Inhibiting BCKDK in Triple Negative Breast Cancer Suppress Protein Translation, Impair Mitochondrial Function, and Potentiate Doxorubicin Cytotoxicity

Home , ACADSB, BCKDHB, BCKDK

bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Inhibiting BCKDK in triple negative breast cancer suppress protein translation, impair 2 mitochondrial function, and potentiate doxorubicin cytotoxicity 3 4 Running Title: Inhibiting BCKDK increase TNBC DOX toxicity

6 Dipsikha Biswas1, Logan Slade1, Luke Duffley1, Neil Mueller1, Khoi Thien Dao1, Angella Mercer1, 7 Yassine El Hiani2, Petra Kienesberger1 and Thomas Pulinilkunnil1*

8 1 Department of Biochemistry and Molecular Biology, Faculty of Medicine, Dalhousie University, 9 Dalhousie Medicine New Brunswick, Saint John, New Brunswick, Canada

10 2Department of Physiology and Biophysics, Faculty of Medicine, Dalhousie University, Halifax, Nova 11 Scotia, Canada

15 *Correspondence to: Thomas Pulinilkunnil

16 Department of Biochemistry and Molecular Biology, Faculty of Medicine, Dalhousie University Dalhousie

17 Medicine New Brunswick, 100 Tucker Park Road, Saint John E2L4L5, New Brunswick, Canada.

18 Telephone: (506) 636-6973; Fax: (506) 636-6001; email: [email protected].

1 Abstract

3 Triple-negative breast cancers (TNBCs) are characterized by poor survival, prognosis and gradual

4 resistance to cytotoxic chemotherapeutics, like doxorubicin (DOX), which is limited by its

5 cardiotoxic and chemoresistant effects that manifest over time. TNBC growth and survival are

6 fuelled by reprogramming branched-chain amino acids (BCAAs) metabolism, which rewires

7 oncogenic gene expression and cell signaling pathways. A regulatory kinase of the rate-limiting

8 enzyme of the BCAA catabolic pathway, branched-chain ketoacid dehydrogenase kinase

9 (BCKDK), have recently been implicated in driving tumor cell proliferation and conferring drug

10 resistance by activating RAS/RAF/MEK/ERK signaling. However it remains unexplored if

11 BCKDK remodels TNBC proliferation, survival and susceptibility to DOX-induced genotoxic

12 stress. TNBC cell lines exhibited reduced BCKDK expression in response to DOX. Genetic and

13 pharmacological inhibition of BCKDK in TNBC cell lines displayed reduced intracellular and

14 secreted BCKAs. Moreover, BCKDK inhibition with concurrent DOX treatment exacerbated

15 apoptosis, caspase activity and loss of TNBC proliferation. Transcriptome analysis of BCKDK

16 silenced cells confirmed a marked upregulation of the apoptotic signaling pathway with increased

17 protein ubiquitylation and compromised mitochondrial metabolism. BCKDK silencing in TNBC

18 downregulated mitochondrial metabolism genes, reduced electron complex protein expression,

19 oxygen consumption and ATP production. Silencing BCKDK in TNBC upregulated sestrin 2 and

20 concurrently decreased protein synthesis and mTORC1 signaling. Inhibiting BCKDK in TNBC

21 remodel BCAA flux, reduces protein translation triggering cell death, ATP insufficiency and

22 susceptibility to genotoxic stress.

1 Keywords: Triple-negative breast cancer, branched-chain ketoacid dehydrogenase kinase,

2 apoptosis, mitochondria, protein synthesis, chemosensitivity

4 Abbreviations: The abbreviations used are: Acadsb, Acyl-CoA dehydrogenase short branched

5 chain; ATM, ataxia-telangiectasia mutated; BCAA, branched-chain amino acid; BCKA, branched-

6 chain keto acid; BCAT1, branched-chain aminotransferase; BCKDHA, Branched chain keto acid

7 dehydrogenase E1 alpha polypeptide; BCKDHB, Branched chain keto acid dehydrogenase E1

8 subunit beta; BCKDK, branched-chain keto acid dehydrogenase kinase; BT2, 3,6-

9 dichlorobenzo[b]thiophene-2-carboxylic acid; eEF2, eukaryotic elongation factor 2; eEF2K,

10 eukaryotic elongation factor 2 kinase; eIF2α, eukaryotic translation initiation factor 2A; HADHA,

11 Hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunit alpha; HIBCH, 3-

12 hydroxyisobutyryl-CoA hydrolase; HIBADH, 3-hydroxyisobutyrate dehydrogenase; HSP90AB1,

13 Heat shock protein HSP 90-beta; KLF15, Kruppel like factor 15; mTORC1, mammalian target of

14 rapamycin; PARP, poly (ADP-ribose) polymerase; PPM1K, Protein phosphatase Mg2+/Mn2+

15 dependent 1K; SESN2, sestrin2.

1 Introduction

2 Triple-negative breast cancers (TNBCs) are an aggressive subtype of breast cancer, constituting

3 10-15% of all breast cancers and is defined by the lack of expression of estrogen receptor (ER),

4 progesterone receptor (PR), and the human epidermal growth factor receptor 2 (HER2)1.

5 Depending on the severity and stage of the disease2, TNBCs are generally treated in different

6 combinations with genotoxic chemotherapies including anthracycline, taxane, docetaxel and

7 cyclophosphamides. Doxorubicin (DOX) is an anthracycline class of cytotoxic chemotherapeutic

8 that induces TNBC remission in about 30% of patients3; however, prolonged treatment with higher

9 doses of DOX result in cardiotoxicity4, limiting its efficacy. Furthermore, breast cancer cells

10 reprogram cellular metabolism and function to evade toxicity of chemotherapeutic agents, thereby

11 exhibiting chemoresistance5. Metabolic maladaptation in TNBCs includes increased reliance on

12 pentose phosphate pathway for NADPH, lysosomal and mitochondrial drug sequestration, altered

13 nutrient metabolism, mitochondrial reprogramming, increased polyamine biosynthesis, glycolysis

14 and glutaminolysis and amino acid uptake for protein biosynthesis5. Therefore, identifying

15 mechanisms by which TNBCs survive DOX cytotoxicity and targeting the vulnerabilities in

16 metabolic pathways that render TNBCs resistant to chemotherapy will enable developing

17 concurrent therapies co-administered with lower doses of DOX.

18 Amino acid uptake and utilization facilitate the uninterrupted synthesis of TCA intermediates and

19 proteins for increased nitrogen and proliferative demand of TNBCs6. Cancer cells tightly control

20 the regulation of branched-chain amino acids (BCAAs), namely leucine, isoleucine and valine,

21 which contribute to 20% of the TCA cycle intermediates. Besides direct incorporation into

22 proteins, catabolism of BCAAs produces intermediates like glutamate vital for driving TNBC

23 growth and survival7. BCAAs are reversibly transaminated to branched-chain ketoacids (BCKAs)

1 by branched-chain aminotransferase (BCAT). BCKAs are further catabolized by branched-chain

2 alpha-ketoacid dehydrogenase (BCKDH), ultimately generating TCA cycle intermediates.

3 BCKDH is inhibited by phosphorylation at Ser 293 residue by the kinase, branched-chain alpha-

4 ketoacid dehydrogenase kinase (BCKDK), while this phosphorylation is reversed by the

5 Mg2+/Mn2+ Dependent 1K (PPM1K) protein phosphatase7. Recently in glioblastomas and

6 pancreatic cancer, a role for BCKAs was uncovered in metabolic homeostasis8, immune

7 suppression, and cell death evasion9. Indeed, degrading BCAAs and oxidizing BCKAs decreases

8 cancer cell survival10. Although the role of BCKDH complex and BCAT isoforms (cytosolic

9 BCAT1 or mitochondrial BCAT2) have been widely explored in different cancers11, the role of

10 BCKDK remains mostly elusive. Recent studies have identified novel phosphorylation sites on

11 BCKDK that stabilize BCKDK activity and influence tumor cell proliferation and survival12,13.

12 Since reprogramming of BCAA metabolism produces intermediates that rewire oncogenic gene

13 expression and cell signaling pathways in TNBC11, the importance of BCKDK in maintenance and

14 utilization of BCAA pool, regulation of BCAA fate and BCAA flux into BCKA in TNBC growth

15 and survival merits investigation.

16 In the current study, we investigated the role of BCKDK in TNBC survival and metabolism and

17 examined whether DOX’s anticancer effect involves targeting the BCAA catabolic pathway. DOX

18 treatment at 2µM suppressed BCKDK and significantly increased BCAA degradation enzyme

19 mRNA and protein levels in two different TNBC cell lines. Transcriptome analysis of MDA-

20 MB231 cells with BCKDK silencing revealed a marked increment in apoptotic signaling and

21 deregulated mitochondrial function and biogenesis. Genetic and pharmacological inhibition of

22 BCKDK sensitized TNBCs to DOX-induced cytotoxicity. BCKDK inhibition suppressed

23 mitochondrial function, reduced nascent protein synthesis and increased sestrin 2 (SESN2)

1 expression, a negative regulator of protein synthesis and cell proliferation. Collectively, our data

2 highlight the role of BCKDK as a novel DOX sensitive target in TNBCs. Targeting BCKDK can

3 be an attractive therapeutic target in regulating TNBC survival, proliferation, and potentiating

4 chemosensitivity to DOX.

1 Results

2 Altered BCAA catabolic enzyme expression in TNBCs is sensitive to DOX treatment

3 BCAA catabolic enzyme expression was examined in two TNBC cell lines, BT549 and MDA-

4 MB231. In agreement with prior studies14, we found a striking upregulation of BCAT1 mRNA

5 levels, specifically in MDA-MB231 cells, while BCAT2 mRNA (Fig 1A) and protein (Fig S1B)

6 levels were significantly downregulated in both BT549 and MDA-MB231 cells when compared

7 with normal breast epithelial cell line, MCF10A. Treatment with 2µM DOX reduced BCAT1 and

8 further reduced BCAT2 mRNA expression in both TNBCs (Fig S1C-D). BCKDHA and BCKDHB

9 protein levels were significantly downregulated in TNBC cell lines (Fig S1B), although their

10 mRNA levels were selectively reduced in the MDA-MB231 cells (Fig 1A). DOX treatment

11 increased BCKDHA mRNA (Fig 1B) and protein (Fig 1D) levels in MDA-MB231 cells. In the

12 TNBC cells, mRNA (Fig S1A) and protein (Fig S1B) expression of KLF15, the transcriptional

13 activator of the BCAA pathway, was significantly suppressed. DOX treatment significantly

14 increased KLF15 mRNA (Fig 1B-C) expression in both TNBCs and KLF15 protein content

15 specifically (Fig 1D) in MDA-MB231 cells. The BCKDH phosphatase, PPM1K, mRNA levels

16 were also decreased in TNBCs (Fig 1A) and were upregulated in response to DOX (Fig 1B-C).

17 mRNA expression of distal enzymes, ACADSB, HADHA, HIBCH were unchanged (Fig S1A) and

18 DOX increased the mRNA expression of HADHA in both TNBCs (Fig S1C-D). These data suggest

19 that in TNBCs BCAA catabolic enzyme expression is suppressed, and treatment with DOX

20 counteracts this suppression, likely affecting BCAA catabolism. Since the role of BCAT and

21 BCKDH were examined in prior studies, we focussed on studying the role of BCKDK, the kinase

22 which regulates the flux of BCKAs towards oxidation by inactive phosphorylation of BCKDH.

23 BCKDK transcripts were markedly upregulated in BT549 and marginally increased in MDA-

1 MB231 cells (Fig 1A). Although BCKDK protein expression was augmented in BT549, it

2 remained comparable (p=0.08) to MCF10A cells in MDA-MB231 cells (Fig S1B). The

3 phosphorylated to total BCKDE1α S293 content was increased in BT549 but not in MDA-MB231

4 and MCF10A cells (Fig S1B). DOX treatment reduced BCKDK mRNA levels in both TNBCs (Fig

5 1B-C). 2µM DOX treatment reduced BCKDK protein levels and the corresponding

6 phosphorylated BCKDE1α S293 levels in MDA-MB231 (Fig 1D) and BT549 cells (Fig S1E).

7 Since BCKDK expression was altered in TNBCs and regulated by DOX, we next examined the

8 significance of BCKDK in TNBC metabolism and proliferation.

9 Downregulating BCKDK inhibits proliferation, promotes apoptosis and potentiates DOX

10 mediated cell death

11 To decipher whether BCKDK silencing alters metabolism and survival of TNBCs, MDA-MB231

12 cells were treated with either control siRNA (siCON) or siRNAs targeting BCKDK exon 2

13 (siBDK#1) or exon 7 (siBDK#2). BCKDK silencing was confirmed by decreased mRNA (Fig

14 S2A) and protein (Fig S2B) levels. Silencing BCKDK using siBDK#2 resulted in reduced cell

15 count at both 24h and 48h following transfection, while siBDK#1 showed reduced cell count at

16 48h (Fig S2C), suggesting that targeting BCKDK influences TNBC survival and proliferation.

17 BCKDK deletion also resulted in significant upregulation of cleaved caspase 3 content (Fig S2B),

18 indicating increased cell death. In an antibody array of apoptosis-associated proteins, levels of the

19 anti-apoptotic protein, survivin were decreased and pro-apoptotic protein, BIM and cytochrome C

20 levels were increased in BCKDK silenced MDA-MB231 cells while potentiating DOX mediated

21 cytochrome C, Fas, BIM and caspase 8 expression (Fig 2A). Further, BCKDK knockdown

22 exacerbated DOX mediated increases in cleaved caspase 3, caspase 7 and PARP levels in MDA-

23 MB231 cells (Fig 2B). In BCKDK depleted cells, exposure to DOX increased phosphorylation of

1 the DNA damage marker, ataxia-telangiectasia mutated (ATM), suggesting increased genomic

2 instability and apoptosis (Fig 2B). Similar increases was observed in cleaved caspase 3/7 and BAX

3 Bax expression in DOX-treated BT549 cells with silencing BCKDK (Fig S2D). To rule out off-

4 target effects of siRNA mediated gene silencing, we also performed the experiments with

5 adenoviral-mediated knockdown of BCKDK in both MDA-MB231 and BT549 cells and observed

6 similar increases in cleaved caspase levels in response to DOX (Fig S2E). Changes in protein

7 levels were also reflected in increased caspase activity in BCKDK depleted MDA-MB231 (Fig

8 2C) and BT549 (Fig S2F) cells treated with DOX. Cytotoxicity of DOX was evident by increased

9 lactate dehydrogenase (LDH) release into the media, which was exacerbated further following

10 BCKDK silencing (Fig 2D).

11 To examine if the increased cytotoxic response of MDA-MB231 cells to DOX following BCKDK

12 knockdown was associated with decreased proliferation, CCK-8 assay was performed. BCKDK

13 knockdown alone reduced cell proliferation by 2d (siBDK#2) or 3d (siBDK#1) which was

14 potentiated further by DOX treatment (Fig 2E). Additionally, MDA-MB231 cells transduced with

15 shBCKDK were treated with DOX for 18h followed by a chase in drug-free media for 48h. Cell

16 viability was determined by using the Presto Blue reagent wherein reduction in resazurin is the

17 surrogate readout for metabolic activity. Knockdown of BCKDK resulted in a significant decrease

18 in metabolic activity that further declined by exposure to DOX (Fig S2G).

19 We next recapitulated the cytotoxic effects of BCKDK silencing by treating cells with the

20 BCKDK inhibitor, BT2, in the presence and absence of DOX. BT2 treatment alone resulted in

21 increased levels of cleaved PARP, caspase 3 and caspase 7 and phosphorylated ATM in MDA-

22 MB231 cells (Fig 3A). BT2 also potentiated DOX’s effects on cleaved caspase3, cleaved PARP

23 and pATM levels in BT549 cells (Fig 3B), with the maximal effects observed at 500 µM. BT2

1 markedly exacerbated DOX mediated increase in caspase activity in MDA-MB231 (Fig 3C) and

2 BT549 (Fig 3D) cells. Moreover, in MDA-MB231 cells BT2 alone increased LDH release, and

3 this effect was further exacerbated in the presence of DOX (Fig 3E).

4 Silencing BCKDK alters TNBC transcriptome

5 We next aimed to identify metabolic and signaling pathway networks regulated by BCKDK to

6 influence TNBC cell death and proliferation. MDA-MB231 cells depleted of BCKDK were

7 subjected to RNA-Seq transcriptomic analysis. Principle component analysis and distance

8 clustering showed that the transcriptome of cells treated with siBDK#1 and siBDK#2 displayed

9 little similarity (Fig 4A-B), thus we classified genes as differentially expressed compared with

10 control if the adjusted P-value was below 0.01 for each siRNA individually and the fold change

11 was occurring in the same direction. This method identified 1024 genes, which were differentially

12 expressed by both BCKDK siRNA’s and over-represented Gene Ontology (GO) were determined

13 using DAVID (Fig 4C). Among the top 20 GO terms identified, activation of response to unfolded

14 protein response (UPR) and positive regulation of ubiquitin-dependent protein catabolism

15 appeared multiple times (Fig 4D). Both these processes play a crucial role in regulating apoptosis

16 by direct or indirect targeting of important regulators of apoptosis and caspases15,16. Indeed, the

17 GO term related to activation of caspases, the cysteine-type endopeptidase, was enriched in the

18 differentially expressed genes from BCKDK knockdown cells (Fig 2D), consistent with our data

19 (Fig 2-3). Genes annotated to this GO term include the initiator or activator caspases, including

20 CASP2, CASP8, CASP9, and CASP10, also called initiator (or apical, or activator) caspases,

21 suggesting that BCKDK silencing is plausibly sensitizing TNBC to cell death. Indeed,

22 differentially expressed genes involved in death receptor signaling was upregulated after BCKDK

23 knockdown, including TNF-related apoptosis-inducing ligand (TRAIL/TNFSF10) and

1 TNFRSF1A associated via death domain (TRADD) (Fig 4E). The expression of pro-apoptotic

2 factor, Bax and Smad3, involved in activating TGF-β-induced apoptosis, increased in BCKDK

3 depleted cells (Fig 4E), indicating augmented pro-apoptotic gene expression.

4 BCKDK inhibition decreases intracellular and secreted BCKAs

5 We next ascertained the consequences of BCKDK silencing in influencing BCKA flux by

6 measuring mRNA expression of BCAA catabolic enzymes that facilitate BCKA oxidation. RNA-

7 Seq analysis revealed that BCKDK knockdown resulted in increased expression of genes involved

8 in BCKA oxidation, specifically BCKDHA, methylmalonyl-CoA epimerase (MCEE), propionyl-

9 CoA carboxylase (PCCB), HADHB and ACADM (Fig S3A). BCKDK depleted cells also displayed

10 reduced expression of BCAT1 and HIBCH (Fig S3A), genes augmented in several cancers17. qPCR

11 analysis revealed significant increases in the BCKDHA and KLF15 mRNA expression in MDA-

12 MB231 (Fig S3B), although it was only observed with siBDK#1. No changes were observed in

13 the mRNA expression of BCAA catabolic enzymes in the BT549 cells (Fig S3C). Contrary to

14 BCKDK silencing, BT2 treatment resulted in significant increases in KLF15, BCKDHA,

15 BCKDHB, HIBCH and HIBADH mRNA levels in BT549 (Fig S3D). We next ascertained whether

16 increased expression of BCAA degradation genes resulted in reduced intracellular accumulation

17 and secretion of BCKAs. Adenoviral knockdown of BCKDK in MDA-MB231 cells resulted in

18 decreased intracellular BCKAs accumulation as revealed by UPLC-MS/MS analysis (Fig S3E). A

19 significant reduction in total BCKAs was observed in the cells and in the media in MDA-MB231

20 cells treated with BT2 (Fig S3F-G). Previous studies have demonstrated the role of tumor secreted

21 BCKAs in proliferation and immune suppression of cancer cells8,9. Our data suggest that the

22 mechanism by which BCKDK depletion arrests TNBC proliferation likely involves increased

23 BCKA oxidation with a concomitant reduction in accumulation and secretion of BCKAs.

1 BCKDK inhibition results in decreased mitochondrial complex protein expression and impaired

2 mitochondrial function

3 We next ascertained if increased cell death following BCKDK silencing is an outcome of

4 mitochondrial dysfunction. Indeed, permeabilized and damaged mitochondria generate reactive

5 oxygen species in a complex I and II-dependent manner18 and when recognized by activated

6 caspase is targeted for destruction. Transcriptomic analysis of BCKDK depleted MDA-MB231

7 cells revealed a marked downregulation of genes involved in mitochondrial function and

8 biogenesis (Fig 5A). For instance, BCKDK silencing reduced nucleoside diphosphate kinase

9 (NME4) and clusterin (Clu) gene expression, which are involved in promoting redistribution of

10 cardiolipin to prevent mitochondrial permeabilization and suppressing Bax dependent release of

11 cytochrome C, respectively (Fig 5A). Genes involved in mitochondrial translation, mitochondrial

12 ribosome construction and abundance of 16S mt-rRNA, such as G elongation factor mitochondrial

13 1 (GFMI), mitochondrial ribosomal protein L19 (MRPL19) and RNA pseudouridine synthase D4

14 (RPUSD4), were also downregulated upon BCKDK silencing (Fig 5A). Furthermore, genes

15 involved in complex I assembly (NADH:ubiquinone oxidoreductase assembly factor 1,

16 NDUFAF1), the formation of complex III (tetratricopeptide repeat domain 19, TTC19),

17 suppression of ROS (superoxide dismutase 2, SOD2) and synthesis of nucleotide triphosphates

18 and mitochondrial respiration (NME4, serine active site containing 1, SERAC) were

19 downregulated in BCKDK depleted cells (Fig 6A). Additionally, mitochondrial genes involved in

20 pyruvate transport (mitochondrial pyruvate carrier 1, MPC1), glycine metabolism (glycine

21 cleavage system protein H, GCSH and glycine c-acetyltransferase, GCAT) and ether lipid

22 synthesis and intracellular cholesterol trafficking (peroxisomal alkylglycerone phosphate

23 synthase, AGPS and mitochondrial SERAC1, respectively) are also downregulated upon BCKDK

1 silencing (Fig 5A). Consistent with the transcriptomic data, silencing BCKDK in MDA-MB231

2 cells resulted in decreased complex I, complex II and complex III protein levels (Fig 5B).

3 Extracellular flux analysis studies revealed reduced basal oxygen consumption rates (OCR) (Fig

4 5C), ATP production (Fig 5D) and non-mitochondrial respiration (Fig S4A) in the BCKDK

5 silenced cells. Inhibiting BCKDK by BT2 also resulted in reduced basal and maximal OCR (Fig

6 5E) and ATP production (Fig 5F), indicating impaired mitochondrial oxidative metabolism. Lower

7 OCR also resulted in reduced spare respiratory capacity (Fig 5G) and proton leak (Fig 5H) as well

8 as low non-mitochondrial respiration (Fig S4B), suggestive of impaired adaptation to metabolic

9 changes. Mitochondrial DOX accumulation can also disrupt mitochondrial architecture and

10 oxidative capacity19. We determined whether BCKDK silencing in MDA-MB231 cells can

11 potentiate DOX’s effects, exacerbating mitochondrial dysfunction. DOX treatment reduced basal

12 OCR (Fig 5C), ATP production (Fig 5D) and non-mitochondrial respiration (Fig S4A), but it was

13 not significantly potentiated upon BCKDK silencing. Unlike TNBCs, normal breast epithelial

14 cells, MCF10A, when treated with BT2, did not show suppressed mitochondrial respiration or

15 ATP production (Fig S4 C-G).

16 Silencing BCKDK reduces protein synthesis in TNBCs

17 Previous studies have demonstrated that induction of ER stress and mitochondrial dysfunction20,

18 induces apoptosis21 and inhibits protein translation and synthesis. As measured by puromycin

19 incorporation, protein synthesis was reduced in BCKDK depleted BT549 cells (Fig 6A). Similarly,

20 azidohomoalaine (AHA) labelled nascent protein levels were decreased upon BCKDK silencing

21 (Fig 6B). A similar reduction in puromycin incorporation was observed in BT549 cells treated

22 with BT2 (Fig 6C). In both BT549 and MDA-MB231 cells, BT2 treatment resulted in a robust

23 increase in sestrin 2 (SESN2) content (Fig 6D), a negative regulator of mammalian target of

1 rapamycin (mTOR) and protein synthesis22. Consistently, mTOR signaling was suppressed in BT2

2 treated cells as observed by reduced phosphorylation on mTOR S2448, P70S6K T389 and S6

3 S240/244 (Fig 6D). Similar increases in SESN2 levels were observed in MDA-MB231 cells

4 adenovirally transduced with shBCKDK in the presence or absence of DOX (Fig 6E). Further,

5 BCKDK inhibition in the presence or absence of DOX was accompanied by increased inactivating

6 phosphorylation of eIF2α at S51 (Fig 6E), suggesting inhibited protein translation. BCKDK

7 inhibition also increased phosphorylation of the elongation factor eEF2 at T56 (Fig 6E), which

8 also correlated with suppressed protein translation.

1 Discussion

2 TNBCs are a heterogeneous breast cancer subtype with early relapse rates, poor prognosis and

3 limited therapeutic options1. Sustaining oncogenesis in TNBCs is dependent on BCAAs which

4 contribute to protein biosynthesis, nitrogen homeostasis, epigenetic regulation, redox balance,

5 immune response and tumor surveillance, growth and metastasis23,24. Numerous types of human

6 cancers exhibit significant reprogramming of the BCAA metabolic pathway11. Nevertheless, the

7 contribution of BCAA catabolic dysregulation in TNBC pathology, cell function and

8 chemoresistance merit investigation. In the current study, we observed consistently suppressed

9 expression of BCAA catabolic enzymes along with increased mRNA expression of the regulatory

10 kinase, BCKDK, in two TNBC cell lines. Transcriptome analysis demonstrated that genes

11 involved in caspase-dependent apoptosis and mitochondrial function were differentially enriched

12 in MDA-MB231 cells with BCKDK silencing. BCKDK inhibition also sensitized TNBCs to DOX-

13 induced apoptosis, mitochondrial ATP insufficiency and suppression of mTOR dependent protein

14 translation by SESN2. For the first time, our data proposes a novel role of BCKDK in TNBC

15 survival, proliferation, and DOX chemosensitivity.

16 Elevated intratumoral BCAAs are reported in several cancers, including breast cancer14,

17 PDAC25,26, HCC27, and NSCLC28, while increased BCKAs secretion is observed in glioblastomas

18 and PDACs8,9. However, not only BCAA metabolites but also BCAA catabolizing enzymes have

19 recently observed to influence cancer outcomes11 . Indeed, upregulation of BCAT1 is reported in

20 glioblastomas, HCC, leukemias, osteosarcomas, ovarian and endometrial cancer27,29,11. BCAT2 is

21 upregulated in PDAC25,26, luminal type breast cancer14, NSCLC28, while it is downregulated in

22 HCC27. BCKDH is upregulated in leukemia and luminal type breast cancer30,14 and suppressed in

23 HCC and PDAC27,28. Concomitantly, BCKDK expression was increased in HCC however, more

1 distal enzymes of BCAA catabolism, ACADS and ACADSB were downregulated27. Our data are

2 in agreement with these findings demonstrating dysregulated BCAA catabolic enzyme expression

3 in MDA-MB231 and BT549 cells, highlighting the involvement of BCAA catabolizing enzymes

4 and metabolites in tumorigenesis.

5 Two studies have recently identified different phosphorylation sites on BCKDK that mediates

6 cancer proliferation, signaling and metastasis12,13. Notably, stabilizing BCKDK by Src promoted

7 migration, invasion and metastasis in colorectal cancers12. Our data in TNBCs demonstrate that

8 inhibiting BCKDK impairs TNBC proliferation and sensitizes TNBCs to DOX toxicity. We

9 propose that the metabolic effects of BCKDK are likely an outcome of altered BCAA-BCKA flux

10 towards oxidation and away from protein synthesis. Indeed, increased intracellular BCAA

11 accumulation upon BCKDHA deletion resulted in increased cell proliferation in an immortalized

12 hepatocyte cell line27. Moreover, increased extracellular secretion of BCKAs in a BCAT1

13 dependent manner supports cell proliferation in PDACs8 and survival of glioblastoma cells by

14 evading immune surveillance9. Our study also demonstrated that pharmacological inhibition of

15 BCKDK was sufficient to reduce both intracellular and secreted BCKAs in MDA-MB231 cells,

16 plausibly contributing to reduced cell proliferation. Since TNBCs are classically treated with

17 combination therapy involving anthracyclines2, such as DOX, it is plausible that targeting

18 BCKDK might trigger a metabolic vulnerability rendering TNBCs susceptible to DOX toxicity.

19 BCKDK is recently reported to act as a regulatory, upstream kinase of MEK31 and ERK1/213,

20 drivers of cell proliferation. Prior studies have inferred that BCKDK stabilization is an alternative

21 mechanism to activate RAS/RAF/MEK/ERK signaling and confer drug resistance31. It remains to

22 be determined whether the accumulation of BCAA metabolites, such as BCKAs, activates

23 oncogenes and drives growth factor signaling in TNBCs. We theorized whether BCKDK silencing

1 with concomitant DOX treatment potentiates BCKA oxidation reducing intracellular BCKAs

2 compromising growth and increasing cell death. In our study, we observed that when TNBCs were

3 exposed to DOX, BCKDK expression was significantly downregulated. Moreover, silencing

4 BCKDK markedly exacerbated cytotoxic effects of DOX in a caspase 3/7 dependent manner and

5 reduced cell proliferation. Although the mechanism by which BCKDK silencing triggers apoptosis

6 induction remains unclear, we theorize that perturbations in protein synthesis and altered

7 mitochondrial metabolism might precede or follow apoptotic events following BCKDK inhibition.

8 Dysregulation of protein translation and anomalous energy metabolism are characteristic of

9 cancers, processes that are sensitive to the effects of chemotherapeutic agents32,33.

10 Chemotherapeutics such as DOX can accumulate within the mitochondria resulting in

11 mitochondrial dysfunction and energy stress by disrupting the electron transport chain (ETC) and

12 reducing mitochondrial oxidative capacity19. Indeed, a decline in ATP synthesis is associated with

13 mitochondrial accumulation of DOX34. We found that BCKDK inhibition markedly reduced ATP

14 production and oxygen consumption in MDA-MB231 cells but did not exacerbate DOX mediated

15 suppression of respiration or energy production. Notably, BCKDK deficient fibroblasts show

16 reduced intracellular ATP and respiration, with corresponding increase in ROS production and

17 mitochondrial hyperfusion35. Interestingly, targeting BCKDHA in PDACs did not affect

18 mitochondrial metabolism or oxygen consumption26, suggesting the effects of BCKDK on

19 mitochondrial function and metabolism might be independent of BCKDHA. Moreover, our

20 transcriptomics data in BCKDK depleted cells identified reduced expression of several

21 mitochondrial genes associated with complex formation, glycine metabolism and cholesterol

22 trafficking. However, it remains to be determined whether targeting BCKDK in TNBCs impairs

1 mitochondrial function due to altered mitochondrial biogenesis, mitophagy mitochondrial DOX

2 accumulation or because of reduced entry of BCAA derived TCA cycle intermediates.

3 Changes in the metabolic microenvironment within TNBCs hyperactivates signaling pathways of

4 protein translation and biosynthesis, enabling uncontrolled growth and survival. Protein synthetic

5 rates are significantly reduced following apoptosis induction, where caspase-dependent proteolytic

6 cleavage and altered phosphorylation states of translation initiation factors (eIF4GI, eIF4GII,

7 eIF4E, eIF4B and eIF2α) influence cancer progression21,36. We found that BCKDK inhibition

8 decreased protein synthesis. Consistent with this finding, upon BCKDK depletion we also

9 observed significant increase in the inhibitory phosphorylation of the elongation factor, eEF2 at

10 Thr 56. eEF2 phosphorylation impedes translocation of peptidyl-tRNA from the A site to the P

11 site on the ribosome. High intracellular Ca2+ levels increase phosphorylation of eEF237. Our

12 transcriptomic data in BCKDK depleted cells revealed enrichment of the GO-term related to

13 NFAT-calcineurin signaling cascade, which in turn is activated by high intracellular Ca2+ levels, a

14 likely trigger for increased eEF2 phosphorylation. eEF2 is downstream of the AMPK-mTORC1

15 axis where it is inhibited by AMPK and activated by mTORC1. Our data in both TNBC cell types

16 showed impaired mTORC1 signaling upon BCKDK inhibition and significant upregulation of

17 SESN2, the negative regulator of mTORC1. The suppressed activation of mTORC1 signaling can

18 explain the inactivation of eEF2. Additionally, reduction in protein synthesis following BCKDK

19 inhibition was associated with increased inhibitory phosphorylation of eIF2α at Ser 51. eIF2α

20 phosphorylation increases in response to stress and cancer36. eIF2α phosphorylation at Ser 51

21 inhibits the eIF4B catalyzed guanine nucleotide exchange function of eIF2 that prevents the

22 formation of the 43S preinitiation complex21. It is unclear whether a decline in mTORC1 driven

1 protein synthesis induces apoptosis38 or increased apoptosis suppress global translation or there is

2 a diminished amino acid availability in TNBCs with inhibited BCKDK.

3 In conclusion, our study highlights BCKDK as a target for potentiating cytotoxic effects of DOX

4 in TNBCs. This study demonstrates that BCKDK inhibition alone is sufficient to induce ATP

5 insufficiency and decrease protein translation, caspase activation processes rendering TNBCs

6 susceptible to cell death. Augmented BCKA levels due to BCKDK silencing can be fated to either

7 get reaminated to BCAAs or get oxidised. Since BCAAs (majorly leucine) is recognised as a key

8 amino acid to activate mTOR, the former fate for BCKAs is unlikely in our study as mTOR

9 signaling is suppressed. However, changes in mTOR signaling may or may not be dependent on

10 changes in BCAA levels. Further studies are required to identify whether BCKDK’s role in

11 TNBCs are mediated via increased BCKA oxidation or is independent of the changes in

12 intracellular BCKA flux. Moreover, it needs to be determined how BCKDK silencing triggers the

13 activation of caspases and the apoptotic pathway and if p53 mediates these effects. DOX’s

14 genotoxic stress also involves DNA intercalation and double-strand breaks; whether BCKDK

15 silencing confers potentiated sensitivity to DOX by incrementing DNA damage remains to be

16 explored. Since BCKDK was recently identified as a cytosolic kinase regulating hepatic lipid

17 accumulation39, future studies will involve the identification of other metabolic pathways that co-

18 operate with BCAA metabolism to promote tumor growth or alternatively compensate for

19 sustaining cell proliferation when BCAA metabolism is inhibited.

1 Materials and Methods

2 Cell lines and culture conditions

3 MCF10A cells (CRL-10317, US) and BT549 cells (HTB-122) were obtained from ATCC. MDA-

4 MB231 cells were a gift from Dr. G. Robichaud (Université de Moncton). The cells were cultured

5 according to our previously published studies3. All the cell lines were maintained at 37°C in a

6 humidified atmosphere of 5% CO2. The cell lines have not been genetically authenticated. For the

7 BT2 experiments, cells were pretreated for 20h with the desired concentration of BT2 (Sigma and

8 Matrix Scientific). BT2 from Sigma was dissolved in cremaphore solution, generously gifted by

9 Ayappan Subbiah (Sevengenes Inc.).

10 Transfections and viral transductions

11 Adenoviral vector expressing BCKDK shRNA (shADV-225358) and the control vector for

12 scrambled shRNA GFP (Cat# 1122) were obtained from Vector Biolabs (USA). Adenoviral

13 infection of cells was done 24h post-plating, and the multiplicity of infection (MOI) was kept

14 constant between the control and experimental constructs. siRNA knockdown of BCKDK was

15 performed using Ambion silencer select siRNA oligonucleotides (Thermo-Fisher Scientific Cat#

16 4390824). The siRNAs used in this paper were siBCKDK#1: #s20126, siBCKDK#2: #s20127,

17 siRNA negative control (Cat# 4390844). Transfection was achieved using Lipofectamine

18 RNAiMAX (Invitrogen) following the manufacturer’s instructions with a concentration of 10 nM

19 of siRNA per plate.

20 Cell lysate processing and immunoblotting

21 Cell pellets were sonicated in ice-cold lysis buffer (containing 20 mM Tris-HCl, pH 7.4, 5 mM

22 EDTA, 10 mM Na4P2O7 (Calbiochem), 100 mM NaF, 1% Nonidet P-40, 2 mM Na3VO4, protease

23 inhibitor (10 ml/ml; Sigma), and phosphatase inhibitor (10 ml/ml; Calbiochem) and centrifuged at

1 16,000g for 15 min. Protein concentration determination, immunoblotting and development of

2 immunoblots were performed as described previously40. List of primary antibodies is described in

3 Table S1. Densitometric analysis was performed using Image laboratory software (Bio-Rad), and

4 the quantifications were normalized by total protein loading using GraphPad software (Clarivate).

5 qPCR analysis

6 mRNA levels of BCAA-catabolizing enzymes were determined using qPCR by employing

7 validated optimal reference gene pairs. Primer information of the target and reference genes are

8 provided in Table S2. RNA isolation, quality control, cDNA synthesis and qPCR analysis were

9 performed as described previously41.

10 Trypan blue exclusion and metabolic activity assays

11 For cell viability, 1 × 104 cells were plated in 24-well plates in triplicates. Total number of viable

12 cells were calculated using the trypan blue exclusion assay. For measuring metabolic activity, cells

13 were plated in 96-well plates and treated with DOX for 18h followed by incubated in drug-free

14 media for 48h. Metabolic activity of viable cells were measured by incubating the cells with Presto

15 Blue (Thermo-Fischer Scientific) for 3h at 37°C. Fluorescent intensity was read at Ex/Em=

16 560/590 on a microplate fluorometer (Synergy H4).

17 CCK8 assay

18 Cell Counting Kit-8 (CCK-8; Sigma Aldrich) was used to measure cell viability according to the

19 manufacturer’s protocol. Briefly, 1 x 105 MDA-MB231 cells were seeded in 96-well plates

20 transfected with the siRNAs for different time points followed by DOX treatment for 18h.

21 Following the treatments, the cells were incubated in CCK-8 solution at 37 ℃ for 4h. Synergy H4

22 microplate reader was used to measure the absorbance at 450nm.

1 Extracellular flux analyzer studies

2 MDA-MB231 or MCF10A cells were plated at a density of 30,000 cells/per well. Mitostress assay

3 was performed according to Khan et al.42 using the Seahorse XFe24 analyzer (Agilent

4 Technologies Inc. CA, USA). Briefly, cells were incubated with XF Assay medium (with 20 mM

5 glucose, 1 mM sodium pyruvate and 1 mM glutamate, without sodium bicarbonate) for 1 h. 1 μM

6 of oligomycin, 1 μM of FCCP [carbonyl cyanide‐4‐ (trifluoromethoxy)phenylhydrazone], and 1

7 μM each of rote‐ none and antimycin A was injected over 100 min to measure the different stages

8 of respiration. The assay was normalized with protein and analyzed with the XFe 2.0.0 software

9 (Agilent Technologies Inc. CA, USA).

10 Apoptosis array

11 Expression of 43 apoptosis-associated proteins in duplicates were determined by the Human

12 apoptosis antibody array (ab134001, Abcam). Membranes were incubated with 250μg of cell

13 extracts, immunoblotted and developed using the chemidoc (BioRad) according to manufacturers’

14 instructions. Densitometric analysis was performed using Image laboratory software (Bio-Rad).

15 Lactate dehydrogenase cytotoxicity assay

16 Lactate dehydrogenase (LDH) cytotoxicity assay was performed using the assay kit (Invitrogen,

17 C20301) according to the manufacturer’s instructions. Briefly, 1x104 cells were plated in a 96 well

18 plate for experimental conditions as well as spontaneous and maximum LDH activity controls. On

19 the day of the assay, 10 μl of 10X lysis buffer was added to wells serving as the maximum LDH

20 activity controls and incubated at 37°C for 45 min. The reaction mixture was added to the collected

21 media in a 1:1 ration and incubated at RT for 30min in the dark, followed by adding a stop solution.

22 The absorbance was measured at 490 nm, and 680 nm and the assay was normalized by protein

23 concentration.

1 Caspase 3 activity assay

2 Caspase activity assay (Caspase-3 Activity Assay Kit #5723, Cell Signaling Technology, Danvers,

3 MA, USA) was performed according the manufacturer’s instructions. Briefly, 1x104 cells were

4 plated in a 96 well plate and were harvested using ice-cold lysis buffer. Lysates from two wells

5 within the same treatment group were combined for all treatment groups and transferred to a black

6 plate. Fluorescence was measured immediately at Ex/Em 380/440 for the 0h reading. A second

7 reading was taken after 1h incubation in the dark at 37°C. The assay was normalized by protein

8 concentration.

9 AHA incorporation assay

10 AHA incorporation analysis was performed using the Click-iT AHA Alexa Fluor 488 Imaging Kit

11 (Thermo-Fisher Scientific) according to the manufacturer’s instructions. Briefly, cells on

12 coverslips were pulse labelled with AHA for 30 min in methionine/cysteine free DMEM media

13 without serum before performing the click labeling. Coverslips were mounted in Prolong anti-fade

14 mounting media (Thermo-Fisher Scientific) and imaged with a Zeiss Axio Observer Z1 equipped

15 with an Apotome.2 structural illumination unit using a 20x Plan-Apochromat objective (NA: 0.8,

16 air). Images were processed for analysis in ImageJ, and image processing was identical for each

17 image set. Images were analyzed in Cell Profiler, and data are represented as the mean ± S.D. per

18 cell of three experiments.

19 RNA-seq analysis and bioinformatics

20 MDA-MB231 cells were transfected with either of two siRNA’s targeting BCKDK or a scrambled

21 siRNA and cultured for 48 h before cells were harvested and RNA extracted using the Qiagen

22 RNeasy mini kit according to the manufacturer’s instructions. RNA-Seq was conducted by McGill

23 University and the Genome Quebec Innovation Center (Montreal, Canada) with the Illumina

1 NovaSeq 6000 S2 PE100 — 50 M platform. Analysis of the transcriptomic data was performed

2 according to our previous published study3. Testing for differential expression was conducted only

3 on genes with an average estimated count of greater than 0.3. Genes were considered significantly

4 differentially expressed if the adjusted P-value was less than 0.01 for both siBCKDK#1 and

5 siBCKDK#2 groups, and the fold change was occurring in the same direction. RNA-Seq data is

6 desposited in NCBI’s Gene Expression Omnibus under the assecession number GSE163297.

7 SunSET method

8 Protein synthesis was measured in vitro by the SunSET assay as described previously 40. Briefly,

9 BT549 cells were either transfected with siBCKDK #1 or #2 and siCon for 48h or with 500µM

10 BT2 for 20h followed by treatment with 1 µM puromycin dihydrochloride (P8833, Sigma) for 30

11 min followed by a 30min chase with complete media. Puromycin incorporation was detected by

12 Western blotting using the monoclonal puromycin antibody.

13 BCKA measurements

14 For BCKA measurements, cells were incubated with serum-free DMEM low glucose, leucine-free

15 medium (D9443, Sigma) for 24h or during the duration of the treatment.

16 Secreted BCKA extraction

17 20 µl of the media and 120 µl of internal standard (4 mg/ml in H2O) containing leucine-d3 (CDN

18 Isotopes), 40 µl of MilliQ water, 60 µl of 4 M perchloric acid (VWR) were combined and vortexed.

19 Proteins were precipitated in two sequential steps, followed by centrifugation at 13,000 rpm for 15

20 min at 4 °C. Supernatants collected from both steps were combined for measuring BCKAs.

21 Intracellular BCKA extraction, BCKA derivatization and quantification were done according to

22 our previously published study40.

1 Statistical analysis

2 Data are expressed as mean±S.D. unless otherwise indicated. Statistical analyses were conducted

3 using Prism software (GraphPad, La Jolla, CA, USA). Comparisons between multiple groups were

4 performed using one-way or two-way analysis of variance followed by a Tukey post hoc test, as

5 appropriate. Data sets with two groups were analyzed using a two-tailed student’s t-test. p values

6 of 0.05 were considered statistically significant.

1 Declaration

2 Funding & Acknowledgement

3 This work was supported by Natural Sciences and Engineering Research Council of Canada

4 (RGPIN-2020-05906), Diabetes Canada (NOD_OG-3-15-5037-TP & NOD_SC-5-16-5054-

5 TP) and New Brunswick Health Research Foundation grants to T.P.; D.B., was funded by

6 postdoctoral fellowships from New Brunswick Health Research Foundation and Dalhousie

7 Medicine New Brunswick.

9 Conflict of Interest

10 The authors declare no conflict of interest.

12 Data availability

13 All data and materials used in the current study are available from the corresponding author upon

14 request. The datasets geberated and/or analysed during the current study are available in NCBI’s

15 NCBI’s Gene Expression Omnibus.

16 Accession ID: GSE163297

17 Databank URL: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE163297

1 References

2 1 Mehanna, J., Haddad, F. G., Eid, R., Lambertini, M. & Kourie, H. R. Triple-negative breast cancer: 3 current perspective on the evolving therapeutic landscape. Int J Womens Health 11, 431-437, 4 doi:10.2147/IJWH.S178349 (2019). 5 2 Harbeck, N. et al. Breast cancer. Nat Rev Dis Primers 5, 66, doi:10.1038/s41572-019-0111-2 6 (2019). 7 3 Slade, L. et al. A lysosome independent role for TFEB in activating DNA repair and inhibiting 8 apoptosis in breast cancer cells. Biochem J 477, 137-160, doi:10.1042/BCJ20190596 (2020). 9 4 Octavia, Y. et al. Doxorubicin-induced cardiomyopathy: from molecular mechanisms to 10 therapeutic strategies. J Mol Cell Cardiol 52, 1213-1225, doi:10.1016/j.yjmcc.2012.03.006 (2012). 11 5 Chen, X., Chen, S. & Yu, D. Metabolic Reprogramming of Chemoresistant Cancer Cells and the 12 Potential Significance of Metabolic Regulation in the Reversal of Cancer Chemoresistance. 13 Metabolites 10, doi:10.3390/metabo10070289 (2020). 14 6 Pavlova, N. N. & Thompson, C. B. The Emerging Hallmarks of Cancer Metabolism. Cell Metab 15 23, 27-47, doi:10.1016/j.cmet.2015.12.006 (2016). 16 7 Biswas, D., Duffley, L. & Pulinilkunnil, T. Role of branched-chain amino acid-catabolizing 17 enzymes in intertissue signaling, metabolic remodeling, and energy homeostasis. FASEB J 33, 18 8711-8731, doi:10.1096/fj.201802842RR (2019). 19 8 Zhu, Z. et al. Tumour-reprogrammed stromal BCAT1 fuels branched-chain ketoacid dependency 20 in stromal-rich PDAC tumours. Nat Metab 2, 775-792, doi:10.1038/s42255-020-0226-5 (2020). 21 9 Silva, L. S. et al. Branched-chain ketoacids secreted by glioblastoma cells via MCT1 modulate 22 macrophage phenotype. EMBO Rep 18, 2172-2185, doi:10.15252/embr.201744154 (2017). 23 10 Ananieva, E. A. & Wilkinson, A. C. Branched-chain amino acid metabolism in cancer. Curr Opin 24 Clin Nutr Metab Care 21, 64-70, doi:10.1097/MCO.0000000000000430 (2018). 25 11 Peng, H., Wang, Y. & Luo, W. Multifaceted role of branched-chain amino acid metabolism in 26 cancer. Oncogene 39, 6747-6756, doi:10.1038/s41388-020-01480-z (2020). 27 12 Tian, Q. et al. Phosphorylation of BCKDK of BCAA catabolism at Y246 by Src promotes 28 metastasis of colorectal cancer. Oncogene 39, 3980-3996, doi:10.1038/s41388-020-1262-z (2020). 29 13 Zhai, M. et al. APN-mediated phosphorylation of BCKDK promotes hepatocellular carcinoma 30 metastasis and proliferation via the ERK signaling pathway. Cell Death Dis 11, 396, 31 doi:10.1038/s41419-020-2610-1 (2020). 32 14 Zhang, L. & Han, J. Branched-chain amino acid transaminase 1 (BCAT1) promotes the growth of 33 breast cancer cells through improving mTOR-mediated mitochondrial biogenesis and function. 34 Biochem Biophys Res Commun 486, 224-231, doi:10.1016/j.bbrc.2017.02.101 (2017).

1 15 Iurlaro, R. & Munoz-Pinedo, C. Cell death induced by endoplasmic reticulum stress. FEBS J 283, 2 2640-2652, doi:10.1111/febs.13598 (2016). 3 16 Zhang, H. G., Wang, J., Yang, X., Hsu, H. C. & Mountz, J. D. Regulation of apoptosis proteins in 4 cancer cells by ubiquitin. Oncogene 23, 2009-2015, doi:10.1038/sj.onc.1207373 (2004). 5 17 Shan, Y. et al. Targeting HIBCH to reprogram valine metabolism for the treatment of colorectal 6 cancer. Cell Death Dis 10, 618, doi:10.1038/s41419-019-1832-6 (2019). 7 18 Ricci, J. E., Gottlieb, R. A. & Green, D. R. Caspase-mediated loss of mitochondrial function and 8 generation of reactive oxygen species during apoptosis. J Cell Biol 160, 65-75, 9 doi:10.1083/jcb.200208089 (2003). 10 19 Guven Celal, Y. S., and Eylem Taskin. in Mitochondrial Diseases 323 (2018). 11 20 Topf, U., Uszczynska-Ratajczak, B. & Chacinska, A. Mitochondrial stress-dependent regulation of 12 cellular protein synthesis. J Cell Sci 132, doi:10.1242/jcs.226258 (2019). 13 21 Clemens, M. J., Bushell, M., Jeffrey, I. W., Pain, V. M. & Morley, S. J. Translation initiation factor 14 modifications and the regulation of protein synthesis in apoptotic cells. Cell Death Differ 7, 603- 15 615, doi:10.1038/sj.cdd.4400695 (2000). 16 22 Wang, L. X., Zhu, X. M. & Yao, Y. M. Sestrin2: Its Potential Role and Regulatory Mechanism in 17 Host Immune Response in Diseases. Front Immunol 10, 2797, doi:10.3389/fimmu.2019.02797 18 (2019). 19 23 Suryawan, A. et al. A molecular model of human branched-chain amino acid metabolism. Am J 20 Clin Nutr 68, 72-81, doi:10.1093/ajcn/68.1.72 (1998). 21 24 Lieu, E. L., Nguyen, T., Rhyne, S. & Kim, J. Amino acids in cancer. Exp Mol Med 52, 15-30, 22 doi:10.1038/s12276-020-0375-3 (2020). 23 25 Li, J. T. et al. BCAT2-mediated BCAA catabolism is critical for development of pancreatic ductal 24 adenocarcinoma. Nat Cell Biol 22, 167-174, doi:10.1038/s41556-019-0455-6 (2020). 25 26 Lee, J. H. et al. Branched-chain amino acids sustain pancreatic cancer growth by regulating lipid 26 metabolism. Exp Mol Med 51, 1-11, doi:10.1038/s12276-019-0350-z (2019). 27 27 Ericksen, R. E. et al. Loss of BCAA Catabolism during Carcinogenesis Enhances mTORC1 28 Activity and Promotes Tumor Development and Progression. Cell Metab 29, 1151-1165 e1156, 29 doi:10.1016/j.cmet.2018.12.020 (2019). 30 28 Mayers, J. R. et al. Tissue of origin dictates branched-chain amino acid metabolism in mutant Kras- 31 driven cancers. Science 353, 1161-1165, doi:10.1126/science.aaf5171 (2016). 32 29 Tonjes, M. et al. BCAT1 promotes cell proliferation through amino acid catabolism in gliomas 33 carrying wild-type IDH1. Nat Med 19, 901-908, doi:10.1038/nm.3217 (2013).

1 30 Ikeda, K. et al. Slc3a2 Mediates Branched-Chain Amino-Acid-Dependent Maintenance of 2 Regulatory T Cells. Cell Rep 21, 1824-1838, doi:10.1016/j.celrep.2017.10.082 (2017). 3 31 Xue, P. et al. BCKDK of BCAA Catabolism Cross-talking With the MAPK Pathway Promotes 4 Tumorigenesis of Colorectal Cancer. EBioMedicine 20, 50-60, doi:10.1016/j.ebiom.2017.05.001 5 (2017). 6 32 Moreno-Sanchez, R. et al. Who controls the ATP supply in cancer cells? Biochemistry lessons to 7 understand cancer energy metabolism. Int J Biochem Cell Biol 50, 10-23, 8 doi:10.1016/j.biocel.2014.01.025 (2014). 9 33 Leibovitch, M. & Topisirovic, I. Dysregulation of mRNA translation and energy metabolism in 10 cancer. Adv Biol Regul 67, 30-39, doi:10.1016/j.jbior.2017.11.001 (2018). 11 34 Dartier, J. et al. ATP-dependent activity and mitochondrial localization of drug efflux pumps in 12 doxorubicin-resistant breast cancer cells. Biochim Biophys Acta Gen Subj 1861, 1075-1084, 13 doi:10.1016/j.bbagen.2017.02.019 (2017). 14 35 Oyarzabal, A. et al. Mitochondrial response to the BCKDK-deficiency: Some clues to understand 15 the positive dietary response in this form of autism. Biochim Biophys Acta 1862, 592-600, 16 doi:10.1016/j.bbadis.2016.01.016 (2016). 17 36 Grzmil, M. & Hemmings, B. A. Translation regulation as a therapeutic target in cancer. Cancer 18 Res 72, 3891-3900, doi:10.1158/0008-5472.CAN-12-0026 (2012). 19 37 Hovland, R. et al. cAMP inhibits translation by inducing Ca2+/calmodulin-independent elongation 20 factor 2 kinase activity in IPC-81 cells. FEBS Lett 444, 97-101, doi:10.1016/s0014-5793(99)00039- 21 3 (1999). 22 38 Castedo, M., Ferri, K. F. & Kroemer, G. Mammalian target of rapamycin (mTOR): pro- and anti- 23 apoptotic. Cell Death Differ 9, 99-100, doi:10.1038/sj.cdd.4400978 (2002). 24 39 White, P. J. et al. The BCKDH Kinase and Phosphatase Integrate BCAA and Lipid Metabolism 25 via Regulation of ATP-Citrate Lyase. Cell Metab 27, 1281-1293 e1287, 26 doi:10.1016/j.cmet.2018.04.015 (2018). 27 40 Biswas, D. et al. Branched-chain ketoacid overload inhibits insulin action in the muscle. J Biol 28 Chem 295, 15597-15621, doi:10.1074/jbc.RA120.013121 (2020). 29 41 Biswas, D. et al. Adverse Outcomes in Obese Cardiac Surgery Patients Correlates With Altered 30 Branched-Chain Amino Acid Catabolism in Adipose Tissue and Heart. Front Endocrinol 31 (Lausanne) 11, 534, doi:10.3389/fendo.2020.00534 (2020). 32 42 Khan, M. W. et al. mTORC2 controls cancer cell survival by modulating gluconeogenesis. Cell 33 Death Discov 1, 15016, doi:10.1038/cddiscovery.2015.16 (2015).

1 Figure Legends

2 Figure 1. DOX suppresses BCKDK expression and augments BCAA oxidation enzyme

3 expression in TNBCs. A) Quantification of BCKDK, BCKDHA, PPM1K, BCKDHB, BCAT2, and

4 BCAT1 BCAA catabolic enzyme mRNA expression corrected to 18S/HSPCB reference genes in

5 MCF10A, BT549 and MDA-MB231 cells. B-C) BCAA catabolic enzyme expression in TNBCs

6 treated with 2µM DOX for 18h. Quantification of BCKDK, BCKDHA, PPM1K, BCKDHB,

7 BCAT2, and KLF15 mRNA expression corrected to 18S/HSPCB reference genes in MDA-MB231

8 (B) and BT549 (C) cells. D) Immunoblot and densitometric analysis of BCKDK, total and

9 phosphorylated BCKDHA E1α Ser 293, DLD and KLF15 in MDA-MB231 cells treated with 1µM

10 and 2µM DOX for 18h. Quantifications are from three independent experiments. Data presented

11 as mean ± S.D. Statistical analysis was performed using a two-way ANOVA followed by a

12 Tukey’s multiple comparison test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated. *MCF10A

13 v/s BT549 or DMSO v/s 1µM DOX, #MCF10A v/s MDA-MB231 or DMSO v/s 2µM DOX.

14 Figure 2. Silencing BCKDK exacerbates DOX mediated cell death in TNBCs. A) Antibody-

15 immobilized PVDF membranes containing 43 different apoptosis associated proteins incubated

16 with extracts of CON (siCON) and BCKDK (siBDK#1) silenced MDA-MB231 cells, treated with

17 or without 2µM DOX for 18h. Differentially expressed proteins (fold change more than 1.5 or

18 less than 0.6) are quantified and indicated on the blots. a=Survivin, b=Bim, c=Cytochrome, d=

19 Fas, e= Caspase8. B) Immunoblot and densitometric analysis of BCKDK, total and cleaved

20 Caspase 3, cleaved Caspase 7, total and cleaved PARP, total and phosphorylated ATM Ser 1981

21 in MDA-MB231 cells transfected with siCON, siBDK#1 or siBDK#2 for 72h followed by 2µM

22 DOX or DMSO for 18h. (C-E) MDA-MB231 cells were transfected with siCON, siBDK#1 or

23 siBDK#2 for 72h followed by 2µM DOX or DMSO treatment for 18h and analyzed for C) Caspase

1 3 activity, D) LDH release in the media, and E) cell proliferation measured by CCK8 assay. Data

2 presented as mean ± S.D. Statistical analysis was performed using a two-way ANOVA followed

3 by a Tukey’s multiple comparison test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

4 Figure 3. BT2 mediated BCKDK inhibition induces cell death and potentiates DOX mediated

5 apoptosis in TNBCs. A) Immunoblot and densitometric analysis of phosphorylated BCKDE1α

6 Ser 293, BCKDK, total and cleaved Caspase 3, cleaved Caspase 7, total and cleaved PARP, total

7 and phosphorylated ATM Ser 1981 in MDA-MB231 cells treated with 500µM BT2 for 20h.

8 Densitometric analysis is from three independent experiments. B) Immunoblot and densitometric

9 analysis of total and phosphorylated BCKDE1α Ser 293, BCKDK, total and cleaved Caspase 3,

10 total and cleaved PARP, total and phosphorylated ATM Ser 1981 in MDA-MB231 cells pre-

11 treated with 500µM BT2 for 20h followed by 2µM DOX or DMSO treatment for 18h. Caspase 3

12 activity measured in MDA-MB231 (C) and BT549 (D) cells and LDH release into the media was

13 measured in MDA-MB231 cells (E) pre-treated with 500µM BT2 for 20h followed by 2µM DOX

14 or DMSO treatment for 18h. Quantifications are from three independent experiments. Data

15 presented as mean ± S.D. Statistical analysis was performed using a two-way ANOVA followed

16 by a Tukey’s multiple comparison test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

17 Figure 4. Differential enrichment of genes related to metabolism and cell death pathways in

18 MDA-MB-231 cells with BCKDK knockdown. A) Principle component analysis plot for all

19 gene expression from MDA-MB231 cells treated with both siRNAs targeting BCKDK. B)

20 Distance matrix clustering for all gene expression from MDA-MB231 cells. C) Heatmap for

21 differentially expressed genes from MDA-MB231 cells. D) Top 20 most significantly enriched

22 gene ontology (GO) terms from BCKDK knockdown induced differentially expressed genes, color

23 represents the fold enrichment statistic, and size represents the percentage of the differentially

1 expressed genes in the gene set compared with the total gene set size. E) Heatmap for genes

2 involved in apoptosis which were significantly differentially regulated by BCKDK knockdown in

3 MDA-MB231 cells.

4 Figure 5. BCKDK inactivation suppresses mitochondrial function and complex protein

5 expression. A) Heatmap for genes involved in mitochondrial biogenesis, structure, and function

6 was significantly differentially regulated by BCKDK knockdown. B) Immunoblot and

7 densitometric quantification of complex I-V proteins in MDA-MB231 whole cell lysates

8 transfected with siCON, siBDK#1 or siBDK#2 for 72h. Data presented as mean ± S.D. Statistical

9 analysis was performed using a two-way ANOVA followed by a Tukey’s multiple comparison

10 test;*p <0.05, **p < 0.01, **** p <0.0001 as indicated. (C-H) Mitochondrial function measured

11 in BCKDK depleted MDA-MB231 cells using extracellular flux analyzer in the presence of 25mM

12 glucose. C) Basal OCR, D) ATP production measured in MDA-MB231 cells transfected with

13 siCON or siBDK#2. E) Basal and maximal OCR, F) ATP production, G) spare capacity, and H)

14 proton leak measured in MDA-MB231 cells treated with 250µM or 500µM BT2 for 20h. Data

15 presented as mean ± S.D. Statistical analysis was performed using a was performed using Student’s

16 t-test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

17 Figure 6. BCKDK inhibition decreases protein synthesis and mTOR signaling in TNBCs. A)

18 BT549 transfected with siCON, siBDK#1 or siBDK#2 for 72h followed by incubation with 1μM

19 puromycin for 30 mins. Immunoblot and densitometric analysis of puromycin incorporation. B)

20 Fluorogram and quantification of AHA incorporation in BT549 cells transfected with siCON,

21 siBDK#1 or siBDK#2. C) Immunoblot and densitometric analysis of puromycin incorporation in

22 BT549 cells treated with 500µM BT2 for 20h followed by incubation with 1μM puromycin for 30

23 mins. The graph represents mean ± S.D., n=3, *p<0.05 was performed using Student’s t-test. D)

1 Immunoblot and densitometric analysis of SESN2, total and phosphorylated mTOR S2448, total

2 and phosphorylated P70S6K T389 and total to phosphorylated S6 S240/244 in BT549 and MDA-

3 MB231 cells treated with 500µM BT2 for 20h. Data presented as mean ± S.D. Statistical analysis

4 was performed using a was performed using Student’s t-test; *p <0.05, **p < 0.01, **** p <0.0001

5 as indicated. Immunoblot and densitometric analysis of total and phosphorylated eIF2α S51,

6 Sesn2, total and phosphorylated eEF2 T56, total and phosphorylated eEF2K S266, total and

7 phosphorylated AKT S473 and AKT T308 in MDA-MB231 cells transduced with shGFP or

8 shBCKDK for 48h followed by 2µM DOX or DMSO treatment for 18h. Quantifications are from

9 three independent experiments. Data presented as mean ± S.D. Statistical analysis was performed

10 using a two-way ANOVA followed by a Tukey’s multiple comparison test; *p <0.05, **p < 0.01,

11 **** p <0.0001 as indicated.

33 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Inhibiting BCKDK in triple negative breast cancer suppress protein translation, impair mitochondrial function, and potentiate doxorubicin cytotoxicity

Running Title: Inhibiting BCKDK increase TNBC DOX toxicity

Dipsikha Biswas1, Logan Slade1, Luke Duffley1, Neil Mueller1, Khoi Thien Dao1, Angella Mercer1, Yassine El Hiani2, Petra Kienesberger1 and Thomas Pulinilkunnil1*

SUPPORTING INFORMATION

Figure S1. BCAA degradation enzyme expression are suppressed in TNBCs and can be revived by DOX treatment. A) Quantification of ACADSB, HADHA, HIBCH, and KLF15 BCAA catabolic enzyme mRNA expression corrected to 18S/HSPCB reference genes in MCF10A, BT549 and MDA-MB231 cells. B) Immunoblot and densitometric analysis of BCKDK, total and phosphorylated BCKDHA E1α Ser 293, BCAT2, BCKDHB, DLD and KLF15 in MCF10A, MDA-MB231 and BT549 cells. C-D) BCAA catabolic enzyme expression in TNBCs treated with 2µM DOX for 18h. Quantification of BCAT1, ACADSB, HADHA, HIBCH, mRNA expression

corrected to 18S/HSPCB reference genes in MDA-MB231 (C) and BT549 (D) cells. E) Immunoblot and densitometric analysis of BCKDK, total and phosphorylated BCKDHA E1α Ser 293 and BCAT2 in BT549 cells treated with 1µM and 2µM DOX for 18h. Data presented as mean ± S.D. Statistical analysis was performed using a two-way ANOVA followed by a Tukey’s multiple comparison test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

Figure S2. BCKDK knockdown increases cell death and reduces proliferation and potentiates DOX’s effects in TNBCs. mRNA quantification of BCKDK (A) and protein expression (B) of BCKDK and cleaved and total caspase 3 expression in MDA-MB231 cells transfected with siRNA targeting BCKDK at exon 4 (siBDK#1) and exon 5 (siBDK#2). C) Cell counts at 0, 24 and 48h of MDA-MB231 cells transfected with siCON, siBDK#1 or siBDK#2. D) Immunoblot and densitometric analysis of BCKDK, total and cleaved Caspase 3, cleaved Caspase 7, total and Bax in BT549 cells transfected with siCON, siBDK#1 or siBDK#2 for 72h followed by 2µM DOX or DMSO for 18h. E) Immunoblot and densitometric analysis of total and cleaved

Caspase 3 in MDA-MB231 and BT549 cells transduced with either shGFP or shBCKDK for 48h followed by 2µM DOX or DMSO for 18h. F) BT549 cells were transfected with siCON, siBDK#1 or siBDK#2 for 72h followed by 2µM DOX or DMSO treatment for 18h and analyzed for Caspase 3 activity. G) Metabolic viability of MDA-MB231 cells transduced with either shGFP or shBCKDK and treated with 2µM DOX or DMSO for 18 h plus 48 h in drug-free medium. Data presented as mean ± S.D. Statistical analysis was performed using a two-way ANOVA followed by a Tukey’s multiple comparison test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

Figure S3. BT2 augments mRNA expression of BCAA catabolic genes and reduces BCKA accumulation and secretion. A) Heatmap for the genes involved in BCAA catabolism which were differentially regulated by BCKDK knockdown. B-C) Quantification of BCKDK, BCKDHA, PPM1K, BCKDHB, BCAT2, BCAT1, ACADSB, HADHA, HIBCH, HIBADH and KLF15 mRNA expression corrected to 18S/HSPCB reference genes in B) MDA-MB231 and C) BT549 cells transfected with siCON, siBDK#1 or siBDK#2 for 72h. D) Quantification of BCKDK,

BCKDHA, PPM1K, BCKDHB, ACADSB, HIBCH, HIBADH and KLF15 mRNA expression corrected to 18S/HSPCB reference genes in BT549 cells treated with 500µM BT2 for 20h. Statistical analysis was performed using a two-way ANOVA followed by a Tukey’s multiple comparison test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated. E) UPLC MS/MS analysis of intracellular BCKAs in MDA-MB231 cells transduced with shBCKDK and shGFP for 48h. Measurement of intracellular (F) and secreted BCKAs in the media (G) by UPLC MS/MS in MDA-MB231 cells treated with 500µM BT2 for 20h. Data presented as mean ± S.D. Statistical analysis was performed using a Student’s t-test; *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

Figure S4. BT2 does not affect mitochondrial respiration in MCF10A cells. A-B) Non mitochondrial respiration measured in MDA-MB231 cells A) transfected with siCON or siBDK#2 for 48h or B) treated with 250µM or 500µM BT2 for 20h, in the presence of 25mM glucose. C-G) Basal and maximal OCR (C), ATP production (D), spare capacity (E), proton leak (F) and non mitochondrial respiration (G) measured in MCF10A cells treated with 250µM or 500µM BT2 for 20h in the presence of 25mM glucose. Data presented as mean ± S.D. Statistical analysis was performed using a was performed using two-way ANOVA followed by a Tukey’s multiple comparison test for (A) or Student’s t-test for (B); *p <0.05, **p < 0.01, **** p <0.0001 as indicated.

Table S1. List of antibodies. (Fine chemicals, if noted otherwise under the Materials and Methods segment, are from Sigma.)

Antibodies Source Identifier BCKDK My Biosource MBS 275719 p-BCKDE1α S293 Bethyl Lab A303-567A BCKDHA My Biosource MBS 275832 BCKDHB Santa Cruz Biotechnology Inc. H-6; sc-374630 BCAT2 Invitrogen PA5-21549 DLD Santa Cruz Biotechnology Inc. G-2; sc-365977 KLF15 Novus Biologicals NBP2-24635 cCaspase 3 D175 Cell Signaling Technologies 9664 cCaspase 7 D198 Cell Signaling Technologies 8438 Caspase 3 Cell Signaling Technologies 9662 cPARP D124 Abcam ab110315 PARP Cell Signaling Technologies 9532 p-ATM S1981 Cell Signaling Technologies 5883 ATM Cell Signaling Technologies 2873 Sestrin 2 Cell Signaling Technologies 8487S p-mTOR S2448 Cell Signaling Technologies 5536 mTORC1 Cell Signaling Technologies 2972 p-P70S6K T389 Cell Signaling Technologies 9234 P70S6K Cell Signaling Technologies 2708 p-S6 S240/244 Cell Signaling Technologies 5364 S6 Cell Signaling Technologies 2217 p-eEF2 T56 Cell Signaling Technologies 2331 eEF2 Cell Signaling Technologies 2332 p-eEF2K S366 Cell Signaling Technologies 3691 eEF2K Cell Signaling Technologies 3692 p-eIF2α S51 Cell Signaling Technologies 9721 eIF2α Cell Signaling Technologies 9722 Puromycin clone 12D10 Millipore-Sigma MABE343 Total OXPHOS cocktail Abcam ab110413 mouse anti-rabbit IgG-HRP Santa Cruz Biotechnology Inc. sc-2357 m-IgG kappa BP-HRP Santa Cruz Biotechnology Inc. sc-516102

S2 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.18.444634; this version posted May 18, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Table S2. List of primers.

Primer Gene name Sequence 5'-3' h- BCKDK-F Branched chain ketoacid GACTTCCCTCCGATCAAGGAC h- BCKDK-R dehydrogenase kinase CTCTCACGTAGGCCCTCTG h-BCKDHA-F Branched chain keto acid CTACAAGAGCATGACACTGCTT h-BCKDHA-R dehydrogenase E1 alpha polypeptide CCCTCCTCACCATAGTTGGTC h- BCKDHB-F Branched chain keto acid TGGAGTCTTTAGATGCACTGTTG h- BCKDHB-R dehydrogenase E1 subunit beta CGCAATTCCGATTCCAAATCCAA h-KLF15 -F Kruppel like factor 15 CGGCTGGAGGTTCTCGCGCTCTG h- KLF15 -R AGGCTGGGGTTCAGGGCGCTTTC h-PPM1K-F Protein phosphatase Mg2+/Mn2+ ATAACCGCATTGATGAGCCAA h-PPM1K-R dependent 1K CGCACCCCACATTTTCCAAG h-BCAT2-F Branched chain amino acid CGCTCCTGTTCGTCATTCTCT h-BCAT2-R transaminase 2 CCCACCTAACTTGTAGTTGCC h-ACADSB-F Acyl-CoA dehydrogenase short GATGGCAAATGTAGACCCTACC h-ACADSB-R branched chain GGAGGTTTTAGTCCTGGTCCC h-HADHA-F Hydroxyacyl-CoA dehydrogenase CTGCCCAAAATGGTGGGTGT h-HADHA-R trifunctional multienzyme complex GGAGGTTTTAGTCCTGGTCCC subunit alpha h-HIBCH-F 3-hydroxyisobutyryl-CoA hydrolase TGGTTCTTGCCAGAAACCTTATG h-HIBCH-R GTAGCCACTCGAAATTGCCCA h-BCAT1-F Branched chain amino acid AGCCCTGCTCTTTGTACTCTT h-BCAT1-R transaminase 1 CCAGGCTCTTACATACTTGGGA h-HIBADH-F 3-hydroxyisobutyrate dehydrogenase TGCTGCCCACCAGTATCAATG h-HIBADH-R GCAGGATCAATAGTGCTGGAATC h-18S-F 18S ribosomal RNA AGAAACGGCTACCACATCCA h-18S-R CACCAGACTTGCCCTCCA h-HSP90AB1-F Heat shock protein HSP 90-beta TCTGGGTATCGGAAAGCAAGCC h-HSP90AB1-R GTGCACTTCCTCAGGCATCTTG