Dysregulated cytokine profile associated with biochemical premature ovarian insufficiency

Abstract Problem Premature ovarian insufficiency (POI) imposes great challenge on female reproduction. Whether immune disturbance in ovarian environment was implicated in POI remains unclear. We aimed to characterize the cytokine profile in follicular fluid (FF) and paired serum in patients with biochemical POI (bPOI). Method of study Multiplex immunoassay containing 45 cytokines was performed for individual FF and paired serum samples from 35 bPOI patients and 37 matched controls. Cytokine profiles were compared between the two groups and cytokines correlated to ovarian reserve, and the rates of day‐3 good‐quality embryos were further analyzed. Results In FF, significantly elevated level of chemokines MIP‐1α (P = .043), CXCL8 (P = .024), IP‐10 (P = .041), and eotaxin‐1 (P = .015) as well as growth factors VEGF‐D (P = .047), BDNF (P = .043), LIF (P = .002), and bFGF (P = .046) was found in bPOI patients compared to controls. Yet RANTES manifested an opposite trend with reduced levels among bPOI patients (P = .006). All these chemokines and growth factors in FF were significantly correlated with ovarian reserve (P < .05). In paired serum, cytokine signature was not likely accordant with that in FF between two groups, except for increased IP‐10 (P = .032) in bPOI patients and its significant correlation to FSH and AFC (P < .05). Among all differentially expressed cytokines, RANTES in FF was correlated with the rate of day‐3 good‐quality embryos (P = .035). Conclusion Altered cytokine profile characterized by increased chemokines and growth factors was associated with early stage of POI, which may fuel the progression of the disease or even play a crucial role in the development of ovarian insufficiency.


| INTRODUC TI ON
Premature ovarian insufficiency (POI), also termed as primary ovarian insufficiency (POI), refers to the decline or cessation of ovarian function prior to age 40. It encompasses the spectrum of ovarian dysfunction from decreased fecundity (occult POI), and elevated follicle-stimulating hormone (FSH) (biochemical POI) to amenorrhea (overt POI). [1][2][3] It is estimated that POI affects 1%-5% of women before the age of 40. 4 As a major cause of infertility, POI represents one of the most challenging issues in female reproduction. Upon confirmed diagnosis, available ovarian function seldom exists.
Additionally, there is no effective strategy to improve or rescue ovarian function and fertility currently. Therefore, it is of great importance to identify risk factors of ovarian insufficiency at earlier stage. However, studies were mainly conducted in women with fully developed overt POI and failed to determine biomarkers or indicators that could accurately predict the status or progress of POI.
Premature ovarian insufficiency is pathophysiologically complicated and etiologically heterogeneous. The known causes include genetics, autoimmunity, infectious, and iatrogenic. Still, most cases remain elusive. 5 The autoimmune disturbance explains 5%-30% of POI cases, and it has been attracting increasingly clinical and scientific concerns. 6 How autoimmunity induces ovarian senescence has not been fully established. No specific, reliable, and non-invasive diagnostic tests for autoimmune POI exist currently. 7 Especially, in contrast to accumulated evidence of peripheral autoimmune dysregulation reported, local ovarian autoimmunity was seldom explored in patients with POI.
Follicular microenvironment is critical for folliculogenesis and acquisition of oocyte competence. 8,9 Cytokines in follicular fluid are increasingly recognized as potential modulators of ovarian function either by autocrine or paracrine mechanisms. A cascade of cytokines, chemokines, and growth factors could mediate the interaction among lympho-hematopoietic cells, somatic cells and oocytes, and involve in follicular growth, steroidogenesis, activation of leukocytes necessary for ovulation, and tissue remodeling during ovulation, luteinization, and luteolysis. 10,11 Inflammation negatively affects the follicular microenvironment and thereby reduces oocyte quality and quantity. 12 Increased levels of pro-inflammatory cytokines, such as interleukin IL-6 and CXCL8, in follicular fluid were associated with reduced oocyte quality. 8,13 Moreover, oocyte quality was decreased in pathologies associated with local follicular inflammation, including aging and polycystic ovary syndrome (PCOS). 14 However, it is currently unclear whether local inflammation in follicular environment induced dysregulated cytokine milieu is associated with the development of POI.
In this study, given bPOI is easier to diagnose and represents the early stage of ovarian insufficiency, the follicular fluid and paired serum from patients with bPOI were collected. Cytokine profiles were characterized in follicular fluid and paired serum in patients with biochemical POI synchronously to determine specific cytokine signature involved, which would be of great value to explain the decline of ovarian function and provide new avenues for early intervention.

| Participants
The study recruited subjects undergoing their first in vitro fertili-

| Serum and follicular fluid collection
Peripheral blood was sampled on day 3 of menstrual cycle, and serum was isolated and stored at −80°C until assayed. Paired follicular fluid from dominant follicles containing oocytes (diameter ≥ 1.4 cm) was collected during oocyte retrieval, separated, and stored at −80°C until used. Only follicular fluid without blood or flushing solution was collected.

| Statistical analyses
Baseline characteristics and cytokine levels were expressed as median with interquartile ranges. Kolmogorov-Smirnov test was used to evaluate the normality of data distribution. t test was used for original data or Napierian logarithm transformed data that were normally distributed, and Mann-Whitney U test was used for the other data to compare between groups. Putting participants from the two groups together, Spearman's correlation was used to estimate the association between cytokines in follicular fluid and the biomarkers of ovarian reserve (FSH, AMH and AFC) as well as between the follicular and serum. According to the Society for Assisted Reproductive Technology (SART) grading system, day-3 good-quality embryos were defined as embryos with six or more blastomere and with fragmentation degree ＜25%. The rate of day-3 good-quality embryo was defined as number of day-3 good-quality embryos divided by the number of 2 pronuclear (2PN) zygotes. 15 Spearman's correlations between follicular fluid and serum cytokines from all participants were also evaluated. P ＜ .05 was considered to be statistically significant. All statistical analyses were performed with SPSS (SPSS Inc, version 21.0).

| RE SULTS
In this study, 35 women with bPOI and 37 controls with normal ovarian reserve were enrolled. There were no significant differences regards to age and body mass index (BMI) between the two groups (P > .05). Patients with bPOI presented with significantly higher FSH and decreased AMH level and AFC as expected (Table 1).
Cytokine profile in FF and paired serum was characterized and compared between patients with bPOI and control women.
In paired serum, IP-10 (P = .032), IL-7 (P = .027), and IL-17A (P = .017) showed significantly higher levels but IL-1α (P = .044) and IL-21 (P = .042) were decreased in patients with bPOI compared with controls. Cytokine signature in serum was not likely accordant with that in paired FF between two groups, except for increased IP-10 in patients with bPOI ( Table 2 and Table S2). Serum IP-10 was also found significantly associated with both FSH and AFC (P < .05) (Table S4).
To elucidate the association of different follicular and serum cytokines with oocyte quality and embryo quality, the correlation with the rate of day-3 good-quality embryos was further analyzed.
Among all the differentially expressed cytokines, RANTES in follicular fluid turned out to be correlated with rate of day-3 good-quality embryos rate (P = .035) ( Table 4). Additionally, ovarian somatic cells, including luteal, stromal, thecal, and granulosa cells, are an important cellular source as well. 11 The abundance or dysregulation may play a role in ovarian insufficiency.

| D ISCUSS I ON
The marked difference specific in FF profile between control and bPOI is striking, suggesting a distinct progression status of ovarian insufficiency in bPOI. Although we cannot accurately describe the contribution of each cytokine, the abnormally elevated chemokine   Abbreviations: BDNF, brain-derived neurotrophic factor; bFGF, basic fibroblast growth factor; IP, interferon inducible protein; LIF, Leukemia inhibitory factor; MIP, macrophage inflammatory protein; VEGF, vascular endothelial growth factor.

TA B L E 3 Correlation between follicular fluid cytokines and biomarkers of ovarian reserve
and reserve. eotaxin-1 mainly engaged in eosinophil and basophil migration. 16 eotaxin-1/CCR3 axis was transiently upregulated in the ovarian theca-interstitial layer in response to LH surge during the terminal stages of follicular development. 21 It induced oxidative stress-related tissue damage by promoting ROS production in a dose-dependent manner by eosinophil activation. [22][23][24] Moreover, eotaxin-1 was increased in plasma and cerebral spinal fluid of healthy aging human and correlated with reduced neurogenesis in aged mice. 25 It could also directly impair Schwann cell remyelination. 26 Therefore, eotaxin-1 might be an age-related factor and contributor for the aging process, including premature aging of oocyte, even except for its chemotactic function. Further researches are warranted to confirm the association and explore the possible mechanism.
Chemokine RANTES played a critical role in attracting macrophage and NK cell migration and intermediating interaction between T cell and dendritic cell. 16   Abbreviations: GRO, growth related oncogene; IL, interleukin; PDGF, platelet-derived growth factor; PIGF, placenta growth factor.
inadequate statistical power. The heterogeneity of individuals might also obscure the trends. Secondly, several cytokines showed very low level especially in follicular fluid, which made results less reliable and more difficult to be statistically significant between two groups. This partially might due to the multiplex immunoassay with magnetic beads used.
Taken together, altered cytokine profile characterized by in-

ACK N OWLED G M ENTS
We are grateful to the subjects for taking part in our research. This

CO N FLI C T O F I NTE R E S T
The authors report no conflict of interest.