Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Jun 20;4(6):101061.
doi: 10.1016/j.xcrm.2023.101061. Epub 2023 Jun 1.

Profiling the metabolome of uterine fluid for early detection of ovarian cancer

Pan Wang  1 , Jihong Ma  1 , Wenjing Li  1 , Qilong Wang  2 , Yinan Xiao  1 , Yuening Jiang  1 , Xiaoyang Gu  1 , Yu Wu  1 , Suwei Dong  3 , Hongyan Guo  4 , Mo Li  5
Affiliations

Profiling the metabolome of uterine fluid for early detection of ovarian cancer

Pan Wang et al. Cell Rep Med. .

Abstract

Ovarian cancer (OC) causes high mortality in women because of ineffective biomarkers for early diagnosis. Here, we perform metabolomics analysis on an initial training set of uterine fluid from 96 gynecological patients. A seven-metabolite-marker panel consisting of vanillylmandelic acid, norepinephrine, phenylalanine, beta-alanine, tyrosine, 12-S-hydroxy-5,8,10-heptadecatrienoic acid, and crithmumdiol is established for detecting early-stage OC. The panel is further validated in an independent sample set from 123 patients, discriminating early OC from controls with an area under the curve (AUC) of 0.957 (95% confidence interval [CI], 0.894-1). Interestingly, we find elevated norepinephrine and decreased vanillylmandelic acid in most OC cells, resulting from excess 4-hydroxyestradiol that antagonizes the catabolism of norepinephrine by catechol-O-methyltransferase. Moreover, exposure to 4-hydroxyestradiol induces cellular DNA damage and genomic instability that could lead to tumorigenesis. Thus, this study not only reveals metabolic features in uterine fluid of gynecological patients but also establishes a noninvasive approach for the early diagnosis of OC.

Keywords: DNA damage; early diagnosis; metabolome; ovarian cancer; uterine fluid.

PubMed Disclaimer

Conflict of interest statement

Declaration of interests The authors declare the following patent related to this work: M.L., H.G., Y.W., P.W., Y.J., J.M., Y.X., and X.G. are listed as inventors on a patent (C.N. patent application no. ZL 2022 1 1651535.6) filed in December 2022 and granted in March 2023.

Figures

None
Graphical abstract
Figure 1
Figure 1
Metabolic features of uterine fluid in gynecological patients (A) Schematic of study design. Uterine fluid samples were harvested from 219 gynecological patients from two cohorts. The training cohort consisted of 26 patients with stage I–II (early) OC (OVE group), 25 patients with stage III–IV (late) OC (OVL group), 25 patients with benign ovarian diseases (OVBs) for the group of “non-cancer controls,” and 20 patients with uterine corpus endometrial cancer (UCEC) for the group of “differentiation control.” The metabolomic profiling of uterine fluid was performed by RPLC-MS and HILIC-MS followed by normalization and analysis. An independent cohort of 123 gynecological patients (32 OVE, 31 OVL, 30 OVB, and 30 UCEC) was recruited as the validation set. (B) Classes and counts of metabolites detected in uterine fluid samples. 566, 437, and 421 compounds were in total recognized in the modes of RPLC ESI positive, RPLC ESI negative, and HILIC ESI positive. (C) Partial least squares discriminant analysis (PLS-DA) of metabolomics data from different groups of patients in the training set. Red, OVE; yellow, OVL; blue, OVB; green, UCEC. Each colored dot indicates an individual patient. (D) Dendrogram of hierarchical clustering of metabolomics data from different groups of patients in the training set. OVE samples and OVB samples formed tight clusters, while OVL samples and UCEC samples were dispersed across all clusters. (E) Venn diagram of numbers of metabolites in different groups of patients in the training set. 724 metabolites were commonly detected in all the groups of patients. See also Figure S1 and Table S1, S2, and S3.
Figure 2
Figure 2
Biomarker discovery for early detection of OC (A) Orthogonal PLS-DA (OPLS-DA) of uterine fluid metabolomics between patients with OVBs and OVE in the training set. Blue, OVB; red, OVE. (B) The validity of OPLS-DA model in (A) was confirmed by the permutation test (permutation times n = 1,000), showing no overfitting. R2Y measures the goodness of fit, while Q2 measures the predictive ability of the model. (C) Heatmap of uterine fluid metabolomics between patients with OVBs and OVE in the training set. A total of 298 differential metabolites (174 up-regulated and 124 down-regulated) between the two groups were screened. The color schema from blue to red represents the relative abundance of metabolites from low to high, respectively. (D) Representative metabolites that were significantly up-/down-regulated from (C) are shown. Tyrosine, 3-aminocaproic acid, phenylalanine, PC(14:0/20:3(5Z,8Z,11Z)), PC(16:0/16:0), pantetheine, mascaroside, arecoline, and NE were mostly up-regulated, while ethyl-hexanol sulfate, VMA, gibberellin A75, and pipecolic acid were generally down-regulated, in all samples. (E) Pathway analysis of differential metabolites between OVB and OVE groups. The top 10 pathways enriched in the OVE group are shown on the y axis, and the x axis represents -log10 (FDR). Dot size indicates the impact value of each pathway. (F) Pathway analysis of differential metabolites between OVE and UCEC groups. The top 10 pathways enriched in the OVE group are shown. (G) Volcano plot analysis of metabolites involved in tyrosine catabolism pathway between OVB and OVE groups. Metabolites in the OVE group that were 4-fold higher or lower than those in OVB group are cut off by the dashed lines. 6 metabolites that were significantly increased or decreased with FDRs <0.05 are labeled in red and blue, respectively. (H) Levels of the six metabolites in (G) in uterine fluid samples were measured by targeted LC-MS. 15 samples were included for each group. Data are presented as medians and interquartile ranges (IQRs). p1, p2, and p3 values were calculated by unpaired two-tailed Wilcoxon rank-sum tests for the comparison of OVB and OVE, the comparison of OVE and OVL, and the comparison of OVE and UCEC, respectively. p values were calculated for comparing all four groups using multiple comparison tests. See also Figure S2 and Table S4.
Figure 3
Figure 3
Metabolic evolution during OC development (A) OPLS-DA of uterine fluid metabolomics between the OVB and OV groups (the combination of OVE and OVL) in the training set. Blue, OVB; red, OV. (B) Fuzzy c-means clustering of uterine fluid metabolomics during sequential stages (OVB, OVE, and OVL) of OC development in the training set. Different changing trends of metabolites are summarized in clusters 1–6. (C) Heatmap of differential metabolites during sequential stages (OVB, OVE, and OVL) of OC development in the training set extracted from clusters 1 and 4 in (B). The differential metabolites with VIP scores >1.5 and FDRs <0.05 (two-tailed Wilcoxon rank-sum tests) are displayed in the heatmap. The color schema from blue to red represents the relative abundance of metabolites from low to high, respectively. (D) Levels of ten representative compounds in uterine fluid samples were measured by targeted LC-MS. Each staged group contained 20 samples. Relative levels of compounds were summarized. Data are presented as medians and IQRs. The centerline of the boxplots represents the median, the box boundaries represent the IQRs, and whiskers span 1.5-fold the IQRs. p1 and p2 values were calculated by unpaired two-tailed Wilcoxon rank-sum tests for the comparison of OVB and OVE and the comparison of OVE and OVL, respectively. p values were calculated for comparing all three groups using multiple comparison tests. See also Figure S3 and Table S5.
Figure 4
Figure 4
A seven-metabolite panel for early diagnosis of OC (A) ROC analysis of the individual metabolite (phenylalanine, beta-alanine, tyrosine, NE, VMA, 12S-HHT, or crithmumdiol) and single CA125 for differentiating patients with OVE from patients with OVBs in the validation set. The value of AUC for each metabolite is shown. (B) Decision curve analysis of the individual metabolite for differentiating patients with OVE from patients with OVBs in the validation set. The “all” line refers to the net benefit (NB) of treating all women, and the “none” line shows the NB of treating no patients. (C) ROC analysis of the seven-metabolite panel versus CA125 and ROMA for differentiating patients with OVE from patients with OVBs in the validation set (p = 0.004, compared with CA125; p = 0.002, compared with ROMA, Delong test). The values of AUCs and their confidence intervals (CIs) for CA125 and ROMA are shown, respectively. (D) Decision curve analysis of the seven-metabolite panel versus CA125 and ROMA for differentiating patients with OVE from patients with OVBs in the validation set. The “all” line refers to the NB of treating all women, and the “none” line shows the NB of treating no patients. (E) ROC analysis of the seven-metabolite panel combined with CA125 for differentiating patients with OVE from patients with OVBs in the validation set. The values of AUC and AUC CI are 0.969 and 0.899–1, respectively. (F) ROC analysis of the seven-metabolite panel combined with or without CA125 for differentiating patients with OVE from patients with UCEC in the validation set. The values of AUCs and AUC CIs are shown. See also Figure S4.
Figure 5
Figure 5
Increased 4-OHE2 in OC cells disturbs the tyrosine catabolism pathway (A) An illustration of the tyrosine catabolism pathway in which tyrosine is catabolized to diverse intermediates including L-DOPA, dopamine, NE, epinephrine, normetanephrine, 3,4-dihydroxymandelate, MHPG, etc., and to the end product of VMA. These processes are catalyzed by several enzymes, among which COMT participates in multiple steps. MHPG, 3-Methoxy-4-hydroxyphenyl glycolaldehyde. (B) Concentrations of NE and VMA in 12 pairs of para-tumor and tumor tissues were measured with an NE ELISA kit and a VMA ELISA kit, respectively. Data are presented as mean values ± SD. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (C) Expression levels of CYP1B1 between normal ovarian tissues (88 available) and serous OC tissues (426 available) were compared through an integrated analysis of TCGA and GTEx databases. The values of log2 (normalized count) are used to measure the gene expression levels. p values were calculated by unpaired two-tailed Welch’s t test. p <0.05 was considered statistically significant. (D) The protein level of CYP1B1 in 12 pairs of para-tumor and tumor tissues examined by western blot. Beta-actin was used as a loading control. (E) 4-OHE2 in para-tumor tissues and tumor tissues was extracted, and the corresponding levels were measured by targeted LC-MS analysis. Data are presented as mean values ± SD. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (F) Schematic diagram of dual functions of COMT in tyrosine catabolism and 4-OHE2 degradation. (G) In vitro binding affinity of 4-OHE2 or NE to COMT was measured by microscale thermophoresis (MST). F norm = F1/F0 (F norm, normalized fluorescence; F1, fluorescence after thermodiffusion; F0, initial fluorescence or fluorescence after T-jump). KD, dissociation constant. (H) Concentrations of NE and VMA in primary ovarian cancer cells and T1074, SKOV3, and A2780 cells treated with or without 5 μM 4-OHE2 for 12 h. Three biologically independent replicates were performed. Data are presented as mean values ± SD. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant.
Figure 6
Figure 6
The effect of 4-OHE2 on genomic stability (A) The level of AP sites in T1074 cells treated with different time periods of 5 μM 4-OHE2 was measured with an AP site assay kit. The antioxidant NAC (5 mM) was added in the last group. Three biologically independent replicates were performed. Data are presented as mean values ± SD. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (B) γH2AX staining of T1074 cells treated with different time periods of 5 μM 4-OHE2. NAC (5 mM) was added in the last group. Scale bar, 5 μm. The γH2AX foci number in each cell was counted. At least 30 cells from three biologically independent replicates were included for each group. Data are presented as mean values ± SEM. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (C) The level of AP sites in primary cells from 12 pairs of para-tumor and tumor tissues was measured with an AP site assay kit. Three biologically independent replicates were performed for each tissue. Data are presented as mean values ± SD. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (D) γH2AX staining of primary cells isolated from 12 pairs of para-tumor and tumor tissues. Scale bar, 5 μm. The γH2AX foci number in each cell was counted. At least 30 cells from three biologically independent replicates were included for each tissue. Data are presented as mean values ± SEM. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (E) Alkali comet assay of primary cells isolated from 12 pairs of para-tumor and tumor tissues. Tail moments of cells represent data from three independent experiments, with at least 30 cells in each group. Data are presented as mean values ± SEM. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. (F) T1074 cells were cultured with or without 1 μM 4-OHE2 and 1 mM NAC for 15 generations. Genomic DNA of the cells was extracted and subjected to whole-genome sequencing (WGS), followed by CNV analysis with CNV sequencing (CNV-seq). The CNV value of each chromosome (1–22) in different groups is shown as the log2 ratio. Violins denote the status of total DNA, and scatters denote DNA segments with absolution log2 ratios >1.5. (G) T1074 cells were cultured with or without 1 μM 4-OHE2 and 1 mM NAC for 15 generations and were subjected to mitotic spreading analysis. Representative phenotypes of chromosome fragments and breakages from the 4-OHE2-treated group are shown. The percentages of chromosome fragments and breakages in each group are summarized in the histogram. At least 30 cells from three biologically independent replicates were included for each group. Data are presented as mean values ± SEM. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. Scale bar, 10 μm. (H) T1074 cells were cultured with or without 1 μM 4-OHE2 and 1 mM NAC for 15 generations followed by malignant transformation assessment with soft agar colony-formation assay. The colony number was counted from three biologically independent replicates. Data are presented as mean values ± SD. p values were calculated by unpaired two-tailed Student’s t tests. p <0.05 was considered statistically significant. See also Figure S5.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Sung H., Ferlay J., Siegel R.L., Laversanne M., Soerjomataram I., Jemal A., Bray F. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA. Cancer J. Clin. 2021;71:209–249. doi: 10.3322/caac.21660. - DOI - PubMed
    1. Armstrong D.K., Alvarez R.D., Bakkum-Gamez J.N., Barroilhet L., Behbakht K., Berchuck A., Chen L.M., Cristea M., DeRosa M., Eisenhauer E.L., et al. Ovarian cancer, version 2.2020, NCCN clinical practice guidelines in oncology. J. Natl. Compr. Canc. Netw. 2021;19:191–226. doi: 10.6004/jnccn.2021.0007. - DOI - PubMed
    1. Lheureux S., Gourley C., Vergote I., Oza A.M. Epithelial ovarian cancer. Lancet. 2019;393:1240–1253. doi: 10.1016/s0140-6736(18)32552-2. - DOI - PubMed
    1. Torre L.A., Trabert B., DeSantis C.E., Miller K.D., Samimi G., Runowicz C.D., Gaudet M.M., Jemal A., Siegel R.L. Ovarian cancer statistics, 2018. CA. Cancer J. Clin. 2018;68:284–296. doi: 10.3322/caac.21456. - DOI - PMC - PubMed
    1. Lheureux S., Braunstein M., Oza A.M. Epithelial ovarian cancer: evolution of management in the era of precision medicine. CA. Cancer J. Clin. 2019;69:280–304. doi: 10.3322/caac.21559. - DOI - PubMed

Publication types

-