Immune phenotype in high- versus low-NRF2 high grade serous ovarian cancer and the impact on prognosis

Immune phenotype in high- versus low-NRF2 high grade serous ovarian cancer and the impact on prognosis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Immune phenotype in high- versus low-NRF2 high grade serous ovarian cancer and the impact on prognosis Samera Hamad, Chelsea Katz, Helen Toma, Kosuke Murakami, Nasrine Bendjilali, and 12 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6925656/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 28 Apr, 2026 Read the published version in Genes & Immunity → Version 1 posted 10 You are reading this latest preprint version Abstract NRF2 modulates tumor immune microenvironment (IMM) in several cancers. NRF2 is activated in about 50% of high grade serous ovarian cancer (HGSOC), the most aggressive type of ovarian cancer. This study aimed to stratify HGSOC patients’ samples by NRF2 levels and identify its impact on immune phenotype and prognosis. We analyzed data from n = 7 scRNA-seq, n = 365 RNA-seq of human HGSOC samples, and n = 240 HGSOC samples from a tumor microarray (TMA). Results showed human HGSOC samples can be classified by NRF2 High and NRF2 Low tumors. RNA-seq data analysis along with IHC labeling showed that NRF2 High HGSOCs are enriched with hallmarks of immune suppressive markers (ISMs). Specifically, NRF2 High tumors are identified as tumors associated macrophages (TAMs) with worst survival (p = 0.038) was observed in CD68 High tumors. NRF2 Low tumors were enriched with immune activated markers such as CD3E and CD80 with a prognostic significance. Immune checkpoints (ICs) are important in both groups. However, their levels and spatial distribution are the factors that define their impact on prognosis in these samples. This study is the first that shows classification of HGSOC based on NRF2 levels and suggests IHC-labeling and genomic evaluation of NRF2 and immune markers in HGSOC to predict prognosis. Biological sciences/Immunology/Gene regulation in immune cells Biological sciences/Genetics/Gene regulation NRF2 pathway high grade serous ovarian cancer immune markers survival Figures Figure 1 Figure 2 Figure 3 Introduction Ovarian cancer is a leading global cause of cancer-related deaths in women, with an estimated 12,740 deaths in the United States in 2024 [ 1 ]. The overall 5-year relative survival for patients diagnosed at early stage is ∼85%, which drops to 31% with distant metastatic disease [ 1 ]. Several histological subtypes of epithelial ovarian cancer have been identified such as low or high grade serous ovarian cancer (LGSOC or HGSOC ), clear cell ovarian cancer (CCOC), endometrioid (EOC), mucinous ovarian cancer (MOC), and borderline ovarian cancer (also called tumors of low malignant potential, LMP, because they do not invade the basement membrane) [ 2 ]. HGSOC is the most common and aggressive type, accounting for ∼70–80% of ovarian cancer deaths, with most cases diagnosed at advanced stages [ 3 ]. The 5-year survival rate of patients with HGSOC is only 34% [ 4 ], with relapse being the main factor responsible for such a high mortality rate [ 5 ]. HGSOC is still treated as a single disease with a combination of surgery plus systemic platinum-based chemotherapy [ 6 ]. HGSOC is comprised of several histological architectures, including solid, glandular, papillary and or pseudo-endometroid architecture [ 7 ]. Further, The Cancer Genome Atlas (TCGA) and other studies have identified four molecular subtypes of HGSOC, these include immunoreactive, differentiated, proliferative, and mesenchymal [ 8 ], which further complicates efforts to find therapies for this deadly cancer. These subtypes are associated with several distinct clinical and pathologic characteristics [ 8 , 9 ], potentially reflecting differences in immune cell [ 10 ] and stromal content [ 11 ], which could impact their responses to treatment [ 12 ]. The NRF2 (also known as nuclear factor erythroid-derived 2-like 2; NFE2L2) signaling pathway is known to provide an important cellular defense mechanism against oxidative and metabolic stress [ 13 ]. In addition to maintaining redox homeostasis, NRF2 pathway also regulates many other cancer hallmarks including proliferation, differentiation, apoptosis, inflammation and metastasis [ 14 – 16 ]. NRF2 is also known as an anti-inflammatory pathway; its activation reduces the expression of several inflammatory genes and pathways [ 17 ], controls PD-L1 expression [ 18 ], and is involved in the functional development of CD8 + T cells [ 19 ]. Therefore, NRF2 activation could be of benefit for immunotherapy [ 20 – 22 ]. Constitutive activation of the NRF2 signaling arises from alterations in any components of the KEAP1-CUL3-RBX1 complex. Most commonly, hotspot NRF2 mutations, localizing to its DLG and ETGE motifs [ 23 ], impair KEAP1 (the negative regulator of NRF2) binding, ubiquitylation and subsequent proteasomal degradation [ 13 ], yielding constitutively-active NRF2-dependent transcription [ 24 ]. Because NRF2 signaling is observed in about 50% of ovarian cancer, with about 37% of tumors demonstrating KEAP1 mutation [ 25 ], we aimed to 1) Identify if the HGSOC can be classified by NRF2 activation, 2) Investigate the role of NRF2 in HGSOC immune phenotype and 3) understand if NRF2-immune phenotype relationship impacts the prognosis. To do so, we used 3 distinct cohort studies: a cohort study with N = 7 (28,850 cells) scRNA-seq data [ 26 ], a cohort study with N = 365 bulk RNA-seq data from TCGA [ 27 ], and a cohort study with N = 240 human HGSOC tumor samples from a tumor microarray (TMA) [ 28 , 29 ]. Results Single cell RNA-seq results reveal high T-cell infiltration in NRF2 Low HGSOC and high myeloid cell infiltration in NRF2 High HGSOC NRF2 modulates tumor immune microenvironment (IMM) [ 30 ]. To identify the putative role of NRF2 activation in the modulation of HGSOC IMM, we analyzed single cell RNA-Seq (scRNA-Seq) data from a cohort study of 7 human HGSOC samples (Supplementary Table 1) [ 26 ]. UMAP analysis clustered all 28,550 cells into 9 compartments corresponding to various cell types, including epithelial, immune, mesenchymal, and endothelial cells (Fig. 1a). While NRF2 is widely expressed in nearly all cell types (Fig. 1b), the highest NRF2 gene-score was observed in epithelial cells followed by myeloid and endothelial cells (Fig. 1c). In addition, NQO1, a key downstream target of NRF2, was enriched in epithelial and endothelial cells only (Fig. 1d), suggesting the activation of NRF2 signaling in these cells occurred through the canonical pathway, the primary mechanism of NRF2 activation that results in NQO1 expression [ 15 , 31 , 32 ]. We then grouped the 7 HGSOC samples by the NRF2 levels and activation score (Fig. 1e-h): high NRF2 activity (NRF2 High , NRF2 activation score > 0, n = 4) and low NRF2 activity (NRF2 Low , NRF2 activation < = 0). Analysis of immune cell composition revealed that T and B cell infiltration was higher in NRF2 Low tumors compared to NRF2 High tumors, whereas NRF2 High tumors had higher myeloid cell infiltration (Fig. 1i). Further differentiation analysis showed T-cell markers such as CD3-epsilon (CD3E), a marker that plays a key role in T-cell development and signal transduction, and CD8A, a marker of cytotoxic T lymphocytes (CTLs) and a positive marker of patient’s outcomes [ 33 ], were both enriched in NRF2 Low (Fig. 1j, and Supplementary Fig. 1a,b). We also identified markers-associated macrophages such as the CD68, a general marker of macrophages [ 34 ] which was enriched in NRF2 High samples (Fig. 1j, and Supplementary Fig. 1c). We then differentiated the macrophages into M1-like macrophages markers that has a role in activating T-cells, named “good” macrophages such as CD80, and M2-like macrophages that is known to support tumor growth, drug resistance and angiogenesis, named “bad” macrophages including the CD163 and CD206 [ 35 , 36 ]. Interestingly, the “bad” macrophage markers (i.e., CD163 and CD206) were enriched in NRF2 High samples, while the “good” macrophage marker (i.e., CD80) was enriched in NRF2 Low samples (Fig. 1j, and Supplementary Fig. 1d-f).These data show that NRF2 activation can modulate the IMM in HGSOC. FOXP3, a marker of regulatory T cell (Treg) that suppresses the immune response by inhibiting other immune cells [ 37 ], was enriched in NRF2 Low tumors compared to NRF2 High tumors (Fig. 1j, and Supplementary Fig. 1g). The NRF2-FOXP3 relationship in cancer is complex and context-dependent; NRF2 was shown to impact FOXP3 expression and Treg function, potentially influencing immune tolerance and tumor growth (Fig. 1j, and Supplementary Fig. 1g) [ 38 ]. We also analyzed samples for cytokines and chemokines, we identified 3 interleukins that associated with immune suppressive microenvironment, these are CXCL8/IL8 and IL10 [ 39 – 41 ], which were both enriched in NRF2 High tumors (Fig. 1k, and Supplementary Fig. 1h,i). Immune checkpoints (ICs) were enriched in NRF2 Low tumors (Fig. 1l, and Supplementary Fig. 1j-l). These results imply that NRF2 High tumors tend to retain immune suppressive microenvironment while the NRF2 Low tumors support the immune-stimulatory microenvironment. Activation of NRF2 signaling correlates with increased expression of immune suppressive markers in human HGSOC from TCGA To validate the NRF2-immune phenotype relationship in a larger HGSOC cohort, we analyzed genomic and bulk RNA-seq data of 365 human HGSOCs from TCGA [ 27 ]. Among them, 286 samples have mutational data available, of which 88% had TP53 mutations (Supplementary Fig. 2), consistent with the known frequency of TP53 mutations in HGSOC. Consistent with previous studies showing NRF2 activation in 27–83% of ovarian cancer [ 25 , 42 ], 53% (194/ 365) of samples had a high NRF2 activation score (Supplementary Table 2). However, genes encoding the components of the KEAP1-NRF2 pathway showed < 1% to no mutations in the pathway elements in these samples (Supplementary Fig. 2). Additionally, differential expression of genes did not show NOQ1 upregulation in samples with high NRF2 score (Supplementary Table 3 ) , a key downstream target of the NRF2 activated through a canonical KEAP1-NRF2 pathway [ 32 , 43 , 44 ]. These findings may suggest that activation of the NRF2 signaling in these samples is mainly through non-canonical pathway, such as NRF2 amplification and overexpression [ 15 ]. These observations are different from the scRNA-seq data where NRF2 was activated by a canonical pathway, which might imply that NRF2 activation and its effects are context-dependent in HGSOC. In fact, NRF2 activation and its effects are heavily dependent on the specific stage of cancer development and the tumor microenvironment [ 45 ]. We did not observe a significant difference in overall survival between NRF2 High and NRF2 Low tumors (Data not shown). To determine how NRF2 impacts the immune phenotype of these samples, we calculated immune gene score using single-sample gene set enrichment analysis (ssGSEA) and a well-established immune gene signature [ 46 , 47 ] for NRF2 high and NRF2 Low HGSOCs (Supplementary Table 2). Surprisingly, NRF2 High tumors were enriched with immune genes compared to NRF2 Low tumors (Fig. 2a). As NRF2 is known to inhibit the inflammation/immune cell accumulation [ 48 ], the high immune cell score in NRF2 high HGSOCs suggests that NRF2 activation modulates the immune cell differentiation and function rather than suppressing their accumulation in HGSOC [ 30 ]. Thus, we used ssGSEA method utilizing gene expression profiling for immune cells quantification and the immune cells-based gene expression to estimate immune cells differentiation states (Fig. 2b, and Supplementary Table 4). In general, NRF2 high tumors were enriched with immune suppressive cells/markers, including central memory CD8 cells, macrophages, mast cells, memory B-cells, inhibitory dendritic cells (iDC), monocytic myeloid-derived dendritic cells (mDC), T follicular helper (TFH), in addition to fibroblast, endothelial cells, and T-helper2 cells (Th2) (Fig. 2b) [ 49 – 52 ]. We did not observe any significant changes in activated immune cells in high versus low NRF2 HGSOC (Data not shown). Additionally, analysis of ICs in high versus low NRF2 HGSOC showed that the PDCD/PD1 was the only IC with significantly higher levels in NRF2 High HGSOC compared to NRF2 Low group (Supplementary Fig. 3a-c, P < 0.05). The pathway enrichment analysis showed that NRF2 High tumors were significantly enriched with hallmark of cancer compared to NRF2 Low tumors (Fig. 2c, and Supplementary Table 5). We also analyzed the RNA-seq data for cytokines and chemokines production in these samples (Fig. 2d). Our analysis highlighted three interleukins that are associated with cancer progression and poor prognosis, these are IL6, CXCL8/IL8 and IL10 (Fig. 2d) [ 39 – 41 ]. These results agree with the immune suppressive microenvironment of the NRF2 High tumors (Fig. 2a-d). These findings indicate that NRF2 activation contributes to the immune suppressive microenvironment and disease progression. Immune-related markers predict prognosis in human HGSOC: A NRF2-dependent phenotype. To assess whether immune markers predict prognosis based on NRF2 status in HGSOC, we performed survival analysis based for differentially expressed immune markers in NRF2 High and NRF2 Low groups (Fig. 2e,f). Survival analysis showed high expression levels of CD3E predicted a better survival in patients with NRF2 Low HGSOC with DFI: P = 0.032 (Fig. 2e). High CTLA4 expression is also prognostic in NRF2 Low tumors (DFI: p = 0.0072; DSS: p = 0.02; OS: p = 0.031; Fig. 2e). Next, we analyzed the expression of M1-macrophage markers such as CD80, and M2-macrophage markers such as CD163 and CD206 in these tumors and investigated if these markers were associated with the prognosis in NRF2 High and/or NRF2 Low groups. We only found that the higher CD80, a M1-marcophages marker that has a role in activating T-cells, the better survival (OS: p = 0.029; DSS: p = 0.013) in NRF2 Low tumors (Fig. 2e) [ 53 ]. Notably, our analysis did not identify any prognostic value for all immune cell markers investigated in NRF2 High tumors, except for CD68, a general marker for macrophages, where patients with tumors that are enriched with CD68 tend to have worse survival (PFI: p = 0.038) (Fig. 2f). Human HGSOC tumors can be stratified by NRF2 High and NRF2 Low levels with different immune phenotype and prognosis. To investigate if NRF2 protein levels can be used to stratify human HGSOC tumor samples, we stained a human tumor microarray (TMA) with N = 240 HGSOC tumor samples [ 28 , 29 ] for NRF2, immune cell markers as well diagnostic markers for HGSOC (Fig. 3a, b). IHC-NRF2 labeling allowed the classification of HGSOC tumors into two groups, NRF2 High and NRF2 Low tumors (Fig. 3a-c). Tumors with H-score > 1 are considered NRF2 High tumors (n = 147), and tumors with NRF2 H-Score ≤ 1 are considered NRF2 Low tumors (n = 92) (Supplementary Table 6a,b). While we did not observe significant change in IHC-score of CD8 T-cells in NRF2 High tumors compared to NRF2 Low tumors, IHC-CD68 was significantly (P < 0.05) higher in NRF2 High tumors compared to NRF2 Low tumors (Fig. 3d,e), indicating that the NRF2 High tumors are tumors-associated macrophages (TAMs), which is generally considered a poor prognostic factor [ 54 ]. These results agree with our sc-RNA-seq data analysis (Fig. 1j) and TCGA data analysis on macrophages, considering CD68 as a general marker of macrophages (Fig. 2b). However, survival of TMA-patients with high CD68 versus low CD68 did not show a significant change in both (NRF2 High and NRF2 Low ) groups (data not shown). As PD1 and PD-L1 expression within the IMM is generally considered a better prognostic marker compared to PD1/PD-L1 expression in tumors [ 55 ], we analyzed PD1 and PD-L1 expression in both IMM and tumors (Fig. 3f,g). PD1 level in IMM was significantly ( P < 0.0001 ) higher than its level on tumor cells (TMRs) in both (NRF2 High and NRF2 Low ) groups (Fig. 3f). PD1 protein was more abundantly expressed in the IMM of NRF2 High tumors than that of NRF2 Low samples (Fig. 3f, P < 0.001). Additionally, PD-L1 was significantly ( P < 0.001 ) enriched in IMM of NRF2 High samples compared to TMRs, while NRF2 Low samples did not show significant change in PD-L1 levels in IMM versus tumor cells (Fig. 3g). Also, the IMM of NRF2 High samples was significantly ( P < 0.05 ) enriched with PD-L1 compared to IMM of NRF2 Low samples (Fig. 3g). The level of FOXP3, a marker of T-regulatory cells (T-reg), in IMM versus TRMs showed no significant changes in both groups (Fig. 3h). However, NRF2 High tumors were significantly ( P < 0.05 ) enriched with FOXP3 compared to NRF2 Low tumors (Fig. 3h). High FOXP3 expression in immune cells can be detrimental, while its expression in cancer cells can be more nuanced and may depend on the specific tumor characteristics [ 56 ]. VIM expressed by tumor cells is often linked to EMT, promoting tumor cell migration and metastasis [ 57 ]. In both (NRF2 High and NRF2 Low ) groups, IHC-VIM was significantly ( P < 0.0001 ) higher in stroma (STR) than in tumors (Fig. 3i). However, STR of NRF2 High HGSOC was significantly ( P < 0.0001 ) enriched with VIM compared to STR of NRF2 Low HGSOC (Fig. 3i). Survival data of N = 210 from a total of 240 HGSOC TMA samples treated with chemotherapies was used to group patients by NRF2 levels; NRF2 High HGSOC had significantly ( P = 0.00804 ) better overall survival than NRF2 Low HGSOC (Fig. 3j). Disease specific-survival (DSS) of patients with NRF2 High tumors (n = 93) was also significantly (P = 0.0133) better than patients with NRF2 Low tumors (n = 104) (Supplementary Fig. 4a), but not the progression free-survival (PFS) (Supplementary Fig. 4b). Discussion We used 3 cohort studies to investigate if NRF2 activation correlates with distinct immune phenotypes in HGSOC and how that impacts the prognosis. Analysis of scRNA-seq HGSOC samples [ 26 ] demonstrated that NRF2 Low tumors were enriched with T-cells compared to NRF2 High tumors which were enriched with myeloid cells instead (Fig. 1j). This is supported by analyzing the bulk transcriptomic data of TCGA HGSOC cohort (n = 365), which showed NRF2 high tumors are enriched with macrophages, the most abundant type of myeloid cells in tumors [ 58 ], compared to NRF2 low tumors (Fig. 1a,b). Furthermore, IHC analysis of a TMA composing of 240 human HGSOC samples [ 28 , 29 ] showed NRF2 high tumors are enriched with CD68 ( a general macrophage marker) compared to NRF2 low tumors (Fig. 3a-d). Therefore, our study suggests that NRF2 high HGSOC is enriched with tumor-associated macrophages (TAMs) compared to NRF2 Low tumors. Notably, HGSOC is known to have abundant TAMs, which was associated with tumor aggressiveness and poor clinical outcomes [ 59 ]. Our results suggest that TAMs in HGSOC could be associated with NRF2 activation. Our Immune phenotyping analysis revealed the potential value of combined NRF2 activation and immune markers in HGSOCs [ 60 ]. TCGA data analysis showed patients with NRF2 Low tumors that enriched with immune stimulatory markers such as CD3E, and CD80 (a M1-macrophages marker) tend to have better survival than those with low levels of these markers (Fig. 2e). These immune stimulatory markers did not show a prognostic significance in NRF2 High tumors. Instead, CD68, an immune suppressive immune marker that was enriched in NRF2 High tumors from all cohort studies showed a significant correlation with prognosis (PFI: p = 0.038) of patients from TCGA, where the higher CD68 expression, the worst survival in NRF2 High tumors was observed. This finding is concordant with a previous study showing that TAMs in ovarian cancer is generally associated with resistance to chemotherapy [ 59 ]. Interestingly, although results from the TMA (n = 240) samples showed NRF2 High tumors are enriched with CD68 (Fig. 3d), no prognostic value of CD68 expression levels was observed in these patients (data not shown). Instead, patients with NRF2 High tumors had significantly better survival than NRF2 Low tumors (Fig. 3j). These tumors were treated with chemotherapy after surgery. Chemotherapy may induce the differentiation of macrophages towards M1-phenotype with anti-tumor characteristics [ 59 ] which could improve the responses to chemotherapy. Also, chemotherapy could activate NRF2 signaling in ovarian cancer [ 61 ] which could modulate the phenotype of macrophages supporting the M1/M2 polarization [ 62 ]. Whether chemotherapy changed the NRF2 expression levels in these tumors, and if these changes influenced the M1/M2-macrophages polarization is not known . Future studies may IHC-stain HGSOC tumors for NRF2, M1-macrophage markers such as CD80, and M2-macrophage markers such as CD163 and CD206, before and after treatment to predict prognosis. Results from scRNA-seq data analysis showed higher ICs levels including PD1, PD-L1, and CTLA4 in (n = 3) NRF2 Low tumors compared to NRF2 High tumors (Fig. 1l), while the TCGA data showed NRF2 High tumors enriched with PD1 (Supplementary Fig. 3b). The TMA patient samples showed the immune-microenvironment (IMM) of NRF2 High samples were significantly enriched with ICs including PD1 ( p < 0.01 ) and PD-L1 ( p < 0.05 ) compared to NRF2 Low samples (Fig. 3f,g). TMA data also showed that the PD-L1 was significantly (p < 0.001) enriched in IMM compared to its levels in (tumors) TMR in NRF2 High samples which we did not observe in the NRF2 Low samples (Fig. 3f,g). Previously published reports showed the NRF2 activation supports the expression of ICs [ 18 , 22 ]. Patients with PD-L1 High -immune cells tend to have better survival compared to patients with PD-L1 High -tumor cells [ 63 ]. Further, a high level of PD-L1 expression within the IMM of HGSOC tumor is often associated with a better response to chemotherapy [ 64 ]. In conclusion, ICs could be enriched in both NRF2 Low and NRF2 High HGSOC, however, CTLA4 seems to be significant in the NRF2 Low tumors with prognostic benefits (Figs. 1l, and 2e). Instead PD1/PD-L1 could be more important in NRF2 High tumors and might impact the prognosis (Fig. 3f,g and supplementary Fig. 3b). The spatial distribution of these markers could also define the prognosis. Studies have shown that the mechanisms of NRF2 activation may impact cancer phenotype [ 43 , 44 ]. Data analysis of scRNA-seq from the 7 human HGSOC tumors showed NQO1, a key marker of NRF2, was expressed in NRF2 High samples (Fig. 1b-d), suggesting the activation of NRF2 signaling through a canonical pathway in these samples [ 65 ]. These samples showed lower expression of immune markers compared to NRF2 Low samples (Fig. 1j) which is expected as anti-inflammatory pathway, normally inhibits inflammation [ 17 ]. In contrast, our mutation profiling of data from RNA-seq of human HGSOC samples from TCGA [ 27 ], show no somatic mutations in KEAP1-NRF2 pathway (Supplementary Fig. 2), with no NQO1 expression (Supplementary Table 2) suggesting that NRF2 was activated through non-canonical pathways in these samples [ 32 , 43 , 44 ]. The NRF2 High tumor samples from the TCGA show high (immunosuppressive markers) ISMs (Fig. 2a-d). Studies have shown that certain non-canonical pathways of NRF2 activation might attenuate its anti-inflammatory action [ 17 , 66 ]. Whether NRF2 activation pathway (canonical vs non canonical) impacts the immune phenotype and responses to therapies in HGSOC, still need to be determined. Future studies may investigate two factors before deciding the type of therapy in HGSOC 1) genomic evaluation of NRF2 pathway in NRF2 High HGSOC , and 2) IHC evaluation of immune markers in NRF2 High & NRF2 Low HGSOC. Our study has several limitations: 1) Bulk RNA-seq had no data on the type of treatment, and thus we were unable to identify the effect of treatment on immunophenotype of NRF2 High and NRF2 Low samples and if that impacts the survival; 2) The TMA, used in this study had no data on tumors before treatment; also 3) We did not have NRF2 genomic data of these samples. Taken together, our results show that HGSOC can be classified clinically based on NRF2 expression levels. Clinical studies have shown that HGSOC can be treated based on its immune phenotypes, meaning that different treatment strategies could be tailored to a specific immune profile of a patient's tumor, with some tumors responding better to immunotherapy depending on the presence or absence of immune infiltrating cells and markers within the tumor microenvironment. Our data analyses of three cohort studies added a new marker (NRF2) that impacts the immune phenotype and prognosis of HGSOC. However, treatment with chemotherapy and or immunotherapy could change the expression levels of NRF2 and thus the responses to treatment. Future studies will investigate the status of NRF2 signaling, and mechanisms of its activation in HGSOC using IHC and WGS before and after treatment to understand the timing and their effects on tumor heterogeneity and immune phenotype. This approach will further stratify HGSOC patients for personalized treatment and reveal the best treatment-based tumor phenotype to avoid resistance to therapies for better prognosis. Methods Single cell RNA-seq data analysis of human HGSOC. Seven human HGSOC samples from the GSE184880 [ 26 ] dataset was downloaded from The National Center for Biotechnology Information website ( https://www.ncbi.nlm.nih.gov/ ). Data processing and core analysis were performed using Seurat v4.9.9 [ 67 ]. More details can be found in the supplementary methods file. Data Analysis of human HGSOC RNA-seq from TCGA Human HGSOC samples (n = 365) from The Cancer Genome Atlas (TCGA) were used to analyze the mutational profile, NRF2 expression levels, immune phenotype, differentially expressed genes and survival. Details can be found in the supplementary methods file. Tissue processing, Hematoxylin & Eosin (H&E), and immunohistochemistry (IHC) labeling of human tissue microarray (TMA) Staining of human tissue microarray (TMA) with Hematoxylin & Eosin (H&E) and labeling with immunohistochemistry (IHC) were performed at Dr. Yemin’s laboratory at the Department of Pathology at Columbia Cancer Center, Vancouver, BC, Canada. Details are found in the supplementary methods file. Data availability A normalized ovarian cancer transcriptomic data used in this study can be obtained from UCSC Xena ( https://xena.ucsc.edu/cite-us ) [ 27 ]. Sample-specific information was obtained from Verhaak et al. [ 8 ], and TCGA Biolinks database [ 68 ]. Organizational scripts are available upon request. Code availability We extensively used publicly available algorithms/methods. Additional code used in this study is available upon request. Statistical analysis We used GraphPad Prism 10 software (GraphPad Prism, RRID:SCR_002798, San Diego, CA, USA) to present the data. Fisher Exact Test and Gehan-Breslow-Wilcoxon tests were used to identify the differences between the NRF2 High and NRF2 Low HGSOC tumor samples. A p value less than 0.05 (typically ≤ 0.05) was considered statistically significant. Declarations Conflict of interest No conflict of interest to report. Author contributions SHH: conceptualization, experimental design, methodology & implementation, data analysis, supervision, funding support, and manuscript writing, review and/revision; HT, CK, KM, NB, GZ, HS, LF, SL, MK, HK, OA, RM, LK, CC, DW, and YW: data analysis and interpretation; YW: experimental design, methodology & implementation, supervision, funding support, and manuscript review and/revision. All authors reviewed and approved the manuscript. Acknowledgment We thank staffs of the MAPCore at the University of British Columbia, BASIC lab of the BC Cancer Research Institute and the Anatomical Pathology department of BC Cancer Agency for performing histology and immunohistochemistry staining. This study was supported by research funds from Cooper University Health Care, Surgery Department at Cooper Medical School of Rowan University, MD Anderson Cancer Center at Cooper, Camden, NJ, Barbara T Foundation, the Canadian Institutes of Health Research (PJT-178179; to Y.W.), Ovarian Cancer Research Alliance (ECIG-2025-3-2014, to Y.W.), and Michael Smith Health Research BC postdoctoral fellowship to L.F. We also appreciate the generous support from the VGH/UBC Hospital Foundation to the Ovarian Cancer Research Centre. References Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA: A Cancer Journal for Clinicians 2024; 74: 12–49. McCluggage WG. Morphological subtypes of ovarian carcinoma: a review with emphasis on new developments and pathogenesis. Pathology 2011; 43: 420–432. Bowtell DD, Böhm S, Ahmed AA, Aspuria PJ, Bast RC, Jr., Beral V et al . Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer. Nat Rev Cancer 2015; 15: 668–679. Vang R, Shih I-M, Kurman RJ. Ovarian Low-grade and High-grade Serous Carcinoma: Pathogenesis, Clinicopathologic and Molecular Biologic Features, and Diagnostic Problems. Advances in Anatomic Pathology 2009; 16: 267–282. Vaughan S, Coward JI, Bast RC, Berchuck A, Berek JS, Brenton JD et al . Rethinking ovarian cancer: recommendations for improving outcomes. Nature Reviews Cancer 2011; 11: 719–725. Tavares V, Marques IS, Melo IG, Assis J, Pereira D, Medeiros R. Paradigm Shift: A Comprehensive Review of Ovarian Cancer Management in an Era of Advancements. Int J Mol Sci 2024; 25. Lisio MA, Fu L, Goyeneche A, Gao ZH, Telleria C. High-Grade Serous Ovarian Cancer: Basic Sciences, Clinical and Therapeutic Standpoints. Int J Mol Sci 2019; 20. Verhaak RG, Tamayo P, Yang JY, Hubbard D, Zhang H, Creighton CJ et al . Prognostically relevant gene signatures of high-grade serous ovarian carcinoma. J Clin Invest 2013; 123: 517–525. Chen GM, Kannan L, Geistlinger L, Kofia V, Safikhani Z, Gendoo DMA et al . Consensus on Molecular Subtypes of High-Grade Serous Ovarian Carcinoma. Clin Cancer Res 2018; 24: 5037–5047. Zhang AW, McPherson A, Milne K, Kroeger DR, Hamilton PT, Miranda A et al . Interfaces of Malignant and Immunologic Clonal Dynamics in Ovarian Cancer. Cell 2018; 173: 1755–1769.e1722. Schwede M, Waldron L, Mok SC, Wei W, Basunia A, Merritt MA et al . The Impact of Stroma Admixture on Molecular Subtypes and Prognostic Gene Signatures in Serous Ovarian Cancer. Cancer Epidemiol Biomarkers Prev 2020; 29: 509–519. Kommoss S, Winterhoff B, Oberg AL, Konecny GE, Wang C, Riska SM et al . Bevacizumab May Differentially Improve Ovarian Cancer Outcome in Patients with Proliferative and Mesenchymal Molecular Subtypes. Clin Cancer Res 2017; 23: 3794–3801. Bryan HK, Olayanju A, Goldring CE, Park BK. The Nrf2 cell defence pathway: Keap1-dependent and -independent mechanisms of regulation. Biochemical pharmacology 2013; 85: 705–717. Taguchi K, Motohashi H, Yamamoto M. Molecular mechanisms of the Keap1-Nrf2 pathway in stress response and cancer evolution. Genes to cells: devoted to molecular & cellular mechanisms 2011; 16: 123–140. Cloer EW, Goldfarb D, Schrank TP, Weissman BE, Major MB. NRF2 Activation in Cancer: From DNA to Protein. Cancer Res 2019; 79: 889–898. Rojo de la Vega M, Chapman E, Zhang DD. NRF2 and the Hallmarks of Cancer. Cancer cell 2018; 34: 21–43. Saha S, Buttari B, Panieri E, Profumo E, Saso L. An Overview of Nrf2 Signaling Pathway and Its Role in Inflammation. Molecules 2020; 25. Zhu B, Tang L, Chen S, Yin C, Peng S, Li X et al . Targeting the upstream transcriptional activator of PD-L1 as an alternative strategy in melanoma therapy. Oncogene 2018; 37: 4941–4954. Sha LK, Sha W, Kuchler L, Daiber A, Giegerich AK, Weigert A et al . Loss of Nrf2 in bone marrow-derived macrophages impairs antigen-driven CD8(+) T cell function by limiting GSH and Cys availability. Free Radic Biol Med 2015; 83: 77–88. Xu X, Yang Y, Liu X, Cao N, Zhang P, Zhao S et al . NFE2L2/KEAP1 Mutations Correlate with Higher Tumor Mutational Burden Value/PD-L1 Expression and Potentiate Improved Clinical Outcome with Immunotherapy. Oncologist 2020; 25: e955-e963. Gandhi L, Rodríguez-Abreu D, Gadgeel S, Esteban E, Felip E, De Angelis F et al . Pembrolizumab plus Chemotherapy in Metastatic Non-Small-Cell Lung Cancer. N Engl J Med 2018; 378: 2078–2092. Duan J, Zhang Y, Chen R, Liang L, Huo Y, Lu S et al . Tumor-immune microenvironment and NRF2 associate with clinical efficacy of PD-1 blockade combined with chemotherapy in lung squamous cell carcinoma. Cell Rep Med 2023; 4: 101302. Shibata T, Ohta T, Tong KI, Kokubu A, Odogawa R, Tsuta K et al . Cancer related mutations in NRF2 impair its recognition by Keap1-Cul3 E3 ligase and promote malignancy. Proceedings of the National Academy of Sciences 2008; 105: 13568–13573. Kansanen E, Kuosmanen SM, Leinonen H, Levonen AL. The Keap1-Nrf2 pathway: Mechanisms of activation and dysregulation in cancer. Redox Biol 2013; 1: 45–49. Konstantinopoulos PA, Spentzos D, Fountzilas E, Francoeur N, Sanisetty S, Grammatikos AP et al . Keap1 mutations and Nrf2 pathway activation in epithelial ovarian cancer. Cancer Res 2011; 71: 5081–5089. Xu J, Fang Y, Chen K, Li S, Tang S, Ren Y et al . Single-Cell RNA Sequencing Reveals the Tissue Architecture in Human High-Grade Serous Ovarian Cancer. Clinical Cancer Research 2022; 28: 3590–3602. Goldman MJ, Craft B, Hastie M, Repečka K, McDade F, Kamath A et al . Visualizing and interpreting cancer genomics data via the Xena platform. Nature Biotechnology 2020; 38: 675–678. Brassard J, Hughes MR, Dean P, Hernaez DC, Thornton S, Banville AC et al . A tumor-restricted glycoform of podocalyxin is a highly selective marker of immunologically cold high-grade serous ovarian carcinoma. Frontiers in oncology 2023; 13: 1286754. Talhouk A, George J, Wang C, Budden T, Tan TZ, Chiu DS et al . Development and Validation of the Gene Expression Predictor of High-grade Serous Ovarian Carcinoma Molecular SubTYPE (PrOTYPE). Clin Cancer Res 2020; 26: 5411–5423. Feng J, Read OJ, Dinkova-Kostova AT. Nrf2 in TIME: The Emerging Role of Nuclear Factor Erythroid 2-Related Factor 2 in the Tumor Immune Microenvironment. Mol Cells 2023; 46: 142–152. Camp ND, James RG, Dawson DW, Yan F, Davison JM, Houck SA et al . Wilms tumor gene on X chromosome (WTX) inhibits degradation of NRF2 protein through competitive binding to KEAP1 protein. J Biol Chem 2012; 287: 6539–6550. Ross D, Siegel D. The diverse functionality of NQO1 and its roles in redox control. Redox Biology 2021; 41: 101950. Sato E, Olson SH, Ahn J, Bundy B, Nishikawa H, Qian F et al . Intraepithelial CD8 + tumor-infiltrating lymphocytes and a high CD8+/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer. Proc Natl Acad Sci U S A 2005; 102: 18538–18543. Klinge U, Dievernich A, Tolba R, Klosterhalfen B, Davies L. CD68 + macrophages as crucial components of the foreign body reaction demonstrate an unconventional pattern of functional markers quantified by analysis with double fluorescence staining. J Biomed Mater Res B Appl Biomater 2020; 108: 3134–3146. Moeini P, Niedźwiedzka-Rystwej P. Tumor-Associated Macrophages: Combination of Therapies, the Approach to Improve Cancer Treatment. Int J Mol Sci 2021; 22. Yao Z, Zhang J, Zhang B, Liang G, Chen X, Yao F et al . Imatinib prevents lung cancer metastasis by inhibiting M2-like polarization of macrophages. Pharmacol Res 2018; 133: 121–131. Wang Y, Zheng J, Islam MS, Yang Y, Hu Y, Chen X. The role of CD4(+)FoxP3(+) regulatory T cells in the immunopathogenesis of COVID-19: implications for treatment. Int J Biol Sci 2021; 17: 1507–1520. Klemm P, Rajendiran A, Fragoulis A, Wruck C, Schippers A, Wagner N et al . Nrf2 expression driven by Foxp3 specific deletion of Keap1 results in loss of immune tolerance in mice. Eur J Immunol 2020; 50: 515–524. Mocellin S, Marincola FM, Young HA. Interleukin-10 and the immune response against cancer: a counterpoint. Journal of leukocyte biology 2005; 78: 1043–1051. Hazini A, Fisher K, Seymour L. Deregulation of HLA-I in cancer and its central importance for immunotherapy. Journal for immunotherapy of cancer 2021; 9. Shi Z, Yang WM, Chen LP, Yang DH, Zhou Q, Zhu J et al . Enhanced chemosensitization in multidrug-resistant human breast cancer cells by inhibition of IL-6 and IL-8 production. Breast Cancer Res Treat 2012; 135: 737–747. Martinez VD, Vucic EA, Thu KL, Pikor LA, Hubaux R, Lam WL. Unique pattern of component gene disruption in the NRF2 inhibitor KEAP1/CUL3/RBX1 E3-ubiquitin ligase complex in serous ovarian cancer. Biomed Res Int 2014; 2014: 159459. Silva-Islas CA, Maldonado PD. Canonical and non-canonical mechanisms of Nrf2 activation. Pharmacol Res 2018; 134: 92–99. Dodson M, Zhang DD. Non-canonical activation of NRF2: New insights and its relevance to disease. Curr Pathobiol Rep 2017; 5: 171–176. Sporn MB, Liby KT. NRF2 and cancer: the good, the bad and the importance of context. Nat Rev Cancer 2012; 12: 564–571. Bhattacharya S, Dunn P, Thomas CG, Smith B, Schaefer H, Chen J et al . ImmPort, toward repurposing of open access immunological assay data for translational and clinical research. Scientific Data 2018; 5: 180015. Breuer K, Foroushani AK, Laird MR, Chen C, Sribnaia A, Lo R et al . InnateDB: systems biology of innate immunity and beyond–recent updates and continuing curation. Nucleic Acids Res 2013; 41: D1228-1233. Ahmed SM, Luo L, Namani A, Wang XJ, Tang X. Nrf2 signaling pathway: Pivotal roles in inflammation. Biochimica et biophysica acta Molecular basis of disease 2017; 1863: 585–597. Ellyard JI, Simson L, Parish CR. Th2-mediated anti-tumour immunity: friend or foe? Tissue Antigens 2007; 70: 1–11. Zamarron BF, Chen W. Dual roles of immune cells and their factors in cancer development and progression. Int J Biol Sci 2011; 7: 651–658. Marques HS, de Brito BB, da Silva FAF, Santos MLC, de Souza JCB, Correia TML et al . Relationship between Th17 immune response and cancer. World J Clin Oncol 2021; 12: 845–867. Tie Y, Tang F, Wei Y-q, Wei X-w. Immunosuppressive cells in cancer: mechanisms and potential therapeutic targets. Journal of Hematology & Oncology 2022; 15: 61. Cheng H, Wang Z, Fu L, Xu T. Macrophage Polarization in the Development and Progression of Ovarian Cancers: An Overview. Frontiers in oncology 2019; 9: 421. Gottlieb CE, Mills AM, Cross JV, Ring KL. Tumor-associated macrophage expression of PD-L1 in implants of high grade serous ovarian carcinoma: A comparison of matched primary and metastatic tumors. Gynecol Oncol 2017; 144: 607–612. Han Y, Liu D, Li L. PD-1/PD-L1 pathway: current researches in cancer. Am J Cancer Res 2020; 10: 727–742. Blanc-Durand F, Clemence Wei Xian L, Tan DSP. Targeting the immune microenvironment for ovarian cancer therapy. Front Immunol 2023; 14: 1328651. Ngan CY, Yamamoto H, Seshimo I, Tsujino T, Man-i M, Ikeda JI et al . Quantitative evaluation of vimentin expression in tumour stroma of colorectal cancer. British journal of cancer 2007; 96: 986–992. van Vlerken-Ysla L, Tyurina YY, Kagan VE, Gabrilovich DI. Functional states of myeloid cells in cancer. Cancer cell 2023; 41: 490–504. El-Arabey AA, Alkhalil SS, Al-Shouli ST, Awadalla ME, Alhamdi HW, Almanaa TN et al . Revisiting macrophages in ovarian cancer microenvironment: development, function and interaction. Med Oncol 2023; 40: 142. Bösmüller HC, Wagner P, Peper JK, Schuster H, Pham DL, Greif K et al . Combined Immunoscore of CD103 and CD3 Identifies Long-Term Survivors in High-Grade Serous Ovarian Cancer. Int J Gynecol Cancer 2016; 26: 671–679. Tossetta G, Fantone S, Montanari E, Marzioni D, Goteri G. Role of NRF2 in Ovarian Cancer. Antioxidants (Basel) 2022; 11. Wang L, He C. Nrf2-mediated anti-inflammatory polarization of macrophages as therapeutic targets for osteoarthritis. Front Immunol 2022; 13: 967193. Mocan LP, Craciun R, Grapa C, Melincovici CS, Rusu I, Al Hajjar N et al . PD-L1 expression on immune cells, but not on tumor cells, is a favorable prognostic factor for patients with intrahepatic cholangiocarcinoma. Cancer Immunol Immunother 2023; 72: 1003–1014. Vanguri R, Benhamida J, Young JH, Li Y, Zivanovic O, Chi D et al . Understanding the impact of chemotherapy on the immune landscape of high-grade serous ovarian cancer. Gynecol Oncol Rep 2022; 39: 100926. Mutter FE, Park BK, Copple IM. Value of monitoring Nrf2 activity for the detection of chemical and oxidative stress. Biochem Soc Trans 2015; 43: 657–662. Izumi Y, Tatsumoto A, Horiuchi N, Arifuku M, Uegomori M, Kume T et al . TPNA10168, an Nrf-2 activator, attenuates inflammatory responses independently of Nrf2 in microglial BV-2 cells: Involvement of the extracellular-signal-regulated kinase pathway. J Pharmacol Sci 2022; 149: 1–10. Hao Y, Hao S, Andersen-Nissen E, Mauck WM, 3rd, Zheng S, Butler A et al . Integrated analysis of multimodal single-cell data. Cell 2021; 184: 3573–3587.e3529. Colaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D et al . TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res 2016; 44: e71. Additional Declarations There is NO conflict of interest to disclose. No conflict of interest to report Supplementary Files SupplementaryMethodsJune2025.pdf Supplementary Methods SIFiguresJune2025.pdf Supplementary Figures SupplementaryTablesNEWFINAL.xlsx Supplementary Tables Cite Share Download PDF Status: Published Journal Publication published 28 Apr, 2026 Read the published version in Genes & Immunity → Version 1 posted Editorial decision: revise 18 Sep, 2025 Review # 2 received at journal 26 Aug, 2025 Reviewer # 3 agreed at journal 26 Aug, 2025 Reviewer # 2 agreed at journal 21 Aug, 2025 Review # 1 received at journal 15 Jul, 2025 Reviewer # 1 agreed at journal 14 Jul, 2025 Reviewers invited by journal 14 Jul, 2025 Submission checks completed at journal 23 Jun, 2025 Editor assigned by journal 18 Jun, 2025 First submitted to journal 18 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6925656","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":485362919,"identity":"3ed820e1-a92a-44a8-ab80-fc37d65c92e2","order_by":0,"name":"Samera Hamad","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0ElEQVRIiWNgGAWjYPCCmgR+EJVQQLyWYwmSDSAtBsRrYU4wOACiidEi33468XFFDVue8fnViR8eGDDI84sdwK/F4EzuZsMzx2SKzW683SwBdJjhzNkJBLQw5G6TbGBjS9x24+wGkJYEg9sEtMj3v93+s+Efc+LmGWc3/yBKC8ON3G2MjW3MiRv4e7cRZ4sB0AuSjX3HEmfc4N1mkWAgQdgv8v25Gz82fKtJ7O8/u/nmjwobeX5pQg6DAwmwSglilYMA/wFSVI+CUTAKRsFIAgDAUUlOoSkewwAAAABJRU5ErkJggg==","orcid":"","institution":"Cooper Medical School of Rowan University","correspondingAuthor":true,"prefix":"","firstName":"Samera","middleName":"","lastName":"Hamad","suffix":""},{"id":485362920,"identity":"90b5e05a-a2fe-4e05-a8b1-6609682d9483","order_by":1,"name":"Chelsea Katz","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Chelsea","middleName":"","lastName":"Katz","suffix":""},{"id":485362921,"identity":"0d7d3346-a627-4624-b482-dc9d01770eb9","order_by":2,"name":"Helen Toma","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Helen","middleName":"","lastName":"Toma","suffix":""},{"id":485362922,"identity":"266f1e5b-389e-444e-ae00-fcfce41c506d","order_by":3,"name":"Kosuke Murakami","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Kosuke","middleName":"","lastName":"Murakami","suffix":""},{"id":485362923,"identity":"8738a4ca-b1dd-4d6b-a018-c532a97bf0a4","order_by":4,"name":"Nasrine Bendjilali","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Nasrine","middleName":"","lastName":"Bendjilali","suffix":""},{"id":485362924,"identity":"dbee7bf2-3f73-4e78-bf5c-f1ae4158bb50","order_by":5,"name":"Gord Zhu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Gord","middleName":"","lastName":"Zhu","suffix":""},{"id":485362925,"identity":"088ed664-4869-4b9f-befe-027dcbaaaa08","order_by":6,"name":"Hadi Shojaei","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Hadi","middleName":"","lastName":"Shojaei","suffix":""},{"id":485362926,"identity":"f7ed043c-717d-4c05-a5ee-f669fc6edcb3","order_by":7,"name":"Lanlan Fang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Lanlan","middleName":"","lastName":"Fang","suffix":""},{"id":485362927,"identity":"131f66dc-2f19-4ccc-a3a9-2b1705d2f0b0","order_by":8,"name":"Samuel Leung","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Samuel","middleName":"","lastName":"Leung","suffix":""},{"id":485362928,"identity":"37c8bfaf-3c1b-46f3-9938-d7b2864f50a8","order_by":9,"name":"Martin Koebel","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Koebel","suffix":""},{"id":485362929,"identity":"e1cedc74-bd55-4910-8b88-159e779d658c","order_by":10,"name":"Huseyin Karaduman","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Huseyin","middleName":"","lastName":"Karaduman","suffix":""},{"id":485362930,"identity":"096880bc-7ef1-4121-b908-b364d5bc464b","order_by":11,"name":"Oliver Abinader","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Oliver","middleName":"","lastName":"Abinader","suffix":""},{"id":485362931,"identity":"01635e56-ccfb-475e-ab49-0a3c2cadcbbf","order_by":12,"name":"Ramkrishna Mitra","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Ramkrishna","middleName":"","lastName":"Mitra","suffix":""},{"id":485362932,"identity":"ff2c7177-5511-42e7-86da-d644c35e353f","order_by":13,"name":"Lauren Krill","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Lauren","middleName":"","lastName":"Krill","suffix":""},{"id":485362933,"identity":"ea925784-9f1f-43ce-9983-014112097b7a","order_by":14,"name":"Christina Chu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Christina","middleName":"","lastName":"Chu","suffix":""},{"id":485362934,"identity":"50729904-9421-4073-8e3a-660e4e73efaf","order_by":15,"name":"David Warshal","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Warshal","suffix":""},{"id":485362935,"identity":"d04e77d8-9f29-4a6d-b3db-b89002a583c3","order_by":16,"name":"Yemin Wang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yemin","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-06-18 19:40:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6925656/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6925656/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41435-026-00400-7","type":"published","date":"2026-04-28T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":87268995,"identity":"2e89af8f-1520-4942-bed7-e150dd58bf63","added_by":"auto","created_at":"2025-07-22 08:12:57","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":660129,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSingle cell RNA-seq data analysis of 7 human HGSOC tumor samples, shows NRF2 expression levels with immune cell distribution and infiltration in high (n=4) versus low (n=3) NRF2 HGSOC. a \u003c/strong\u003eAnnotated UMAP of HGSOC, total: 28,550 cells classified into 9 compartments, \u003cstrong\u003eb\u003c/strong\u003e feature plots of NFE2L2/NRF2 gene expression levels, with \u003cstrong\u003ec\u003c/strong\u003e its gene activation score, and \u003cstrong\u003ed\u003c/strong\u003e NRF2-key downstream target NQO1, \u003cstrong\u003ee \u003c/strong\u003eViolin plot shows NRF2 expression levels and \u003cstrong\u003ef\u003c/strong\u003e its gene signature in high versus low NRF2 HGSOC samples, \u003cstrong\u003eg\u003c/strong\u003e Annotated UMAP of HGSOC, 28,550 cells were classified into 9 compartments in high (n=4) versus low (n=3) NRF2 tumors, with \u003cstrong\u003eh \u003c/strong\u003eNRF2 levels distributed in all the 9 compartments in both groups, \u003cstrong\u003eI\u003c/strong\u003e percentages of immune cells in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC samples presented in \u003cstrong\u003e(g)\u003c/strong\u003e,\u003cstrong\u003e j \u003c/strong\u003edot plot of the expression levels of immune markers associated T-cells and macrophages in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC samples, along with the expression of \u003cstrong\u003ek \u003c/strong\u003einterleukins, and \u003cstrong\u003el\u003c/strong\u003e immune checkpoints expression levels in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC samples\u003c/p\u003e","description":"","filename":"11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/32e56e0eec1b6978bcb4ef7c.jpg"},{"id":87272577,"identity":"4b998a4f-0eb3-498f-afcd-88b9af44c41f","added_by":"auto","created_at":"2025-07-22 08:28:57","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":978524,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRNA-seq data analysis of n=365 HGSOC tumor samples from TCGA immune phenotype in high (n=194) versus low (n=171) NRF2 HGSOC.\u003c/strong\u003e \u003cstrong\u003ea\u003c/strong\u003e Immune gene signature in high versus low NRF2 HGSOC tumors, \u003cstrong\u003eb\u003c/strong\u003e differentiation analysis shows significantly higher accumulation of immune cells with suppressive functions in tumors with high NRF2 expression compared to tumors with low NRF2 expression, \u003cstrong\u003ec\u003c/strong\u003e selected top significantly enriched hallmark of cancer in high versus low NRF2 HGSOC, \u003cstrong\u003ed\u003c/strong\u003e IL6, CXCL8/IL8 and IL10 show higher expression levels in high NRF2 tumors compared to low NRF2 tumors, \u003cstrong\u003ee\u003c/strong\u003e Survival of NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC patients with high versus low immune markers, and \u003cstrong\u003eh\u003c/strong\u003e Survival of NRF2\u003csup\u003eHigh\u003c/sup\u003e HGSOC patients with high versus low CD68; *P\u0026lt; 0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001, ****P\u0026lt;0.0001\u003c/p\u003e","description":"","filename":"12.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/e493d1e8096e52d6efa327f6.jpg"},{"id":87268997,"identity":"cdfe0571-2e38-45b7-af4a-31e7280e4f56","added_by":"auto","created_at":"2025-07-22 08:12:57","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1065549,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTwo representative HGSOC tumor samples from n=240 TMA human tumors used in this study. a \u003c/strong\u003eNRF2\u003csup\u003eHigh\u003c/sup\u003e HGSOC, versus \u003cstrong\u003eb\u003c/strong\u003e NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC tumor sample stained for H\u0026amp;E and labeled for IHC-NRF2, differentiation markers, immune markers, immune checkpoints, and vimentin (VIM). Scale bar=100µm, images were taken by Olympus BX43 microscope, \u003cstrong\u003ec\u003c/strong\u003e IHC Quantification (H-score) of NRF2, \u003cstrong\u003ed \u003c/strong\u003eCD68, \u003cstrong\u003ee\u003c/strong\u003e CD8 in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow\u003c/sup\u003e tumors, \u003cstrong\u003ef\u003c/strong\u003e IHC-quantification of PDCD1 (PD1) in immune microenvironment (IMM) versus tumor (TMR) cells in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow \u003c/sup\u003eHGSOC samples, \u003cstrong\u003eg\u003c/strong\u003e IHC-quantification of CD274 distributed in IMM versus TMR cells in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow \u003c/sup\u003eHGSOC samples, \u003cstrong\u003eh\u003c/strong\u003e H-score of FOXP3 in TMR versus IMM in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow \u003c/sup\u003eHGSOCs, and \u003cstrong\u003ei\u003c/strong\u003e vimentin (VIM) H-score in TMR cells versus stroma (STR) in NRF2\u003csup\u003eHigh\u003c/sup\u003e versus NRF2\u003csup\u003eLow \u003c/sup\u003eHGSOC, \u003cstrong\u003ej \u003c/strong\u003eOverall survival (OS) of n=210 patients with NRF2\u003csup\u003eHigh\u003c/sup\u003e (H-score \u0026gt;1) and NRF2\u003csup\u003eLow\u003c/sup\u003e (H-score \u0026lt;=1) HGSOC.*P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001, ****P\u0026lt;0.0001\u003c/p\u003e","description":"","filename":"13.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/45e7bfeb1dc6e952ca4eb9e7.jpg"},{"id":108494915,"identity":"fb8e9e85-0ed1-4c7e-a073-22b3713dee1c","added_by":"auto","created_at":"2026-05-05 10:07:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3120214,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/74c9bd7e-8a92-4a47-b489-69613414ada1.pdf"},{"id":87270559,"identity":"7876769c-25f1-4ca1-a57d-132db39912ce","added_by":"auto","created_at":"2025-07-22 08:20:57","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":69225,"visible":true,"origin":"","legend":"Supplementary Methods","description":"","filename":"SupplementaryMethodsJune2025.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/cf80565c947871fb42ef29ed.pdf"},{"id":87268998,"identity":"de3f897c-afaf-444d-87a1-432654d95c3d","added_by":"auto","created_at":"2025-07-22 08:12:57","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1808509,"visible":true,"origin":"","legend":"Supplementary Figures","description":"","filename":"SIFiguresJune2025.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/36331ca989fa9489587ef692.pdf"},{"id":87269011,"identity":"559230ca-7a4b-4fd3-b402-bbf11d5455aa","added_by":"auto","created_at":"2025-07-22 08:12:58","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":24524015,"visible":true,"origin":"","legend":"Supplementary Tables","description":"","filename":"SupplementaryTablesNEWFINAL.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6925656/v1/f2db55b8b9ef405c50f8d31d.xlsx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose.\nNo conflict of interest to report","formattedTitle":"Immune phenotype in high- versus low-NRF2 high grade serous ovarian cancer and the impact on prognosis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOvarian cancer is a leading global cause of cancer-related deaths in women, with an estimated 12,740 deaths in the United States in 2024 [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The overall 5-year relative survival for patients diagnosed at early stage is \u0026sim;85%, which drops to 31% with distant metastatic disease [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Several histological subtypes of epithelial ovarian cancer have been identified such as low or \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003ehigh grade serous ovarian cancer\u003c/span\u003e (LGSOC or \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eHGSOC\u003c/span\u003e), clear cell ovarian cancer (CCOC), endometrioid (EOC), mucinous ovarian cancer (MOC), and borderline ovarian cancer (also called tumors of low malignant potential, LMP, because they do not invade the basement membrane) [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. HGSOC is the most common and aggressive type, accounting for \u0026sim;70\u0026ndash;80% of ovarian cancer deaths, with most cases diagnosed at advanced stages [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. The 5-year survival rate of patients with HGSOC is only 34% [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], with relapse being the main factor responsible for such a high mortality rate [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHGSOC is still treated as a single disease with a combination of surgery plus systemic platinum-based chemotherapy [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. HGSOC is comprised of several histological architectures, including solid, glandular, papillary and or pseudo-endometroid architecture [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Further, The Cancer Genome Atlas (TCGA) and other studies have identified four molecular subtypes of HGSOC, these include immunoreactive, differentiated, proliferative, and mesenchymal [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], which further complicates efforts to find therapies for this deadly cancer. These subtypes are associated with several distinct clinical and pathologic characteristics [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], potentially reflecting differences in immune cell [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] and stromal content [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], which could impact their responses to treatment [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe NRF2 (also known as nuclear factor erythroid-derived 2-like 2; NFE2L2) signaling pathway is known to provide an important cellular defense mechanism against oxidative and metabolic stress [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. In addition to maintaining redox homeostasis, NRF2 pathway also regulates many other cancer hallmarks including proliferation, differentiation, apoptosis, inflammation and metastasis [\u003cspan additionalcitationids=\"CR15\" citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. NRF2 is also known as an anti-inflammatory pathway; its activation reduces the expression of several inflammatory genes and pathways [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], controls PD-L1 expression [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], and is involved in the functional development of CD8\u0026thinsp;+\u0026thinsp;T cells [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Therefore, \u003cem\u003eNRF2\u003c/em\u003e activation could be of benefit for immunotherapy [\u003cspan additionalcitationids=\"CR21\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Constitutive activation of the NRF2 signaling arises from alterations in any components of the KEAP1-CUL3-RBX1 complex. Most commonly, hotspot \u003cem\u003eNRF2\u003c/em\u003e mutations, localizing to its DLG and ETGE motifs [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], impair KEAP1 (the negative regulator of NRF2) binding, ubiquitylation and subsequent proteasomal degradation [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], yielding constitutively-active NRF2-dependent transcription [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eBecause NRF2 signaling is observed in about 50% of ovarian cancer, with about 37% of tumors demonstrating KEAP1 mutation [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e], we aimed to 1) Identify if the HGSOC can be classified by NRF2 activation, 2) Investigate the role of NRF2 in HGSOC immune phenotype and 3) understand if NRF2-immune phenotype relationship impacts the prognosis. To do so, we used 3 distinct cohort studies: a cohort study with N\u0026thinsp;=\u0026thinsp;7 (28,850 cells) scRNA-seq data [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e], a cohort study with N\u0026thinsp;=\u0026thinsp;365 bulk RNA-seq data from TCGA [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], and a cohort study with N\u0026thinsp;=\u0026thinsp;240 human HGSOC tumor samples from a tumor microarray (TMA) [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cb\u003eSingle cell RNA-seq results reveal high T-cell infiltration in NRF2\u003c/b\u003e \u003csup\u003e \u003cb\u003eLow\u003c/b\u003e \u003c/sup\u003e \u003cb\u003eHGSOC and high myeloid cell infiltration in NRF2\u003c/b\u003e\u003csup\u003e\u003cb\u003eHigh\u003c/b\u003e\u003c/sup\u003e \u003cb\u003eHGSOC\u003c/b\u003e\u003c/p\u003e \u003cp\u003eNRF2 modulates tumor immune microenvironment (IMM) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. To identify the putative role of NRF2 activation in the modulation of HGSOC IMM, we analyzed single cell RNA-Seq (scRNA-Seq) data from a cohort study of 7 human HGSOC samples (Supplementary Table\u0026nbsp;1) [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. UMAP analysis clustered all 28,550 cells into 9 compartments corresponding to various cell types, including epithelial, immune, mesenchymal, and endothelial cells (Fig.\u0026nbsp;1a). While NRF2 is widely expressed in nearly all cell types (Fig.\u0026nbsp;1b), the highest NRF2 gene-score was observed in epithelial cells followed by myeloid and endothelial cells (Fig.\u0026nbsp;1c). In addition, NQO1, a key downstream target of NRF2, was enriched in epithelial and endothelial cells only (Fig.\u0026nbsp;1d), suggesting the activation of NRF2 signaling in these cells occurred through the canonical pathway, the primary mechanism of NRF2 activation that results in NQO1 expression [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. We then grouped the 7 HGSOC samples by the NRF2 levels and activation score (Fig.\u0026nbsp;1e-h): high NRF2 activity (NRF2\u003csup\u003eHigh\u003c/sup\u003e, NRF2 activation score\u0026thinsp;\u0026gt;\u0026thinsp;0, n\u0026thinsp;=\u0026thinsp;4) and low NRF2 activity (NRF2\u003csup\u003eLow\u003c/sup\u003e, NRF2 activation\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;0). Analysis of immune cell composition revealed that T and B cell infiltration was higher in NRF2\u003csup\u003eLow\u003c/sup\u003e tumors compared to NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors, whereas NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors had higher myeloid cell infiltration (Fig.\u0026nbsp;1i). Further differentiation analysis showed T-cell markers such as CD3-epsilon (CD3E), a marker that plays a key role in T-cell development and signal transduction, and CD8A, a marker of cytotoxic T lymphocytes (CTLs) and a positive marker of patient\u0026rsquo;s outcomes [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e], were both enriched in NRF2\u003csup\u003eLow\u003c/sup\u003e (Fig.\u0026nbsp;1j, and Supplementary Fig.\u0026nbsp;1a,b). We also identified markers-associated macrophages such as the CD68, a general marker of macrophages [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e] which was enriched in NRF2\u003csup\u003eHigh\u003c/sup\u003e samples (Fig.\u0026nbsp;1j, and Supplementary Fig.\u0026nbsp;1c). We then differentiated the macrophages into M1-like macrophages markers that has a role in activating T-cells, named \u0026ldquo;good\u0026rdquo; macrophages such as CD80, and M2-like macrophages that is known to support tumor growth, drug resistance and angiogenesis, named \u0026ldquo;bad\u0026rdquo; macrophages including the CD163 and CD206 [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Interestingly, the \u0026ldquo;bad\u0026rdquo; macrophage markers (i.e., CD163 and CD206) were enriched in NRF2\u003csup\u003eHigh\u003c/sup\u003e samples, while the \u0026ldquo;good\u0026rdquo; macrophage marker (i.e., CD80) was enriched in NRF2\u003csup\u003eLow\u003c/sup\u003e samples (Fig.\u0026nbsp;1j, and Supplementary Fig.\u0026nbsp;1d-f).These data show that NRF2 activation can modulate the IMM in HGSOC.\u003c/p\u003e \u003cp\u003eFOXP3, a marker of regulatory T cell (Treg) that suppresses the immune response by inhibiting other immune cells [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e], was enriched in NRF2\u003csup\u003eLow\u003c/sup\u003e tumors compared to NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors (Fig.\u0026nbsp;1j, and Supplementary Fig.\u0026nbsp;1g). The NRF2-FOXP3 relationship in cancer is complex and context-dependent; NRF2 was shown to impact FOXP3 expression and Treg function, potentially influencing immune tolerance and tumor growth (Fig.\u0026nbsp;1j, and Supplementary Fig.\u0026nbsp;1g) [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWe also analyzed samples for cytokines and chemokines, we identified 3 interleukins that associated with immune suppressive microenvironment, these are CXCL8/IL8 and IL10 [\u003cspan additionalcitationids=\"CR40\" citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e], which were both enriched in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors (Fig.\u0026nbsp;1k, and Supplementary Fig.\u0026nbsp;1h,i). Immune checkpoints (ICs) were enriched in NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;1l, and Supplementary Fig.\u0026nbsp;1j-l). These results imply that NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors tend to retain immune suppressive microenvironment while the NRF2\u003csup\u003eLow\u003c/sup\u003e tumors support the immune-stimulatory microenvironment.\u003c/p\u003e \u003cp\u003e \u003cb\u003eActivation of NRF2 signaling correlates with increased expression of immune suppressive markers in human HGSOC from TCGA\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo validate the NRF2-immune phenotype relationship in a larger HGSOC cohort, we analyzed genomic and bulk RNA-seq data of 365 human HGSOCs from TCGA [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Among them, 286 samples have mutational data available, of which 88% had \u003cem\u003eTP53\u003c/em\u003e mutations (Supplementary Fig.\u0026nbsp;2), consistent with the known frequency of TP53 mutations in HGSOC. Consistent with previous studies showing NRF2 activation in 27\u0026ndash;83% of ovarian cancer [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], 53% (194/ 365) of samples had a high NRF2 activation score (Supplementary Table\u0026nbsp;2). However, genes encoding the components of the KEAP1-NRF2 pathway showed\u0026thinsp;\u0026lt;\u0026thinsp;1% to no mutations in the pathway elements in these samples (Supplementary Fig.\u0026nbsp;2). Additionally, differential expression of genes did not show NOQ1 upregulation in samples with high NRF2 score (Supplementary Table\u0026nbsp;3\u003cb\u003e)\u003c/b\u003e, a key downstream target of the NRF2 activated through a canonical KEAP1-NRF2 pathway [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. These findings may suggest that activation of the NRF2 signaling in these samples is mainly through non-canonical pathway, such as NRF2 amplification and overexpression [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. These observations are different from the scRNA-seq data where NRF2 was activated by a canonical pathway, which might imply that NRF2 activation and its effects are context-dependent in HGSOC. In fact, NRF2 activation and its effects are heavily dependent on the specific stage of cancer development and the tumor microenvironment [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWe did not observe a significant difference in overall survival between NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Data not shown). To determine how NRF2 impacts the immune phenotype of these samples, we calculated immune gene score using single-sample gene set enrichment analysis (ssGSEA) and a well-established immune gene signature [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e] for NRF2\u003csup\u003ehigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOCs (Supplementary Table\u0026nbsp;2). Surprisingly, NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors were enriched with immune genes compared to NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;2a). As NRF2 is known to inhibit the inflammation/immune cell accumulation [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e], the high immune cell score in NRF2\u003csup\u003ehigh\u003c/sup\u003e HGSOCs suggests that NRF2 activation modulates the immune cell differentiation and function rather than suppressing their accumulation in HGSOC [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Thus, we used ssGSEA method utilizing gene expression profiling for immune cells quantification and the immune cells-based gene expression to estimate immune cells differentiation states (Fig.\u0026nbsp;2b, and Supplementary Table\u0026nbsp;4). In general, NRF2\u003csup\u003ehigh\u003c/sup\u003e tumors were enriched with immune suppressive cells/markers, including central memory CD8 cells, macrophages, mast cells, memory B-cells, inhibitory dendritic cells (iDC), monocytic myeloid-derived dendritic cells (mDC), T follicular helper (TFH), in addition to fibroblast, endothelial cells, and T-helper2 cells (Th2) (Fig.\u0026nbsp;2b) [\u003cspan additionalcitationids=\"CR50 CR51\" citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. We did not observe any significant changes in activated immune cells in high versus low NRF2 HGSOC (Data not shown). Additionally, analysis of ICs in high versus low NRF2 HGSOC showed that the PDCD/PD1 was the only IC with significantly higher levels in NRF2\u003csup\u003eHigh\u003c/sup\u003e HGSOC compared to NRF2\u003csup\u003eLow\u003c/sup\u003e group (Supplementary Fig.\u0026nbsp;3a-c, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003eThe pathway enrichment analysis showed that NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors were significantly enriched with hallmark of cancer compared to NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;2c, and Supplementary Table\u0026nbsp;5). We also analyzed the RNA-seq data for cytokines and chemokines production in these samples (Fig.\u0026nbsp;2d). Our analysis highlighted three interleukins that are associated with cancer progression and poor prognosis, these are IL6, CXCL8/IL8 and IL10 (Fig.\u0026nbsp;2d) [\u003cspan additionalcitationids=\"CR40\" citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. These results agree with the immune suppressive microenvironment of the NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors (Fig.\u0026nbsp;2a-d). These findings indicate that NRF2 activation contributes to the immune suppressive microenvironment and disease progression.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImmune-related markers predict prognosis in human HGSOC: A NRF2-dependent phenotype.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo assess whether immune markers predict prognosis based on NRF2 status in HGSOC, we performed survival analysis based for differentially expressed immune markers in NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e groups (Fig.\u0026nbsp;2e,f).\u003c/p\u003e \u003cp\u003eSurvival analysis showed high expression levels of CD3E predicted a better survival in patients with NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC with DFI: P\u0026thinsp;=\u0026thinsp;0.032 (Fig.\u0026nbsp;2e). High CTLA4 expression is also prognostic in NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (DFI: p\u0026thinsp;=\u0026thinsp;0.0072; DSS: p\u0026thinsp;=\u0026thinsp;0.02; OS: p\u0026thinsp;=\u0026thinsp;0.031; Fig.\u0026nbsp;2e). Next, we analyzed the expression of M1-macrophage markers such as CD80, and M2-macrophage markers such as CD163 and CD206 in these tumors and investigated if these markers were associated with the prognosis in NRF2\u003csup\u003eHigh\u003c/sup\u003e and/or NRF2\u003csup\u003eLow\u003c/sup\u003e groups. We only found that the higher CD80, a M1-marcophages marker that has a role in activating T-cells, the better survival (OS: p\u0026thinsp;=\u0026thinsp;0.029; DSS: p\u0026thinsp;=\u0026thinsp;0.013) in NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;2e) [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. Notably, our analysis did not identify any prognostic value for all immune cell markers investigated in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors, except for CD68, a general marker for macrophages, where patients with tumors that are enriched with CD68 tend to have worse survival (PFI: p\u0026thinsp;=\u0026thinsp;0.038) (Fig.\u0026nbsp;2f).\u003c/p\u003e \u003cp\u003e \u003cb\u003eHuman HGSOC tumors can be stratified by NRF2\u003c/b\u003e \u003csup\u003e \u003cb\u003eHigh\u003c/b\u003e \u003c/sup\u003e \u003cb\u003eand NRF2\u003c/b\u003e\u003csup\u003e\u003cb\u003eLow\u003c/b\u003e\u003c/sup\u003e \u003cb\u003elevels with different immune phenotype and prognosis.\u003c/b\u003e\u003c/p\u003e \u003cp\u003eTo investigate if NRF2 protein levels can be used to stratify human HGSOC tumor samples, we stained a human tumor microarray (TMA) with N\u0026thinsp;=\u0026thinsp;240 HGSOC tumor samples [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] for NRF2, immune cell markers as well diagnostic markers for HGSOC (Fig.\u0026nbsp;3a, b). IHC-NRF2 labeling allowed the classification of HGSOC tumors into two groups, NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;3a-c). Tumors with H-score\u0026thinsp;\u0026gt;\u0026thinsp;1 are considered NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors (n\u0026thinsp;=\u0026thinsp;147), and tumors with NRF2 H-Score \u0026le; 1 are considered NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (n\u0026thinsp;=\u0026thinsp;92) (Supplementary Table\u0026nbsp;6a,b). While we did not observe significant change in IHC-score of CD8 T-cells in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors compared to NRF2\u003csup\u003eLow\u003c/sup\u003e tumors, IHC-CD68 was significantly (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) higher in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors compared to NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;3d,e), indicating that the NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors are tumors-associated macrophages (TAMs), which is generally considered a poor prognostic factor [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. These results agree with our sc-RNA-seq data analysis (Fig.\u0026nbsp;1j) and TCGA data analysis on macrophages, considering CD68 as a general marker of macrophages (Fig.\u0026nbsp;2b). However, survival of TMA-patients with high CD68 versus low CD68 did not show a significant change in both (NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e) groups (data not shown).\u003c/p\u003e \u003cp\u003eAs PD1 and PD-L1 expression within the IMM is generally considered a better prognostic marker compared to PD1/PD-L1 expression in tumors [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e], we analyzed PD1 and PD-L1 expression in both IMM and tumors (Fig.\u0026nbsp;3f,g). PD1 level in IMM was significantly (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/em\u003e) higher than its level on tumor cells (TMRs) in both (NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e) groups (Fig.\u0026nbsp;3f). PD1 protein was more abundantly expressed in the IMM of NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors than that of NRF2\u003csup\u003eLow\u003c/sup\u003e samples (Fig.\u0026nbsp;3f, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Additionally, PD-L1 was significantly (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/em\u003e) enriched in IMM of NRF2\u003csup\u003eHigh\u003c/sup\u003e samples compared to TMRs, while NRF2\u003csup\u003eLow\u003c/sup\u003e samples did not show significant change in PD-L1 levels in IMM versus tumor cells (Fig.\u0026nbsp;3g). Also, the IMM of NRF2\u003csup\u003eHigh\u003c/sup\u003e samples was significantly (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.05\u003c/em\u003e) enriched with PD-L1 compared to IMM of NRF2\u003csup\u003eLow\u003c/sup\u003e samples (Fig.\u0026nbsp;3g).\u003c/p\u003e \u003cp\u003eThe level of FOXP3, a marker of T-regulatory cells (T-reg), in IMM versus TRMs showed no significant changes in both groups (Fig.\u0026nbsp;3h). However, NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors were significantly (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.05\u003c/em\u003e) enriched with FOXP3 compared to NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;3h). High FOXP3 expression in immune cells can be detrimental, while its expression in cancer cells can be more nuanced and may depend on the specific tumor characteristics [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. VIM expressed by tumor cells is often linked to EMT, promoting tumor cell migration and metastasis [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. In both (NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e) groups, IHC-VIM was significantly (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/em\u003e) higher in stroma (STR) than in tumors (Fig.\u0026nbsp;3i). However, STR of NRF2\u003csup\u003eHigh\u003c/sup\u003e HGSOC was significantly (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/em\u003e) enriched with VIM compared to STR of NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC (Fig.\u0026nbsp;3i).\u003c/p\u003e \u003cp\u003eSurvival data of N\u0026thinsp;=\u0026thinsp;210 from a total of 240 HGSOC TMA samples treated with chemotherapies was used to group patients by NRF2 levels; NRF2\u003csup\u003eHigh\u003c/sup\u003e HGSOC had significantly (\u003cem\u003eP\u0026thinsp;=\u0026thinsp;0.00804\u003c/em\u003e) better overall survival than NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC (Fig.\u0026nbsp;3j). Disease specific-survival (DSS) of patients with NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors (n\u0026thinsp;=\u0026thinsp;93) was also significantly (P\u0026thinsp;=\u0026thinsp;0.0133) better than patients with NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (n\u0026thinsp;=\u0026thinsp;104) (Supplementary Fig.\u0026nbsp;4a), but not the progression free-survival (PFS) (Supplementary Fig.\u0026nbsp;4b).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe used 3 cohort studies to investigate if NRF2 activation correlates with distinct immune phenotypes in HGSOC and how that impacts the prognosis. Analysis of scRNA-seq HGSOC samples [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] demonstrated that NRF2\u003csup\u003eLow\u003c/sup\u003e tumors were enriched with T-cells compared to NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors which were enriched with myeloid cells instead (Fig.\u0026nbsp;1j). This is supported by analyzing the bulk transcriptomic data of TCGA HGSOC cohort (n\u0026thinsp;=\u0026thinsp;365), which showed NRF2\u003csup\u003ehigh\u003c/sup\u003e tumors are enriched with macrophages, the most abundant type of myeloid cells in tumors [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e], compared to NRF2\u003csup\u003elow\u003c/sup\u003e tumors (Fig.\u0026nbsp;1a,b). Furthermore, IHC analysis of a TMA composing of 240 human HGSOC samples [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] showed NRF2\u003csup\u003ehigh\u003c/sup\u003e tumors are enriched with CD68 ( a general macrophage marker) compared to NRF2\u003csup\u003elow\u003c/sup\u003e tumors (Fig.\u0026nbsp;3a-d). Therefore, our study suggests that NRF2\u003csup\u003ehigh\u003c/sup\u003e HGSOC is enriched with tumor-associated macrophages (TAMs) compared to NRF2\u003csup\u003eLow\u003c/sup\u003e tumors. Notably, HGSOC is known to have abundant TAMs, which was associated with tumor aggressiveness and poor clinical outcomes [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. Our results \u003cem\u003esuggest that TAMs in HGSOC could be associated with NRF2 activation.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eOur Immune phenotyping analysis revealed the potential value of combined NRF2 activation and immune markers in HGSOCs [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. TCGA data analysis showed patients with NRF2\u003csup\u003eLow\u003c/sup\u003e tumors that enriched with immune stimulatory markers such as CD3E, and CD80 (a M1-macrophages marker) tend to have better survival than those with low levels of these markers (Fig.\u0026nbsp;2e). These immune stimulatory markers did not show a prognostic significance in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors. Instead, CD68, an immune suppressive immune marker that was enriched in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors from all cohort studies showed a significant correlation with prognosis (PFI: p\u0026thinsp;=\u0026thinsp;0.038) of patients from TCGA, where the higher CD68 expression, the worst survival in NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors was observed. This finding is concordant with a previous study showing that TAMs in ovarian cancer is generally associated with resistance to chemotherapy [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. Interestingly, although results from the TMA (n\u0026thinsp;=\u0026thinsp;240) samples showed NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors are enriched with CD68 (Fig.\u0026nbsp;3d), no prognostic value of CD68 expression levels was observed in these patients (data not shown). Instead, patients with NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors had significantly better survival than NRF2\u003csup\u003eLow\u003c/sup\u003e tumors (Fig.\u0026nbsp;3j). These tumors were treated with chemotherapy after surgery. Chemotherapy may induce the differentiation of macrophages towards M1-phenotype with anti-tumor characteristics [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e] which could improve the responses to chemotherapy. Also, chemotherapy could activate NRF2 signaling in ovarian cancer [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e] which could modulate the phenotype of macrophages supporting the M1/M2 polarization [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e]. \u003cem\u003eWhether chemotherapy changed the NRF2 expression levels in these tumors, and if these changes influenced the M1/M2-macrophages polarization is not known\u003c/em\u003e. \u003cem\u003eFuture studies may IHC-stain HGSOC tumors for NRF2, M1-macrophage markers such as CD80, and M2-macrophage markers such as CD163 and CD206, before and after treatment to predict prognosis.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eResults from scRNA-seq data analysis showed higher ICs levels including PD1, PD-L1, and CTLA4 in (n\u0026thinsp;=\u0026thinsp;3) NRF2\u003csup\u003eLow\u003c/sup\u003e tumors compared to NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors (Fig.\u0026nbsp;1l), while the TCGA data showed NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors enriched with PD1 (Supplementary Fig.\u0026nbsp;3b). The TMA patient samples showed the immune-microenvironment (IMM) of NRF2\u003csup\u003eHigh\u003c/sup\u003e samples were significantly enriched with ICs including PD1 (\u003cem\u003ep\u0026thinsp;\u0026lt;\u0026thinsp;0.01\u003c/em\u003e) and PD-L1 (\u003cem\u003ep\u0026thinsp;\u0026lt;\u0026thinsp;0.05\u003c/em\u003e) compared to NRF2\u003csup\u003eLow\u003c/sup\u003e samples (Fig.\u0026nbsp;3f,g). TMA data also showed that the PD-L1 was significantly (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) enriched in IMM compared to its levels in (tumors) TMR in NRF2\u003csup\u003eHigh\u003c/sup\u003e samples which we did not observe in the NRF2\u003csup\u003eLow\u003c/sup\u003e samples (Fig.\u0026nbsp;3f,g). Previously published reports showed the NRF2 activation supports the expression of ICs [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Patients with PD-L1\u003csup\u003eHigh\u003c/sup\u003e-immune cells tend to have better survival compared to patients with PD-L1\u003csup\u003eHigh\u003c/sup\u003e-tumor cells [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. Further, a high level of PD-L1 expression within the IMM of HGSOC tumor is often associated with a better response to chemotherapy [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. \u003cem\u003eIn conclusion, ICs could be enriched in both NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eLow\u003c/em\u003e\u003c/sup\u003e \u003cem\u003eand NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eHigh\u003c/em\u003e\u003c/sup\u003e \u003cem\u003eHGSOC, however, CTLA4 seems to be significant in the NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eLow\u003c/em\u003e\u003c/sup\u003e \u003cem\u003etumors with prognostic benefits (Figs.\u0026nbsp;1l, and 2e). Instead PD1/PD-L1 could be more important in NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eHigh\u003c/em\u003e\u003c/sup\u003e \u003cem\u003etumors and might impact the prognosis (Fig.\u0026nbsp;3f,g and supplementary Fig.\u0026nbsp;3b). The spatial distribution of these markers could also define the prognosis.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eStudies have shown that the mechanisms of NRF2 activation may impact cancer phenotype [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Data analysis of scRNA-seq from the 7 human HGSOC tumors showed NQO1, a key marker of NRF2, was expressed in NRF2\u003csup\u003eHigh\u003c/sup\u003e samples (Fig.\u0026nbsp;1b-d), suggesting the activation of NRF2 signaling through a canonical pathway in these samples [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e]. These samples showed lower expression of immune markers compared to NRF2\u003csup\u003eLow\u003c/sup\u003e samples (Fig.\u0026nbsp;1j) which is expected as anti-inflammatory pathway, normally inhibits inflammation [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In contrast, our mutation profiling of data from RNA-seq of human HGSOC samples from TCGA [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], show no somatic mutations in KEAP1-NRF2 pathway (Supplementary Fig.\u0026nbsp;2), with no NQO1 expression (Supplementary Table\u0026nbsp;2) suggesting that NRF2 was activated through non-canonical pathways in these samples [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. The NRF2\u003csup\u003eHigh\u003c/sup\u003e tumor samples from the TCGA show high (immunosuppressive markers) ISMs (Fig.\u0026nbsp;2a-d). Studies have shown that certain non-canonical pathways of NRF2 activation might attenuate its anti-inflammatory action [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. \u003cem\u003eWhether NRF2 activation pathway (canonical vs non canonical) impacts the immune phenotype and responses to therapies in HGSOC, still need to be determined.\u003c/em\u003e Future studies may investigate two factors before deciding the type of therapy in HGSOC \u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003e1)\u003c/span\u003e \u003cem\u003egenomic evaluation of NRF2 pathway in NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eHigh\u003c/em\u003e\u003c/sup\u003e \u003cem\u003eHGSOC\u003c/em\u003e, and \u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003e2)\u003c/span\u003e \u003cem\u003eIHC evaluation of immune markers in NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eHigh\u003c/em\u003e\u003c/sup\u003e \u003cem\u003e\u0026amp; NRF2\u003c/em\u003e\u003csup\u003e\u003cem\u003eLow\u003c/em\u003e\u003c/sup\u003e \u003cem\u003eHGSOC.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eOur study has several limitations: 1) Bulk RNA-seq had no data on the type of treatment, and thus we were unable to identify the effect of treatment on immunophenotype of NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e samples and if that impacts the survival; 2) The TMA, used in this study had no data on tumors before treatment; also 3) We did not have NRF2 genomic data of these samples.\u003c/p\u003e \u003cp\u003eTaken together, our results show that HGSOC can be classified clinically based on NRF2 expression levels. Clinical studies have shown that HGSOC can be treated based on its immune phenotypes, meaning that different treatment strategies could be tailored to a specific immune profile of a patient's tumor, with some tumors responding better to immunotherapy depending on the presence or absence of immune infiltrating cells and markers within the tumor microenvironment. Our data analyses of three cohort studies added a new marker (NRF2) that impacts the immune phenotype and prognosis of HGSOC. However, treatment with chemotherapy and or immunotherapy could change the expression levels of NRF2 and thus the responses to treatment. Future studies will investigate the status of NRF2 signaling, and mechanisms of its activation in HGSOC using IHC and WGS before and after treatment to understand the timing and their effects on tumor heterogeneity and immune phenotype. This approach will further stratify HGSOC patients for personalized treatment and reveal the best treatment-based tumor phenotype to avoid resistance to therapies for better prognosis.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e \u003cb\u003eSingle cell RNA-seq data analysis of human HGSOC.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eSeven human HGSOC samples from the GSE184880 [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] dataset was downloaded from The National Center for Biotechnology Information website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Data processing and core analysis were performed using Seurat v4.9.9 [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e]. More details can be found in the supplementary methods file.\u003c/p\u003e\n\u003ch3\u003eData Analysis of human HGSOC RNA-seq from TCGA\u003c/h3\u003e\n\u003cp\u003eHuman HGSOC samples (n\u0026thinsp;=\u0026thinsp;365) from The Cancer Genome Atlas (TCGA) were used to analyze the mutational profile, NRF2 expression levels, immune phenotype, differentially expressed genes and survival. Details can be found in the supplementary methods file.\u003c/p\u003e\n\u003ch3\u003eTissue processing, Hematoxylin \u0026 Eosin (H\u0026E), and immunohistochemistry (IHC) labeling of human tissue microarray (TMA)\u003c/h3\u003e\n\u003cp\u003eStaining of human tissue microarray (TMA) with Hematoxylin \u0026amp; Eosin (H\u0026amp;E) and labeling with immunohistochemistry (IHC) were performed at Dr. Yemin\u0026rsquo;s laboratory at the Department of Pathology at Columbia Cancer Center, Vancouver, BC, Canada. Details are found in the supplementary methods file.\u003c/p\u003e\n\u003ch3\u003eData availability\u003c/h3\u003e\n\u003cp\u003eA normalized ovarian cancer transcriptomic data used in this study can be obtained from UCSC Xena (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://xena.ucsc.edu/cite-us\u003c/span\u003e\u003cspan address=\"https://xena.ucsc.edu/cite-us\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Sample-specific information was obtained from \u003cem\u003eVerhaak et al.\u003c/em\u003e [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], and TCGA Biolinks database [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. Organizational scripts are available upon request.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003eWe extensively used publicly available algorithms/methods. Additional code used in this study is available upon request.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eWe used GraphPad Prism 10 software (GraphPad Prism, RRID:SCR_002798, San Diego, CA, USA) to present the data. Fisher Exact Test and Gehan-Breslow-Wilcoxon tests were used to identify the differences between the NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e HGSOC tumor samples. A p value less than 0.05 (typically\u0026thinsp;\u0026le;\u0026thinsp;0.05) was considered statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflict of interest\u003c/h2\u003e \u003cp\u003eNo conflict of interest to report.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eSHH: conceptualization, experimental design, methodology \u0026amp; implementation, data analysis, supervision, funding support, and manuscript writing, review and/revision; HT, CK, KM, NB, GZ, HS, LF, SL, MK, HK, OA, RM, LK, CC, DW, and YW: data analysis and interpretation; YW: experimental design, methodology \u0026amp; implementation, supervision, funding support, and manuscript review and/revision. All authors reviewed and approved the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgment\u003c/h2\u003e \u003cp\u003eWe thank staffs of the MAPCore at the University of British Columbia, BASIC lab of the BC Cancer Research Institute and the Anatomical Pathology department of BC Cancer Agency for performing histology and immunohistochemistry staining. This study was supported by research funds from Cooper University Health Care, Surgery Department at Cooper Medical School of Rowan University, MD Anderson Cancer Center at Cooper, Camden, NJ, Barbara T Foundation, the Canadian Institutes of Health Research (PJT-178179; to Y.W.), Ovarian Cancer Research Alliance (ECIG-2025-3-2014, to Y.W.), and Michael Smith Health Research BC postdoctoral fellowship to L.F. We also appreciate the generous support from the VGH/UBC Hospital Foundation to the Ovarian Cancer Research Centre.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSiegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. \u003cem\u003eCA: A Cancer Journal for Clinicians\u003c/em\u003e 2024; 74: 12\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCluggage WG. Morphological subtypes of ovarian carcinoma: a review with emphasis on new developments and pathogenesis. \u003cem\u003ePathology\u003c/em\u003e 2011; 43: 420\u0026ndash;432.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBowtell DD, B\u0026ouml;hm S, Ahmed AA, Aspuria PJ, Bast RC, Jr., Beral V \u003cem\u003eet al\u003c/em\u003e. Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer. \u003cem\u003eNat Rev Cancer\u003c/em\u003e 2015; 15: 668\u0026ndash;679.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVang R, Shih I-M, Kurman RJ. Ovarian Low-grade and High-grade Serous Carcinoma: Pathogenesis, Clinicopathologic and Molecular Biologic Features, and Diagnostic Problems. \u003cem\u003eAdvances in Anatomic Pathology\u003c/em\u003e 2009; 16: 267\u0026ndash;282.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVaughan S, Coward JI, Bast RC, Berchuck A, Berek JS, Brenton JD \u003cem\u003eet al\u003c/em\u003e. Rethinking ovarian cancer: recommendations for improving outcomes. \u003cem\u003eNature Reviews Cancer\u003c/em\u003e 2011; 11: 719\u0026ndash;725.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTavares V, Marques IS, Melo IG, Assis J, Pereira D, Medeiros R. Paradigm Shift: A Comprehensive Review of Ovarian Cancer Management in an Era of Advancements. \u003cem\u003eInt J Mol Sci\u003c/em\u003e 2024; 25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLisio MA, Fu L, Goyeneche A, Gao ZH, Telleria C. High-Grade Serous Ovarian Cancer: Basic Sciences, Clinical and Therapeutic Standpoints. \u003cem\u003eInt J Mol Sci\u003c/em\u003e 2019; 20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerhaak RG, Tamayo P, Yang JY, Hubbard D, Zhang H, Creighton CJ \u003cem\u003eet al\u003c/em\u003e. Prognostically relevant gene signatures of high-grade serous ovarian carcinoma. \u003cem\u003eJ Clin Invest\u003c/em\u003e 2013; 123: 517\u0026ndash;525.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen GM, Kannan L, Geistlinger L, Kofia V, Safikhani Z, Gendoo DMA \u003cem\u003eet al\u003c/em\u003e. Consensus on Molecular Subtypes of High-Grade Serous Ovarian Carcinoma. \u003cem\u003eClin Cancer Res\u003c/em\u003e 2018; 24: 5037\u0026ndash;5047.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang AW, McPherson A, Milne K, Kroeger DR, Hamilton PT, Miranda A \u003cem\u003eet al\u003c/em\u003e. Interfaces of Malignant and Immunologic Clonal Dynamics in Ovarian Cancer. \u003cem\u003eCell\u003c/em\u003e 2018; 173: 1755\u0026ndash;1769.e1722.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwede M, Waldron L, Mok SC, Wei W, Basunia A, Merritt MA \u003cem\u003eet al\u003c/em\u003e. The Impact of Stroma Admixture on Molecular Subtypes and Prognostic Gene Signatures in Serous Ovarian Cancer. \u003cem\u003eCancer Epidemiol Biomarkers Prev\u003c/em\u003e 2020; 29: 509\u0026ndash;519.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKommoss S, Winterhoff B, Oberg AL, Konecny GE, Wang C, Riska SM \u003cem\u003eet al\u003c/em\u003e. Bevacizumab May Differentially Improve Ovarian Cancer Outcome in Patients with Proliferative and Mesenchymal Molecular Subtypes. \u003cem\u003eClin Cancer Res\u003c/em\u003e 2017; 23: 3794\u0026ndash;3801.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBryan HK, Olayanju A, Goldring CE, Park BK. The Nrf2 cell defence pathway: Keap1-dependent and -independent mechanisms of regulation. \u003cem\u003eBiochemical pharmacology\u003c/em\u003e 2013; 85: 705\u0026ndash;717.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaguchi K, Motohashi H, Yamamoto M. Molecular mechanisms of the Keap1-Nrf2 pathway in stress response and cancer evolution. \u003cem\u003eGenes to cells: devoted to molecular \u0026amp; cellular mechanisms\u003c/em\u003e 2011; 16: 123\u0026ndash;140.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCloer EW, Goldfarb D, Schrank TP, Weissman BE, Major MB. NRF2 Activation in Cancer: From DNA to Protein. \u003cem\u003eCancer Res\u003c/em\u003e 2019; 79: 889\u0026ndash;898.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRojo de la Vega M, Chapman E, Zhang DD. NRF2 and the Hallmarks of Cancer. \u003cem\u003eCancer cell\u003c/em\u003e 2018; 34: 21\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaha S, Buttari B, Panieri E, Profumo E, Saso L. An Overview of Nrf2 Signaling Pathway and Its Role in Inflammation. \u003cem\u003eMolecules\u003c/em\u003e 2020; 25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu B, Tang L, Chen S, Yin C, Peng S, Li X \u003cem\u003eet al\u003c/em\u003e. Targeting the upstream transcriptional activator of PD-L1 as an alternative strategy in melanoma therapy. \u003cem\u003eOncogene\u003c/em\u003e 2018; 37: 4941\u0026ndash;4954.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSha LK, Sha W, Kuchler L, Daiber A, Giegerich AK, Weigert A \u003cem\u003eet al\u003c/em\u003e. Loss of Nrf2 in bone marrow-derived macrophages impairs antigen-driven CD8(+) T cell function by limiting GSH and Cys availability. \u003cem\u003eFree Radic Biol Med\u003c/em\u003e 2015; 83: 77\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu X, Yang Y, Liu X, Cao N, Zhang P, Zhao S \u003cem\u003eet al\u003c/em\u003e. NFE2L2/KEAP1 Mutations Correlate with Higher Tumor Mutational Burden Value/PD-L1 Expression and Potentiate Improved Clinical Outcome with Immunotherapy. \u003cem\u003eOncologist\u003c/em\u003e 2020; 25: e955-e963.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGandhi L, Rodr\u0026iacute;guez-Abreu D, Gadgeel S, Esteban E, Felip E, De Angelis F \u003cem\u003eet al\u003c/em\u003e. Pembrolizumab plus Chemotherapy in Metastatic Non-Small-Cell Lung Cancer. \u003cem\u003eN Engl J Med\u003c/em\u003e 2018; 378: 2078\u0026ndash;2092.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuan J, Zhang Y, Chen R, Liang L, Huo Y, Lu S \u003cem\u003eet al\u003c/em\u003e. Tumor-immune microenvironment and NRF2 associate with clinical efficacy of PD-1 blockade combined with chemotherapy in lung squamous cell carcinoma. \u003cem\u003eCell Rep Med\u003c/em\u003e 2023; 4: 101302.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShibata T, Ohta T, Tong KI, Kokubu A, Odogawa R, Tsuta K \u003cem\u003eet al\u003c/em\u003e. Cancer related mutations in NRF2 impair its recognition by Keap1-Cul3 E3 ligase and promote malignancy. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 2008; 105: 13568\u0026ndash;13573.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKansanen E, Kuosmanen SM, Leinonen H, Levonen AL. The Keap1-Nrf2 pathway: Mechanisms of activation and dysregulation in cancer. \u003cem\u003eRedox Biol\u003c/em\u003e 2013; 1: 45\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKonstantinopoulos PA, Spentzos D, Fountzilas E, Francoeur N, Sanisetty S, Grammatikos AP \u003cem\u003eet al\u003c/em\u003e. Keap1 mutations and Nrf2 pathway activation in epithelial ovarian cancer. \u003cem\u003eCancer Res\u003c/em\u003e 2011; 71: 5081\u0026ndash;5089.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu J, Fang Y, Chen K, Li S, Tang S, Ren Y \u003cem\u003eet al\u003c/em\u003e. Single-Cell RNA Sequencing Reveals the Tissue Architecture in Human High-Grade Serous Ovarian Cancer. \u003cem\u003eClinical Cancer Research\u003c/em\u003e 2022; 28: 3590\u0026ndash;3602.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoldman MJ, Craft B, Hastie M, Repečka K, McDade F, Kamath A \u003cem\u003eet al\u003c/em\u003e. Visualizing and interpreting cancer genomics data via the Xena platform. \u003cem\u003eNature Biotechnology\u003c/em\u003e 2020; 38: 675\u0026ndash;678.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrassard J, Hughes MR, Dean P, Hernaez DC, Thornton S, Banville AC \u003cem\u003eet al\u003c/em\u003e. A tumor-restricted glycoform of podocalyxin is a highly selective marker of immunologically cold high-grade serous ovarian carcinoma. \u003cem\u003eFrontiers in oncology\u003c/em\u003e 2023; 13: 1286754.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalhouk A, George J, Wang C, Budden T, Tan TZ, Chiu DS \u003cem\u003eet al\u003c/em\u003e. Development and Validation of the Gene Expression Predictor of High-grade Serous Ovarian Carcinoma Molecular SubTYPE (PrOTYPE). \u003cem\u003eClin Cancer Res\u003c/em\u003e 2020; 26: 5411\u0026ndash;5423.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng J, Read OJ, Dinkova-Kostova AT. Nrf2 in TIME: The Emerging Role of Nuclear Factor Erythroid 2-Related Factor 2 in the Tumor Immune Microenvironment. \u003cem\u003eMol Cells\u003c/em\u003e 2023; 46: 142\u0026ndash;152.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCamp ND, James RG, Dawson DW, Yan F, Davison JM, Houck SA \u003cem\u003eet al\u003c/em\u003e. Wilms tumor gene on X chromosome (WTX) inhibits degradation of NRF2 protein through competitive binding to KEAP1 protein. \u003cem\u003eJ Biol Chem\u003c/em\u003e 2012; 287: 6539\u0026ndash;6550.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoss D, Siegel D. The diverse functionality of NQO1 and its roles in redox control. \u003cem\u003eRedox Biology\u003c/em\u003e 2021; 41: 101950.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSato E, Olson SH, Ahn J, Bundy B, Nishikawa H, Qian F \u003cem\u003eet al\u003c/em\u003e. Intraepithelial CD8\u0026thinsp;+\u0026thinsp;tumor-infiltrating lymphocytes and a high CD8+/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e 2005; 102: 18538\u0026ndash;18543.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlinge U, Dievernich A, Tolba R, Klosterhalfen B, Davies L. CD68\u0026thinsp;+\u0026thinsp;macrophages as crucial components of the foreign body reaction demonstrate an unconventional pattern of functional markers quantified by analysis with double fluorescence staining. \u003cem\u003eJ Biomed Mater Res B Appl Biomater\u003c/em\u003e 2020; 108: 3134\u0026ndash;3146.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoeini P, Niedźwiedzka-Rystwej P. Tumor-Associated Macrophages: Combination of Therapies, the Approach to Improve Cancer Treatment. \u003cem\u003eInt J Mol Sci\u003c/em\u003e 2021; 22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao Z, Zhang J, Zhang B, Liang G, Chen X, Yao F \u003cem\u003eet al\u003c/em\u003e. Imatinib prevents lung cancer metastasis by inhibiting M2-like polarization of macrophages. \u003cem\u003ePharmacol Res\u003c/em\u003e 2018; 133: 121\u0026ndash;131.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y, Zheng J, Islam MS, Yang Y, Hu Y, Chen X. The role of CD4(+)FoxP3(+) regulatory T cells in the immunopathogenesis of COVID-19: implications for treatment. \u003cem\u003eInt J Biol Sci\u003c/em\u003e 2021; 17: 1507\u0026ndash;1520.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlemm P, Rajendiran A, Fragoulis A, Wruck C, Schippers A, Wagner N \u003cem\u003eet al\u003c/em\u003e. Nrf2 expression driven by Foxp3 specific deletion of Keap1 results in loss of immune tolerance in mice. \u003cem\u003eEur J Immunol\u003c/em\u003e 2020; 50: 515\u0026ndash;524.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMocellin S, Marincola FM, Young HA. Interleukin-10 and the immune response against cancer: a counterpoint. \u003cem\u003eJournal of leukocyte biology\u003c/em\u003e 2005; 78: 1043\u0026ndash;1051.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHazini A, Fisher K, Seymour L. Deregulation of HLA-I in cancer and its central importance for immunotherapy. \u003cem\u003eJournal for immunotherapy of cancer\u003c/em\u003e 2021; 9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi Z, Yang WM, Chen LP, Yang DH, Zhou Q, Zhu J \u003cem\u003eet al\u003c/em\u003e. Enhanced chemosensitization in multidrug-resistant human breast cancer cells by inhibition of IL-6 and IL-8 production. \u003cem\u003eBreast Cancer Res Treat\u003c/em\u003e 2012; 135: 737\u0026ndash;747.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartinez VD, Vucic EA, Thu KL, Pikor LA, Hubaux R, Lam WL. Unique pattern of component gene disruption in the NRF2 inhibitor KEAP1/CUL3/RBX1 E3-ubiquitin ligase complex in serous ovarian cancer. \u003cem\u003eBiomed Res Int\u003c/em\u003e 2014; 2014: 159459.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilva-Islas CA, Maldonado PD. Canonical and non-canonical mechanisms of Nrf2 activation. \u003cem\u003ePharmacol Res\u003c/em\u003e 2018; 134: 92\u0026ndash;99.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDodson M, Zhang DD. Non-canonical activation of NRF2: New insights and its relevance to disease. \u003cem\u003eCurr Pathobiol Rep\u003c/em\u003e 2017; 5: 171\u0026ndash;176.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSporn MB, Liby KT. NRF2 and cancer: the good, the bad and the importance of context. \u003cem\u003eNat Rev Cancer\u003c/em\u003e 2012; 12: 564\u0026ndash;571.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhattacharya S, Dunn P, Thomas CG, Smith B, Schaefer H, Chen J \u003cem\u003eet al\u003c/em\u003e. ImmPort, toward repurposing of open access immunological assay data for translational and clinical research. \u003cem\u003eScientific Data\u003c/em\u003e 2018; 5: 180015.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreuer K, Foroushani AK, Laird MR, Chen C, Sribnaia A, Lo R \u003cem\u003eet al\u003c/em\u003e. InnateDB: systems biology of innate immunity and beyond\u0026ndash;recent updates and continuing curation. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 2013; 41: D1228-1233.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmed SM, Luo L, Namani A, Wang XJ, Tang X. Nrf2 signaling pathway: Pivotal roles in inflammation. \u003cem\u003eBiochimica et biophysica acta Molecular basis of disease\u003c/em\u003e 2017; 1863: 585\u0026ndash;597.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEllyard JI, Simson L, Parish CR. Th2-mediated anti-tumour immunity: friend or foe? \u003cem\u003eTissue Antigens\u003c/em\u003e 2007; 70: 1\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZamarron BF, Chen W. Dual roles of immune cells and their factors in cancer development and progression. \u003cem\u003eInt J Biol Sci\u003c/em\u003e 2011; 7: 651\u0026ndash;658.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarques HS, de Brito BB, da Silva FAF, Santos MLC, de Souza JCB, Correia TML \u003cem\u003eet al\u003c/em\u003e. Relationship between Th17 immune response and cancer. \u003cem\u003eWorld J Clin Oncol\u003c/em\u003e 2021; 12: 845\u0026ndash;867.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTie Y, Tang F, Wei Y-q, Wei X-w. Immunosuppressive cells in cancer: mechanisms and potential therapeutic targets. \u003cem\u003eJournal of Hematology \u0026amp; Oncology\u003c/em\u003e 2022; 15: 61.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng H, Wang Z, Fu L, Xu T. Macrophage Polarization in the Development and Progression of Ovarian Cancers: An Overview. \u003cem\u003eFrontiers in oncology\u003c/em\u003e 2019; 9: 421.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGottlieb CE, Mills AM, Cross JV, Ring KL. Tumor-associated macrophage expression of PD-L1 in implants of high grade serous ovarian carcinoma: A comparison of matched primary and metastatic tumors. \u003cem\u003eGynecol Oncol\u003c/em\u003e 2017; 144: 607\u0026ndash;612.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan Y, Liu D, Li L. PD-1/PD-L1 pathway: current researches in cancer. \u003cem\u003eAm J Cancer Res\u003c/em\u003e 2020; 10: 727\u0026ndash;742.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlanc-Durand F, Clemence Wei Xian L, Tan DSP. Targeting the immune microenvironment for ovarian cancer therapy. \u003cem\u003eFront Immunol\u003c/em\u003e 2023; 14: 1328651.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNgan CY, Yamamoto H, Seshimo I, Tsujino T, Man-i M, Ikeda JI \u003cem\u003eet al\u003c/em\u003e. Quantitative evaluation of vimentin expression in tumour stroma of colorectal cancer. \u003cem\u003eBritish journal of cancer\u003c/em\u003e 2007; 96: 986\u0026ndash;992.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Vlerken-Ysla L, Tyurina YY, Kagan VE, Gabrilovich DI. Functional states of myeloid cells in cancer. \u003cem\u003eCancer cell\u003c/em\u003e 2023; 41: 490\u0026ndash;504.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEl-Arabey AA, Alkhalil SS, Al-Shouli ST, Awadalla ME, Alhamdi HW, Almanaa TN \u003cem\u003eet al\u003c/em\u003e. Revisiting macrophages in ovarian cancer microenvironment: development, function and interaction. \u003cem\u003eMed Oncol\u003c/em\u003e 2023; 40: 142.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026ouml;sm\u0026uuml;ller HC, Wagner P, Peper JK, Schuster H, Pham DL, Greif K \u003cem\u003eet al\u003c/em\u003e. Combined Immunoscore of CD103 and CD3 Identifies Long-Term Survivors in High-Grade Serous Ovarian Cancer. \u003cem\u003eInt J Gynecol Cancer\u003c/em\u003e 2016; 26: 671\u0026ndash;679.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTossetta G, Fantone S, Montanari E, Marzioni D, Goteri G. Role of NRF2 in Ovarian Cancer. \u003cem\u003eAntioxidants (Basel)\u003c/em\u003e 2022; 11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang L, He C. Nrf2-mediated anti-inflammatory polarization of macrophages as therapeutic targets for osteoarthritis. \u003cem\u003eFront Immunol\u003c/em\u003e 2022; 13: 967193.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMocan LP, Craciun R, Grapa C, Melincovici CS, Rusu I, Al Hajjar N \u003cem\u003eet al\u003c/em\u003e. PD-L1 expression on immune cells, but not on tumor cells, is a favorable prognostic factor for patients with intrahepatic cholangiocarcinoma. \u003cem\u003eCancer Immunol Immunother\u003c/em\u003e 2023; 72: 1003\u0026ndash;1014.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVanguri R, Benhamida J, Young JH, Li Y, Zivanovic O, Chi D \u003cem\u003eet al\u003c/em\u003e. Understanding the impact of chemotherapy on the immune landscape of high-grade serous ovarian cancer. \u003cem\u003eGynecol Oncol Rep\u003c/em\u003e 2022; 39: 100926.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMutter FE, Park BK, Copple IM. Value of monitoring Nrf2 activity for the detection of chemical and oxidative stress. \u003cem\u003eBiochem Soc Trans\u003c/em\u003e 2015; 43: 657\u0026ndash;662.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIzumi Y, Tatsumoto A, Horiuchi N, Arifuku M, Uegomori M, Kume T \u003cem\u003eet al\u003c/em\u003e. TPNA10168, an Nrf-2 activator, attenuates inflammatory responses independently of Nrf2 in microglial BV-2 cells: Involvement of the extracellular-signal-regulated kinase pathway. \u003cem\u003eJ Pharmacol Sci\u003c/em\u003e 2022; 149: 1\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao Y, Hao S, Andersen-Nissen E, Mauck WM, 3rd, Zheng S, Butler A \u003cem\u003eet al\u003c/em\u003e. Integrated analysis of multimodal single-cell data. \u003cem\u003eCell\u003c/em\u003e 2021; 184: 3573\u0026ndash;3587.e3529.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D \u003cem\u003eet al\u003c/em\u003e. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 2016; 44: e71.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"genes-and-immunity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"genes","sideBox":"Learn more about [Genes \u0026 Immunity](http://www.nature.com/gene/)","snPcode":"41435","submissionUrl":"https://mts-gene.nature.com/cgi-bin/main.plex","title":"Genes \u0026 Immunity","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"NRF2 pathway, high grade serous ovarian cancer, immune markers, survival ","lastPublishedDoi":"10.21203/rs.3.rs-6925656/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6925656/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e \u003cem\u003eNRF2\u003c/em\u003e modulates tumor immune microenvironment (IMM) in several cancers. NRF2 is activated in about 50% of high grade serous ovarian cancer (HGSOC), the most aggressive type of ovarian cancer. This study aimed to stratify HGSOC patients\u0026rsquo; samples by NRF2 levels and identify its impact on immune phenotype and prognosis. We analyzed data from n\u0026thinsp;=\u0026thinsp;7 scRNA-seq, n\u0026thinsp;=\u0026thinsp;365 RNA-seq of human HGSOC samples, and n\u0026thinsp;=\u0026thinsp;240 HGSOC samples from a tumor microarray (TMA). Results showed human HGSOC samples can be classified by NRF2\u003csup\u003eHigh\u003c/sup\u003e and NRF2\u003csup\u003eLow\u003c/sup\u003e tumors. RNA-seq data analysis along with IHC labeling showed that NRF2\u003csup\u003eHigh\u003c/sup\u003e HGSOCs are enriched with hallmarks of immune suppressive markers (ISMs). Specifically, NRF2\u003csup\u003eHigh\u003c/sup\u003e tumors are identified as tumors associated macrophages (TAMs) with worst survival (p\u0026thinsp;=\u0026thinsp;0.038) was observed in CD68\u003csup\u003eHigh\u003c/sup\u003e tumors. NRF2\u003csup\u003eLow\u003c/sup\u003e tumors were enriched with immune activated markers such as CD3E and CD80 with a prognostic significance. Immune checkpoints (ICs) are important in both groups. However, their levels and spatial distribution are the factors that define their impact on prognosis in these samples. This study is the first that shows classification of HGSOC based on NRF2 levels and suggests IHC-labeling and genomic evaluation of NRF2 and immune markers in HGSOC to predict prognosis.\u003c/p\u003e","manuscriptTitle":"Immune phenotype in high- versus low-NRF2 high grade serous ovarian cancer and the impact on prognosis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-22 08:12:53","doi":"10.21203/rs.3.rs-6925656/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2025-09-18T21:52:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-08-26T20:41:14+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-08-26T04:44:54+00:00","index":3,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-08-21T19:24:58+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-07-15T11:10:08+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-07-14T20:12:26+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2025-07-14T20:01:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-23T14:39:57+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-18T19:37:52+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genes \u0026 Immunity","date":"2025-06-18T19:37:51+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"genes-and-immunity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"genes","sideBox":"Learn more about [Genes \u0026 Immunity](http://www.nature.com/gene/)","snPcode":"41435","submissionUrl":"https://mts-gene.nature.com/cgi-bin/main.plex","title":"Genes \u0026 Immunity","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ae1b5b14-1ee0-46c1-a5cf-106b6aa74b07","owner":[],"postedDate":"July 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":51525377,"name":"Biological sciences/Immunology/Gene regulation in immune cells"},{"id":51525378,"name":"Biological sciences/Genetics/Gene regulation"}],"tags":[],"updatedAt":"2026-05-05T09:52:52+00:00","versionOfRecord":{"articleIdentity":"rs-6925656","link":"https://doi.org/10.1038/s41435-026-00400-7","journal":{"identity":"genes-and-immunity","isVorOnly":false,"title":"Genes \u0026 Immunity"},"publishedOn":"2026-04-28 04:00:00","publishedOnDateReadable":"April 28th, 2026"},"versionCreatedAt":"2025-07-22 08:12:53","video":"","vorDoi":"10.1038/s41435-026-00400-7","vorDoiUrl":"https://doi.org/10.1038/s41435-026-00400-7","workflowStages":[]},"version":"v1","identity":"rs-6925656","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6925656","identity":"rs-6925656","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}