Harnessing Id1 as a biomarker in a plasmid reporter system for cervical cancer

Abstract

Stagnancy of ten-year cervical cancer (CC) incidence in the U.S., despite screening advancements, suggests the need for new CC screening technologies. This study in preclinical CC models evaluated a diagnostic plasmid that induces expression of a reporter (secreted embryonic alkaline phosphatase, SEAP) through the control of cancer-specific promoter sequence (inhibitor of differentiation 1, Id1). The plasmid (pId1-SEAP) was used to transfect CC cells in vitro and characterize SEAP production based on Id1 expression. Western Blot and immunohistochemistry were used to establish Id1 expression in cell models and human tissues. Timed transfections in various conditions were used to correlate Id1 and SEAP expression. CC cell lines expressed increased normalized baseline Id1 (HeLa 3.0 ± 0.13, SiHa 2.9 ± 0.27, both P < 0.0001) compared to non-cancer 3T3 fibroblasts (1.0 ± 0.0). Normal cervical tissues had a mean Id1 staining value of 3E4 ± 3E4, while early- and late-stage CC tissues had increased mean Id1 staining (3E5 ± 1E5 P < 0.0001 and 2E5 ± 1E5 P = 0.0002, respectively). HeLa and SiHa lines produced increased normalized SEAP (0.63 ± 0.25 and 0.50 ± 0.10, P < 0.05) compared to 3T3 cells, both with pId1-SEAP (0.16 ± 0.058). As few as 12,500 pId1-SEAP transfected HeLa cells resulted in increased SEAP (3E4 ± 3E3 P = 0.004) compared to background (1E4 ± 4E2). SiHa xenograft ex vivo tumor transfected with 25 µg/µL pId1-SEAP produced significantly greater SEAP (21.7 ± 8.6, P = 0.0003) relative to muscle transfected in the same conditions (0.94 ± 0.24). pId1-SEAP can transfect CC cells to produce SEAP proportionally to endogenous Id1 expression, demonstrating its potential for additional studies in CC models.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-49073-2.

Introduction

The American Cancer Society estimates 13,360 new cases of cervical cancer will be diagnosed in the U.S. in 2025. However, over 90% of cervical cancer is preventable with vaccination against human papillomavirus (HPV) infection or with routine participation in early detection programs that screen for the presence of HPV infection or the presence of abnormal cervical cells. The five-year survival rate for localized cervical cancer is 91%, and regular participation in screening according to recommended guidelines increases the likelihood of detecting precancer or localized cervical cancer that can be treated and cured. Even with existing screening and prevention efforts, patients in rural and medically-underserved areas of the country are disproportionately diagnosed at later stages and consequently succumb to this preventable disease¹. Cervical cancer incidence has not changed considerably within the past 10 years² which suggests that current screening and prevention methods have reached their maximum efficacy and new efforts are needed to continue to decrease disease incidence. Current recommended screening efforts rely on detection of HPV DNA or abnormal cells in a prepared Pap cytology smear, both of which analyze cells collected from a cervical swab. HPV testing only indicates the presence of a current HPV infection, which could be transient and clear spontaneously without treatment or intervention³. In contrast, Pap cytology detects abnormal cells that could be malignant, alerting to the presence of possible precancerous conditions, such as low- or high-grade squamous intraepithelial lesions (LSIL, HSIL)⁴. While HPV testing is the most sensitive of current cervical cancer screening methods, it lacks specificity in younger populations due to its inability to distinguish between transient and persistent, transforming HPV infections^5,6.

Inhibitor of differentiation 1 (Id1) is a class V helix-loop-helix (HLH) protein that serves as a negative regulator of basic HLH transcription factor DNA binding through its ability to heterodimerize with other HLH proteins and its lack of DNA binding domain. As part of the Id protein family, Id1 has been implicated in normal processes such as immune cell development, skin injury, and other inflammatory conditions^7–9. Id1 dysregulation has also been implicated in several cancer types and has been strongly correlated with the grade, stage, and aggressiveness of cervical cancer, with Id1 expression increasing proportionately during the progression from normal to precancer to cancer^7,10–13. Additionally, Id1 expression has been closely related to active transforming HPV infections, due to the link between increased Id1 and HPV E6 expression and the role of Id1 in promoting cervical carcinogenesis through its interaction with HPV E6 and E7 targeting p53 and Rb tumor suppressor genes⁹. Therefore, Id1 presents a promising target for differentiating between dormant and actively transforming HPV infections. Previous research has evaluated the use of Id1 in a plasmid-based reporter system for diagnostic applications in breast and prostate cancer, using the reporter protein secreted embryonic alkaline phosphatase (SEAP)^14,15. SEAP has several characteristics that support its role as a reporter protein, namely its heat stability, resistance to phosphatase inhibitor L-homoarginine, and ability to be detected at concentrations on the order of picograms/mL¹⁶. Combining Id1 regulatory sequences into an expression plasmid for SEAP could allow for the cancer-specific controlled production of SEAP as a reporter for transforming HPV infection based on Id1 regulatory activity through the downstream interaction of HPV E6 protein.

In this study, we evaluated a pId1-SEAP plasmid vector for the cancer-specific production of reporter SEAP in the presence of Id1 overexpression in preclinical cervical cancer models. In vitro transfections were performed using pId1-SEAP in cervical cancer cell lines, with a previously characterized breast cancer cell line as a positive control and a normal fibroblast cell line as a negative control. pId1-SEAP is hypothesized to transfect cervical cancer cells and result in proportionate expression of reporter SEAP according to endogenous Id1 expression levels. pId1-SEAP function was characterized in cervical cancer cells by examining the time course of SEAP expression post-transfection and evaluating the minimum number of transfected cells required to produce detectable levels of SEAP above background. SEAP production was also evaluated in cervical cancer xenograft ex vivo tumor sections exposed to pId1-SEAP.

Results

Expression of Id1 in human cervical cancer cell lines and tissues

Three cervical cancer cell lines, one previously characterized breast cancer cell line, and one fibroblast cell line were probed for Id1 expression to determine the relative expression of Id1 (normalized to Beta-actin) in cancer models compared to a non-cancer model. Cervical cancer cell lines HeLa, SiHa, and CaSki expressed significantly increased normalized Id1 (3.0 ± 0.13, 2.9 ± 0.27, and 3.1 ± 0.19 respectively, P < 0.05) compared to negative control normal fibroblast 3T3 cells (Fig. 1a and b). The positive control breast cancer line MDA-MB-231 had a normalized Id1 expression of 1.5 ± 0.11, which was also significantly increased compared to 3T3 cells (P < 0.05). Cervical cancer patient tissues in a microarray, including over 100 cases of stage I-IV disease, were also probed for Id1 expression (Fig. 1c). Normal tissues had an average Id1 staining value of 3.4E4 ± 3.5E4, while early-stage (stage I-II) cancers had an average staining value of 3.0E5 ± 1.7E5 and late-stage (stage III-IV) cancers had an average staining value of 2.7E5 ± 1.7E5. Both early and late-stage cancer tissues had significantly increased Id1 staining compared to normal tissue (P < 0.05, Fig. 1d).

Fig. 1.

Id1 is expressed in human cervical cancer cell lines and tissues. a) Western blot (image cropped) for Id1 from cell lines with Beta-actin as a loading control, b) ImageJ densitometry quantification results of protein bands from a, c) image of cervical tissue microarray immunohistochemically stained for Id1 expression, with normal tissues on bottom row, d) QuPath quantification results of Id1 staining in tissues from c. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05.

Comparison of SEAP production in cervical cancer cells transfected with pId1-SEAP to endogenous Id1 expression

pCMV-SEAP was used to optimize transfection conditions for each of the cervical cancer cell lines by comparing various plasmid DNA:jetPEI ratios for highest SEAP output and highest cell viability at 48 h post-transfection (Supplemental Fig. 1a-c). For each cell line, one plasmid:jetPEI ratio was chosen based on the highest viability and SEAP results; the final ratios for each cell line are displayed in Supplemental Fig. 1d. These optimal ratios were used for all subsequent transfections on these cell lines. SEAP measured after pCMV-SEAP transfection for 48 h was as follows for each cell line (Fig. 2a): 3T3, 1.8E5 ± 1.2E5 counts; MDA-MB-231, 7.2E4 ± 6.6E4 counts; HeLa, 5.9E5 ± 1.5E5 counts; SiHa, 2.0E4 ± 5.0E3 counts; CaSki, 6.5E4 ± 1.5E3 counts. These values represent each cell line’s ability to be transfected and produce SEAP from the plasmid with no specific promoter control and allow for transfection efficiency normalization of other investigational plasmids in each cell line. The optimal amount of pId1-SEAP plasmid DNA was found by transfecting each cervical cancer cell line with a range of plasmid concentrations, with the 0.001 µg/µL condition producing the highest SEAP counts in each cell line (HeLa 8.8E5 ± 1.3E5, SiHa 2.3E4 ± 6.2E3, CaSki 9.5E3 ± 2.5E3) compared to the non-transfected control (HeLa 7.8E3 ± 7.8E2, SiHa 5.5E3 ± 7.3E2, CaSki 4.1E3 ± 6.9E2) (Supplemental Fig. 2a-c).

Fig. 2.

Id1 expression correlates with SEAP production in cervical cancer cells transfected with pId1-SEAP. a) SEAP counts from media of cell lines at 48 h post-transfection with pCMV-SEAP, b) SEAP counts (normalized to pCMV-SEAP data from a and background subtracted) from media of cell lines at 48 h post-transfection with negative control pSEAP2-basic or pId1-SEAP, c) linear regression and Pearson’s correlation results between normalized SEAP values from b and normalized Id1 values from Fig. 1, and d) SEAP production rates over 72 h from cervical cancer cell lines transfected with negative control pSEAP2-basic, positive control pCMV-SEAP, pId1-SEAP, or not transfected with plasmid (Sum Total Significance Results: HeLa, pId1-SEAP vs. pSEAP2-basic P = 0.0060; SiHa, pId1-SEAP vs. pSEAP2-basic P = 0.2276; CaSki, pId1-SEAP vs. pSEAP2-basic P = 0.5294). ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05.

Figure 2b shows normalized, background-subtracted SEAP counts from cells transfected with the control plasmid pSEAP2-basic lacking a promoter sequence and cells transfected with pId1-SEAP relying on the Id1 promoter for transcriptional control. HeLa and SiHa lines produced significantly increased SEAP following transfection with pId1-SEAP compared to negative control 3T3 cells transfected with pId1-SEAP (3T3: 0.16 ± 0.058, HeLa: 0.63 ± 0.25, SiHa: 0.50 ± 0.10, P < 0.05). CaSki cells did not produce an increased amount of normalized SEAP after transfection with pId1-SEAP (0.048 ± 0.011) compared to 3T3 cells transfected with pId1-SEAP. While not shown on the graph in Fig. 2b, SEAP values from pSEAP2-basic and pId1-SEAP transfection were compared within each cell line; HeLa and SiHa had significantly increased SEAP values with pId1-SEAP transfection compared to negative control pSEAP2-basic transfection (HeLa 0.63 ± 0.25 vs. 0.0020 ± 0.0020, SiHa 0.50 ± 0.10 vs. 0.093 ± 0.11, P < 0.05). A nonsignificant correlation was found between normalized Id1 and SEAP expression values (Pearson’s correlation r = 0.51, P = 0.38) from HeLa, SiHa, CaSki, 3T3, and MDA-MB-231 cell lines transfected with pId1-SEAP (Fig. 2c). Removing CaSki from the correlation as a biological outlier (high Id1 expression, low SEAP expression) yielded a significant correlation between normalized Id1 and SEAP expression among the other cell lines (graph not shown, r = 0.99, P = 0.012).

Rates of SEAP production over a 72-hour experiment were determined from HeLa, SiHa, and CaSki cells without transfection or transfected with pSEAP2-basic, pCMV-SEAP, and pId1-SEAP (Fig. 2d). SEAP rates per hour were compared for each 12-hour segment following exposure to plasmid. The three cell lines expressed increased SEAP levels after pId1-SEAP transfection compared to pSEAP2-basic at different timepoints throughout the experiment. Total SEAP produced over 72 h was significantly greater for HeLa cells transfected with pId1-SEAP compared to pSEAP2-basic (1.1E5 ± 4.4E4 vs. 1.9E3 ± 3.4E2, respectively; P = 0.0060), with peak expression occurring at 60 h post-pId1-SEAP transfection. Peak SEAP expression in SiHa cells occurred at 60 h post-pId1-SEAP transfection; total SEAP production in SiHa cells over 72 h was not significantly different following transfection with pId1-SEAP compared to pSEAP2-basic (2.9E3 ± 7.6E2 vs. 2.0E3 ± 6.6E2, respectively; P = 0.23). CaSki reached peak SEAP expression at 72 h post-pId1-SEAP transfection, although total SEAP production over 72 h was not significantly different between CaSki cells transfected with pId1-SEAP or pSEAP2-basic (3.4E3 ± 8.3E2 vs. 1.5E3 ± 4.0E2, respectively; P = 0.53). A preliminary version of this experiment showed the accumulation of SEAP in media from HeLa and CaSki cells transfected with pCMV-SEAP and pId1-SEAP; accumulation was significantly increased over the first 12 h post-transfection with pId1-SEAP for CaSki cells only (Supplemental Fig. 3a-b). To determine if SEAP expression rates were influenced by differences in cell line growth, the percentage change in cell number was quantified for each cell line, which showed that SiHa and CaSki had low cell growth compared to HeLa throughout the transfection period (data not shown). Additionally, lysates were harvested from transfected cells at 48 h post-transfection and measured for Id1 expression using Western blot. Quantification and normalization of Id1 expression to Beta-actin showed no significant differences in Id1 expression between plasmid transfection conditions (data not shown).

Threshold for minimum number of cells exposed to pId1-SEAP required to produce significantly increased levels of SEAP

Next, we explored the amounts of cells exposed to plasmid required for significantly increased SEAP levels. Figure 3a shows the experimental design for co-culturing plasmid-exposed cervical cancer cells (using either pCMV-SEAP or pId1-SEAP) with cells that were not exposed to plasmid to simulate an environment wherein not every cell is successfully transfected with plasmid DNA. Cells were mixed at varying ratios to create a stepwise decrease in the number of total cells exposed to plasmid diluted with cells that were not exposed to plasmid, ranging from 100% of plasmid-exposed cells and decreasing by half with each dilution to as low as 6.25% of plasmid-exposed cells plated. pCMV-SEAP again served as a positive control, and results in Supplemental Fig. 4a-c show that as few as 50,000 HeLa cells exposed to pCMV-SEAP (a), 25,000 SiHa cells exposed to pCMV-SEAP (b), and 200,000 CaSki cells exposed to pCMV-SEAP (c) were needed to see a significant SEAP value compared to background levels from cells not exposed to plasmid. When pId1-SEAP was used in the co-culture experiment (Fig. 3b), as few as 12,500 plasmid-exposed HeLa cells, 25,000 plasmid-exposed SiHa cells, and 50,000 plasmid-exposed CaSki cells produced significantly greater SEAP levels (HeLa: 3.6E4 ± 3.2E3 P = 0.0041, SiHa: 1.8E4 ± 2.7E3 P = 0.039, CaSki: 5.2E3 ± 1.8E2 P = 0.033) above levels from cells that were not exposed to plasmid (HeLa: 1.1E4 ± 4.0E2, SiHa: 1.0E4 ± 1.0E2, CaSki: 4.8E3 ± 1.1E2).

Fig. 3.

Threshold for minimum number of cells exposed to pId1-SEAP required to produce positive levels of SEAP. a) Experimental design for co-culture of plasmid-exposed cells and naïve cells (Created with BioRender.com), and b) SEAP counts from media of cervical cancer cell lines 48 h after plating co-culture of naïve and pId1-SEAP-exposed cell dilutions. *, P < 0.05 compared to 0% transfected cells.

Feasibility of tissue transfection with pId1-SEAP

Finally, we developed a preliminary ex vivo method to determine the feasibility of transfecting tissue with pId1-SEAP. Figure 4a shows the experimental design to determine if tumor tissue exposed to pId1-SEAP results in SEAP production in vitro. SiHa cell xenograft tumors grown in the flank of athymic nude mice were harvested and exposed to pId1-SEAP ex vivo to simulate transfection of 3-D tumor tissue. Amounts of tumor tissue were varied to create a baseline (1x) tumor tissue condition, along with a 2x tissue and 3x tissue condition, corresponding to a stepwise increase in tumor weight for each condition (Fig. 4c). pId1-SEAP concentrations of 25 µg/µL or 75 µg/µL were used in these studies to determine whether an increased amount of plasmid yielded greater SEAP signals in 3-D ex vivo tissues. Muscle tissue was also harvested from the mice and exposed to pId1-SEAP at 25 µg/µL for a representative background measurement of SEAP produced from tissue with low levels of endogenous Id1 expression¹⁷. As shown in Fig. 4b, a 3x amount of tumor tissue produced a significantly increased fold change in SEAP counts (fold change relative to tumor tissue not exposed to pId1-SEAP, 21.7 ± 8.6) compared to muscle tissue fold change (0.94 ± 0.24, P = 0.0003) at 25 µg/µL pId1-SEAP concentration; the fold change in SEAP counts of 3x tumor tissue exposed to 75 µg/µL pId1-SEAP was also significantly higher (19.5 ± 6.4, P = 0.0004) relative to the muscle tissue fold change. No significant differences were seen between fold change in SEAP count for the different concentrations of pId1-SEAP evaluated. ATP concentrations measured from tissue lysates indicated viability of the ex vivo tissues used in these studies at 48 h post-harvest compared to the low ATP concentration seen from tumor tissue cultured in PBS as a negative control for viability (Fig. 4d). Representative images of histological tissue sections from these studies confirm that SiHa tumor xenografts were malignant by H&E staining (Fig. 4e) and were positive for Id1 expression by IHC staining (Fig. 4f). Similar results were seen in a replicate experiment using HeLa xenograft tumors (Supplemental Fig. 5a-e), where a 3x amount of tumor tissue exposed to 25 µg/µL pId1-SEAP yielded a significantly increased fold change in SEAP counts compared to fold change in muscle tissue (3.5 ± 2.1 vs. 1.0 ± 0.15, respectively; P = 0.025).

Fig. 4.

Feasibility of tissue transfection with pId1-SEAP. a) Experimental design for harvest of subcutaneous flank cervical tumors and ex vivo exposure of varying amounts of tumor tissue to pId1-SEAP in varying concentrations (created with BioRender.com), b) Fold change in SEAP counts at 24 h for 1x, 2x, or 3x amounts of plasmid-exposed ex vivo SiHa tumor tissue compared to non-plasmid-exposed SiHa tumor tissue with muscle as representative background tissue, c) Corresponding weight (g) at 48 h of ex vivo tissues in b, d) Corresponding ATP concentration (M) at 48 h of ex vivo tissues in b measured by ATPlite assay (PBS only evaluated in the 1x Tumor group), e) H&E image of representative SiHa tumor tissue imaged at 20X magnification, f) Id1 IHC image of representative SiHa tumor tissue imaged at 20X magnification with brown staining representing Id1 expression. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05.

Discussion

The goal of this study was to determine the ability of pId1-SEAP to transfect cervical cancer cells in vitro and to further characterize SEAP expression capacity based on endogenous Id1 expression. Previous preclinical research has suggested Id1 as a promising biomarker for cervical cancer, correlating Id1 expression with increasing stage, grade, and aggressiveness of the disease^10–13. However, the clinical utility of Id1 as a cervical cancer biomarker has not yet been confirmed. Previous studies have established Id1 dysregulation as an important factor in creating and maintaining cancer-specific phenotypes through its role in implementing a stem-cell-like regulatory program in several malignancies. Id1 dysregulation has also been shown to play a role in tumor angiogenesis, metastasis, and therapeutic resistance mechanisms^8,9. Research on the role of Id1 in these mechanisms in cervical cancer is limited. Schindl et al. and Maw et al. were some of the first to show that high Id1 expression is a possible prognostic factor in cervical cancer, reporting significant correlations between increased Id1 expression in patient samples and poorer overall survival^11,12. However, the underlying mechanistic interactions of Id1 overexpression with cervical cancer cell processes have not been clearly elucidated, with some studies focused on the tumorigenic link between HPV infection and increased Id1 expression, and others focused on the role of Id1 overexpression in tumor progression and inflammation through NF-κB and ANXA1 interactions⁹. Considering that the overexpression of Id1 plays a central role in regulating multiple processes that influence cancer cell transformation and survival, Id1 is a compelling target to include in diagnostic tests to better inform upon prognostic factors such as the increased risk of precancerous progression, the increased risk of aggressive tumor characteristics, and the decreased odds of survival.

The novelty of our approach centers around the combination of Id1, an established marker of cancerous transformation, and SEAP, a readily detectable secreted protein, in a plasmid reporter system to indicate the presence of cervical cancer cells. Stieh et al. reported the use of an engineered human immunodeficiency virus (HIV) dual-reporter system encoding luciferase and mCherry genes to monitor major areas of infection and viral dissemination in the female reproductive tract, and while these reporters work in a similar way to our SEAP-encoding plasmid, the controlling promoter (CMV) in their study was not specific to cancer cells¹⁸. Orrù et al. described a similar technique for using a reporter plasmid in cervical cancer cells to identify regulators of HPV16 late gene expression¹⁹; however, their model was not intended to indicate malignant transformation in cells through reporter expression but rather indicated genetic regulators. Shiku et al. also explored a similar system in HeLa cells with SEAP as the main reporter protein expressed from plasmids with various responsive elements to control reporter production, although their focus was in developing a more sensitive whole-cell electrochemical assay system rather than using the SEAP-encoding plasmids to specifically target oncogenic gene expression in cervical cancer cells²⁰. To date, no studies have proposed the use of a reporter plasmid system as a potential diagnostic tool in cervical cancer and have only used their reporter systems as experimental tools to study specific gene function. However, the successful application of reporter plasmids in vitro in these previous studies, along with the data presented in our study, support the idea that a reporter plasmid system could successfully produce detectable levels of protein secreted from transfected cervical cancer cells, and in the case of our pId1-SEAP system, do so in a manner proportional to expression of the cancer-specific target gene, Id1.

The present study confirmed Id1 overexpression in cervical cancer models and in human cervical cancer tissues using Western Blot and immunohistochemistry, both compared to normal, non-cancer conditions. All three cervical cancer cell lines (HeLa, SiHa, and CaSki) expressed increased Id1 compared to 3T3 fibroblasts and MDA-MB-231 breast cancer cells (Fig. 1). Id1 overexpression compared to 3T3 cells was expected based on the non-cancer nature of the cell line. This is corroborated by early-stage cervical cancer tissues exhibiting increased Id1 expression compared to normal tissue (Fig. 1), as Id1 has been shown to increase with disease progression from normal to precancer to invasive carcinoma¹⁰. MDA-MB-231 cells were previously shown to express Id1 but were not compared to a non-cancer control nor to a cervical cancer model¹⁴, so the difference in Id1 expression between breast and cervical cancer cells was unknown. HPV infection in cervical cancer cells could contribute to increased Id1 levels in these cervical cancer cell lines compared to MDA-MB-231 cells^9,21–23.

pId1-SEAP successfully transfected HeLa, SiHa, and CaSki cervical cancer cells and resulted in increased SEAP expression compared to pId1-SEAP-transfected 3T3 fibroblasts. Id1 and SEAP expression were shown to be positively correlated with pId1-SEAP transfection in all but one of the cell lines. As pId1-SEAP is non-replicative, we sought to investigate how long cells could produce SEAP and SEAP levels peak after a single plasmid exposure. Time-course SEAP production characterization showed that cervical cancer cells produced sustained SEAP expression through 72 h following a 12-hour exposure to pId1-SEAP. While all cervical cancer cell lines used in the study successfully produced SEAP following transfection with pId1-SEAP, HeLa had higher output levels of SEAP compared to the other two cell lines (Fig. 2). Even with normalization for transfection efficiency using pCMV-SEAP, HeLa consistently outperformed SiHa and CaSki cells in SEAP production post-transfection. This increase in SEAP production in HeLa cells is likely due to the cell line’s high transfection efficiency and tolerance paired with the previously demonstrated relationship between increased Id1 and HPV E6/E7 expression²³. While all three cervical cell lines used in our studies are either HPV16 or HPV18 positive, increased HPV E6 genome copy number and increased mRNA expression has been shown in HeLa cells compared to SiHa and CaSki cell lines²⁴. This could potentially explain the difference in SEAP expression capacity between each cell line due to differential HPV types, genome copy numbers, and relative HPV E6/E7 expression influencing the interaction with endogenous Id1 expression and thereby influencing SEAP production from pId1-SEAP controlled by the same Id1 regulatory sequences. Future studies using a larger panel of HPV-positive and HPV-negative (e.g., C33A) cervical cancer cell models would be needed to adequately address this question.

By modeling a condition where not all cells were transfected with pId1-SEAP, as few as 12,500 cervical cancer cells (HeLa) exposed to the plasmid were needed to produce a significantly increased SEAP signal compared to background SEAP production. While a higher proportion of plasmid-exposed SiHa and CaSki cells was required to produce a significantly increased SEAP signal over background, the transfected proportion required for significant SEAP production remained less than 25% of the total population for these cell lines. Positive SEAP production capacity in this in vitro mixture of pId1-SEAP-exposed and naïve cells can inform potential performance in heterogeneous tumor models where some cells would take up the plasmid and other cells would not. For direct examination using Pap cytology, at least 5,000 squamous cells are required for adequate visualization in a liquid-based preparation, and the number increases to 8,000–12,000 cells for a conventional preparation²⁵, although only a few abnormal cells within the sample are required to determine a positive Pap test. Our finding that only 12,500 pId1-SEAP-transfected malignant cells produce a positive SEAP signal merits further work to evaluate the sensitivity of this approach to detect cervical cancer.

In agreement with the cell-based data, the present studies also showed that pId1-SEAP can transfect 3-D tumor tissue ex vivo and produce positive SEAP levels compared to normal muscle tissue exposed to the plasmid. It should be noted that the amount of tumor tissue was much less than the amount of muscle tissue in this study (Fig. 4). Translation of the plasmid approach to an in vivo anatomical environment would face a similar scenario, where most tissue on the cervix would be non-cancerous relative to a proportionately small area containing cervical cancer; thus, it would be important to consider the amount of background SEAP production in normal tissues surrounding the cervical lesion. The significantly greater SEAP production from tumor tissue relative to normal tissue observed in our ex vivo study is notable considering rapid loss of tissue viability immediately following resection, resulting in reduced SEAP production capacity of the tissues. Relative to our ex vivo experiment, in vivo tissues are anticipated to maintain higher capacity to produce the SEAP reporter, resulting in potentially greater differences in SEAP production from pId1-SEAP-exposed tumor tissue compared to normal tissue. The collective results from these experiments demonstrate the feasibility of this pId1-SEAP system in transfecting human cervical cancer cells and producing the SEAP reporter in the presence of an overexpressed Id1 promoter.

These studies represent the initial exploration of pId1-SEAP function in vitro in the context of preclinical cervical cancer models and do not compare pId1-SEAP to available cervical cancer diagnostic methods such as Pap cytology. Our studies did not evaluate cumulative SEAP production over time due to the replacement of media at 24 h post-resection, nor did we evaluate any in vivo models. Exploring SEAP production rates and accumulation over time would be especially important during in vivo studies using patient derived xenografts that more closely reflect the heterogeneity of human disease compared to in vitro cell models. Future directions include preclinical in vitro and in vivo characterization of pId1-SEAP function in normal human keratinocytes and additional cervical cancer models with differing HPV infection status (e.g., HPV30-positive HT-3, HPV-negative C33A). Additional worthwhile studies include using pId1-SEAP to detect precancer cervical disease in animal models and preclinical comparisons of sensitivity and specificity of pId1-SEAP to established screening methods for detecting cervical cancer. These studies would form the basis for potential future translation of pId1-SEAP into a self-applied system for topical transfection of the cervix. Following transfection, malignant or pre-malignant cells with upregulation of Id1 would release SEAP into the cervical fluid for detection as part of a novel cervical cancer screening application. Id1’s relationship with HPV infection (a hallmark risk factor for cervical cancer) and its implication in cervical tumor initiation, progression, and metastasis^9,10 warrant continued investigations of the clinical value of Id1 as a diagnostic and prognostic marker for cervical cancer. Continued development of an Id1-SEAP transfection and detection system would allow the biological characteristics of Id1 to be harnessed for diagnostic tests that could probe internal cellular mechanisms without requiring removal of cells from the cervix for external analysis.

Methods

Cell culture

Plasmid transfection and function was characterized in three cervical cancer cell lines: HeLa, CaSki (both provided by Dr. Nilam Sanjib Banerjee), and SiHa (American Type Culture Collection, Manassas, VA, USA). Two cell lines were used as controls: MDA-MB-231 as a positive control due to its previous characterization of Id1 expression and plasmid transfection¹⁴ and 3T3 fibroblasts as negative controls due to their null expression of Id1. HeLa, SiHa, and 3T3 cells were maintained in DMEM (10-013-CV, Corning Inc., Corning, NY, USA) with 10% fetal bovine serum (FBS) and Plasmocin prophylactic (InvivoGen, San Diego, CA, USA). CaSki cells were maintained in RPMI-1640 (10-040-CV, Corning Inc., Corning, NY, USA) with 10% FBS and Plasmocin prophylactic. All cells were cultured at 37 °C and 5% CO₂. Cells were allowed to grow to 70–90% confluency before passaging, and no passages were carried beyond passage 25.

Western blot

Protein lysates were generated from cells with RIPA buffer containing Triton X-100 and Pierce Protease Inhibitor Tablets (Thermo Fisher Scientific, Waltham, MA, USA). Lysate concentration was measured using Pierce Rapid Gold BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) and 20 µg of protein was size-separated using NuPage 4–12% Bis-Tris electrophoresis gels (Invitrogen, Carlsbad, CA, USA) before transfer to PVDF membranes (MilliporeSigma, Burlington, MA, USA). Membranes were first blocked with 5% milk in Tris buffered saline with 0.5% Tween 20 (TBST) and probed overnight with rabbit monoclonal anti-human Id1 (F2M1J) (Cell Signaling Technology, Danvers, MA, USA), followed by horseradish peroxidase (HRP)-conjugated mouse anti-rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA, USA). In between antibodies and before detection, membranes were washed with TBST 3–5 times for a total of 15 min. Id1 protein was detected with chemiluminescent substrate (WesternBright ECL, Advansta Inc, San Jose, CA, USA). Beta-actin was probed on all membranes as a loading control using HRP-conjugated mouse monoclonal anti-Beta-actin (Santa Cruz Biotechnology, Santa Cruz, CA, USA). All samples were run as biological triplicates. Densitometry was performed using ImageJ software²⁶ as described previously¹⁵, and Id1 expression for each cell line was normalized to Id1 expression from 3T3 cells.

Immunohistochemistry

A slide with a formalin-fixed paraffin-embedded (FFPE) human cervical cancer tissue microarray (CR1101, TissueArray.com LLC, Derwood, MD, USA) was deparaffinized and antigen was retrieved by heating for 10 min at 90 °C in 1X citrate buffer pH 6. Tissues were blocked with 5% bovine serum albumin (BSA) in TBST at room temperature, incubated overnight at 4 °C with mouse monoclonal anti-Id1 (B-8) (Santa Cruz Biotechnology, Santa Cruz, California), and then incubated for 1 h at room temperature with HRP-conjugated anti-mouse IgG. Tissues were washed with TBST in between antibodies and before visualization. Diaminobenzidine (DAB) was used as the chromogen to visualize Id1 and tissues were counterstained with hematoxylin. Tissues were then imaged on an EVOS m7000 microscope (Thermo Fisher Scientific, Waltham, MA, USA) at 20X magnification using brightfield, and positive Id1 staining was quantified using QuPath open-source software V.0.5.1²⁷.

In vitro transfections of cell lines

For cell transfection experiments, cells (2.0E5–5.0E5) were plated in triplicate 24 h prior to transfection with various plasmids using jetPEI (linear polyethylenimine derivative) (Sartorius Polyplus, Illkirch-Graffenstaden, France). Plasmids used included: pSEAP2-basic (Takara Bio USA, San Jose, CA, USA) as a negative control as it lacks eukaryotic promoter and enhancer sequences for SEAP expression; pCMV-SEAP as a positive control with the use of the ubiquitous cytomegalovirus (CMV) promoter; and pId1-SEAP as the experimental cancer-specific plasmid. pCMV-SEAP and pId1-SEAP were obtained from stocks generated as previously described¹⁴. Transfection was performed according to the forward protocol from Polyplus by mixing 1 µg plasmid DNA and jetPEI solutions at varying ratios before applying to cells in phenol-free complete media. Cells were allowed to incubate with the plasmid solution in the media for at least 12 h before the media was replaced to avoid accumulation of SEAP in unchanged media. While each experiment had a different protocol for media sampling times, every sample was stored at -20 °C until it could be measured for SEAP using the Great EscAPe Fluorescent SEAP Assay (Takara Bio USA, San Jose, CA, USA) in 96-well plate format. When appropriate, SEAP counts from pSEAP2-basic and pId1-SEAP conditions were first background subtracted using SEAP counts from non-transfected cells, and values were further normalized using division by corresponding SEAP counts from pCMV-SEAP. For one experiment, media was sampled every 12 h over a 72-hour period and rates of SEAP production were calculated by dividing the SEAP counts from a 12-hour period by 12 to represent the amount of SEAP produced per hour. The rates of SEAP production were calculated as a sum total for statistical comparison by summing the rates for each cell line over 72 h. When naïve and plasmid-exposed cells were co-cultured, the total number of plasmid-exposed cells was calculated by multiplying the dilution percentage by the total number of cells plated in one well.

Viability Assay by ATP Concentration

Intracellular ATP concentration was used as a measure of cellular viability at 48 h post-transfection with pCMV-SEAP in varying plasmid to jetPEI ratios. ATPlite luminescence assay (Revvity, Waltham, MA, USA) was used to measure intracellular ATP concentration by first lysing cells in ATP-stabilizing buffer and then performing the assay procedure according to standard kit protocol in a 96-well plate format. Standard ATP solutions were made and run in triplicate alongside experimental samples for creation of a standard curve to relate luminescence counts to ATP concentration.

In vivo tumor growth and ex vivo transfection of tumor tissue

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Alabama at Birmingham and performed in compliance with guidelines from the Public Health Service Policy and Animal Welfare Act of the United States. These methods and related results are reported in accordance with relevant ARRIVE guidelines. Ten female athymic nude mice (4–6 weeks old, Charles River Laboratories, Charleston, SC, USA) were injected subcutaneously on both flanks with 5.0 × 10⁶ HeLa or 1.0 × 10⁷ SiHa cells (five mice per cell line). Tumors were allowed to grow for 5–6 weeks until tumor size was approximately 1.5 cm in the greatest dimension (length). Mice were anesthetized by inhalation of 5% isoflurane mixed with oxygen gas; fully anesthetized mice were euthanized by cervical dislocation. Tumors were then harvested under sterile conditions and surgical technique. Tumor cores were punched using sterile biopsy tools and then delivered onto a sterile surface for horizontal bisection to standardize tissue volume. Bisected cores were each treated as one piece of tumor material and were added to prepared wells of a sterile culture plate containing various plasmid transfection solutions in phenol-free media (DMEM with 10% FBS). Transfection solutions included pId1-SEAP at either 25 µg/µL or 75 µg/µL mixed with in vivo JetPEI (linear polyethylenimine derivative at higher concentrations for tissue penetration) (Sartorius Polyplus, Illkirch-Graffenstaden, France) according to recommended ratio of 7 µL PEI for each 1 µg of plasmid DNA. Remaining tumor tissue that could not be punched further was fixed in formalin and processed for embedding in paraffin. Muscle tissue from each mouse was collected after tumors were harvested and placed into control wells prepared with pId1-SEAP transfection solution. Tissues were incubated at 37 °C and 5% CO₂. Wells with uncontaminated tissue media were sampled at 24- and 48-hours post-transfection, with media replacement at 24 h. Tissue media samples were stored at -20 °C until later analysis for SEAP using the Great EscAPe Fluorescent SEAP Assay (Takara Bio USA, San Jose, CA, USA) in 96-well plate format. Formalin-fixed paraffin-embedded tissues were sliced, mounted, and stained with hematoxylin and eosin by the Comparative Pathology Laboratory at the University of Alabama at Birmingham; unstained slides were also obtained and stained for Id1 as previously described in the Immunohistochemistry methods section.

Calculations and statistical analysis

Results are reported as the mean plus or minus the standard deviation from the mean of biological triplicates. Data were compared and tested for statistical significance using one-way or two-way analysis of variance (ANOVA) with either Bonferroni’s, Šídák’s, Tukey’s, or Dunnett’s multiple comparison test as appropriate (Prism GraphPad software, version 10.0). Pearson’s test was used for correlation analysis and linear regression was used to model the relationship graphically. P < 0.05 was considered significant.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

A.V.E., B.B.K., Y.H., and J.M.W. conceptualized and outlined the manuscript in its earliest stages. All experiments were performed primarily by A.V.E. and Y.H., with support from A.M.W., S.J., and K.A.T. E.E.H. contributed to animal experiment design and institutional animal resource program protocol creation and management. A.M.W. helped design and perform animal experiments, along with additional support from S.J. and K.A.T. A.V.E., B.B.K., and J.M.W. drafted the manuscript, and obtained initial inputs on content from K.T., R.C.A., and A.G.K. as clinical specialists. A.V.E. created all figures, with further refinement by A.M.W., S.J., K.A.T., B.B.K., and J.M.W. All authors reviewed the manuscript and made edits for quality and content.

Data availability

All data generated and analyzed during this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.