Close this search box.

Activated fibroblasts modify keratinocyte stem niche through TET1 and IL-6 to promote their rapid transformation in a mouse model of prenatal arsenic exposure – Scientific Reports

Prenatal arsenic exposure disrupts EMT-associated proteins in the skins of offspring

Our previous study suggested that only prenatal As exposure sufficiently enhances the effects of 7, 12-Dimethylbenz[a]anthracene/12-O-Tetradecanoylphorbol-13-acetate (DMBA/TPA), resulting in increased tumor numbers with reduced latency18. Tumors developed under the effects of prenatal exposure contained basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) types, with increased levels of PCNA, MMP 9, and Vimentin at 18 weeks. In this study, we hypothesized that prenatal As exposure might induce EMT in the skins of the offspring at an early stage which may lead to the development of aggressive tumors when re-challenged with DMBA/TPA. Following the same experimental regimen, we validated if EMT markers are elevated in response to prenatal exposure. EMT proteins, including E-cadherin, N-cadherin, Vimentin, and Fibronectin, were checked at 18 weeks (Fig. 1B–E) and postnatal day 2 (PND 2) (Fig. 1G–J). At termination, i.e., 18 weeks EMT proteins including E-cadherin, N-cadherin, Vimentin and Fibronectin were checked in tumour sections. In prenatally exposed animals, a decrease in the levels of E-cadherin (MFI 4.1 ± 0.6), with increase in N-cadherin (MFI 8.5 ± 1.2), Vimentin (MFI 4.4 ± 0.2) and Fibronectin (MFI 8.6 ± 1.2) was observed in comparison to control group animals with the expression levels in E-cadherin (MFI 7.0 ± 0.5), N-cadherin (MFI 1.4 ± 0.2), Vimentin (MFI 2.3 ± 0.2) and Fibronectin (MFI 0.8 ± 0.1). Further, to validate if these markers were elevated in response to prenatal exposure, before receiving any secondary hit, we checked these markers at an early stage, i.e., PND 2. The results confirmed an early EMT switch due to prenatal exposure as observed by the low levels of E-cadherin (MFI 1.7 ± 0.3) compared to control mice (MFI 11 ± 1.9). Also, upregulation in the expression of N-cadherin (MFI 8.5 ± 1.6), Vimentin (MFI 7.0 ± 08) and Fibronectin (MFI 3.2 ± 0.4) was clear with reference to control animals, i.e., (MFI 2.9 ± 0.5), (MFI 4.0 ± 0.5) and (MFI 1.4 ± 0.3) respectively.

Figure 1
figure 1

Prenatal arsenic exposure promotes EMT in mouse skin. (A) The figure depicts the experimental schedule, where mice were exposed to arsenic from 15 days before mating and throughout gestation until GD 21. Subsequently, the offspring received DMBA/TPA treatment, commencing at 6 weeks of age and persisting until 18 weeks. Assessment of EMT was done in skin tissue sections of offspring at 18 weeks (BF) and PND 2 (G-K). Representative images showing the level of E-cadherin, N-cadherin, Fibronectin, Vimentin, and α-SMA proteins involved in EMT; n = 5; Scale = 50 µm. The respective graph shows positive staining intensities presented as MFI; α-SMA expression is presented as positive cell counts per field; n = 5; Scale = 100 µm. Error bars indicate mean ± SEM. The significance levels were determined as *p < 0.05, **p < 0.01, and ***p < 0.001 using an unpaired Student’s t-test (two-tailed).

Prenatal arsenic exposure induces dermal fibroblast activation in mice

Our previous findings depicted disturbed basal cell dynamics, including increased proliferative rate, further causing accelerated tumorigenesis in the offspring. As basal cells are actively dividing progenitors, there may be some discrepancies in the functions of parent KSCs. In skin malignancies, stem cells may acquire EMT properties under the modified niche environment. As DFs play an essential role in stem cell niche modification during wounding and tumor conditions, we checked for their activation, comparing α-SMA positive cells in the dermis of prenatally exposed and control mice both at 18 weeks (Fig. 1F) and PND 2 (Fig. 1K). At 18 weeks, prenatally exposed offspring showed higher positive cells (26.2 ± 2.6) than control offspring (6.2 ± 1.0). Similarly, at PND 2 As group demonstrated a rise in α-SMA positive cell counts (37 ± 2.7) compared to unexposed pups (9.6 ± 1.2). These observations inferred a rise in AFs in prenatally exposed offspring at PND 2 and 18 weeks.

Activation of fibroblasts was also confirmed in vitro (Fig. 2). Following prenatal As exposure, DFs were isolated at PND 2 and were checked for their transformation into AFs by cell-specific markers, α-SMA, Collagen IV, and Fibronectin. Cultured fibroblasts showed a high number of α-SMA (MFI 10.5 ± 1.4) positive cells (Fig. 2A) in prenatally exposed group compared to the control (MFI 0.7 ± 0.1) cells. Additionally, the formation of stress fibers is responsible for mediating contractile function, demonstrating induced activation of fibroblasts in prenatally conditioned cells. AFs are characterized by increased synthesis of ECM (Extra Cellular Matrix) components, including stromal collagens.

Figure 2
figure 2

Prenatal arsenic exposure promotes the transition of DFs into activated phenotypes in vitro. The transition of DFs induced by prenatal exposure was assessed on primary cells isolated at PND 2 (AC). Representative images for EMT proteins, α-SMA, Collagen IV, and Fibronectin are presented. An increased number of α-SMA-stained stress fibers signify the differentiation of DFs into AFs. Graphical representation of MFI in prenatally exposed cells compared to control cells. (D) Immunoblots represent the quantitative analysis of total EMT proteins in cultured cell lysates. Error bars represent mean ± SEM; n = 3; Scale = 50 µm. Significance calculated as *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 by unpaired Student’s T-test (two-tailed).

A significant increase in the levels of Collagen IV (MFI 10.9 ± 0.7) concerning control cells (MFI 4.7 ± 0.7) confirmed induced activation of fibroblasts caused by prenatal stress (Fig. 2B). Another ECM protein, Fibronectin, known to be strongly involved in metastasis, was observed to be upregulated (MFI 5.8 ± 0.8) in prenatally conditioned fibroblasts compared to control cells (MFI 0.9 ± 0.4) (Fig. 2C). These results were also validated in whole cell lysates by immunoblotting (Fig. 2D), where levels of total proteins α-SMA, Collagen IV, and Fibronectin increased by 1.4 ± 0.08, 2.4 ± 0.2, 2.2 ± 0.1fold respectively.

Prenatally exposed DFs promote colony formation and EMT in KSCs via IL-6-enriched secretome

Cancer-associated stroma produces and releases diverse proteins, including IL-6, that promote tumor development and invasion21. Figure 3A represents islolation of primary DFs at PND 2. These fibroblasts are then subjected to a process of media conditioning, and subsequently, keratinocyte stem cells are cultured in this conditioned media. To assess the degree of expression and cellular source of IL-6, we used ELISA to identify its expression in both lysate and supernatant acquired from cultured DFs. As indicated in Fig. 3B, IL-6 levels in the prenatally exposed cell lysates were considerably higher (MFI 342.2 ± 31.6) than in unexposed control cells (MFI 91 ± 14.4). Similarly, IL-6 expression was substantially greater in cell supernatant procured from previously conditioned cells (MFI 7919 ± 350.4) vs. control cells (MFI 4086 ± 50.4) as presented in Fig. 3C.

Figure 3
figure 3

Activated DFs derived IL-6 rich secretome induces survival and EMT in KSCs. Both prenatally exposed DFs and unexposed KSCs were isolated from the offspring at PND 2 for in vitro experiments. (A) Primary DFs were isolated from the F1 generation at PND 2, followed by conditioning the media and culturing unexposed primary KSCs in it. The expression of IL-6 in both cell (B) lysate and (C) supernatant of prenatally exposed DFs was measured by ELISA assay and is reported as net MFI; n = 3. (D) Representative images display colony formation in KSCs cultured in DF-CM on days 2, 4, and 6, captured by phase contrast microscope; scale = 200 µm. (E) Survival in KSCs was quantified by determining the percentage of TUNEL positive cells per field. (F) Immunoblot quantification of EMT markers in cultured KSCs displays protein levels of N-cadherin, Vimentin, Fibronectin, and E-cadherin in IL-6 enriched EMEM-CM. The Bax to Bcl-xL ratio depicts the survival rate in KSCs. Graphs represent the quantification of total proteins in cell lysates of primary KSCs; n = 3. Error bars in all graphs denote the mean value with standard error of the mean (SEM). Statistical significance was calculated using an unpaired Student’s T-test (two-tailed) and is indicated as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

IL-6 is a known promoter for EMT and proliferation in epidermal stem cells. Since prenatal As exposure induced a high synthesis and release of IL-6 in DFs, we used this secretome to study niche modifications in unexposed primary KSCs. The first morphological difference observed was the time difference in colony formation between KSCs cultured with and without prenatally exposed dermal fibroblasts conditioned media. Media conditioned over prenatally exposed DFs induced a faster colony formation in KSCs than those cultured in media from unexposed fibroblasts. The snapshots were taken focusing same fields; representative images of days 2, 4, and 6 post-cell seeding are shown in Fig. 3D. In tumorigenesis, stem cell survival is an important aspect; therefore, the survival rate of KSCs was examined by TUNEL assay (Fig. 3E). Representative images demonstrate a reduced number of TUNEL-positive (apoptotic) cells in KSCs with prenatally exposed conditioned media (2 ± 0.3%) vs unexposed cells (10.4 ± 0.5%), showing a high survival rate acquired by KSCs induced by activated DFs.

Unexposed KSCs were cultured in conditioned media to examine the role of prenatally AFs in mediating EMT in skin tumors. As shown in Fig. 3F, the culture of KSCs in EMEM-CM acquired from prenatally exposed cells significantly reduced E-cadherin expression by 0.5 ± 0.06 fold while increasing N-cadherin expression by 1.6 ± 1.2 fold, confirming an alteration in their microenvironment. Also, cell marker Vimentin and Fibronectin, responsible for migration, were upregulated by 2.9 ± 1.8 and 1.8 ± 0.1 fold, respectively. The survival rate in KSCs cultured in EMEM-CM was also checked by drawing Bax to Bcl-xL ratio. A reduced ratio validated a high survival rate in these cells.

Activated fibroblasts derived IL-6 enriched conditioned media promotes GP130/JAK2/STAT3 activation in KSCs

The conventional IL-6 signal transduction pathway begins with interaction with IL-6R and phosphorylation of STAT3 through JAK2. To investigate the involvement of the IL-6/JAK2/STAT3 pathway in AFs-induced EMT alterations in KSCs, we first studied the activation of the IL-6/JAK2/STAT3 pathway in KSCs following their culture in conditioned media.

An essential component of the IL-6 receptor is GP130, which is needed for the receptor’s proper functioning, allowing its downstream cascade to be activated. Activation of the JAK-STAT pathway was validated at protein levels with upregulated levels of GP130 (1.7 ± 0.1), JAK2 (1.6 ± 0.1), and STAT3 (1.6 ± 0.1) in KSCs cultured in prenatally exposed conditioned media (Fig. 4A). ICC results in Fig. 4C–H also demonstrate upregulated levels of GP130, confirming an overexpression of IL-6 receptors in KSCs. Also, high phosphorylated protein levels of JAK2 and STAT3 established the activation of this pathway in KSCs cultured in prenatally exposed EMEM-CM.

Figure 4
figure 4

Prenatally activated DFs stimulates JAK2-STAT3 signaling in KSCs. Evaluation of IL-6/STAT3 pathway was done by western blotting and ICC. (A) Total proteins were measured by western blot analysis to confirm activated pathway. Graphical representation of densitometry quantification show IL-6/STAT3 pathway proteins; n = 3. (B) Immunoblots showing pSTAT3 expression in nuclear extracts of KSCs. GAPDH and Histone 3(H3) represent cytosolic and nuclear controls respectively. (CE) ICC images of GP130, pJAK2, and pSTAT3 in KSCs cultured in EMEM-CM. Images indicate an increased nuclear translocation of pSTAT3 in KSCs cultured with prenatally conditioned DF-CM. (F, G) Graphs represent the Mean Fluorescence Intensity (MFI) of cell-specific markers and nuclei to cytosol ratio (H) for pSTAT3 translocation, with a sample size of n = 5. The scale bar in ICC images corresponds to 50 µm.

Under basal conditions, STAT3 may be found in both the cytoplasm and nucleus, which shows that STAT3 constantly moves between the two cellular compartments. Activation of STAT3 involves multiple steps, including phosphorylation by JAK2, followed by the dimerization of pSTAT3 and translocation in the nucleus, DNA binding, and expression. Increased import of STAT3 in the nucleus has been implicated in cancers as it activates genes specific to cell cycle progression, EMT, and survival. Increased translocation was confirmed by protein levels of pSTAT3 in cytosolic and nuclear fractions of KSCs (Fig. 4B). These results were also validated by ICC results (Fig. 4E, H). pSTAT3 was observed to be localized more in the nuclei of KSCs than in cytoplasmic regions, providing evidence of active translocation due to dysregulated niche.

Blockade of GP130/JAK2/STAT3 pathway impairs proliferation, migration, and EMT in KSCs

We investigated the role of the IL-6-induced GP130/JAK2/STAT3 pathway in the proliferation, migration, and EMT of KSCs, as these properties are known to promote cell metastasis in cancer. We confirmed the influence of AFs in modifying the colony-forming potential, proliferation, migration, and EMT properties of unexposed KSCs.

To validate these findings, we used SC144, a specific inhibitor of IL-6-induced STAT3 activation, to block the activation of phosphorylated STAT3 (pSTAT3). Colony forming ability in KSCs was checked by clonogenic assay (Fig. 5A) where number of colonies in conditioned group (As) were ≥ 92 compared to control with colony no. ≥ 67. However, the colony formation rate declined after pSTAT3 inhibition in both control and As groups, i.e., ≥ 50 and ≥ 66, respectively. Figure 5B illustrates high migratory rate of KSCs, as seen in representative images taken at 18 h, where complete healing of the wound was observed in KSCs cultured in prenatally exposed DFs culture media. Increased migration was further reduced after inhibiting pSTAT3, confirming the involvement of activated STAT3 signaling.

Figure 5
figure 5

Blocking IL-6-STAT3 activation impairs colony formation, proliferation, and migration in KSCs. Acquired proliferative and migratory properties in KSCs were evaluated by clonogenic, scratch wound and BrdU assay. (A) Representative images of clonogenic assay showing before and after pSTAT3 blockade in KSCs. Images were captured using ChemiDoc™ MP (BioRad). (B) Representative images of scratch wound assay at 18 h in KSCs with EMEM-CM, before and after pSTAT3 inhibition. (C) BrdU proliferation assay showing a number of dividing cells in KSCs cultured in EMEM-CM before and after pSTAT3 inhibition by SC144. Graphs represent the quantification of respective assays; n = 5; scale = 200 µm. Error bars represent mean ± SEM. *p < 0.05, ****p < 0.0001 according to two-way analysis of variance (ANOVA) and Tukey’s multiple comparison tests. (D) Progression of EMT in KSCs validated by immunoblotting. Total protein separated and developed against pSTAT3, N-cadherin, Vimentin, Fibronectin, and E-cadherin. Graphs show quantification of total protein levels in primary KSCs; n = 3. **p < 0.01; ***p < 0.001; ****p < 0.0001 computed using two-way ANOVA and Tukey’s multiple comparison tests.

The proliferative rate induced in KSCs by EMEM-CM was validated by BrdU assay in primary KSCs (Fig. 5C). The results depicted an increased uptake of BrdU by KSCs demonstrating a high division rate (11.2 ± 0.7%) in these cells when cultured in IL-6-rich media vs. control cells (3 ± 0.8%). However, the number of BrdU positive cells reduced significantly after pSTAT3 inhibition confirming induction of proliferation of KSCs (3.4 ± 0.5%) than in control (2.0 ± 0.3%) KSCs due to activated IL-6- STAT3 axis following prenatal As exposure.

To confirm the involvement of the IL-6-STAT3 axis in promoting EMT, we re-evaluated the protein levels after inhibiting pSTAT3 with SC144 (Fig. 5D). We observed an upregulation of pSTAT3 by 1.7 ± 0.1 fold in KSCs cultured in prenatally conditioned media. However, the expression decreased to 0.41 ± 0.09 after inhibition with SC144, compared to 0.28 ± 0.06 in the control group treated with SC144. The levels of EMT proteins were also rescued after pSTAT3 inhibition, confirming the role of IL-6-induced activation of the JAK2-STAT3 pathway in KSCs. The observed levels of E-cadherin in the prenatally exposed group were 0.4 ± 0.08 fold, with no significant changes after pSTAT3 inhibition, i.e., 0.43 ± 0.03 fold. Similarly, the upregulated levels of N-cadherin, which were 1.8 ± 0.1 fold in KSCs cultured in prenatally conditioned fibroblast media, decreased to 1.2 ± 0.14 fold. The protein levels of Vimentin also decreased from 1.7 ± 0.14 fold in the prenatally exposed group to 0.8 ± 0.09 fold in the group treated with both prenatally conditioned media and SC144. Fibronectin levels decreased from 1.8 ± 0.08 to 1.1 ± 0.07 fold after inhibiting the IL-6-STAT3 cascade.

TET1 increases IL-6 production by the accumulation of 5-hmC at the promoter in prenatally exposed primary DFs

Epigenetic modifications at promoter sites have been linked to the excessive production of multiple inflammatory cytokines, with changes in methylation levels being a possible cause. In our study, we focused on prenatal exposure and examined the mRNA levels of specific genes that may regulate the epigenetic landscape of IL-6 in prenatally challenged DFs. We found a significant increase in the levels of TET1 in prenatally exposed DFs. TET1 is responsible for converting 5-mC to 5-hmC, which makes the target region more active by reducing methyl marks. Increased 5-hmC conversion was validated via immunostaining on primary DFs (Fig. 6A). Prenatally exposed DFs showed a higher transition of 5-mC to 5-hmC compared to unexposed cells. The number of cells with positive 5-mC modification decreased in previously exposed DFs with significant rise in 5-hmC marks. We also confirmed an upregulation (1.8 ± 0.06 fold) of TET1 at protein level through western blotting in prenatally exposed DFs, compared to the control group (Fig. 6B). We also observed a significant upregulation of IL-6 mRNA levels by 8.1 ± 1.4 fold in the prenatally exposed cells (Fig. 6C).

Figure 6
figure 6

TET1 mediates IL-6 overexpression via increased 5-hmC accumulation at the gene promoter. MeDIP analysis of 5-hmC and 5-mC at IL-6 promoter for TET1 activity. (A) Representative images show conversion of 5-mC to 5-hmC in prenatally exposed DFs. The graph represents numbers of 5-hmC and 5-mC positive cells in primary DFs. (B) Immunoblot analysis of total TET1 protein in primary DFs. (C) mRNA levels of IL-6 in prenatally exposed primary DFs presented as fold change. (D, E) MeDIP results are represented as a 5-hmC: 5-mC ratio adjusted to input. Results are reported as relative enrichment (n = 3) with respect to the control group covering two promoter sites. Scale bars = 50 µm. Data are presented as the mean ± SEM. Significance calculated as *p < 0.05; **p < 0.01; ***p < 0.001 by an unpaired Student’s T-test (two-tailed). (F) Schematic summary depicting the impact of prenatal As exposure on increased skin carcinogenesis. Prenatal exposure results in an increased expression of TET1 in dermal fibroblasts. This leads to a simultaneous rise in 5-hmC levels at the promoter region of Il-6, along with its release in the niche of KSCs. As a result, the increased activity of the downstream JAK2/STAT3 signaling pathway in KSCs leads to heightened cell growth and promotion of EMT. This disruption in the cellular microenvironment contributes to a more severe tumor response in the offspring.

Furthermore, we assessed the ratio of 5-hmC to 5-mC in a 2 kb upstream flank of the IL-6 gene promoter to examine any alterations due to disturbed TET1 function (Fig. 6D, E). We performed MeDIP on primary DFs isolated at PND 2 and found an increase of 8.4 ± 1.5% in the 5-hmC/5-mC ratio compared to the input in prenatally exposed DFs, while control DFs showed a decrease in the ratio with 0.5 ± 0.1% (Fig. 6D). Another site showed a similar pattern, though with less prominent modifications, with an increase of 1.6 ± 0.16% in the prenatally exposed group compared to the decreased ratio in control cells of 0.4 ± 0.09% (Fig. 6E). These results confirm an increased rate of conversion in 5-hmC at the Il-6 promoter, leading to increased gene expression, subsequent high synthesis, and release of IL-6 by DFs. This further contributes to the transformation of the microenvironment and primes KSCs for enhanced skin tumorigenesis.

Latest Intelligence