by erastin plus tubacin (Fig. 4A). RT-qPCR confirmed that these zinc-related genes were strongly induced by erastin plus tubacin, but not by either erastin or tubacin alone (Fig. 5A,B). Similarly, CAY10603 in combination with erastin strongly activated the zinc-related gene response (Fig. 5C). Zinc is known to cause ROS production and induce a mixed type of cell death, including apoptosis and necrosis 40,41 . Therefore, we examined whether the level of zinc is altered by erastin plus tubacin. Indeed, FluoZin™-3, a Zn 2+ -selective indicator, detected that the level of labile zinc was highly increased in cells by erastin and tubacin, but not significantly by either erastin or tubacin alone ( Fig. 5D and Fig. S4A). Labile zinc mostly accumulated in the nucleus, as it co-localized with nuclear DNA. The total cellular amount of zinc measured by ICP-OES, including both labile and bound zinc, was not significantly altered under any treatments (Fig. 5E), indicating that the labile zinc originated from zinc cellular proteins or compartments; it was not imported from culture media. However, chelation of labile zinc with TPEN 42 did not protect cells from cell death induced by erastin plus tubacin ( Fig. 5F and Fig. S4B), indicating that labile zinc is not the causative agent of cell death.

Inhibition of PKC suppresses the synthetic lethality of tubacin.
Considering that zinc functions as a structural or modulatory component of many regulatory and signaling proteins 43,44 , the increase of labile zinc in cells is probably an indication of cellular protein function or signaling alterations. Particularly, zinc release from protein kinase C (PKC) is a common event during PKC activation by reactive oxygen species (ROS) during cell death 45,46 . Therefore, we examined the role of PKC in the synthetic lethality of tubacin. We found that Gö 6983, a PKC inhibitor with a broad inhibitory spectrum 47 , rescued cells from cell death induced by erastin and tubacin, while Gö 6976, a selective inhibitor of PKC α/β , had no protective role ( Fig. 6A and Fig. S5A). Immunoblotting and RT-qPCR analysis confirmed that Gö 6983 abolished phosphorylation of PKC substrates and the induction of cell death genes induced by erastin plus tubacin (Fig. 6B,C and Fig. S5B). Moreover, Gö 6983 sig- www.nature.com/scientificreports/ nificantly suppressed the zinc-related gene response and decreased labile zinc in cells (Fig. 6D,E and Fig. S5C). These results suggested that activation of PKC, but not PKC α/β , is required for the tubacin-mediated cell death.
PKCγ is required for the tubacin-mediated synthetic lethality. To examine which member of PKC family is required for the synthetic lethality of tubacin, two additional PKC inhibitors, bisindolylmaleimide I and sotrastaurin, with slightly different inhibitory spectrums were used. Similar to Gö 6983, bisindolylmaleimide I strongly protected cells from cell death induced by erastin plus tubacin, but sotrastaurin did not (Fig. 7A,B). Based on the protective role and inhibitory spectrums of various PKC inhibitors, PKCγ stood out as the candidate kinase since it is inhibited by Gö 6983 and bisindolylmaleimide I, but not by Gö 6976 and sotrastaurin. Additionally, PKCγ was transiently phosphorylated in the early phase and associated with the phosphorylation of PKC substrates during the erastin plus tubacin treatment. On the other hand, the phosphorylation of PKC δ/θ occurred in the later phase (Fig. 7C). Knockdown of PKCγ by shPKCγ significantly alleviated cell death induced by erastin plus tubacin in non-mesenchymal TNBC cells (Fig. 7D,F and Fig. S6A). Consistent with the protective role of PKCγ, its knockdown suppressed induction of cell death markers (PARP1, pho-H2AX, and BNIP3) ( Fig. 7E and Fig. S6A,B). Taken together, activation of PKCγ is required for the synthetic lethality of tubacin.

Discussion
TNBC is the most challenging subtype of breast cancer and overall its survival rate is low 48 . Developing novel and precise targeted therapies is urgently needed. Targeting cancer metabolism emerges as a promising strategy to treat some cancers with metabolic deregulations. Particularly, cysteine depletion can eradicate a subset of TNBC with highly mesenchymal states (EMT) 49 . However, many TNBCs with epithelial features, as well as luminal breast cancers are cysteine-independent and unresponsive to cysteine deprivation. Epigenetic alterations can change tumor identity, metabolism, heterogeneity, and result in drug resistances in cancer 35,37 . Many epigenetic activators or inhibitors are employed as therapeutic adjuvants to enhance chemotherapy efficacy and overcome www.nature.com/scientificreports/ drug resistance 32,33 . In line with these ideas, we identified that HDAC6 inhibitors sensitize non-mesenchymal TNBC cells and some luminal cancer cells to cysteine deprivation. HDAC6 is a unique member of the HDAC family that can regulate cell proliferation, metastasis, and invasion in tumors, and can also drive tumor progression and confer drug resistance in some cancers [50][51][52][53] . Although HDAC6 is likely a reasonable molecular target in our context, we have shown that knockout of HDAC6 protein expression does not mimic the effects of HDAC6 inhibitors; endogenous HDAC6 removal does not synergize with erastin to induce cell death in the non-mesenchymal TNBC. These suggest that the synergistic effect of HDAC6 inhibitors occurs via a new cellular molecule or pathway that is entirely independent of HDAC6. We aim in future studies to identify such genes or pathways as they can be valuable direct targets to be exploited in combination with cysteine deprivation for cancer treatment.
Although the intrinsic target of HDAC6 inhibitors in this study remains unknown, we found that tubacin plus erastin triggered many stress and apoptotic gene responses, one of which mimics the gene response of photodynamic therapy (PDT) 39 . The synthetic lethality of tubacin exhibited as a mixed type of cell death, including apoptosis, necroptosis, and ferroptosis. Ferroptosis is an iron-dependent cell death driven by lipid peroxidation with necrotic features [54][55][56] . Our transcriptional profiling data didn't show any iron-related gene response; instead, the expression of several zinc-related genes and the increase of labile zinc was prominently induced by erastin plus tubacin.
Zinc is a trace but essential metal micronutrient and is integral to many enzymes and regulatory proteins, and functions as a signaling messenger in cells 57,58 . Although cellular zinc homeostasis is disturbed by tubacin plus erastin, we observed that the total amount of zinc, including labile and bound forms, was not significantly changed in treated cells; also, chelation of labile zinc didn't protect cells from death. These suggested that labile zinc is not the direct causative agent for cell death, instead, it is an indicator of intracellular signaling perturbations. It's known that reactive oxygen species (ROS) can cause zinc releasing from oxidized metallothioneins and zinc-fingers of abnormal proteins 59 . For example, activation of PKC α/β by ROS can release zinc directly from PKC α/β zinc fingers into the cytoplasm 45,46 . Indeed, inhibition of PKC abolished the release of labile zinc and cell death in our study. Specifically, activation of PKCγ is required for the synthetic lethality of tubacin. As one of the conventional PKC isozymes, PKCγ is mainly present in the brain and has rarely been explored in studies on cancer 60 . Recent studies showed that activation of PKCγ increases the migratory capacity of colon cancer 61,62 . Our study identified a new role of PKCγ in breast cancer, whereby the PKCγ signaling mediates the synthetic lethality of tubacin in non-mesenchymal TNBC cells.
Taken together, many TNBC cells are non-mesenchymal and cysteine-independent. HDAC6 inhibitors identified in our study, particularly tubacin, can overcome the resistance of cysteine deprivation in non-mesenchymal TNBC and promote the synthetic lethality of cysteine deprivation. These inhibitors execute their lethal effects independent of their canonical target, the HDAC6 protein. Instead, HDAC6 inhibitors, in synergism with erastin, trigger an extensive gene transcriptional program to induce cell death via the PKCγ signaling. HDAC6 www.nature.com/scientificreports/ inhibitors will be immensely valuable as adjuvants in the application of targeted cysteine-dependence therapy to treat various types of breast cancer.

Methods
Cell culture and reagents. All breast tumor cells and 293 T cells were purchased from ATCC and maintained as per standard protocols in an incubator with 95% humidity and 5% CO 2 at 37 °C. Cells were cultured in DMEM with 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cysteine deficient medium was prepared according to the previous report 63    www.nature.com/scientificreports/ Gene expression profiling and geneset enrichment analysis (GSEA). The gene expression profile in HCC38 cells treated with either control, erastin, tubacin, or erastin plus tubacin in triplicate for 24 h was analyzed by the GeneChip™ Human Gene 2.1 ST 24-Array Plate (ThermoFisher Scientific). The data were deposited in the GEO database (GSE154425). Probe intensities were normalized by RMA module. Gene expression changes in treated HCC38 cells were derived by zero-transformation (Δlog 2 ) against those in the control condition. Probe sets that varied by twofold in at least 3 samples were selected for hierarchical clustering. The pathway enrichment was analyzed by Gene Set Enrichment Analysis (GSEA) module using the G2 annotated-genesets with default criteria of 1000 permutations.
RNA extraction and real-time RT-PCR. RNA was extracted from cells by RNeasy kit (Invitrogen). Total RNA (2 µg) was reverse-transcribed to cDNA and the quantitative PCR was performed using SYBR Green PCR master mixture (Applied Biosystems). The relative difference in gene expression was normalized with the Actin B gene expression using the ΔΔCT method. All primers in this study were listed in supplementary table 2.
Protein immunoblotting analysis. Proteins were extracted from cells using RIPA extraction and lysis buffer (Sigma) with the protease and phosphatase inhibitor cocktail (ThermoFisher Scientific). Protein concentrations were determined by BCA protein assay. Equal amounts of protein were loaded for immunoblot analysis. Signals were detected by the ECL plus Western blotting detection system (Amersham) and visualized by LAS-4000 lumino image analyzer.
Cell viability and cytotoxicity. Cell viability was measured by either counting of trypan blue negative cells, relative ATP levels using CellTiter-Glo assay kit (Promega), or evaluation by crystal violet staining. Cell cytotoxicity was measured as protease release using CytoTox-Fluor cytotoxicity assay kit (Promega).
Clonogenic assay. Cells were seeded at a density of 5 × 10 3 cells per well in 12-well plates and incubated for two days. Then, the cells were treated with drugs for three days, and replaced with fresh culture medium for additional two or three weeks at 37 °C. Finally, colonies were stained with 1% crystal violet at room temperature.

Microscopic imaging.
To analyze intracellular zinc levels, 7 × 10 3 cells were seeded in the 96-well plate under various treatments. At the end, cells were stained by a 2 μM FluoZin-3 molecular probe (ThermoFisher Scientific) at 37 °C for 30 min. Dialyzed FBS medium was used to ensure that the medium was zinc-free. Additionally, cells were also stained by 1 µM of 4′,6-diamidino-2-phenylindole (DAPI) for 5 min. Cells were then imaged by ZOE™ Fluorescent Cell Imager microscopy.
Cellular zinc level measurement. Total cellular zinc, including labile and bound zinc, was determined by ICP-OES (CHORI Elemental Analysis Facility) 64 . Briefly, 3 × 10 6 HCC38 cells under different treatments were collected and fully dissolved in 0.25 mL OmniTrace 70% HNO 3 (EMD Chemicals) by microwave digestion. Samples were diluted to 5% HNO 3 with OmniTrace water and analyzed with the use of a Vista Pro ICP-OES (Varian Vista Pro). Zinc was measured at the 213-nm wavelength with a detection range between 0.005 and 5 ppm. All associated reagents and plasticware were certified as trace metal-free or tested for trace metal contamination. Zinc concentrations were normalized by total protein mass. All samples were analyzed in triplicate.
Statistical analyses. The significance of differences between groups was determined using a t test. Statistical analysis was performed using GraphPad Prism version 8.3.1, GraphPad Software, San Diego, California USA, www. graph pad. com. A p-value < 0.05 was considered statistically significant. Data were presented in figures as mean ± standard deviation (SD).

Data availability
The datasets generated in the current study are available in the Gene Expression Omnibus (GEO) database (https:// www. ncbi. nlm. nih. gov/ geo/). www.nature.com/scientificreports/