Sohlh1 and Lhx8 are prominent biomarkers to estimate the primordial follicle pool in mice
Reproductive Biology and Endocrinology volume 21, Article number: 46 (2023)
Efficient evaluation of the primordial follicle pool (PFP) of mammalian models is an essential subject in biomedical research relating to ovarian physiology and pathogenesis. Our recent study has identified a gene signature including Sohlh1, Nobox, Lhx8, Tbpl2, Stk31, Padi6, and Vrtn strongly correlated with ovarian reserve by using bioinformatics analysis. Aimed to investigate the validity of these candidate biomarkers for evaluating the PFP, we utilized an OR comparison model to decode the relationship between the numbers of PFP and candidate biomarkers in the present study. Our results suggest that these biomarkers Sohlh1, Nobox, Lhx8, Tbpl2, Stk31, Padi6, and Vrtn possess independent potential to evaluate the number of the PFP. And the combination of Sohlh1 and Lhx8 can be used as the optimal biomarkers for rapid assessment of the PFP in the murine ovary. Our findings provide a new perspective for evaluating the PFP of the ovary in animal studies and the clinic.
It is well-known that the primordial follicle pool in mammals represents the ovarian reserve (OR) that is fixed after birth and not renewed . As time goes on, the primordial follicle pool (PFP) gradually decreases with continued atresia and recruitment until depleting after menopause . The depletion of the PFP accelerates when the ovaries are exposed to chemotherapy, environmental chemicals, and numerous pathological factors, leading to primary ovarian insufficiency (POI).
As the basic functional unit of the ovary, the primordial follicle is a critical subject for studies related to ovarian physiology and pathogenesis [4,5,6]. The measurement of PFP is accomplished based on microscopic recognition and individual counting by the visual observation of researchers on serial tissue Sects. [7,8,9]. The workload for follicle counting is tremendous in animal studies with large sample requirements. Critically, this method is quite subjective, and the number of PFP reported by different teams varies dramatically. It is considered that this variation in the follicular number reported reflects inherent biases of this counting technique . Furthermore, with the application of new assisted reproductive technologies (ART) such as ovarian tissue cryopreservation (OTC), there is a challenge to assess PFP level efficiently and accurately in ovarian tissue slices in clinical practice [10,11,12]. Therefore, an objective, precise, and simplified approach to assessing PFP at the histological level is urgently needed.
High throughput sequencing has advanced dramatically in recent years, providing new approaches for identifying biomarkers . Our recently published study has found a gene signature strongly correlated with OR based on the transcriptomic data of bulk and single-cell RNA-seq. Genes including Sohlh1, Nobox, Lhx8, Tbpl2, Stk31, Padi6, and Vrtn could be potential biomarkers to evaluate the PFP in humans and mice ovarian tissues .
In the present study, we aimed to investigate the validity of these seven candidate biomarkers and identify the optimal combination of biomarkers for evaluating the PFP, hoping to provide a novel and complementary method for the rapid assessment of the PFP.
Materials and methods
Animals and sample preparation
Different aged C57BL/6L female mice with specific pathogen-free conditions were purchased from Vital River Laboratory Animal Technology Corp. The mice were raised in an environment with a temperature of between 18 and 23 °C and humidity of between 40 and 60% under 12-hour light/dark cycles. According to the previous study, bilateral ovariectomy was performed under the microscope according to the previous study . Both ovaries of each mouse were dissected intact, and one side was randomly selected for serial histological analysis, while the entire opposite ovary was subjected to RNA extraction. Three mice were excluded due to failure of sample collecting or RNA extraction. Animal experiments were approved by the Animal Ethics Committee of Shanghai Tenth People’s Hospital (No. SHDSYY-2021-Y0688).
Quantitative real-time PCR
The total RNA extraction was completed with RNAiso Plus (TaKaRa, Shiga, Japan). RNA quality and quantity were measured by NanoDrop2000 (Thermo Scientific, Wilmington, DE, USA). The RNA was reverse-transcribed into complementary DNA using PrimeScript RTMaster Mix (TaKaRa, Japan) according to the manufacturer’s instructions. Quantitative real-time PCRs (qRT-PCR) were performed on QuantStudio Dx (ABI, America) with SYBR Premix ExTaq kit (Takara, Shiga, Japan). The thermal cycler conditions were as follows: 30 s at 95.0 °C for cDNA denatured, followed by 40 cycles of 15 s at 95 ̊C and 60 ̊C for the 34 s. Verification of specific product amplification was performed by dissociation curve analysis. The mRNA relative expression was calculated by the 2^−ΔCt method with Gapdh used as an internal control [16, 17].Primer sequences for each gene were listed in Supplementary Table 1.
Histological analysis and follicle counting
Ovaries were fixed in 4% buffered paraformaldehyde for one day, followed by paraffin embedding. Each murine ovary was serially sectioned at 5-µm thickness and mounted on glass slides. The slides were stained with hematoxylin and eosin (Merck, Darmstadt, Germany) and analyzed for morphological grading and follicle density count. To avoid miscounting, only follicles surrounded by a single flattened granulosa layer were also scored as primordial follicles , and only grade1 primordial follicles were counted in .
Data analysis was performed using SPSS statistical software (IBM, New York, NY, United States). Quantitative data were expressed as the standard deviation of the mean. Data were tested for normal distribution using the Kolmogorov-Smirnov test. The data analysis between the two groups was assessed with Student’s t-test or Mann–Whitney U test according to the normality of the data distribution. Receiver Operating Characteristic (ROC) curve was calculated in SPSS with Mann-Whitney U (Wilcoxon) test. The cut-off point was calculated with the Youden index based on the ROC analysis. Bivariate correlations were analyzed with the Pearson correlation test. Univariate and multivariate stepwise linear regression analyses performed with the independent variables reference correlation analysis results, including all candidate genes. The variance inflation factor (VIF) was used to measure the collinearity of variables, and variables with a VIF smaller than 5 were included in the analysis. P values < 0.05 were considered statistically significant.
Establishment of ovarian reserve comparison murine model
It is well known that ovarian reserve is progressively depleted with age in mammals. Over the past decades, rodents have been widely used in ovarian-related biological research as an essential component of mammalian models. In this study, we attempt to establish a comparative model of OR depending on the age of the mice . A total of 37 mice were divided into the high OR group (week-old 12.93 ± 1.66) and the low OR group (week-old 36.00 ± 0.67) based on previous estimates of mouse reproductive age. As shown in Fig. 1a, primordial follicle counts were completed by histology and morphological analysis, and only grade1 primordial follicles were counted in . The number of primordial follicles in the high OR group was significantly more than that in the low OR group (1433.95 ± 73.47 vs. 530.54 ± 49.24; P < 0.001) (Fig. 1b), which indicates the effective establishment of the animal model for OR comparison.
Validation of potential biomarkers to evaluate OR capability
To investigate whether the expression of Lhx8, Nobox, Sohlh1, Tbpl2, Stk31, Padi6, and Vrtn possess the capacity to evaluate OR levels, qRT-PCR was performed to verify their expression in the contralateral ovary from each mouse. The relative expression level of all these genes was significantly higher in the high OR group than in the low OR group (Fig. 2a). ROC curve analysis was further performed to assess the capacity of each of these biomarkers to evaluate high or low OR. As shown in Fig. 2b; Table 1, the AUC for each gene was significantly greater than 0.5 (p < 0.05). The AUC for Sohlh1 was 0.993 with a standard error of 0.010, and the sensitivity and specificity of the Sohlh1 in evaluating OR capacity were 92.6% and 100%, respectively, at the chosen cut-off value of 0.00015527. These results suggested that all these genes have the capability to evaluate OR levels in mice.
Correlation of mRNA expression levels with the primordial follicle pool
Next, the correlation was further analyzed between the expression of candidate biomarkers and the number of the PFP counted in each mouse. All seven candidate biomarkers were significantly correlated with PFs (Fig. 3), yet their correlation coefficients varied. Among them, Nobox, Sohlh1, and Lhx8 had a strong correlation (r > 0.7), whereas Tbpl2 and Vrtn had a weaker correlation.
Univariate linear analysis of candidate biomarkers to evaluate PFP
As shown in Fig. 3, the linear fit does not seem an optimal fit between gene expression form (2−ΔCt) and PFP. To investigate the optimal form of gene expression for linear regression analysis with PFP, the best-fit curves between the relative expression of candidate genes (2−ΔCt) and PFP were assessed via Curve Estimation using SPSS software. The results showed that the logarithmic model had the best goodness of fit between the relative expression of each gene and the number of PFs (Sup. Table 2). We depicted the expression levels of each gene in the form of -ΔCt to obtain more accurate fitting results in the subsequent linear regression analysis. Univariate analysis revealed that each of these potential biomarkers had a significant (p < 0.05) influence on PFP (Table 2). Among them, sohlh1 possessed the highest fitting coefficient (R2 = 0. 544) and Tbpl2 possessed the lowest (R2 = 0. 191).
Stepwise regression analysis to identify the best model for evaluating the PFP
To identify the best model of these seven biomarkers for evaluating PFP, a multivariate linear stepwise regression analysis was performed with these seven genes (-ΔCt) as predictors and the number of the PFP as the dependent variable. The stepwise approach allowed us to identify the best combination of predictive variables by including all statistically significant independent variables according to their degree of contribution. Our analysis showed that the combination of Sohlh1 and Lhx8, each acted as significant independent predictors (p < 0.05 and VIF < 5), exhibited the best fitting coefficient (R2 = 0. 597) for evaluating PFP among these seven biomarkers (Table 3, Sup. Table 3).
In the present study, we utilized an OR comparison model in mice to further investigate the validity of seven potential biomarkers identified by our previous bioinformatics analysis for assessing PFs, including Sohlh1, Nobox, Lhx8, Tbpl2, Stk31, Padi6 and Vrtn. The results showed significant differences in the expression of the seven candidate genes in the OR comparison model. AUC analysis showed these seven biomarkers possessed the independent potential to assess OR levels. Further regression analysis between these biomarkers with PF showed that all seven genes had the independent potential to evaluate PFP, and the combination of Sohlh1 and Lhx8 had the best fit with PFs, suggesting that Sohlh1 and Lhx8 can be used as the optimal combinational biomarkers for rapid assessment of PFP in the mouse ovary.
The PFP is an essential subject in studies related to ovarian physiology and pathogenesis [4,5,6]. In clinical practice, the assessment of ovarian PFP in patients is mainly done indirectly by imaging and serology such as AMH, while there is no efficient method for the direct evaluation of ovarian tissue applied in clinical OTC, in vitro activation (IVA) and other ART [21,22,23]. In clinical OTC studies, ovarian tissues are processed to a specific size (e.g., 5 mm*5mm*1mm) and then frozen . Several years later, before these ovarian tissues are rethawed with a series of treatments and re-transplanted back into the patient, the OR levels in the tissues are unknown to the physicians and patients, and the effectiveness of the transplantation is even less known. Generally, the assessment of the exact number of PFP in ovarian tissues based on microscopic counting is considerably subjective, and the number of follicles reported by different teams varies dramatically , which makes it conceivably challenging to apply for rapid assessment in clinical applications and animal studies. Therefore, a rapid and effective approach is urgently needed to evaluate OR levels in clinics. Our previous work has initially screened candidate genes in evaluating OR levels histologically , and this study further validated the feasibility of seven candidate genes to assess histological OR levels in combination with the classical histological assessment of PFP.
Quantitative real-time PCR (qRT-PCR) as a technology with high sensitivity and reproducibility has been applied in many areas, including clinical diagnostics, pharmacology, toxicology, and food safety [24,25,26]. In clinical diagnostics such as viral or bacterial infections, qRT-PCR techniques have been developed as diagnostic tools for rapid and accurate detection compared to traditional methods. In particular, the recent prevalence of the COVID-19 pathogen has reinforced the role of qRT-PCR as an irreplaceable tool for the early diagnosis of this disease, also highlighting the potential applicability of qRT-PCR that has been overlooked . Therefore, in the present study, the qRT-PCR assay was used to verify further the validity of candidate biomarkers for PFP assessment based on our previous bioinformatics analysis. Our results confirmed that all seven genes could independently evaluate the number of PFP, and the combination of Sohlh1 and Lhx8 had an optimal fit with PFs, suggesting that Sohlh1 and Lhx8 could be used as the optimal biomarkers for rapid assessment of PFs in murine ovarian tissue.
Previous studies reveal that SOHLH1 and LHX8, as crucial oocyte-specific transcription factors, are essential in regulating postnatal folliculogenesis, especially in the earliest stages of primordial follicle activation [28,29,30]. Sohlh1, located on human chromosome 9 and mouse chromosome 2, encodes a basic helix–loop–helix transcription factor with homologs in humans and mice. Lhx8 is located on human chromosome 1 and mouse chromosome 3, and its encoded protein is a transcription factor containing a cysteine-histidine LIM structural domain involved in the morphogenesis of several organs . In the murine ovary, both Sohlh1 and Lhx8 are expressed specifically in both embryonic and postnatal female germ cells starting at embryonic day 13.5 (E13.5). Despite a slight difference in the expressions of Sohlh1 and Lhx8 in oocytes at different stages of the postnatal follicle, both were relatively highly expressed at early stages (primordial and primary stages). Conditional deficiency of either Sohlh1 or Lhx8 in the oocytes of primordial follicles causes massive primordial oocyte activation and postnatal oocyte depletion leading to infertility in mice .
As for the regulatory relationship between Sohlh1 and Lhx8, ChIP experiments of newborn mouse ovaries conducted by Pangas et al. in earlier years suggested that Lhx8 is likely one of the direct downstream target genes of SOHLH1 through the E box elements in their promoters . While a recent study conducted by Wang et al. shows that SOHLH1 and LHX8 can cross-regulate each other and physically bind to each other with Figla to form a nuclear complex in oocytes to regulate transcriptional activity during early oocyte growth and differentiation .
The above studies on the molecular mechanisms of Sohlh1 and Lhx8 further demonstrated the reliability and feasibility of Sohlh1 and Lhx8 could use as prominent biomarkers to estimate PFP in our present findings. However, the explicit mechanism of maintaining quiescent follicles by these transcription factors still needs further study. We believe it is feasible for the validity of such assessment tools in models like the chemically induced premature ovarian failure (POF) model or OTC model, but of course, the exact validity still needs to be verified in each specific model with a considerable number of mice. Besides, although the genes in this study are highly homologous to humans, the feasibility of the present findings in human ovarian tissue slices requires additional validation with large sample sizes.
In conclusion, the results of this study suggest that Sohlh1, Nobox, Lhx8, Tbpl2, Stk31, Padi6 and Vrtn these seven biomarkers possess independent potential to evaluate PF numbers, and the combination of Sohlh1 and Lhx8 can be used as the optimal biomarkers for rapid assessment of PFs in the mouse ovary. Our findings provide a new perspective for evaluating PFP of ovaries in animal studies and clinics.
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Bhardwaj JK, Paliwal A, Saraf P, Sachdeva SN. Role of autophagy in follicular development and maintenance of primordial follicular pool in the ovary. J Cell Physiol. 2022;237:1157–70.
Ruth KS, Day FR, Hussain J, Martinez-Marchal A, Aiken CE, Azad A, Thompson DJ, Knoblochova L, Abe H, Tarry-Adkins JL, et al. Genetic insights into biological mechanisms governing human ovarian ageing. Nature. 2021;596:393–7.
La Marca A, Mastellari E. Fertility preservation for genetic diseases leading to premature ovarian insufficiency (POI). J Assist Reprod Genet. 2021;38:759–77.
Terren C, Munaut C. Molecular Basis Associated with the Control of Primordial Follicle Activation During Transplantation of Cryopreserved Ovarian Tissue. Reprod Sci 2020.
Shimamoto S, Nishimura Y, Nagamatsu G, Hamada N, Kita H, Hikabe O, Hamazaki N, Hayashi K. Hypoxia induces the dormant state in oocytes through expression of Foxo3. Proc Natl Acad Sci U S A. 2019;116:12321–6.
Ferst JG, Rovani MT, Dau AMP, Gasperin BG, Antoniazzi AQ, Bordignon V, Oliveira DE, Goncalves PBD, Ferreira R. Activation of PPARG inhibits dominant follicle development in cattle. Theriogenology. 2020;142:276–83.
Sarma UC, Winship AL, Hutt KJ. Comparison of methods for quantifying primordial follicles in the mouse ovary. J Ovarian Res. 2020;13:121.
Depmann M, Faddy MJ, van der Schouw YT, Peeters PH, Broer SL, Kelsey TW, Nelson SM, Broekmans FJ. The relationship between variation in size of the primordial follicle Pool and Age at Natural Menopause. J Clin Endocrinol Metab. 2015;100:E845–851.
Myers M, Britt KL, Wreford NG, Ebling FJ, Kerr JB. Methods for quantifying follicular numbers within the mouse ovary. Reproduction. 2004;127:569–80.
Roness H, Meirow D. FERTILITY PRESERVATION: follicle reserve loss in ovarian tissue transplantation. Reproduction. 2019;158:F35–F44.
Anderson RA, Wallace WHB, Telfer EE. Ovarian tissue cryopreservation for fertility preservation: clinical and research perspectives. Hum Reprod Open. 2017;2017:hox001.
Liu W, Zhang J, Wang L, Liang S, Xu B, Ying X, Li J. The protective effects of rapamycin pretreatment on ovarian damage during ovarian tissue cryopreservation and transplantation. Biochem Biophys Res Commun. 2021;534:780–6.
Ginsburg GS. Genomics-Inspired biomarkers and Diagnostics-Where are they? Clin Chem. 2017;63:255–7.
Liu L, Liu B, Li K, Wang C, Xie Y, Luo N, Wang L, Sun Y, Huang W, Cheng Z, Liu S. Identification of biomarkers for Predicting Ovarian Reserve of primordial follicle via Transcriptomic Analysis. Front Genet. 2022;13:879974.
Souza V, Mendes E, Casaro M, Antiorio A, Oliveira F, Ferreira C. Description of Ovariectomy Protocol in mice. Methods in molecular biology (Clifton NJ). 2019;1916:303–9.
Ma Y, Jin J, Tong X, Yang W, Ren P, Dai Y, Pan Y, Zhang Y, Zhang S. ADAMTS1 and HSPG2 mRNA levels in cumulus cells are related to human oocyte quality and controlled ovarian hyperstimulation outcomes. J Assist Reprod Genet. 2020;37:657–67.
Resch T, Hackl H, Esser H, Gunther J, Schwelberger H, Ritschl PV, Ebner S, Maglione M, Mellitzer V, Biebl M, et al. Expression of MICA in Zero Hour Biopsies predicts Graft Survival after Liver Transplantation. Front Immunol. 2021;12:606146.
Lv Y, Cao R, Liu H, Su X, Lu G, Ma J, Chan W. Single-Oocyte Gene Expression Suggests That Curcumin Can Protect the Ovarian Reserve by Regulating the PTEN-AKT-FOXO3a Pathway. Int J Mol Sci 2021, 22.
Kong H, Kim S, Lee J, Youm H, Lee J, Suh C, Kim S. Effect of exogenous Anti-Müllerian hormone treatment on Cryopreserved and transplanted mouse ovaries. Reproductive Sci (Thousand Oaks Calif). 2016;23:51–60.
Dutta S, Sengupta P. Men and mice: relating their ages. Life Sci. 2016;152:244–8.
Tanaka Y, Hsueh AJ, Kawamura K. Surgical approaches of drug-free in vitro activation and laparoscopic ovarian incision to treat patients with ovarian infertility. Fertil Steril. 2020;114:1355–7.
Hinkle K, Orwig KE, Valli-Pulaski H, Taylor S, van Leeuwen K, Carpentieri D, Walsh A. Cryopreservation of ovarian tissue for Pediatric Fertility. Biopreserv Biobank. 2021;19:130–5.
Sugishita Y, Meng L, Suzuki-Takahashi Y, Nishimura S, Furuyama S, Uekawa A, Tozawa-Ono A, Migitaka-Igarashi J, Koizumi T, Seino H, et al. Quantification of residual cryoprotectants and cytotoxicity in thawed bovine ovarian tissues after slow freezing or vitrification. Hum Reprod. 2022;37:522–33.
Agrimonti C, Bottari B, Sardaro MLS, Marmiroli N. Application of real-time PCR (qPCR) for characterization of microbial populations and type of milk in dairy food products. Crit Rev Food Sci Nutr. 2019;59:423–42.
Galluzzi L, Ceccarelli M, Diotallevi A, Menotta M, Magnani M. Real-time PCR applications for diagnosis of leishmaniasis. Parasit Vectors. 2018;11:273.
Padhi BK, Pelletier G, Shwed PS. A bioinformatics workflow for the evaluation of RT-qPCR primer specificity: application for the assessment of gene expression data reliability in toxicological studies. Regul Toxicol Pharmacol. 2020;111:104575.
Udugama B, Kadhiresan P, Kozlowski HN, Malekjahani A, Osborne M, Li VYC, Chen H, Mubareka S, Gubbay JB, Chan WCW. Diagnosing COVID-19: the Disease and Tools for detection. ACS Nano. 2020;14:3822–35.
Wang Z, Liu C, Zhao Y, Dean J. FIGLA, LHX8 and SOHLH1 transcription factor networks regulate mouse oocyte growth and differentiation. Nucleic Acids Res. 2020;48:3525–41.
Shin YH, Ren Y, Suzuki H, Golnoski KJ, Ahn HW, Mico V, Rajkovic A. Transcription factors SOHLH1 and SOHLH2 coordinate oocyte differentiation without affecting meiosis I. J Clin Invest. 2017;127:2106–17.
D’Ignazio L, Michel M, Beyer M, Thompson K, Forabosco A, Schlessinger D, Pelosi E. Lhx8 ablation leads to massive autophagy of mouse oocytes associated with DNA damage. Biol Reprod. 2018;98:532–42.
Pangas SA, Choi Y, Ballow DJ, Zhao Y, Westphal H, Matzuk MM, Rajkovic A. Oogenesis requires germ cell-specific transcriptional regulators Sohlh1 and Lhx8. Proc Natl Acad Sci U S A. 2006;103:8090–5.
Pangas S, Choi Y, Ballow D, Zhao Y, Westphal H, Matzuk M, Rajkovic A. Oogenesis requires germ cell-specific transcriptional regulators Sohlh1 and Lhx8. Proc Natl Acad Sci USA. 2006;103:8090–5.
The authors declare no competing interests.
Ethics approval and consent to participate
Animal experiments were approved by the Animal Ethics Committee of Shanghai Tenth People’s Hospital (No. SHDSYY-2021-Y0688).
Consent for publication
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
About this article
Cite this article
Liu, L., Liu, B., Wang, L. et al. Sohlh1 and Lhx8 are prominent biomarkers to estimate the primordial follicle pool in mice. Reprod Biol Endocrinol 21, 46 (2023). https://doi.org/10.1186/s12958-023-01097-3
- Primordial follicle pool
- Clinical assessment