Downregulation of miR‐193a/b‐3p during HPV‐induced cervical carcinogenesis contributes to anchorage‐independent growth through PI3K–AKT pathway regulators

Cervical cancer is caused by a persistent infection with high‐risk types of human papillomavirus (HPV) and an accumulation of (epi)genetic alterations in the host cell. Acquisition of anchorage‐independent growth represents a critical hallmark during HPV‐induced carcinogenesis, thereby yielding the most valuable biomarkers for early diagnosis and therapeutic targets. In a previous study, we found that miR‐193a‐3p and miR‐193b‐3p were involved in anchorage‐independent growth. This study aimed to delineate the role of miR‐193a/b‐3p in HPV‐induced carcinogenesis and to identify their target genes related to anchorage‐independent growth. Cell viability and colony formation were assessed in SiHa cancer cells and HPV‐16 and ‐18 immortalized keratinocytes upon miR‐193a/b‐3p overexpression. Both microRNAs reduced cell growth of all three cell lines in low‐attachment conditions and showed a minor effect in adherent conditions. Online target‐predicting programs and publicly available expression data were used to find candidate messenger RNA (mRNA) targets of miR‐193a/b‐3p. Seven targets showed reduced mRNA expression upon miR‐193a/b‐3p overexpression. For three targets, Western blot analysis was also performed, all showing a reduced protein expression. A direct interaction was confirmed using luciferase assays for six genes: LAMC1, PTK2, STMN1, KRAS, SOS2, and PPP2R5C, which are phosphatidylinositol 3‑kinase/protein kinase B (PI3K–AKT) regulators. All six targets were overexpressed in cervical cancers and/or precursor lesions. Together with an observed downregulation of phosphorylated‐AKT upon miR‐193a/b‐3p overexpression, this underlines the biological relevance of miR‐193a/b‐3p downregulation during HPV‐induced cervical carcinogenesis. In conclusion, the downregulation of miR‐193a‐3p and miR‐193b‐3p is functionally involved in the acquisition of HPV‐induced anchorage independence by targeting regulators of the PI3K–AKT pathway.

miR-193a-3p and miR-193b-3p is functionally involved in the acquisition of HPV-induced anchorage independence by targeting regulators of the PI3K-AKT pathway.

| INTRODUCTION
Cervical cancer is the fourth most diagnosed malignancy and the fourth leading cause of cancer-related death among women worldwide. 1 Approximately 604 000 women were diagnosed and 340 000 died because of cervical cancer in 2020. Due to the widely applied cervical cancer screening by cytological examination and/or human papillomavirus (HPV) DNA testing, the incidence and mortality rate is largely reduced in developed countries. 2 However, in less developed countries and regions, cervical cancer still endangers women's health.
The major risk factor of cervical cancer is persistent infection with high-risk types of HPV (hr-HPV). 3 Hr-HPV is detected in virtually all squamous cell carcinoma (SCC) and in the majority of adenocarcinoma (AdCa). Cervical SCC is preceded by high-grade cervical intraepithelial neoplasia (high-grade CIN). Fortunately, the majority of hr-HPV infections are cleared by the immune system. In some cases, hr-HPV infection persists, which, together with the accumulation of genetic and epigenetic changes in the host cell, can lead to cancer development. 4 Amongst the epigenetic alterations contributing to malignant transformation, changes in microRNA (miRNA) expression are associated with cervical carcinogenesis. 5 miRNAs are short noncoding RNA that modulate posttranscriptional gene silencing. Due to the multitargeting feature, miRNAs regulate many biological and pathological processes.
One of the critical biological alterations during early cancer development is the acquisition of anchorage-independent growth.
Anchorage dependence of epithelial cells refers to the need for a stable surface for cell growth. Since cancer cells can grow anchorage independently, anchorage independence is often regarded as a hallmark of cancer and reflects a complete transformation in vitro. 6 Studies on cancer cell lines showed that miRNAs play an important role in regulating anchorage-independent growth by targeting various individual genes. 7 However, little data exist on the molecular events driving the acquisition of anchorage independence in anchoragedependent epithelial cells, such as early-stage precancerous cells.
More insight into the molecular origin of this phenomenon improves the understanding of carcinogenesis.
In previous studies, we have established in vitro models based on HPV16 and HPV18 transfected human foreskin keratinocytes that upon longitudinal culturing represent different stages during HPVinduced malignant transformation. By the analysis of genetic and expression profiles at different stages of transformation, we showed that altered miRNA expression was associated with anchorageindependent growth. 8-10 Identification of biologically relevant miRNAs by a functional screen on SiHa cervical cancer cells, showed that miR-193a-3p and miR-193b-3p were involved in anchorage-independent growth. 11 To further investigate the role of miR-193a-3p and miR-193b-3p in HPV-induced carcinogenesis, the present study aimed to examine their functional involvement in HPV-transformed keratinocytes and to identify and validate their target genes in relation to anchorage-independent growth.

| Cell lines
The establishment of HPV-16 and HPV18-transformed keratinocytes (FK16A and FK18B, respectively) has been described previously. 12 FK16A and FK18B cell lines were cultured at 37°C and 5% CO 2 in serum-free Keratinocyte Growth Medium (Thermo Fisher Scientific) supplemented with bovine pituitary extract (25 mg), epidermal growth factor human recombinant (3.2 µg), penicillin (100 U/mL), streptomycin (100 µg/mL), and L-glutamine (2 mM) (Thermo Fisher Scientific). Cervical cancer cell line SiHa and renal epithelium cell line HEK293 were obtained from American Type Culture Collection and authenticated by short tandem repeat (STR) testing and cultured as described previously. 13

| Cell viability assay and anchorage-independent growth
Anchorage-independent growth was assessed using conventional soft agar assays and on ultra-low attachment (ULA) plates as described before. 11 For colony formation assay, 2500 cells were transfected and seeded in a medium containing 0.35% agarose (Sea-plague agarose; Lonza Group Ltd.) on a surface of 0.6% bottom agarose in a 48-well plate in duplicate. After 3 weeks, colonies exceeding ∼50 cells were counted. Each experiment was performed at least two times. Representative experiments are shown.

| External expression data of cervical tissue specimens
Expression levels of miR-193a-3p and miR-193b-3p were obtained and analyzed from our previously published microarray data of normal cervical epithelium (n = 10), high-grade CIN lesion (hgCIN, n = 18), and SCCs (n = 10). This data is available from GEO through series accession number GSE30656. 10 Global mRNA expression data of the same tissue specimens is available through GSE138080. 12    which was significantly reduced in both high-grade CIN and SCC compared to normal samples, is considered an early continuous downregulation as described previously ( Figure 1A). 12 To determine whether miR-193a/b-3p expression alters upon the acquisition of anchorage independence, we measured the expression levels in early (anchorage-dependent) and late (anchorage-independent) passages of FK16A and FK18B cells. In accordance with the tissue data, the expression levels of miR-193a-3p were lower in late-passage FK16A and FK18B cells compared to early-passage cells ( Figure 1B).

| Functional involvement of miR-193a-3p and miR-193b-3p in vitro
In earlier research, we showed that both miR-193a-3p and miR-193b-3p reduce anchorage-independent growth in SiHa cervical cancer cells. 11 To further understand their role in the regulation of anchorage-independent growth, we investigated the effect of over-

| Target prediction for miR-193a-3p and miR-193b-3p
To further investigate the mechanism of how miR-193a-3p and miR-193b-3p reduce anchorage-independent growth during HPVinduced malignant transformation, gene target prediction was performed using the online programs Starbase and miRDB. We focused on candidate gene targets involved in anchorage independence. These commonly used target prediction tools rely on the miRNA seed sequence. Since miR-193a-3p and miR-193b-3p share the same seed sequence, gene target prediction resulted in the same list of genes from both online tools. Therefore, we proceeded with the target prediction procedure for miR-193a-3p. Target prediction was performed as outlined in Figure 3.

| Confirmation of increased target expression in cervical tissue specimens
In addition to mRNA expression obtained from online data sets as supplied in Starbase and GEO, we also obtained miRNAs and mRNAs expression of the target genes from previously published in-house expression profiles. 10 To study the direct interaction between miRNAs and these seven targets, dual-reporter luciferase assays were performed. Cotransfection of 3′UTR vector with either miR-193a-3p or miR-193b-3p mimic in HEK293 cells decreased luciferase activity compared to cotransfection of nontargeting control mimic (negative control #1) or empty pmirGLO vector in seven out of eight targets. All targets except PRKACB showed a reduction of luciferase signal upon miR-193a-3p and miR-193b-3p overexpression ( Figure 5B). To further validate our method, two targets, LAMC1 and SOS2, were chosen for verification by the site-directed mutagenesis of the seed sequence in the 3′UTR. lack of further confirmation by in-house expression data. In contrast to the wild-type 3′UTR sequences, no reduction in signal was observed with the mutated LAMC1-3′UTR or SOS2-3′UTR ( Figure 5C). This confirms a direct interaction between LAMC1 and SOS2 and both miRNA-193a-3p and miR193b-3p.
For a subset of candidate targets also protein levels were measured upon miR-193a-3p or miR-193b-3p overexpression in anchorage-independent FK18B cells. In line with our mRNA expression results, protein expression of all 3 targets tested (LAMC1, PTK2, and STMN1) was reduced in both AD and ULA conditions ( Figure 5D). Since all are regulators of the PI3K-AKT pathway, we further analyzed total AKT and p-AKT protein expression. p-AKT levels were reduced upon miR-193a-3p or miR-193b-3p overexpression, while total AKT was not altered ( Figure 5D). This indicates an inhibition of the PI3K-AKT pathway by miR-193a-3p and miR-193b-3p.

| DISCUSSION
In this study, we investigated the functional involvement of inhibitor. Epigenetically silencing of miR-193a(-3p) was also observed in gastric cancer 29 and HER2-positive breast cancer. 30 At present, we have no evidence for MIR193 hypermethylation in cervical cancer cells and HPV-immortalized keratinocytes. 9 The promoter region of MIR193B does not contain CpG islands. Huang et al. 31 reported that miR-193b is silenced by N6-methyladenosine in cervical cancer.
However, Jiménez-Wences et al. 32   of invasive endocervical AdCa. 48 These findings suggest that upregulation of STMN1 might not only be involved in the development of precancerous lesions but also promote the progression to invasive cancer. Hence, STMN1 overexpression in precancerous lesions might be a valuable marker for cancer risk stratification.
Another way of PI3K activation is through Ras signaling. KRAS belongs to the RAS family of small GTPases, of which the activation could lead to various cellular outcomes such as cell proliferation, differentiation, growth, apoptosis, and cell survival. 49 Sheffels et al. 50 reported that SOS2 expression is critical in mediating mutant KRASdriven transformation. The frequency of somatic KRAS mutation in cervical cancer is relatively low, with only 0.1% in SCC. 51 In the immunohistochemical stainings, we found overexpression of SOS2 in cervical cancer, while the overexpression of KRAS was less pronounced. The fact that SOS2 was identified as a direct target of miR193a-3p and miR193b-3p indicates that altered miRNA expression may contribute to its overexpression in cervical cancer. The regulation of the PI3K-AKT pathway by SOS2 in the context of cervical cancer warrants further study.
PPP2R5C is the B56γ subunit of protein phosphatase 2A (PP2A), which is a crucial regulator of the cellular signaling pathways, proliferation, cell cycle checkpoints, and apoptosis. 52 Its inhibiting effect on AKT is studied in other cancer types. PPP2R5C was shown to be important in SV40-mediated human cell transformation.
Suppression of the PP2A B56γ subunit in HEK TER cell lines promotes anchorage-independent growth, 53 which is controversial to our observations. Our in-house mRNA expression data and expression data from GSE7410 showed that PPP2R5C was overexpressed in high-grade CIN compared to normal tissues 20,22 and immunohistochemical staining from HPA showed an overexpression of PPP2R5C in cervical cancers, suggesting its involvement in transformation. The function of PPP2R5C remains unclear in HPV-induced cervical carcinogenesis and needs further investigation.

| CONCLUSION
Our study shows that the acquisition of anchorage-independent growth during HPV-induced malignant transformation is at least in part mediated through miR-193a-3p and miR-193b-3p downregulation. Their functional role was proven in our unique cell model of HPV-induced carcinogenesis, showing a specific suppression of anchorage-independent growth upon miR-193a-3p and miR-193b-3p overexpression. Six genes, LAMC1, PTK2, STMN1, KRAS, SOS2, and PPP2R5C, were identified and validated as target genes of these miRNAs that regulate anchorage-independent growth. The clinical F I G U R E 6 Concept of the effect of miR-193a/b-3p on PI3K-AKT pathway. PI3K-AKT, phosphoinositide 3-kinase/protein kinase B; RTK, receptor tyrosine kinase.
relevance of these genes is supported by the upregulation of mRNA