Remember me
Research ArticleImmunologyInflammationOncology
Open Access | 
10.1172/jci.insight.190244
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Tassinari, V. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Kaciulis, M. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Petrai, S. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Stabile, H. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Pernazza, A. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Leopizzi, M. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Di Maio, V. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Belleudi, F. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Ranieri, D. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Mancini, V. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Palaia, I. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Tanzi, F. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Lospinoso Severini, L. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Ruggeri, S. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Greco, M. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Bernardini, G. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Zingoni, A. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Cippitelli, M. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Cerboni, C. in: PubMed | Google Scholar
1Department of Molecular Medicine,
2Department of Medical-Surgical Sciences and Biotechnologies, and
3Department of Clinical and Molecular Medicine, Sapienza University of Rome, Rome, Italy.
4Department of Life Sciences, Health and Professions, Link Campus University, Rome, Italy.
5Department of Maternal and Child Health and Urological Sciences and
6Pasteur Institute–Cenci Bolognetti Foundation, Sapienza University of Rome, Rome, Italy.
Address correspondence to: Cristina Cerboni or Alessandra Soriani, Department of Molecular Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 Rome, Italy. Phone: 39.0649255151; Email: cristina.cerboni@uniroma1.it (CC); alessandra.soriani@uniroma1.it (AS).
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Find articles by Soriani, A. in: PubMed | Google Scholar
Authorship note: VT, MK, CC, and AS contributed equally to this work. VT is deceased.
Published July 8, 2025 - More info
Published in Volume 10, Issue 13 on July 8, 2025
AbstractADAR1 edits double-stranded RNAs (dsRNAs) by deaminating adenosines into inosines, preventing aberrant activation of innate immunity by endogenous dsRNAs, which may resemble viral structures. Several tumors exploit ADAR1 to evade immune surveillance; indeed, its deletion reduces tumor viability and reshapes infiltrating leukocytes. Here we investigated the role of ADAR1 in immune evasion mechanisms during cervical cancer (CC) progression. Patients’ biopsy samples showed higher ADAR1 expression already in premalignant lesions (squamous intraepithelial lesions [SIL]) and a substantially reduced percentage of infiltrating CD7+ innate cells in in situ and invasive carcinomas compared with normal mucosa, with CD56+ NK cells showing phenotypic alterations that may have affected their functional responses. In CC-derived cell lines (SiHa, CaSki), ADAR1 silencing reduced cell proliferation, an effect further enhanced by exogenous IFN-β administration. It also induced proinflammatory gene expression, as demonstrated by RNA-Seq analysis, and conditioned supernatants collected from these cells activated several NK cell effector functions. NK cell infiltration and activation were also confirmed in organotypic 3D tissue models of SiHa cells knocked out for ADAR1. In conclusion, ADAR1 expression increased with CC progression and was accompanied by alterations in tumor-infiltrating NK cells, but its silencing in CC-derived cell lines potentiated antitumor NK cell activities. Thus, ADAR1 inhibition may represent a therapeutic perspective for CC and possibly other malignancies.
IntroductionCervical cancer (CC) is the fourth leading cause of cancer-related mortality in women (1). It takes years or decades to develop, and it is distinguished by a spontaneous continuous progression, starting with a persistent intraepithelial human papillomavirus (HPV) infection in almost all cases, evolving to squamous intraepithelial lesions (SILs) and then into invading tumors and metastasis (2). Although prophylactic vaccines are available, a large portion of the population remains unvaccinated, and the vaccine does not prevent cancer development in individuals already exposed to the virus (3).
From an immunological perspective, CC is classified as an immune-infiltrated yet immunosuppressive malignancy. Despite the presence of an antitumor immune response, its effectiveness is often compromised during disease progression, primarily due to HPV-mediated modulation of the tumor microenvironment (TME) (3, 4).
Indeed, effective evasion of immune recognition seems to be the hallmark of HPV infections already at earlier stages, as the virus is almost invisible to the immune system due to an exclusively intraepithelial infectious cycle — with no viremic phase and no virus-induced cell death — and viral replication and release are not associated with inflammation. In addition, HPV downregulates innate immune signaling pathways; proinflammatory cytokines, particularly type I interferons (IFN-Is), are not released and the signals for antigen-presenting cell activation and recruitment are either not present or inadequate. This immune ignorance results in chronic infections that persist over weeks and months. Progression to high-grade SIL (HSIL) is associated with further deregulation of immunologically relevant molecules — particularly chemotactic chemokines and their receptors — on keratinocytes and endothelial cells of the underlying microvasculature, limiting or preventing infiltration of cytotoxic effectors into the lesion (5, 6).
NK cells are innate lymphocytes with a critical cytotoxic and immunoregulatory role, and although they are disseminated in the uterine mucosa, their role in the natural history of HPV infection and HPV-driven tumorigenesis is not entirely clear. It was reported that they emerge at an early stage in HPV-infected lesions and NK cells are present at higher level in premalignant lesions with a lower viral load (7). In studies of individuals with confirmed SILs of different grades, lesion regression strongly correlated with early infiltration of intraepithelial effector cytotoxic cells (granzymeB+CD8+ [GzmB+CD8+] and GzmB+CD56+) (8, 9). More recent studies applying single-cell multi-omics technologies provided new in-depth maps of the complex CC ecosystem (10–14). One common characteristic was the observation of an increased NK cell infiltrate in the tumor area, often (but not always) accompanied by an enriched cytotoxicity signature (e.g., GZMB, GZMH, PRF1), higher expression of genes involved in migration and adhesion, and lower levels of inhibitory molecules (e.g., TIGIT, CTLA-4). Moreover, expression of some of these genes was associated with a better prognosis (11–13), consistent with the observation that patients with CC with a high level of intratumoral NK cells had a decreased risk of progression (10). Interestingly, in one of these studies, the heterogeneity of malignant cells was investigated in relation to TME. Among the different cellular states uncovered, one was characterized by a bidirectional tumor stroma–immune system interaction involving NK and T cells through IFN signaling (14). However, to the best of our knowledge, beyond such transcriptomic approaches, other studies analyzing the phenotype of infiltrating NK cell subsets or of other innate lymphoid cells (ILCs) isolated from patients’ biopsy samples are lacking.
Over the last 15 years, several populations of ILCs have been described and classified into 5 groups (NK cells, ILC1, ILC2, ILC3, and LTi) according to their transcription factors (TFs) and cytokine production profile (15). In particular, NK cells are historically identified as CD56+CD16+/–CD127–EOMES+ (13). Although ILCs are now emerging as important players in numerous tumor types, their role has not been thoroughly investigated in either intraepithelial lesions or invasive carcinomas (ICs) of the cervix.
Antitumor responses are also dependent on IFNs, and loss of IFN signaling results in resistance to immune checkpoint therapies (16–19). Expression of IFN-I can be induced by long, fully base-paired dsRNAs deriving from viruses, but also by certain endogenous self RNAs that could aberrantly stimulate innate immune responses (20, 21). To prevent an erroneous activation of cytosolic sensors by self dsRNAs, ADAR1 — a member of the adenosine deaminase acting on RNA (ADAR) family of enzymes — edits such dsRNAs (20–22) by deaminating adenosines to inosines (A-to-I conversion), which are subsequently read out as guanosines, leading to transcriptomic and proteomic changes (20). Indeed, in many types of tumors, a substantial amount of mutational load is due to RNA editing/hyperediting by ADAR1 (21, 23). Moreover, recent studies demonstrate that ADAR1 deletion reduces cancer cell viability via IFN-I pathway activation (24, 25), and in ADAR1-null tumors, there is a global reshaping of immune cell profiles, suggesting that inflammation caused by ADAR1 deletion can bypass the loss of tumor-specific CD8+ T cells (19, 26). This evidence is in line with the fact that ADAR1 is considered a “master regulator” of cytoplasmic innate immunity, as it prevents autoimmunity (22). Indeed, in humans, loss-of-function mutations of ADAR1 can confer severe interferonopathies and autoimmune diseases (27).
There are few studies on the role of ADAR1 in HPV-driven tumorigenesis. An increase in expression was associated with CC progression (28, 29), and its ablation correlated with a proinflammatory phenotype (30), while a particular ADAR1 haplotype was related to recurrent dysplasia in patients coinfected with HPV/HIV (31). However, whether ADAR1 affects NK/ILC effector functions is not known, and to our knowledge, the interplay among ADAR1, IFN-I, and NK cells has not been previously addressed in any tumor model, including CC. Thus, in our study we investigated whether ADAR1 plays a role in CC tumorigenesis via its ability to dampen IFN-I signaling and production, with potential effects on NK/ILC-mediated innate immune responses as well.
ResultsADAR1 expression correlates with disease progression in CC. To determine the importance of ADAR1 in the progression of CC, we first asked whether ADAR1 expression can represent a prognostic factor for patients’ survival. TCGA data revealed that high ADAR1 levels were predictive of poor overall patient survival (Figure 1A), and R2: Genomics Analysis and Visualization Platform data also showed a significant progressive increase in ADAR1 expression from normal tissues to SIL to IC (Figure 1B). These initial findings were confirmed by real-time PCR (RT-PCR) performed on mRNA extracted from formalin-fixed, paraffin-embedded tissues derived from normal mucosa, low-grade SIL (LSIL), or IC biopsies. Indeed, ADAR1 expression increased markedly during disease progression (Figure 1C). Additional analysis of TCGA data for ADAR1 expression in ICs at different stages of the disease revealed a tendency toward increased ADAR1 expression from stage I to stage IV, which was associated with a significant decrease in interferon-stimulated gene (ISG) expression in more advanced tumors (Figure 1D and Supplemental Table 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.190244DS1).
Figure 1ADAR1 overexpression characterizes CC progression. (A) Kaplan-Meier survival plot for CC patients (n = 292) stratified by low (red line) and high (blue line) ADAR1 expression (cutoff mode: scan). Data were obtained from TCGA. For survival analysis, statistical significance was assessed with the log-rank test. (B) ADAR1 expression was assayed on single HG_U133A arrays. Gene Expression Omnibus database (GEO GSE7803). (C) Total RNA was extracted from paraffin-embedded biopsy samples, and ADAR1 expression was analyzed by RT-PCR. (D) ADAR expression (left panel) and ISG core score (right panel) at different clinical stages in the TCGA data. Between-group P values were computed using Wilcoxon’s rank-sum test. bonf, Bonferroni post hoc test; f.i, fold increase; Pt. patient Ctr, normal mucosa; IC, invasive CC.
Expression of ADAR1 in CC progression was further investigated by IHC on a panel of paraffin-embedded tissue samples (Figure 2). While ADAR1 was almost undetectable in normal mucosa (Figure 2A, a and b), its expression increased noticeably during disease progression, with clear positivity already in the LSIL (Figure 2A, panels c and d). Thus, we compared ADAR1 expression with the proliferative index pattern of Ki-67, which is routinely used to aid diagnosis in morphologically difficult cases, for example between LSIL and reactive or metaplastic lesions (32). In certain cases, we found similar expression patterns of ADAR1 and Ki-67, with basal and parabasal layer positivity in LSILs, which indicated that ADAR1 positivity was related to an increase in proliferative index, although it was limited to the lower epithelial layers (Figure 2B). ADAR1 expression further increased during disease progression (Figure 2A, panels e–j) and extended to the upper layer of squamous epithelium, and Ki-67 showed a similar distribution pattern, alongside p16, a surrogate marker of HPV infection and high-grade dysplasia (data not shown).
Figure 2Expression of ADAR1 in normal cervical mucosa, premalignant lesions, and CCs. (A) Expression of ADAR1 on normal mucosa (immunostains; original magnification, 5× and 20×) (a and b), premalignant lesions (LSIL) (c and d), and invasive CCs (e–j) (immunostains; original magnification, 8× top row; 20× bottom row) was analyzed by IHC. (B) Representative images of IHC staining of ADAR1 compared with Ki-67 expression on LSIL lesions (immunostains; original magnification, 8× and 10×). (C) Quantification (%) of cytoplasmic (cyt) and nuclear (nuc) ADAR1 expression in the different lesions (17 LSIL, 10 HSIL/CIS, 41 IC). (D) Quantification (%) of total ADAR1 in G2–G3 IC stained sections. P values were calculated by ANOVA. (E) Expression of total ADAR1+ cells in G2 and G3 lesions with low numbers of infiltrating CD56+ cells (CD56+ <5). The P value was calculated by ANOVA. *P < 0.05, **P < 0.01.
We also investigated whether the increasing expression of ADAR1 was accompanied by changes in its staining pattern, since ADAR1 is localized in the nucleus and/or the cytoplasm. We observed that in HSIL/in situ carcinomas (HSIL/CISs) and in ICs, ADAR1 showed cytoplasmic positivity significantly different from that in LSILs (Figure 2C). Moreover, there was a decrease in nuclear positivity, particularly between LSIL and CIS. Within the same group of lesions, the increase in cytoplasmic versus nuclear expression was statistically significant in CIS and IC, but not in LSIL (Figure 2C, P = 0.004 in CIS; P = 0.017 in IC). ADAR1 expression was even more evident in high-grade (grade 3 [G3]) compared with G2 CC (Figure 2D), although our consideration was semiquantitative.
Considering our interest in tumor-infiltrating innate lymphocytes, in more aggressive tumors we also examined the presence of cells expressing CD56, a well-known NK cell marker. Separating tumors on the basis of a low (CD56+ <5) or not low (CD56+ >5) number of positive cells, we observed a significantly higher percentage of ADAR1+ cells in G3 tumors in the group with low CD56+ cells (Figure 2E). Together, these results showed increasing expression of ADAR1 during disease progression from LSIL to HSIL/CIS and IC. Moreover, in samples from the latter group of patients, ADAR1 expression was even higher in G3 compared with G2 tumors, with a significant difference also maintained when lesions with low CD56+ cells were analyzed.
Decreased levels of tumor-infiltrating ILCs are observed in CC. To further investigate immune cells infiltrating cervical lesions, we analyzed the proportion of ILCs isolated from fresh biopsy samples obtained from different groups of patients (Figure 3). In situ and invasive carcinomas were grouped together and compared with LSILs, as well as with normal mucosa used as control. Total leukocytes were identified as CD45+ cells, and within them innate lymphocytes were gated as CD7+ and negative for T cell (CD3, CD4, CD5), B cell (CD19), and monocyte (CD14) markers. Although an increased percentage of infiltrating CD45+ cells characterized both LSIL (44%) and CIS/IC (37%) compared with normal mucosa (32%), the frequency of innate CD7+ lymphocytes was significantly decreased in CIS/IC (from ~8% in normal mucosa, to ~7% in LSIL, and to ~3% in CIS/IC) (Figure 3, B and C). We further characterized CD7+ cells to discriminate between NK, ILC1, ILC2, and ILC3 cell populations. In humans, the distinction between NK and ILC1 cells can be challenging, but it is widely accepted that tissue ILC1, ILC2, and ILC3 cells can be identified by the expression of IL-7Ra (CD127) in combination with group-specific TFs: EOMES, GATA3, and RORγt, respectively. Analysis of CD7+ infiltrating lymphocytes showed a very low percentage (0%–5%) of CD127+ cells regardless of the biopsy sample analyzed and undetectable GATA3 and RORγt expression, thus ruling out the presence of ILC2 and ILC3 populations. On the other hand, approximately 60% of CD7+ cells coexpressed the NK marker CD56 and the NK TF EOMES, leading to their identification as NK cells, although no marked differences were detected between control mucosa, LSIL, and CIS/IC samples (Figure 3, A and D, and data not shown). However, we noticed a significantly lower percentage when more aggressive (G3) ICs were analyzed (~60% in G3 versus ~80% in G2) (Figure 3E).
Figure 3Analysis of innate immune cells infiltrating LSIL, HSIL/CIS, and IC. Fresh biopsy samples obtained from different groups of patients were analyzed for the presence of an innate immune cell infiltrate. (A) Gating strategy from representative FACS plots, showing the isolation of (left to right panel) live cells, CD45+ cells (gate excludes monocytes), and Lin–CD7+ ILCs (middle right panel), further divided according to CD56, CD127, EOMES, and CD103 expression. ILCs were thus defined as Lin–CD45+CD7+ and then separated into NK/ILC1, ILC2, and ILC3 subpopulations (see the main text for more details). (B and C) Bar graphs represent the percentage of CD45+ (B) and CD7+ (C) cells among each group of patients. HSIL/CIS and IC biopsy samples were grouped together (CIS/IC, red triangles) and compared with LSIL or normal mucosa used as control. (D and E) Bar graphs represent the percentage of CD56+ cells among the Lin–CD45+CD7+cells in the different groups of patients (D) and in G2/G3 tumors (E). (F) Bar graphs represent the percentage of distinct subsets expressing or not CD56 and CD103. Histograms represent mean ± SEM. *P < 0.05, **P < 0.01. Two-way ANOVA was used for multiple comparisons. Each symbol represents a single biopsy sample.
NK cells can recirculate from blood to tissues, where they can be quickly recruited during viral infection or tumor growth, and, as with other lymphocytes, their tissue retention is facilitated by CD69, CD103 (αE integrin), and CD49a expression. Therefore, to investigate the possible tissue-resident nature of infiltrating NK cells, we further analyzed CD7+ cells for expression of CD103 in combination with CD56 (Figure 3F). There was a general rise in CD103+ cells in CIS/IC biopsy samples, with a statistically significant difference reached in the CD56– subset (with a 3-fold increase compared with normal mucosa or LSIL). This increase was accompanied by a statistically significant decrease in the CD103–CD56– subset (from ~55% in LSIL to ~20% in CIS/IC). These cells were, however, NK cells, as they maintained expression of the typical NK markers EOMES and CD94.
We further broadened the phenotypic characterization of the CD56/CD103 subset and assessed expression of some activating/inhibitory receptors (CD16, NKG2D, NKp46, NKp44, NKp30, KIR, CD94, Tigit) and adhesion (CD49a) and cytotoxic molecules (GzmB and GzmK). The major modulations were observed in the CD56+ cell subsets, where an increase in the expression of CD49a, CD94 and NKp44 alongside a decrease in the activating receptor CD16 and in GzmB expression characterized CIS/IC-infiltrating cells (Figure 4).
Comments (0)