FOXP3 Isoforms Expression in Cervical Cancer: Evidence about the Cancer-Related Properties of FOXP3Δ2Δ7 in Keratinocytes

Simple Summary FOXP3 is a critical transcription factor that works as a master regulator of the lymphoid lineage. The expression of FOXP3 was also observed in tumor cells; in cervical cancer, FOXP3 increases as the tumor progresses. However, the biological role of FOXP3 in cervical pathology is not well understood. In addition, FOXP3 has isoforms that could have different biological properties. In this work, the expression of FOXP3 and its isoforms were evaluated. It was found that the isoform FOXP3Δ2Δ7 is expressed in the cervical cancer-derived cell line SiHa. The transduction of this isoform in nontumorigenic keratinocytes induces proliferation, cell division, and migration. RNAseq analysis indicated that the FOXP3Δ2Δ7 isoform induces the expression of different protooncogenes and modulates essential pathways related to the immune response and the tumorigenic process. Abstract Cervical cancer (CC) is the fourth most common type of cancer among women; the main predisposing factor is persistent infection by high-risk human papillomavirus (hr-HPV), mainly the 16 or 18 genotypes. Both hr-HPVs are known to manipulate the cellular machinery and the immune system to favor cell transformation. FOXP3, a critical transcription factor involved in the biology of regulatory T cells, has been detected as highly expressed in the tumor cells of CC patients. However, its biological role in CC, particularly in the keratinocytes, remained unclarified. Therefore, this work aimed to uncover the effect of FOXP3 on the biology of the tumoral cells. First, public databases were analyzed to identify the FOXP3 expression levels and the transcribed isoforms in CC and normal tissue samples. The study’s findings demonstrated an increased expression of FOXP3 in HPV16+ CC samples. Additionally, the FOXP3Δ2 variant was detected as the most frequent splicing isoform in tumoral cells, with a high differential expression level in metastatic samples. However, the analysis of FOXP3 expression in different CC cell lines, HPV+ and HPV-, suggests no relationship between the presence of HPV and FOXP3 expression. Since the variant FOXP3Δ2Δ7 was found highly expressed in the HPV16+ SiHa cell line, a model with constitutive expression of FOXP3Δ2Δ7 was established to evaluate its role in proliferation, migration, and cell division. Finally, RNAseq was performed to identify differentially expressed genes and enriched pathways modulated by FOXP3Δ2Δ7. The exogenous expression of FOXP3Δ2Δ7 promotes cell division, proliferation, and migration. The transcriptomic analyses highlight the upregulation of multiple genes with protumor activities. Moreover, immunological and oncogenic pathways were detected as highly enriched. These data support the hypothesis that FOXP3Δ2Δ7 in epithelial cells induces cancer-related hallmarks and provides information about the molecular events triggered by this isoform, which could be important for developing CC.


Introduction
Cervical cancer (CC) is one of the most preventable and treatable types of cancer due to the extended precancerous phase that lasts decades, yet it is still recognized as a health burden in low-and middle-income countries where it is ranked as the fourth most common type of cancer [1,2] and the second leading cause of cancer death, both in women and men [3]. The most frequent histopathologic subtype of CC is squamous cell carcinoma (SCC), which accounts for nearly 75-85% of the cases, and it is derived from the transformation of squamous cells at the external portion of the cervix, termed ectocervix, and the adenocarcinoma (ADC), with approximately 10-25% of CC cases, which is a consequence of columnar epithelial cells transformation at the endocervix, the internal portion of the organ [4][5][6]; there is also a third subtype named as adenosquamous carcinoma (ADSC), which accounts for merely 2-3% and is originated by a merge of both squamous and glandular cells [5].
The main risk factor for CC is persistent infection by high-risk human papillomavirus (hr-HPV), such as the 16 and 18 genotypes, which both are associated with approximately 70% of cases worldwide [7,8]; HPV16 is frequently associated with SCC, while HPV18 is associated with ADC [9]. Hr-HPVs stimulate cell transformation over decades through the constant expression of their oncoproteins to activate immunological and carcinogenic pathways to favor tumorigenesis and modulate local immunity [10][11][12][13].
FOXP3, forkhead box p3, is a member of the transcription factor family known as forkhead/winged helix; its locus is in the short arm of the X chromosome, which transcribes for an mRNA of 11 protein-coding exons. Interestingly, by alternative splicing, five isoforms could arise from the FOXP3 mRNA; these are FOXP3-FL, FOXP3∆2, FOXP3∆7, FOXP3∆2∆3, and FOXP3∆2∆7 [14,15]. FOXP3 is recognized as a master regulator responsible for immune tolerance through the modulation of genetic and functional programs in regulatory T cells (Treg) [16]. The increase in Tregs in cancer is associated with a worse prognosis due to their immunosuppressive functions. The suppressive properties of Tregs rely on stable FOXP3 expression [16][17][18]. The two major isoforms of FOXP3 reported in Tregs are FOXP3-FL and FOXP3∆2, which are expressed in approximately equal amounts; the main functional difference between both isoforms is derived from the inability of FOXP3∆2 to inhibit RORα and RORγt functions [19][20][21]. A recently published study suggested that the expression of FOXP3 is closely related to the occurrence and growth of cervical cancer [22]. The expression of FOXP3 correlates with the prognosis of cervical cancer, and it is significantly higher in cancer than in cervical intraepithelial neoplasia or chronic cervicitis [22]. However, it remains unanswered whether there is a correlation between Tregs FOXP3+ and hr-HPV infection [23]. Nowadays, it is known that the FOXP3 expression is not exclusive of lymphoid lineage as it has been detected in epithelial cells from the breast, lung, prostate [24], and retinal tissue [25], as well as in cancer cells [26,27].
The role of FOXP3 in tumor cells is, however, controversial. It has been observed in breast and prostate cancer cells that FOXP3 expression is associated with antitumoral roles [28][29][30][31]. Furthermore, in ovarian cancer cells, inhibitory properties in proliferation, migration, and invasion were observed when FOXP3 was upregulated [32]. Conversely, the FOXP3 increase has also been linked to protumor functions in pancreatic, colorectal, gastric, bladder, thyroid, cervical, and nonsmall cell lung cancer [33][34][35][36][37][38]. In CC, the FOXP3 expression increased as the lesion progressed [39]. In addition, the knockdown of FOXP3 diminishes the growth of the SiHa cells [40]. However, whether the FOXP3 expression is regulated by HPV, which isoform is expressed in normal and tumoral keratinocytes, and the implications of FOXP3 expression in nontumorigenic epithelial cells are questions that remain unanswered. Therefore, this work aimed to evaluate the role of FOXP3 in the biology of keratinocytes and its contribution to the development of CC. 1 Primer used for gene cloning. 2  Finally, the plasmid pLVX-FOXP3∆2∆7 was used to obtain lentiviral particles.
After 5 min of incubation at room temperature, both solutions, diluted plasmids and diluted Lipofectamine 2000 reagent, were gently mixed and incubated for 20 min at room temperature. Finally, the plasmid-lipofectamine complexes were added to the cells, and the plate was mixed by gently rocking back and forth.
After 48 h, the supernatants of transfected cells were collected and filtered through a 0.45-µm PES filter to eliminate detached cells, aliquoted, and subsequently stored at −80 • C until use. Viral titer was determined by qPCR from 100 µL of supernatant.

HaCaT Cell Transduction
HaCaT cells were infected with at least 5 IFU per cell of lentiviral particles containing the FOXP3∆2∆7 ORF. After 72 h of infection, 0.5 µg/mL of puromycin (Cat. No. A1113803, Gibco, Waltham, MA, USA) was added to the cell culture. The cells were left under these conditions for another 72 h. After that, the cells were cultured in media with 0.1 µg/mL of puromycin for two more weeks or until no surviving cells were observed in the noninfected control plate. Following this procedure, two cell lines were obtained: HaCaT-LVX (transduced with the empty vector as control) and HaCaT-FOXP3∆2∆7.  50,000 cells were seeded and incubated until the next day. A 5 mM working solution was prepared in DMEM media without FBS. The cells were incubated with the working solution for an hour, then the media was changed, and fresh DMEM with 10% FBS was added. The cells were kept growing for 72 h until the fluorescence measurement by flow cytometry was performed. The cells were measured right before the incubation for the initial time detection.

Cell Migration Assay
For the cell migration assay, 800,000 cells were seeded in a 6-well plate (Cat. No. 140675, Thermo Fisher Scientific Inc., Waltham, MA, USA) for 24 h until full confluence was reached. A new media with 10 µg/mL mitomycin C (Cat. No. 10107409001, Roche Diagnostics, Basel, Switzerland) was added to avoid proliferation bias. After two hours of incubation with mitomycin C, three straight lines and one horizontal line, as a reference, were manually created with a 200-µL sterile pipette tip on each cell group and its duplicate; the plate was washed twice with PBS to remove nonadherent cells, and finally, fresh media was added.
Selection of the differentially expressed genes (DEGs) was defined by −1.5 ≤ Log2(fold change) ≥ 1.5 and p-value < 0.05 as selection criteria.

Gene Set Enrichment Analysis
From the data obtained by the DESeq2 tool, two matrices were generated, the first in a tab-delimited text file with the normalized expression data in FPKM and the second in a cls file with the identification of the phenotypes. Both matrices were loaded in GSEA software (version 4.2.2) [49], the "Hallmark" gene set database (h.all.v2022.1.Hs.symbols.gmt) was selected, and for the Chip Platform, the "Human_Gene_Symbol_with_Remapping_MsigDB.v7.5.1.chip" option was chosen. One thousand phenotype permutations were established, and the Signal2Noise parameter was selected in the metric for ranking genes configuration. Finally, an FDR value < 0.25 was chosen to identify significant enriched pathways.

Statistics
All experiments were performed with at least two independent replicates; the twoway analysis of variance (ANOVA) and unpaired Student t tests were used to calculate differences between groups. A p-value < 0.05 was considered statistically significant.

Augmented FOXP3 Expression in Cervical Cancer
To identify the expression profile of FOXP3 in CC samples, an expression analysis in OncoDB (https://oncodb.org/ accessed on 7 September 2022) [50] was performed. Results from CC samples (n = 304) compared against normal tissue (n = 22) showed a significant increase in FOXP3 mRNA expression in tumor samples (Figure 1a). To identify if FOXP3 expression was related to the three most common CC subtypes, data collected from "Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma (TCGA, Firehoses Legacy)" (n = 310) in cBioPortal was assessed (https://www.cbioportal.org/ accessed on 7 September 2022) [51,52]. The findings depict the FOXP3 mRNA expression level by CC subtype as well as the frequency of positive samples for FOXP3; deriving from those values, FOXP3 was detected in 98.8% of samples with SCC, 100% of ADC samples, and 80% of ADSC; moreover, FOXP3 expression average in SCC was higher compared with the other two subtypes (Figure 1b).

Augmented FOXP3 Expression in Cervical Cancer
To identify the expression profile of FOXP3 in CC samples, an expression analysis in OncoDB (https://oncodb.org/ accessed on 7 September 2022) [50] was performed. Results from CC samples (n = 304) compared against normal tissue (n = 22) showed a significant increase in FOXP3 mRNA expression in tumor samples (Figure 1a). To identify if FOXP3 expression was related to the three most common CC subtypes, data collected from "Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma (TCGA, Firehoses Legacy)" (n = 310) in cBioPortal was assessed (https://www.cbioportal.org/ accessed on 7 September 2022) [51,52]. The findings depict the FOXP3 mRNA expression level by CC subtype as well as the frequency of positive samples for FOXP3; deriving from those values, FOXP3 was detected in 98.8% of samples with SCC, 100% of ADC samples, and 80% of ADSC; moreover, FOXP3 expression average in SCC was higher compared with the other two subtypes (Figure 1b).

FOXP3Δ2 Is the Most Prevalent Variant in Cervical Cancer Samples, and Its Expression Is Highly Differential in Metastatic Stages
To deepen into the FOXP3 mRNAs isoforms expressed in cervical cancer, data from TCGA-CESC was analyzed in ISOexpresso (http://wiki.tgilab.org/ISOexpresso/ accessed on 10 October 2022) [53] and TSVdb (http://tsvdb.com/ accessed on 10 October 2022) [54] web-based platforms. Results obtained from the analyses showed that tumor samples express three variants with a differential frequency, FOXP3Δ2 with a frequency of 74%, FOXP3Δ7 with 17%, and FOXP3 X1 with 9% ( Figure 2a). TSVdb results showed that the most frequent variant, FOXP3Δ2, was expressed with a high difference between normal and metastatic tissue ( Figure 2b). The FOXP3Δ7 variant was observed with slightly increased expression in primary solid tumors, but no difference was observed for normal and metastatic samples ( Figure 2c). Unfortunately, the FOXP3Δ2Δ7 variant is not included in these databases. Therefore, no information related to this isoform was available in either database.

FOXP3∆2 Is the Most Prevalent Variant in Cervical Cancer Samples, and Its Expression Is Highly Differential in Metastatic Stages
To deepen into the FOXP3 mRNAs isoforms expressed in cervical cancer, data from TCGA-CESC was analyzed in ISOexpresso (http://wiki.tgilab.org/ISOexpresso/ accessed on 10 October 2022) [53] and TSVdb (http://tsvdb.com/ accessed on 10 October 2022) [54] web-based platforms. Results obtained from the analyses showed that tumor samples express three variants with a differential frequency, FOXP3∆2 with a frequency of 74%, FOXP3∆7 with 17%, and FOXP3 X1 with 9% ( Figure 2a). TSVdb results showed that the most frequent variant, FOXP3∆2, was expressed with a high difference between normal and metastatic tissue (Figure 2b). The FOXP3∆7 variant was observed with slightly increased expression in primary solid tumors, but no difference was observed for normal and metastatic samples (Figure 2c). Unfortunately, the FOXP3∆2∆7 variant is not included in these databases. Therefore, no information related to this isoform was available in either database.

HPV16 Infection Could Increase FOXP3 Expression Levels in Cervical Cancer
The main etiological factor for developing CC is persistent infection with high-risk HPV, such as 16 or 18 genotypes. To identify if there is an association between the HPV presence and the FOXP3 increase, FOXP3 expression levels in CC samples HPV16+ and HPV18+ from OncoDB were analyzed. Results obtained from OncoDB showed an increase in FOXP3 expression in the presence of both genotypes but with a marked difference in HPV16+ samples, in which a greater tendency of overexpression was observed compared to the HPV18+ and HPV(-) CC samples (Figure 3a).
After that, it was of interest to determine the status of FOXP3 expression in CC-de-

HPV16 Infection Could Increase FOXP3 Expression Levels in Cervical Cancer
The main etiological factor for developing CC is persistent infection with high-risk HPV, such as 16 or 18 genotypes. To identify if there is an association between the HPV presence and the FOXP3 increase, FOXP3 expression levels in CC samples HPV16+ and HPV18+ from OncoDB were analyzed. Results obtained from OncoDB showed an increase in FOXP3 expression in the presence of both genotypes but with a marked difference in HPV16+ samples, in which a greater tendency of overexpression was observed compared to the HPV18+ and HPV(-) CC samples (Figure 3a).

HPV16 Infection Could Increase FOXP3 Expression Levels in Cervical Cancer
The main etiological factor for developing CC is persistent infection with high-risk HPV, such as 16 or 18 genotypes. To identify if there is an association between the HPV presence and the FOXP3 increase, FOXP3 expression levels in CC samples HPV16+ and HPV18+ from OncoDB were analyzed. Results obtained from OncoDB showed an increase in FOXP3 expression in the presence of both genotypes but with a marked difference in HPV16+ samples, in which a greater tendency of overexpression was observed compared to the HPV18+ and HPV(-) CC samples (Figure 3a).
After that, it was of interest to determine the status of FOXP3 expression in CC-de-   The evidence so far supports that FOXP3 expression tends to increase in cervical cancer and HPV-positive cervical samples. However, in the panel of CC-derived cell lines analyzed, only the SiHa and C33A cell lines seem to maintain this feature. To have some insights into the molecular implications of FOXP3 expression, we selected the SiHa cell line. Therefore, the FOXP3 ORF from this cell line was isolated, cloned, and sequenced. The findings showed that SiHa cells express the FOXP3∆2∆7 variant (Figure 4b), an isoform of FOXP3 are poorly studied; then, the derived question was to know whether the FOXP3∆2∆7 varies the shapes and behavior of nontumorigenic keratinocytes. The evidence so far supports that FOXP3 expression tends to increase in cervical cancer and HPV-positive cervical samples. However, in the panel of CC-derived cell lines analyzed, only the SiHa and C33A cell lines seem to maintain this feature. To have some insights into the molecular implications of FOXP3 expression, we selected the SiHa cell line. Therefore, the FOXP3 ORF from this cell line was isolated, cloned, and sequenced. The findings showed that SiHa cells express the FOXP3Δ2Δ7 variant (Figure 4b), an isoform of FOXP3 are poorly studied; then, the derived question was to know whether the FOXP3Δ2Δ7 varies the shapes and behavior of nontumorigenic keratinocytes.

Exogenous Expression of FOXP3Δ2Δ7 Promotes Cell Proliferation, Division, and Migration
Once the FOXP3Δ2Δ7 ORF from SiHa was subcloned and stably transduced into Ha-CaT cells, validation of the lentiviral transduction was assessed by qPCR and RNAseq. The results demonstrated a significant increase in FOXP3's relative expression in HaCaT-FOXP3Δ2Δ7 (Figure 4a). Furthermore, the RNAseq analysis of the transduced model shows abundant reads covering the genomic region corresponding to the FOXP3 gene. It was also possible to observe the lack of exons 2 and 7, confirming that the mRNA isoform expressed by HaCaT-transduced cells was the FOXP3Δ2Δ7 version ( Figure 4b).
To identify the effect of FOXP3Δ2Δ7 exogenous expression in cell proliferation, cell index determination by impedance with the xCELLigence RTCA platform was performed ( Figure 5a); results exhibit an increase in cell index when compared with control. Additionally, cell division tracking assays revealed a diminished fluorescence in FOXP3Δ2Δ7expressing cells, which indicates a higher cell division rate (Figure 5b). Further, wound healing assays showed that FOXP3Δ2Δ7 significantly promoted the migration process (Figure 5c).

Exogenous Expression of FOXP3∆2∆7 Promotes Cell Proliferation, Division, and Migration
Once the FOXP3∆2∆7 ORF from SiHa was subcloned and stably transduced into HaCaT cells, validation of the lentiviral transduction was assessed by qPCR and RNAseq. The results demonstrated a significant increase in FOXP3's relative expression in HaCaT-FOXP3∆2∆7 (Figure 4a). Furthermore, the RNAseq analysis of the transduced model shows abundant reads covering the genomic region corresponding to the FOXP3 gene. It was also possible to observe the lack of exons 2 and 7, confirming that the mRNA isoform expressed by HaCaT-transduced cells was the FOXP3∆2∆7 version (Figure 4b).
To identify the effect of FOXP3∆2∆7 exogenous expression in cell proliferation, cell index determination by impedance with the xCELLigence RTCA platform was performed ( Figure 5a); results exhibit an increase in cell index when compared with control. Additionally, cell division tracking assays revealed a diminished fluorescence in FOXP3∆2∆7expressing cells, which indicates a higher cell division rate (Figure 5b). Further, wound healing assays showed that FOXP3∆2∆7 significantly promoted the migration process (Figure 5c).

Exogenous Expression of FOXP3Δ2Δ7 Induces the Transcription of Protumoral Genes and the Enrichment of Immunological and Oncogenic Pathways
Next-generation sequencing of total mRNA was performed to comprehend the molecular mechanisms behind the protumorigenic activities induced by the exogenous expression of FOXP3Δ2Δ7. RNAseq assays highlighted differentially expressed genes. Using the selection criteria of −1.5 ≤ Log2 (fold change) ≥ 1.5 and a p-value < 0.05, the results showed 23 upmodulated genes and 27 downmodulated genes. As expected, FOXP3 was the most overexpressed gene with a Log2(fold change) of 6.84, and among the 23 upregulated genes, numerous are associated with protumorigenic functions (Figure 6a). To validate the RNAseq results, the genes SATB1, C1R, GLI2, LAMP3, NSG1, and HSPB8 were selected and analyzed by qPCR (Figure 6b,c); the results show a global trend towards the upregulation of pro-oncogenic genes.

Exogenous Expression of FOXP3∆2∆7 Induces the Transcription of Protumoral Genes and the Enrichment of Immunological and Oncogenic Pathways
Next-generation sequencing of total mRNA was performed to comprehend the molecular mechanisms behind the protumorigenic activities induced by the exogenous expression of FOXP3∆2∆7. RNAseq assays highlighted differentially expressed genes. Using the selection criteria of −1.5 ≤ Log2 (fold change) ≥ 1.5 and a p-value < 0.05, the results showed 23 upmodulated genes and 27 downmodulated genes. As expected, FOXP3 was the most overexpressed gene with a Log2(fold change) of 6.84, and among the 23 upregulated genes, numerous are associated with protumorigenic functions (Figure 6a). To validate the RNAseq results, the genes SATB1, C1R, GLI2, LAMP3, NSG1, and HSPB8 were selected and analyzed by qPCR (Figure 6b,c); the results show a global trend towards the upregulation of pro-oncogenic genes.
Furthermore, for the recognition of FOXP3∆2∆7 modulated pathways, enrichment analysis was performed by using the GSEA software. For a confident selection of modulated pathways, an FDR q-value < 0.25 was used. Among the pathways modulated by FOXP3∆2∆7, several are involved in the immune response (e.g., complement, inflammatory response, IL2/STAT5 signaling, IL6/JAK-STAT signaling), and others have been associated with oncogenic processes (e.g., KRAS, WNT/beta-catenin, hypoxia) (Figure 7). Furthermore, for the recognition of FOXP3Δ2Δ7 modulated pathways, enrichment analysis was performed by using the GSEA software. For a confident selection of modulated pathways, an FDR q-value < 0.25 was used. Among the pathways modulated by FOXP3Δ2Δ7, several are involved in the immune response (e.g., complement, inflammatory response, IL2/STAT5 signaling, IL6/JAK-STAT signaling), and others have been associated with oncogenic processes (e.g., KRAS, WNT/beta-catenin, hypoxia) ( Figure 7).

Discussion
FOXP3, a member of the transcriptional factor family forkhead/winged-helix, is commonly known as a master regulator of regulatory T CD4+ CD25+ cells, which promotes homeostasis and immunological tolerance. These functions are established through the transcriptional activation or repression of approximately 700 genes and miRNAs related to the TCR pathway, cell communication, and transcriptional regulation [55].

Discussion
FOXP3, a member of the transcriptional factor family forkhead/winged-helix, is commonly known as a master regulator of regulatory T CD4+ CD25+ cells, which promotes homeostasis and immunological tolerance. These functions are established through the transcriptional activation or repression of approximately 700 genes and miRNAs related to the TCR pathway, cell communication, and transcriptional regulation [55].
Nowadays, it is known that FOXP3 expression occurs not only in hematopoietic cells but also in cancer cells. Karanikas et al. reported high expression levels of FOXP3 in 25 cancer-derived cell lines [26]. Several studies have reported the FOXP3 expression increase in tumor cells, and in some cases with tumor suppressor activity, as in ovarian, prostate, and breast cancer [29,32,56], but also as an oncoprotein, as in gastric, bladder, lung, and cervical cancer, among others [37,38,40,57].
Since 2012, when the first report about FOXP3's increase in CC was published [39], its biological role in the development of cervical pathology has been unclear. Initial findings showed that FOXP3 overexpression was especially marked in CC samples [22,39,40]; in this work, such increased expression was also evident in HPV+ CC samples, particularly within the most frequent genotype, HPV16. These findings suggest that the increase in FOXP3 does not occur solely due to the presence of HPV but could also be dependent on the infecting genotype; Zeng et al., in 2012, established the hypothesis that the virus uses FOXP3 increase as an essential mechanism to control the immune system and to preserve the active infection [39]. To confirm or deny this last postulate, it is necessary to carry out more elaborated approaches.
On the other hand, the analysis of FOXP3 variants' frequencies highlights the FOXP3∆2 isoform as the most common in CC samples. To the best of our knowledge, this is the first work analyzing FOXP3 variants in cervical cancer, particularly the FOXP3∆2∆7 variant, which was identified in the SiHa cell line. The relevant feature of losing the coding exon two region, or ∆2, is the absence of a repression domain, which is believed to have tumor suppressor functions due to a possible transcriptional restriction of oncogenes [58]. Therefore, isoforms lacking could allow free transcription of normally repressed genes. On the other hand, the full-length variant, in which the coding exon two is present, should have opposite activities, and this is confirmed by the observations using breast and ovarian cancer cells, in which the introduction of the FOXP3-FL variant decreases tumor growth, inhibits proliferation, and decreases migration and invasion [29,32]. Additionally, the coding exon two loss was associated with greater aggressiveness and chemo-resistance in bladder cancer [57]. Moreover, in nonsmall cell lung cancer, it was related to increased proliferation, migration, and invasion [38]. The loss of the coding exon 7 (∆7) was previously documented in an isoform of FOXP3 [15], FOXP3∆7 lacks an 81-bp region that contains part of the leucine zipper domain of the protein, which is necessary for protein-protein interactions, and it was postulated that mutations in the exon 7 prevent protein dimerization and, therefore, DNA binding [59]. Mutations in exon 7 were associated with an autoimmune disease called IPEX [60,61]. Nevertheless, Mailer et al., working with the FOXP3∆2∆7, found that the lack of exon 7 does not affect protein dimerization or interaction with RUNX1, NFAT, or NF-kB [62]. However, the exogenous expression of FOXP3∆2∆7 failed to induce the typical Treg-associated phenotype in lymphocytes [62]. The latter indicates that FOXP3∆2∆7 has a different function compared with FOXP3-FL. In this work, we report the expression of FOXP3∆2∆7 in SiHa cells. The lentiviral transduction of FOXP3∆2∆7 in nontumorigenic keratinocytes stimulates cell proliferation and division. These results agree with the observations of Luo et al. [40]. They induced the silencing of FOXP3 by employing RNA interference in SiHa cells and observed a decrease in proliferation, migration, and invasion [40]. Evidence of an FOXP3 oncogenic behavior was also observed in gastric cancer [37]. Similarly to our work, the exogenous expression of FOXP3 induces both proliferation and migration, however, the authors do not provide information about the FOXP3 isoform [37]. All this evidence indicates the importance of determining not only the expression of FOXP3 but also the isoform that the cell is expressing.
Functional analyzes establish clear evidence of the protumoral capacity of FOXP3∆2∆7 in vitro, to delve into the molecular mechanisms underlying the biological activities of this isoform transcriptomic analysis were performed. Among the overexpressed genes, SATB1 highly increased. This gene is involved in the epithelial-mesenchymal transition [63][64][65][66], and its overexpression has been associated with a worse prognosis in solid tumors [67]. In CC, it was identified as highly expressed in those women who presented advanced stages of the disease and through a Kaplan-Meier analysis, it was possible to associate it as a worse prognosis marker [68].
Other gene upregulated was GLI2, it has been identified as a potent oncogene that mediates the hedgehog signaling pathway [69][70][71]; in CC it was shown that its subexpression inhibited the growth and migration of cell lines derived from cancer, and it was also associated as a poor prognostic marker based on TCGA data [72].
Regarding C1R, there is no information related to CC; however, what has been published to date relates this gene to tumor growth, vascularization, and invasion in cutaneous squamous cell carcinoma [73,74]. High levels of C1R were reported in nonsmall cell lung cancer compared to control groups. Interestingly, the high levels of this gene were associated with an increased risk of death [75].
LAMP3 was associated with increased migration in an overexpression model derived from a cervical cancer cell line, and an in vivo metastasis assay demonstrated that 9 of 11 LAMP3 overexpressing mouse models elicited a process of distal metastasis with invasion of the lymphovascular space. Additionally, they demonstrated that in 100% of the CC samples, the gene was found to be overexpressed [76].
The enrichment analysis highlights important pathways modulated by FOXP3∆2∆7 expression, some of them closely related to the immune response. It has been reported that FOXP3 inhibits the activity of NFAT and NF-κB, two important transcription factors essential for cytokine expression [77]. These results agree with those reported in melanoma, which indicates that FOXP3 in cancer cells modulates the expression of molecules associated with immunity, in addition to contributing to the recruitment of Treg lymphocytes, which establishes a suppressive activity that favors tumor progression [78].
On the other hand, several of the signaling pathways identified as enriched are recognized for their oncogenic potential. The KRAS signaling pathway is highly involved in cell proliferation and division; it belongs to the RAS/MAPK pathway, and in the oncogenic context, KRAS has very important effects in the carcinogenesis process of endometrial, colorectal, bladder, breast, and cervical cancer [79][80][81][82][83][84][85][86].
Regarding the WNT/β-catenin signaling pathway, it is involved in both the proliferation process and migration; in CC, it is recognized as a practically necessary pathway for cell transformation, which leads to an increase in the aforementioned processes [87], additionally, it is interesting to mention that the SATB1 gene is related to the process of migration, and with the WNT/β-catenin pathway, this has been reported in colorectal cancer by promoting tumorigenesis, progression, and epithelial-mesenchymal transition [63,88].
Ultimately, it is important to mention that FOXP3∆2∆7 modulates the hypoxia signaling pathway, this pathway involves genes activated in response to low oxygen concentrations. In cancer, this phenomenon is highly relevant for the establishment of cancer hallmarks and the activation of cellular processes that include proliferation, survival, tumor invasion, and metastasis [89][90][91]. In CC, as expected, high degrees of hypoxia are associated with a poor prognosis [92] and, interestingly, also with resistance to chemotherapeutics [93,94]. As well, it is pertinent to acknowledge the role of LAMP3 in the pathway, as Mujcic et al. demonstrated in a cohort of human cervix tumors an increased expression of the gene as a consequence of hypoxia but furthermore, they found that LAMP3 is an essential regulator of hypoxia-driven nodal metastasis [95].
Future research should aim to evaluate the role of FOXP3∆2∆7 on the immune response regulation in the immune cells and tumor cells at the tumoral niche in CC.

Conclusions
The behavior of the FOXP3 expression during cervical cancer development was observed to increase. FOXP3 expression was also increased in HPV-positive samples, especially in those positive for HPV16. An analysis of the isoforms expressed in this pathology indicates the importance of the loss of the exon coding two of FOXP3. Functional studies suggest that isoform FOXP3∆2∆7 presents a protumorigenic behavior when introduced into keratinocytes since it induces cell proliferation, migration, and division. Additionally, the transcriptomic analysis highlights the importance of this isoform in the modulation of pathways, such as interferon, apoptosis, TNF, hypoxia, IL-2, IL-6, and the inflammatory response, among others. The latter reveals the critical role the FOXP3 isoforms could play in modulating the immune response. Subsequent studies evaluating FOXP3 isoforms and their effect on immune cells will provide a better understanding of the immunomodulatory effects of FOXP3 expression in tumor cells.