Serum GDF9 and BMP15 as Potential Markers of Ovarian Function in Women With and Without Polycystic Ovary Syndrome

Angelique H. Riepsamen; Mark W. Donoghoe; Inthrani R. Indran; Leah Hechtman; David M. Robertson; Robert B. Gilchrist; William L. Ledger; Eu-Leong Yong

Disclosures

Clin Endocrinol. 2023;98(4):567-577.

Abstract and Introduction

Abstract

Objective: Growth differentiation factor-9 (GDF9) and bone morphogenetic protein-15 (BMP15) are critical paracrine regulators of female fertility and are predominantly expressed by oocytes. However, it is unknown if serum concentrations reflect changes in ovarian function and/or reproductive endocrine disorders. This study aimed to determine if serum GDF9/BMP15 are associated with ovarian, pituitary, oestrogenic, androgenic and metabolic characteristics and the ovarian pathologies, polycystic ovarian morphology (PCOM) and polycystic ovary syndrome (PCOS).

Design: Women aged 21–45 years (n = 381) were included from a cross-sectional study at the National University Hospital, Singapore.

Patients: Participants were volunteers and patients with possible PCOS.

Measurements: Anthropometric measurements, transvaginal ultrasound scans and serum sampling were performed and a questionnairecompleted. Serum GDF9 and BMP15 concentrations were matched with menstrual cycle length, ovarian protein and steroid hormone production, pituitary hormone production and metabolic assessments in women with PCOM or PCOS and those with neither (control).

Results: Serum GDF9 and BMP15 were detectable in 40% and 41% of women, respectively and were positively correlated with each other (r = 0.08, p = 0.003). GDF9, but not BMP15, was positively correlated with ovarian volume (p = 0.02) and antral follicle count (AFC) (p = 0.004), but not with anti-Müllerian hormone (p = 0.05). However, serum GDF9 and BMP15 concentrations were not significantly different between control, PCOM and PCOS women, nor associated with androgenic or metabolic PCOS features. However, the relationship between GDF9 and AFC differed between control, PCOM and PCOS women (p = 0.02).

Conclusions: Serum GDF9 and BMP15 concentrations somewhat reflect ovarian but not androgenic or metabolic characteristics of PCOS, with increased GDF9 reflecting high AFC as seen in PCOM/PCOS.

Introduction

Polycystic ovary syndrome (PCOS) is a common endocrine disorder affecting 5%–17% of women of reproductive age and is the major cause of anovulatory infertility and menstrual irregularities.^[1] In addition, between 40% and 80% of women with PCOS are obese or overweight and are at risk of insulin resistance and metabolic syndrome. PCOS is characterized by chronic oligo-/anovulation, hyperandrogenism and polycystic ovarian morphology (PCOM), with the recently endorsed Rotterdam criteria requiring at least two of these features for the diagnosis of PCOS.^[2] Thus, altered folliculogenesis is a key ovarian feature of PCOS, with signalling between the oocyte and granulosa cells reportedly impaired in women with PCOS and oocyte quality being aberrant (reviewed in).^[3] Therefore, understanding how oocyte and ovarian factors are altered in women with PCOS may assist in identifying potential diagnostic targets and improving clinical management.

The oocyte-secreted factors growth differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15), both members of the TGF-β superfamily, are critical for folliculogenesis and oocyte quality.^[4] As paracrine growth factors, GDF9 and BMP15 are involved in the regulating the fate of cells within the ovarian follicle, including control of proliferation, metabolism and differentiation of granulosa and cumulus cells,^[4] in turn affecting oocyte development and quality.^[5] Genetic studies show that gene deletions and inactivating mutations in GDF9 and BMP15 cause sterility and/or subfertility in a range of mammals, while some heterozygous mutations can lead to increased ovulation rate in mono-ovular species.^[6,7] Further, it has been postulated that the ratio of GDF9:BMP15 regulates mammalian ovulation rate and fecundity.^[8]

GDF9 and BMP15 are potential diagnostic markers of PCOS as they are secreted almost exclusively by the oocyte^[8] and their expression may be aberrant in women with PCOS (reviewed in).^[9] For example, increased GDF9 and BMP15 gene expression was observed in MII oocytes following ovarian stimulation in women with PCOS compared to controls,^[10] and delayed GDF9 (but not BMP15) expression during folliculogenesis was observed in ovarian sections of women with PCOS compared to normally cycling women.^[11] However, other studies have shown no difference in these oocyte-secreted factors between women with and without PCOS.^[12] To date there is limited knowledge of the diagnostic utility of serum concentrations of GDF9 and BMP15 in women, as validated assays sufficiently sensitive to measure these paracrine growth factors in serum have only recently been developed.^[13] Therefore, the aim of this study was to measure GDF9 and BMP15 in serum samples collected from women of reproductive age to determine if these oocyte-secreted factors are associated with ovarian, pituitary, oestrogenic, androgenic and metabolic characteristics, as well as the ovarian pathologies, PCOM and PCOS.

1 2 3 4

Next Section

Abstract and Introduction
Materials and Methods
Results
Discussion
References

Table 1. Ovarian, endocrine and metabolic characteristics associated with polycystic ovary syndrome (PCOS) comparing women with undetectable and detectable serum concentrations of GDF9 and BMP15
Characteristics	All women	GDF9				BMP15
Characteristics	All women	Undetectable (≤15.4 pg/ml)	Detectable (>15.4 pg/ml)	p value	Adjusted p value	Undetectable (≤23 pg/ml)	Detectable (>23 pg/ml)	p value	Adjusted p value
N	381	227	154			223	157
Participant
Age (years)	31.4 ± 5.0	31.9 ± 5.1	30.7 ± 4.7	0.016	0.380	31.8 ± 5.0	30.9 ± 4.9	0.104	0.963
No. live births	0.6 ± 0.9	0.57 ± 0.92	0.53 ± 0.91	0.635	1.000	0.57 ± 0.94	0.53 ± 0.88	0.690	1.000
Ovarian
Regular menstrual cycles,^{^} n (%)	243 (64%)	158 (70%)	85 (55%)	0.004	0.116	152 (68%)	90 (57%)	0.031	0.605
Ovarian volume^a	6.04 ± 2.96	5.69 ± 2.64	6.54 ± 3.32	0.006	0.164	5.77 ± 2.82	6.42 ± 3.12	0.035	0.658
AFC^a	18.8 ± 12.4	17.0 ± 10.1	21.6 ± 14.9	0.0004	0.012	17.7 ± 11.6	20.4 ± 13.5	0.041	0.718
AMH (pmol/L)	48.2 ± 36.6	44.0 ± 33.4	54.3 ± 40.3	0.008	0.207	46.1 ± 35.6	51.1 ± 38.0	0.201	0.999
Pituitary
FSH (IU/L)	7.43 ± 2.50	7.57 ± 2.68	7.22 ± 2.20	0.185	0.998	7.52 ± 2.72	7.31 ± 2.17	0.424	1.000
LH (IU/L)	5.91 ± 4.33	5.67 ± 4.09	6.26 ± 4.65	0.207	0.999	5.50 ± 3.73	6.50 ± 5.05	0.033	0.634
Prolactin (mI U/L)	275.7 ± 137.6	266.5 ± 123.9	289.1 ± 155.2	0.116	0.975	278.7 ± 141.6	270.2 ± 131.9	0.554	1.000
Oestrogenic
Estrone (E1, pg/ml)	127.2 ± 48.7	124.0 ± 40.2	132.5 ± 60.3	0.164	0.995	123.0 ± 45.4	132.6 ± 52.8	0.110	0.970
Estradiol (E2, nmol/L)	178.1 ± 65.2	177.0 ± 67.4	179.8 ± 62.0	0.689	1.000	175.0 ± 57.9	182.8 ± 74.6	0.249	1.000
Androgenic
Testosterone (nmol/L)	1.43 ± 0.74	1.37 ± 0.71	1.52 ± 0.78	0.055	0.819	1.37 ± 0.70	1.51 ± 0.77	0.086	0.932
DHT (nmol/L)	1.41 ± 1.00	1.33 ± 0.85	1.52 ± 1.19	0.067	0.876	1.46 ± 1.08	1.33 ± 0.87	0.215	0.999
Androstenedione (ADT) (nmol/L)	8.56 ± 6.31	8.02 ± 4.35	9.35 ± 8.35	0.045	0.749	7.74 ± 4.11	9.75 ± 8.40	0.002	0.068
DHEAS (μmol/L)	5.48 ± 2.29	5.35 ± 2.32	5.67 ± 2.24	0.178	0.997	5.36 ± 2.35	5.64 ± 2.20	0.236	1.000
SHBG (nmol/L)	53.7 ± 30.4	56.2 ± 32.2	50.1 ± 27.3	0.053	0.808	54.4 ± 28.9	52.9 ± 32.6	0.617	1.000
Free androgen index (FAI)	4.07 ± 4.37	3.89 ± 4.51	4.34 ± 4.17	0.330	1.000	3.85 ± 4.27	4.33 ± 4.47	0.293	1.000
mFG score (1–9)	1.65 ± 2.67	1.56 ± 2.16	1.77 ± 3.28	0.454	1.000	1.66 ± 2.82	1.62 ± 2.44	0.882	1.000
Androgen receptor activity	162.5 ± 108.4	152.8 ± 98.1	177.8 ± 121.9	0.036	0.671	153.4 ± 101.7	175.7 ± 117.0	0.060	0.842
Metabolic
BMI	23.7 ± 5.3	23.3 ± 4.4	24.4 ± 6.25	0.040	0.707	23.8 ± 5.34	23.7 ± 5.1	0.964	1.000
WHR	0.79 ± 0.06	0.78 ± 0.06	0.79 ± 0.06	0.191	0.998	0.78 ± 0.06	0.79 ± 0.06	0.127	0.983
Insulin (mu/L)	8.95 ± 8.78	8.67 ± 8.38	9.37 ± 9.35	0.445	1.000	8.36 ± 7.94	9.77 ± 9.84	0.123	0.981
Glucose (mmol/L)	4.76 ± 0.61	4.78 ± 0.67	4.74 ± 0.51	0.541	1.000	4.79 ± 0.68	4.72 ± 0.50	0.286	1.000
Triglycerides (mmol/L)	1.02 ± 0.73	0.97 ± 0.47	1.10 ± 0.99	0.076	0.908	0.99 ± 0.79	1.07 ± 0.64	0.307	1.000
Cholesterol (mmol/L)	4.83 ± 0.86	4.77 ± 0.75	4.91 ± 0.99	0.114	0.973	4.80 ± 0.87	4.87 ± 0.84	0.483	1.000
HDL (mmol/L)	1.47 ± 0.36	1.49 ± 0.36	1.45 ± 0.34	0.297	1.000	1.47 ± 0.36	1.48 ± 0.35	0.720	1.000
LDL (mmol/L)	2.90 ± 0.79	2.82 ± 0.70	3.02 ± 0.90	0.017	0.401	2.90 ± 0.83	2.91 ± 0.74	0.887	1.000
Cholesterol: HDL ratio	3.45 ± 1.00	3.38 ± 0.99	3.55 ± 1.02	0.108	0.967	3.45 ± 1.02	3.45 ± 0.98	0.998	1.000
HOMA-IR	1.94 ± 2.06	1.86 ± 1.80	2.04 ± 2.39	0.403	1.000	1.81 ± 1.75	2.12 ± 2.43	0.145	0.991
Ovarian pathologies
PCOM,^b n (%)	187 (49%)	101 (44%)	86 (56%)	0.030	0.594	99 (44%)	88 (56%)	0.025	0.535
PCOS, n (%)	130 (34%)	71 (31%)	59 (38%)	0.155	0.994	70 (31%)	60 (38%)	0.167	0.996

Note: Data are mean ± SD, with comparisons of undetectable and detectable groups analysed by Mann–Whitney for continuous data, and proportional data are n (%) analysed by χ ² test. To account for multiple comparisons, the Holm–Sidak method was used to calculate adjusted p values, where <0.05 was considered significant (highlighted in bold). ^{^}25–34 days.
Abbreviations: AFC, antral follicle count; AMH, anti-Müllerian hormone; BMI, body mass index; BMP15, bone morphogenetic protein-15; DHEAS, dehydroepiandrosterone-sulphate; DHT, dihydrotestosterone; FSH, follicle stimulating hormone; GDF9, growth differentiation factor-9; HDL, high-density lipoprotein; HOMA-IR, homoeostatic model assessment for insulin resistance; LDL, low-density lipoprotein; LH, luteinizing hormone; mFG, modified Ferriman–Gallwey; PCOM, polycystic ovarian morphology; PCOS, polycystic ovary syndrome; SHBG, sex hormone-binding globulin; WHR, waist to hip ratio.
^aAveraged from left and right ovaries.
^bAll PCOM: ≥22 follicles in either ovary, and/or ovarian volume ≥8.44 ml in either ovary (including PCOS).

Table 2. Correlations between GDF9 and BMP15 serum biomarker concentrations and ovarian characteristics in women with and without PCOM and/or PCOS
	GDF9				BMP15
	n	r (95% CI)	p	p (int)	n	r (95% CI)	p	p (int)
Ovarian volume
Control	192	–0.02 (–0.10 to 0.06)	0.65		191	0.01 (–0.07 to 0.10)	0.73
PCOM	58	0.11 (–0.05 to 0.28)	0.17	0.20	58	0.08 (–0.09 to 0.24)	0.37	0.81
PCOS	129	0.08 (–0.03 to 0.18)	0.15		129	0.02 (–0.09 to 0.13)	0.71
AFC
Control	192	–0.01 (–0.09 to 0.08)	0.90	0.02	191	0.00 (–0.08 to 0.08)	0.94
PCOM	57	0.24 (0.08–0.40)	0.004		57	0.03 (–0.14 to 0.19)	0.74	0.96
PCOS	121	0.11 (0.00–0.22)	0.05		121	0.01 (–0.10 to 0.12)	0.85
AMH
Control	189	–0.03 (–0.11 to 0.05)	0.46		188	0.03 (–0.06 to 0.11)	0.54
PCOM	59	0.03 (–0.13 to 0.19)	0.75	0.14	59	–0.10 (–0.27 to 0.06)	0.21	0.37
PCOS	123	0.11 (0.00–0.21)	0.06		123	0.00 (–0.11 to 0.11)	1.00

Note: Data shown are the estimated Kendall's tau correlation coefficient (r) and 95% CI with associated p value for each slope. The p(int) term is the result from a global Wald interaction test for differences in correlations between the mutually exclusive groups of control, PCOM and PCOS. p values <0.05 were considered significant (highlighted in bold). Ovarian volume and AFC are averaged from the left and right ovaries (see Supporting Information Material for corresponding figures).
Abbreviations: AFC, antral follicle count; AMH, anti-Müllerian hormone; BMP15, bone morphogenetic protein-15; CI, confidence intervals; GDF9, growth differentiation factor-9; PCOM, polycystic ovarian morphology; PCOS, polycystic ovary syndrome.

Table 3. GDF9, BMP15 and AMH mean serum concentrations and GDF9:BMP15 ratios in control, PCOM and PCOS groups
	Control (n = 192)	PCOM (n = 58)	PCOS (n = 129)	p Value
GDF9 pg/ml, mean (95% CI)	6.46 (4.32–9.67)	9.93 (5.34–18.49)	11.30 (7.33–17.43)	0.11
BMP15 pg/ml, mean (95% CI)	12.79 (9.14–17.91)	17.56 (10.44–29.52)	18.30 (12.71–26.33)	0.26
AMH pg/ml, mean (95% CI)	2.58 (2.30–2.88)	6.09 (4.99–7.45)	10.11 (8.80–11.61)	<0.001
GDF9:BMP15 ratio (95% CI)⁺	(n = 33)	(n = 15)	(n = 30)
	0.90 (0.52–2.01)	0.48 (0.26–0.82)	1.08 (0.56–1.92)	0.25

Note: Estimated geometric mean pg/ml (95% CI). +analysed by Kruskal–Wallis; samples with both GDF9 and BMP15 serum concentrations above their respective limits of quantification. p values <0.05 were considered significant (highlighted in bold).
Abbreviations: AMH, anti-Müllerian hormone; BMP15, bone morphogenetic protein-15; CI, confidence intervals; GDF9, growth differentiation factor-9; PCOM, polycystic ovarian morphology; PCOS, polycystic ovary syndrome.

Table 4. Correlations between serum biomarker concentrations of GDF9 and BMP15 in women with and without PCOM and/or PCOS, and all women combined
	GDF9
	n	r (95% CI)	p	p (int)
BMP15
Control	191	0.11 (0.04–0.17)	0.002
PCOM	59	0.09 (–0.06 to 0.24)	0.23	0.39
PCOS	130	0.02 (–0.07 to 0.12)	0.63
All women	380	0.08 (0.03–0.13)	0.003	n/a

Note: Data shown are the estimated Kendall's tau correlation coefficient (r) and 95% CI with associated p value for each slope. The p(int) term is the result from a global Wald interaction test for differences in correlations between the mutually exclusive groups of control, PCOM and PCOS (see Supporting Information Material for corresponding figure). p values <0.05 were considered significant (highlighted in bold).
Abbreviations: BMP15, bone morphogenetic protein-15; CI, confidence intervals; GDF9, growth differentiation factor-9; PCOM, polycystic ovarian morphology; PCOS, polycystic ovary syndrome.

processing....

Serum GDF9 and BMP15 as Potential Markers of Ovarian Function in Women With and Without Polycystic Ovary Syndrome

Abstract and Introduction

Abstract

Introduction

Tables

Tables

Tables

Tables

References

Authors and Disclosures

Authors and Disclosures

Email This

Feedback