Background
Ovarian senescence is a normal age-associated phenomenon, but increasingly younger women are affected by diminished ovarian reserves or premature ovarian insufficiency. There is an urgent need for developing therapies to improve ovarian function in these patients. In this context, previous studies suggest that stem cell–secreted factors could have regenerative properties in the ovaries.
Objective
This study aimed to test the ability of various human plasma sources, enriched in stem cell–secreted factors, and the mechanisms behind their regenerative properties, to repair ovarian damage and to promote follicular development.
Study Design
In the first phase, the effects of human plasma enriched in bone marrow stem cell soluble factors by granulocyte colony–stimulating factor mobilization, umbilical cord blood plasma, and their activated forms on ovarian niche, follicle development, and breeding performance were assessed in mouse models of chemotherapy-induced ovarian damage (n=7 per group). In addition, the proteomic profile of each plasma was analyzed to find putative proteins and mechanism involved in their regenerative properties in ovarian tissue. In the second phase, the most effective plasma treatment was validated in human ovarian cortex xenografted in immunodeficient mice (n=4 per group).
Results
Infusion of human plasma enriched bone marrow stem cell soluble factors by granulocyte colony–stimulating factor mobilization or of umbilical cord blood plasma–induced varying degrees of microvessel formation and cell proliferation and reduced apoptosis in ovarian tissue to rescue follicular development and fertility in mouse models of ovarian damage. Plasma activation enhanced these effects. Activated granulocyte colony–stimulating factor plasma was the most potent inducing ovarian rescue in both mice and human ovaries, and proteomic analysis indicated that its effects may be mediated by soluble factors related to cell cycle and apoptosis, gene expression, signal transduction, cell communication, response to stress, and DNA repair of double-strand breaks, the most common form of age-induced damage in oocytes.
Conclusion
Our findings suggested that stem cell–secreted factors present in both granulocyte colony–stimulating factor–mobilized and umbilical cord blood plasma could be an effective treatment for increasing the reproductive outcomes in women with impaired ovarian function owing to several causes. The activated granulocyte colony–stimulating factor plasma, which is already enriched in both stem cell–secreted factors and platelet-enclosed growth factors, seems to be the most promising treatment because of its most potent restorative effects on the ovary together with the autologous source.
Introduction
Aging involves a progressive loss of physiological integrity, leading to impaired function and increased vulnerability to death. The ovary experiences an early loss of functional integrity because of the highly demanding physiological requirements needed to function. , In addition to aging, poor ovarian response (POR) and other conditions of acquired damage can readily impair female fertility. In parallel with the increasing number of younger women who face conditions that impair ovarian function, socioeconomic changes are prompting delayed childbearing. Preservation of fertility through approaches that improve ovarian function presents an increasingly urgent need for ensuring reproductive success.
Why was this study conducted?
This study aimed to overcome the high prevalence of impaired ovarian reserve, for which the only practical treatment option is oocyte donation.
Key findings
Infusion of plasma enriched in soluble, secreted stem cell factors induced microvessel formation and cell proliferation in mouse and human ovarian tissues, promoting follicle development and restoring fertility.
What does this add to what is known?
Our results suggested that secreted stem cell factors may offer an effective treatment for patients with diminished ovarian reserve.
Despite this knowledge, damaged anovulatory ovaries may retain residual quiescent follicles that could be rescued in an optimal ovarian niche to give mature oocytes and contribute to the birth of healthy children. Previously, we showed that human bone marrow stem cells (BMDSC) promote follicular growth in both murine and human impaired ovaries by increasing ovarian vascularization and stromal cell proliferation and reducing cell death. Furthermore, we tested the therapeutic potential of such treatment in a prospective pilot study in women with POR. In the study, autologous stem cell ovarian transplant (ASCOT) improved ovarian function biomarkers in 81.3% of women, allowing pregnancies and live births in women with poor prognosis, suggesting that ASCOT optimized the growth of existing follicles, possibly mediated by specific BMDSC soluble factors, such as fibroblast growth factor-2 (FGF-2) and thrombospondin-1 (THBS1).
Paracrine signaling, through the secretion of soluble factors, such as cytokines, chemokines, and growth factors with key functions in tissue repair, is a main mechanism by which adult stem cells exert their regenerative effects. In fact, the administration of young growth factor–enriched plasma, umbilical cord blood (UCB) plasma, or plasma-specific proteins into damaged and aged organisms has recently been shown to stimulate nervous tissue regeneration and repair. These findings suggest that tissue damage and aging may be induced, in part, by changes in the composition of circulating factors present in blood and could be reversed by providing these factors.
Therefore, our study aimed to evaluate the role of noncellular components present in different sources of stem cell–rich blood as a possible tool to improve ovarian function.
Material and Methods
Ethical approval
All study procedures were approved and conducted according to the institutional review board of Hospital Universitario y Politécnico La Fe, Valencia, Spain (2014/0147), and the institutional review board and the ethics committee of the University of Valencia, Valencia, Spain (A1510574679987).
Plasma collection
Peripheral blood (PB) from 10 women with POR was obtained in EDTA K 2 BD Vacutainer tubes (BD Diagnostics, Spain). Moreover, an aliquot of the apheresis of these women with POR after a 5-day pharmacologic treatment with granulocyte colony–stimulating factor (G-CSF) was also collected as previously described. G-CSF treatment allows BMDSC to proliferate and mobilize to PB, where it can be collected by apheresis. In addition, UCB from 6 newborn girls was collected after birth at Hospital La Fe in K2EDTA BD Vacutainer tubes (BD Diagnostics). After collection, the plasma fraction of apheresis (G-CSF plasma) and blood samples (PB and UCB plasma) were isolated by a 1600-g centrifugation for 10 minutes at 4°C, pooled and stored at −80°C.
Portions of our 3 pooled plasma types were activated using 5% calcium chloride 0.1 M (CaCl 2 ; Sigma-Aldrich, St. Louis, MO) to release growth factors and signaling molecules enclosed in platelet α-granules, to obtain the activated PB (PBa), activated G-CSF (G-CSFa), and activated UCB (UCBa) plasma.
Assaying the regenerative effects of plasma in mice
Diminished ovarian reserve (DOR) and premature ovarian insufficiency (POI) were induced in 77 8-week-old female nonobese diabetic (NOD) and severe combined immunodeficiency (SCID) mice (Charles River, France) by a reduced (1.2 mg/kg busulfan [Bu]–12 mg/kg cyclophosphamide [Cy]) or standard (12 mg/kg Bu–120 mg/kg Cy) chemotherapy (ChT) dose, respectively. After 1 week, mice with DOR and mice with POI were randomly divided into 7 groups (n=11 per group) and administered with (1) 100 μL of PBS (ChT group), (2) 100 μL of PB plasma (PB control group), (3) 100 μL of PBa plasma (PBa group), (4) 100 μL of G-CSF–mobilized plasma (G-CSF group), (5) 100 μL of G-CSFa–mobilized plasma (G-CSFa group), (6) 100 μL of human UCB plasma (UCB group), or (7) 100 μL of UCBa plasma (UCBa group). Infusion was performed via tail vein with a 27 G needle, every other day for 2 weeks. The volume and duration of plasma treatments were established on the basis of previous animal studies , and taking into account our previous results with BMDSC , in a similar experimental setup. After plasma treatment, animals underwent ovarian hyperstimulation with 10 IU of pregnant mare serum gonadotropin (Sigma-Aldrich, St. Louis, MO) followed 48 hours later by 10 IU of human chorionic gonadotropin (hCG; Sigma-Aldrich, St. Louis, MO) and were mated with males. Furthermore, 7 mice per group were euthanized 36 hours after hCG administration to collect ovaries (1 fixed in neutral buffered formalin and the remaining stored at −80°C) and oviducts for assessing the short-term effects. The remaining animals were used to assess the long-term breeding performance ( Supplemental Figure 1 and Supplemental Material ).
Plasma proteome assessment
A total of 6 pooled group plasma samples were processed and analyzed by sequential windowed acquisition of all theoretical mass spectra as described previously, and a stepwise in-detail protocol is provided in Supplemental Table 1 .
Fold change (FC) values were calculated in 3 comparisons: (1) G-CSF and UCB groups vs PB control group, (2) G-CSFa and UCBa groups vs PBa group, and (3) each activated sample to its respective nonactivated fraction. Proteins with a Log 2 (FC) value above 1.5 were considered up-regulated, and those with a Log 2 (FC) value below 1.5 were considered down-regulated. Pathway analysis was performed using the PANTHER database, focusing on Reactome pathway results.
In vitro analysis of DNA damage and repair in mouse ovarian tissue
Notably, 12-week-old CD1 mouse ovaries were isolated, cultured for 24 hours, and then randomized to the following groups (n=6 per group): (1) ChT group, treated with 1.2 μM 4-hydroperoxy Cy + 0.12 μM Bu; (2) PB group, treated with ChT and PB plasma; (3) G-CSF group, treated with ChT and G-CSF plasma; and (4) UCB group, treated with ChT and UCB plasma. The ovaries were collected in 12 hours or 24 hours and pooled by group to analyze DNA damage by Western blot and DNA repair by quantitative reverse transcription polymerase chain reaction (RT-qPCR) as described in Supplemental Material .
Assaying the regenerative effects of activated granulocyte colony–stimulating factor plasma in xenografted human ovarian tissue
Here, 6 8-week old female NOD and SCID mice (Charles River, France) were randomly allocated to the control or G-CSFa group, anesthetized by isoflurane and ovariectomized following standard procedures. During the same surgery, 2 human ovarian cortex fragments were intraperitoneally grafted in each mouse. Ovarian fragments from the same patient were xenografted in both experimental groups to avoid patient-specific responses. After 1 week, the mice in the control group were administered 100 μl of saline, whereas the mice in the G-CSFa group were treated with 100 μl of G-CSFa plasma every other day for 2 weeks via the tail vein. Animals were then sacrificed and ovarian grafts (4 in the control group and 6 in the G-CSFa group) recovered and divided in 2 parts, one was immediately fixed in 4% paraformaldehyde to assess follicle development and ovarian stroma, whereas the remaining was stored at −80°C to evaluate proteomic changes ( Supplemental Material ).
The time point for analysis was established on the basis of our previous studies with BMDSC, , in which 14 days was shown to be the best time point to observe the highest regenerative effects in human ovarian tissue.
Statistics
Kruskal-Wallis tests followed by Mann-Whitney U tests for 2-by-2 comparisons were performed using GraphPad Prism (GraphPad Software Inc., San Diego, CA; version 8.12). P values of <.05 were considered statistically significant. Data are presented as violin plots with median and quartile values or bar charts with mean±standard deviation.
Results
Different plasma sources, enriched in stem cell–secreted factors, promoted follicle development in chemotherapy-induced ovarian damage mouse models
In the DOR model, ChT treatment reduced the total follicle number to 66.5% of wild-type levels. Treatment with PB, PBa, G-CSF, or UCB did not rescue this depletion ( Figure 1 , A–D; Supplemental Figure 2 ). However, G-CSFa and UCBa partially reduce the ChT-induced follicular depletion, recovering follicle numbers to 94% and 81% of the wild-type reference value, respectively ( Figure 1 , A). This rescue seemed to be largely driven by restoration of primordial follicles, with values similar to that observed in the wild-type group ( Figure 1 , B). Moreover, both activated plasma increased the number of late preantral follicles above wild-type levels, but only G-CSFa restored the number of early antral follicles ( Figure 1 , C–E). In fact, CSFa and UCBa treatments significantly increased the number of metaphase II (MII) oocytes and embryos recovered from the oviducts. Specifically, G-CSFa–treated mice produced 10-fold and 7-fold more MII oocytes and embryos, respectively ( P =.049 and P =NS) ( Figure 2 , A and B), and UCBa–treated animals produced 11-fold and 10-fold more oocytes and embryos, respectively ( P =.020 and P =.021).
In the POI model, ChT treatment reduced the ovarian reserve to 17.5% of wild-type levels and was not rescued by PB, PBa, or G-CSF treatments. Interestingly, UCB-treated mice showed a higher number of total follicles than ChT-treated mice ( P =.034) ( Figure 1 , A) mainly because of the primordial population ( P =.019) ( Figure 1 , B), although growing follicles, especially the late preantral population ( P =.032), also increased. G-CSFa and UCBa were more potent than their nonactivated forms at rescuing ChT-induced follicle loss, recovering values by increasing primordial, secondary, and late preantral populations. In fact, late preantral population values in these groups were similar or even higher than wild-type values ( Figure 1 , C). In addition, G-CSFa was also able to increase the number of antral stage follicles ( Figure 1 , D). A slight increase in primordial, secondary, and late preantral populations was also observed after PBa administration, although it was not as significant as that of the G-CSFa or UCBa effects ( Figure 1 , B and C; Supplemental Figure 2 ). The administration of UCB, UCBa, and G-CSFa plasma also had positive effects on the number of MII oocytes ( P =.049, P =.049, and P =.021 compared with ChT, respectively), and both G-CSFa and UCBa plasma increased the MII cohort to wild-type levels ( Figure 2 , A).
Long-term fertility was rescued by stem cell–secreted factor–based therapy
Female mice with DOR from all treatment groups achieved pregnancies and delivered pups at similar rates ( Figure 2 , C). However, both UCB and G-CSF treatments increased litter size ( P =.015 and P =.002 compared with the ChT group, respectively) ( Figure 2 , D). This increase was more significant when activated forms were used, especially for the UCBa group where litter size was similar to the wild-type value (10±1 pups).
In the POI model, all animals in the ChT and PB control groups failed to achieve pregnancy after several mating attempts. However, 40% of the G-CSF- and UCB-treated mice obtained pregnancy and offspring ( Figure 2 , C and D). The use of G-CFSa and UCBa increased pregnancy rates to 80% and 67%, respectively, with an average litter size of 8±1 pups in both cases. In addition, the administration of PBa plasma resulted in pregnancies and live births; however, the effect of PBa plasma to result in pregnancy or live birth was reduced ( Figure 2 , B–D).
Stem cell soluble factor–rich plasma regenerated the ovarian stroma
G-CSF and UCB plasma dramatically increased cell proliferation in ovarian tissue in both models, with the proliferative cells mainly identified as granulosa (DOR, G-CSF [ P =.0004] and UCB [ P =.0024]; POI, G-CSF [ P =NS] and UCB [ P =.004]; compared with the ChT group) ( Figure 3 , A and B). G-CSFa and UCBa plasma had more potent effects on proliferation than the nonactivated forms in the DOR (G-CSFa vs G-CSF [ P =.036]; UCBa vs UCB [ P =NS]) and the POI (G-CSFa vs G-CSF [ P =.006]; UCBa vs UCB [ P =.02]) models. In addition, PBa plasma increased cell proliferation but less dramatically than the G-CSFa and UCBa plasma ( Figure 3 , A).