The mitochondrial RNA polymerase POLRMT promotes skin squamous cell carcinoma cell growth

RNA polymerase mitochondrial (POLRMT) expression and the potential biological functions in skin squamous cell carcinoma (SCC) were explored. We showed that POLRMT is significantly elevated in skin SCC. Genetic depletion of POLRMT, using shRNA-induced knockdown or CRISPR/Cas9-mediated knockout (KO), resulted in profound anti-skin SCC cell activity. In patient-derived primary skin SCC cells or immortalized lines (A431 and SCC-9), POLRMT shRNA or KO potently suppressed mitochondrial DNA (mtDNA) transcription and suppressed cell viability, proliferation and migration. POLRMT shRNA or KO impaired mitochondrial functions in different skin SCC cells, leading to production of ROS (reactive oxygen species), depolarization of mitochondria and depletion of ATP. Moreover, mitochondrial apoptosis cascade was induced in POLRMT-depleted skin SCC cells. IMT1, a POLRMT inhibitor, largely inhibited proliferation and migration, while inducing depolarization of mitochondria and apoptosis in primary skin SCC cells. Contrarily, ectopic overexpression of POLRMT increased mtDNA transcription and augmented skin SCC cell growth. Importantly, POLRMT shRNA adeno-associated virus injection robustly hindered growth of the subcutaneous A431 xenografts in mice. In the POLRMT shRNA virus-treated A431 xenograft tissues, POLRMT depletion, mtDNA transcription inhibition, cell apoptosis, lipid peroxidation and ATP depletion were detected. Together, overexpressed POLRMT increases mtDNA transcription and promotes skin SCC growth.


RESULTS
POLRMT expression is elevated in skin SCC First, the qRT-PCR results showed that the POLRMT mRNA expression in twenty skin SCC tumor tissues was about three folds of that in the normal skin tissues (Fig. 1A). Moreover, POLRMT protein is upregulated in skin SCC tissues of the two representative SCC patients (Patient-1# to Patient-2#, Fig. 1B). Quantification results of all twenty patients' POLRMT blotting data further confirmed that POLRMT protein is significantly elevated in the tumor tissues (Fig. 1B). We found that the mRNA expression of POLRMT-dependent mitochondrial transcripts, including NDUFB8, UQCRC2 and COXI [9,22], was also significantly elevated in the skin SCC tumor tissues (Fig. 1C).
POLRMT expression in skin SCC cells and normal skin cells was examined. In A431 cells and SCC-9 cells as well as in primary skin SCC cells [derived from Patient-1# ("C1" cells) and Patient-2# ("C2" cells), the mRNA and protein expression of POLRMT was robustly higher than it in the primary human skin keratinocytes and fibroblasts ( Fig. 1D-F)]. These results together confirmed elevated POLRMT expression in skin SCC.
POLRMT silencing or KO disrupts mitochondrial functions and provokes apoptosis in primary skin SCC cells Considering that POLRMT is essential for mtDNA transcription and mitochondrial integrity [18][19][20][21], we examined whether POLRMT depletion affected mitochondrial functions in skin SCC cells. As shown, POLRMT shRNA or KO (see Fig. 2) led to dramatic ROS production in C1 primary skin SCC cells, elevating CellROX red fluorescence intensity [25] (Fig. 3A). Moreover, the accumulation of green JC-1 monomers indicated depolarization of mitochondria in shPOLRMT and koPOLRMT C1 primary cells (Fig. 3B). In addition, TBAR intensity was increased in POLRMT-depleted cells, supporting lipid peroxidation (Fig. 3C). The cellular ATP contents were decreased as well following POLRMT silencing/KO (Fig. 3D). Thus POLRMT depletion disrupted mitochondrial functions in skin SCC cells.
In human cancer cells mitochondrial dysfunction will lead to mitochondrial apoptosis cascade induction [26][27][28][29]. We found that the caspase-3 activity was boosted in shPOLRMT and koPOLRMT C1 primary cells (Fig. 3E). The percentage of TUNEL-positive nuclei e was increased in shPOLRMT and koPOLRMT C1 SCC cells (Fig. 3F), supporting significant apoptosis activation. As expected, c-sh+Cas9 control treatment (see To further support our hypothesis, we showed that IMT1, a firstin-class specific and noncompetitive POLRMT inhibitor [9,30], robustly suppressed proliferation (Fig. 3G) and migration (Fig. 3H) in C1 cells. Moreover, the POLRMT inhibitor caused depolarization of mitochondria and green JC-1 monomers increasing (Fig. 3I). Apoptosis induction, evidenced by the increased TUNEL nuclei (Fig. 3J), was observed in IMT1-treated C1 SCC cells. These results further supported an important function of POLRMT in skin SCC cells.
POLRMT shRNA exerts anti-cancer activity in other skin SCC cells Next, we examined whether POLRMT depletion could affect functions of other skin SCC cells. In primary skin SCC cells derived from another patient, C2, and in the established lines (A431 and SCC-9 [31]), POLRMT shRNA lentiviral particles were added. POLRMT mRNA and protein expression as well as NDUFB8, UQCRC2 and COXI mRNA expression in skin SCC tumor tissues ("Tum") and matched adjacent normal skin tissues ("Nor") from a set of twenty (n = 20) primary skin SCC patients was tested by qRT-PCR and Western blotting assays (A-C). POLRMT mRNA and protein expression in the listed skin SCC cells, primary human skin fibroblasts ("Fibroblasts") and keratinocytes ("Keratinocytes") was tested (D-F). Data were mean ± standard deviation (SD). *P < 0.05 versus "Nor" tissues/"Fibroblasts".
Puromycin was then included to select stable cells, namely "shPOLRMT" cells. As shown POLRMT mRNA was silenced in the shPOLRMT cells (Fig. 4A). In the SCC cells, EdU staining and "Transwell" assay results showed that POLRMT shRNA remarkably suppressed cell proliferation ( The POLRMT shRNA lentiviral particles were added to skin keratinocytes and fibroblasts, resulting in significant POLRMT mRNA silencing ("shPOLRMT") after selection (Fig. 4G). As shown, shPOLRMT treatment was unable to increase the caspase-3 activity (Fig. 4H) and TUNE-nuclei percentage (Fig. 4I). Therefore, POLRMT silencing failed to provoke apoptosis in non-cancerous skin cells, supporting a cancer cell specific effect.

DISCUSSION
About 4% of skin SCC will metastasize, and these patients have a poor prognosis with 5-year survival rate <25% [1,2,32]. The novel and more efficient targeted therapies are urgently needed especially for skin SCC patients with advanced, metastatic and recurrent tumors [3][4][5]. Clinical studies have shown that immune CD1 inhibitors and epidermal growth factor receptor (EGFR) blockers, among others, displayed promising efficiency for skin SCC patients [33][34][35].
Recent studies have supported POLRMT as a potential oncogenic gene and therapeutic target of human cancer. Several inhibitors of mitochondrial transcription (IMTs) targeting POLRMT were developed [9]. IMTs impaired mtDNA transcription and inhibited oxidative phosphorylation, suppressing growth of different cancer xenografts [9]. Zhou et al., reported that POLRMT is overexpressed in non-small cell lung cancer [22]. Genetic silencing or depletion of POLRMT resulted in robust anti-NSCLC activity. Han et al., have recently shown that POLRMT overexpression in osteosarcoma is vital for cancer cell growth, representing as a novel molecular target [23].
POLRMT is a promising and novel therapeutic target of skin SCC. POLRMT expression is remarkably elevated in skin SCC tissues and cells. While in normal skin tissues and non-cancerous skin cells its expression is relatively low. Genetic depletion of POLRMT, using shRNA-induced knockdown or CRISPR/Cas9 method, resulted in robust anti-skin SCC cell activity. POLRMT shRNA or KO potently suppressed mtDNA transcription and inhibited skin SCC cell viability and proliferation. IMT1, a POLRMT inhibitor, largely inhibited skin SCC cell proliferation and migration, while inducing depolarization of mitochondria and apoptosis in primary skin SCC cells. Contrarily, further increasing POLRMT expression by a lentiviral construct enhanced mtDNA transcription and augmented skin SCC cell growth. Importantly, injection of AAV-packed POLRMT shRNA robustly inhibited subcutaneous A431 xenograft growth.
Like all other tumor cells, rapidly growing skin SCC cells require elevated mitochondrial function to provide energy. Lee et al., reported aberrant skin cancer cell proliferation was possibly due to enhanced mitochondrial biogenesis [36]. The mtDNA copy number as well as mitochondrial biogenesis-associated genes are elevated in skin cancers [36]. Dynamin-related protein 1 (Drp1) is upregulated in skin SCC. Silencing Drp1 inhibited skin SCC cell growth and led to G2 cell cycle arrest. Moreover, Drp1 silencing resulted in an elongated, hyper-fused mitochondrial network in skin SCC cells [37].
Here we show that overexpression POLRMT is vital for mitochondrial functions in skin SCC cells. Conversely, POLRMT shRNA or KO not only hindered mtDNA transcription, but also disrupted mitochondrial functions, causing significant ROS production, lipid peroxidation, depolarization of mitochondria and ATP depletion in skin SCC cells. Importantly, mitochondrial apoptosis cascade was induced in POLRMT-depleted skin SCC cells. In POLRMT-silenced A431 xenograft tissues, decreased mtDNA transcription, apoptosis induction, lipid peroxidation and ATP depletion were observed. Thus, overexpressed POLRMT promotes mtDNA transcription and increases mitochondrial functions, thereby accelerating skin SCC cell growth.

Cell culture
Patient-derived primary skin SCC cells from two written-informed consent patients, "C1" and "C2", the immortalized cell lines (A431 and SCC9), the primary human skin fibroblasts and human skin keratinocytes were provided from Dr. Liu [31]. The detailed protocols for cell culture were reported early [31,38,39]. The protocols were approved by the Ethics Committee of The Second Affiliated Hospital of Soochow University, in according to the principles of Declaration of Helsinki.

Human tissues
Skin SCC tumor tissues along with the matched adjacent normal skin tissues were from a total of twenty (20) primary skin SCC patients. All patients underwent tumor resection surgeries at authors' institution and each provided written-informed consent.

Silencing or overexpression of POLRMT
The lentiviral particles encoding the POLRMT shRNA or the POLRMTexpressing construct were provided by Dr. Shi [22]. Skin SCC cells were maintained at 50% confluence and infected with the lentiviral particles. Puromycin was included for six passages. POLRMT expression was always tested.

Fluorescence dye assays
In brief, according to the manufacture's protocols, skin SCC cells were cultivated at 50-60% confluence, stained with the designated fluorescence dyes (JC-1, CellROX, TUNEL, EdU and DAPI) and cultured for applied time periods. Fluorescence images were captured under a fluorescence microscopy (Leica).
Other functional assays, including the caspase-3 activity assay, Western blotting assay, mRNA detection by qRT-PCR were described in early studies [22,31]. Measuring ATP contents was described previously [40]. The intensity of lipid peroxidation in cell and tissue lysates was examined through a thiobarbituric acid reactive substance (TBAR) kit (Cayman Chemical, MI). The reagent colorimetrically (at 540 nm) quantified lipid peroxidation and the formation malondialdehyde (MDA) [41]. The uncropped blotting images were presented in Fig. S1.

Xenograft studies
As described [22,38,39], A431 skin SCC cells, at 6 × 10 6 cells of each mouse [22], were injected subcutaneously (s.c.) to nude mice's flanks (half male hale female, 5-to 6-week old, 18.4-19.1 g). All mice were provided and kept in the Laboratory Animal Center of Soochow University. Three weeks after cell implantation, A431 xenograft tumors were established and each tumor's volume was close to 100 mm 3 . Afterwards, A431 xenograft mice were assigned into two random groups, intratumorally injected with POLRMT shRNA adeno-associated virus (AAV) or scramble control shRNA AAV ("AAV-csh") (from Dr. Shi [22]). Virus was daily injected for 4 days. Tumor dimensions were measured by the described formula [23]. The protocols of handling experimental mice were according to institutional animal care and use committee (IACUC) and the Ethic Committee of Soochow University.

Statistical analyses
The in vitro experiments were repeated five times and similar results were obtained each time, and in vitro data were mean ± standard deviation (SD, n = 5). Statistical analyses were described previously [22]. Fig. 6 POLRMT shRNA virus injection hinders subcutaneous A431 xenograft growth. The nude mice bearing the subcutaneous A431 xenografts were subject to intratumoral injection of the POLRMT-shRNA-expressing AAV ("AAV-shPOLRMT") or scramble control shRNA AAV ("AAV-c-sh"). Virus was injected daily for four consecutive days. Tumor volumes (A) and mice body weights (D) were recorded weekly for six consecutive weeks ("Day-0" to "Day-42"). The estimated daily tumor growth, in mm 3 per day, was calculated as described (B). All experimental animals were sacrificed by decapitation at "Day-42", and xenograft tumors were carefully isolated and weighed individually (C). Expression of listed genes and proteins in the tumor lysates of the described tumors was tested by qRT-PCR (E and G) and Western blotting (F and H) assays. TBAR activity (I) and ATP contents (J) were tested as well. Data were mean ± standard deviation (SD). *P < 0.05 versus "AAV-c-sh" group.