Complement factors are secreted in human follicular fluid by granulosa cells and are possible oocyte maturation factors


Dr Anny Usheva, 330 Brookline Ave, CLS 701, Boston, MA 02215, USA. Email:


Aims:  In this study, we identify components of the complement system present in human follicular fluid that affect oocyte development and maturation.

Material and Methods:  Using bottom-up liquid chromatography/mass spectrometry/mass spectrometry, we identified complement factors as consistently present in human follicular fluid from 15 different subjects.

Results:  According to our gene-chip data, these complement factors are actively produced by granulosa cells.

Conclusions:  By applying the computational Ingenuity Pathway Analysis software and database we have identified complement pathways that play a role in oocyte maturation and follicular development.


Human follicular fluid (hFF), in close proximity to maturing oocytes, is essential for the biological process of oocyte maturation in the growing follicles. The proteins of hFF contain substances implicated in oocyte meiosis, differentiation, ovulation, and finally fertilization.1

Evidence suggests that hFF is formed mainly as a consequence of both selective transport from plasma and metabolic activities of granulosa cells.2 Conventional knowledge maintains that blood plasma is the main source of proteins in hFF. Recent data, however, suggest that granulosa cells produce and release several proteins and molecules in follicular fluid (FF).3 Recently, our group, as well as several other groups, reported that a large network of interconnected signaling pathways in hFF is involved in the oocyte maturation.4

The complement system, composed of over 30 proteins, acts as an intricate immune surveillance system to discriminate among healthy host tissue, cellular debris, apoptotic cells and foreign intruders, and tunes its response accordingly. Complement proteins also participate in diverse processes, such as synapse maturation, angiogenesis, tissue regeneration and lipid metabolism.5 Complement factor B, and factors I, C3 and C4 have been found in the hFF by the proteomics approach.6 Another study using an alternative bottom-up approach found that complement factors C1S, C3, C4A, C7, C9, and factor B are in hFF.7 Neither of these studies, however, addresses the source of these complement factors in hFF.

Our hypothesis is that the granulosa cells actively participate in oocyte development by synthesizing complement factors to support oocyte maturation and its genomic preservation.

The aim of this study is to determine the components of the complement system in hFF obtained from women undergoing IVF by using liquid chromatography (LC)/mass spectrometry (MS)/MS. Further, we compare these results with gene-chip data for granulosa cells in the same patients to determine sources of these complement factors. Finally, the Ingenuity Pathway Analysis (IPA) help identify the most probable complement pathways involved in controlling human oocyte maturation.


Patient population

Healthy women aged 18–29 years who were undergoing oocyte retrieval for oocyte donation or for treatment of male-factor infertility were included in this study. FF samples were collected from 15 women undergoing IVF with controlled ovarian hyperstimulation at Boston IVF. The stimulation protocol was previously described.8 A follicle was labeled as ‘lead’ and used in the experiments if the mean diameter was ≥16 mm. We included 15 healthy women, aged 18–29 years, 10 patients undergoing oocyte retrieval for oocyte donation, and five for male-factor infertility (Table 1). After written informed consent was obtained, participants completed a brief health and reproductive history questionnaire. The study has been approved by the Beth Israel Deaconess Medical Center institutional review board.

Table 1. Summary characteristics of the 15 healthy donors from Boston IVF whose follicular fluid was used in the analysis
 Age (years)BMI (kg/m2)Peak E2 (pg/mL)Total gonadotrophin dosage (IU)No. of folliclesNo. of oocytes retrieved
  1. Healthy women aged 18–29 years who were undergoing oocyte retrieval for oocyte donation or for treatment of male-factor infertility were included in this study. Follicular fluid samples were collected with controlled ovarian hyperstimulation. A follicle was labeled as ‘lead’ and used in the experiments if the mean diameter was ≥16 mm. Ten patients were undergoing oocyte retrieval for oocyte donation, and five for male-factor infertility. BMI, body mass index.


FF and granulosa cell collection

We used the standard stimulation protocol, which included downregulation with either a gonadotrophin-releasing hormone agonist or antagonist followed by hyperstimulation with recombinant follicle stimulating hormone. When more than three follicles over 18 mm were visible on vaginal ultrasound, patients were given human chorionic gonadotrophin (hCG), and vaginal oocyte retrieval was performed 36 h later. During the procedure, an ultrasound was used to measure all follicles in two dimensions prior to aspiration and collection. The fluid from 3–5 lead follicles was then pooled and used for the analysis. The blood-free FF was centrifuged at 13 000 rpm for 20 min to separate cells, debris and FF. The supernatant was frozen and used for FF protein analysis. The pellet was resuspended in 400 µL of trypsin and incubated for 10 min. After that, the pellets were overlaid on 750 µL 50% pureception (CooperSurgical) Eppendorf tubes and were centrifuged at 8000 r.p.m. for 15 min. The cell layer in between layers of cellular debris and matrix proteins was collected and put in 500-µL of trypsin-neutralizing solution followed by a second spinning at 8000 r.p.m. for 15 min. After completely removing the supernatant, 600 µL of RLT buffer was added to the granulosa cell pellet for total RNA isolation and gene-chip analysis as previously described.9

LC/MS/MS analysis of protein in the FF

Pooled hFF from 15 patients was depleted of immunoglobulins, using a top-12 depletion column (GenWay IgY 12 HC LC10). The individual hFF flow-through samples had been pooled and concentrated on a 7KD MWCO spin filter. Samples were then reduced, alkylated and digested with trypsin. Following digestion, HyperSep C18 solid phase extraction media was used as a clean-up step before fractionation with cation exchange chromatography. The resulting 24 fractions were then run on a linear ion trap orbitrap hybrid mass spectrometer (LTQ-Orbitrap) instrument. Samples in 5% (v/v) Acetonitrile 0.2% (v/v) formic acid were injected with a Thermo-Fisher Scientific MicroAutosampler onto a 75-µL × 25-cm fused silica capillary column packed with Hypersil GOLD C18 3-µm media, in a 250-µL/min (1000:1 split to column) gradient of 5% (v/v) Acetonitrile, 0.2% (v/v) Formic acid to 30% (v/v) Acetonitrile, 0.2% (v/v) formic acid over the course of 180 min with a total run length of 240 min. The LTQ-Orbitrap is run in a top-4 configuration at 60K resolution for a full scan, with monoisotopic precursor selection enabled, and +1, and unassigned charge state rejected. The analysis on the LTQ-Orbitrap instrument was carried out with higher-energy C-trap dissociation (HCD) fragmentation, 15K resolution for the MS/MS events, with a normalized collision energy level of 45.

RNA isolation and gene-chip analysis of granulosa cells

Total RNA was extracted using RNeasy Mini Kit, QIAshredder, and was eluted in a final volume of 60 µL. Concentration and quality were assessed using NanoDrop ND-1000 spectrophotometer. Gene-chip analysis has been conducted by using Agilent microarray gene chip applying the protocol of the supplier.10 Two independent RNA pools from the selected 15 patients were prepared and analyzed independently. Each pool contained an equal amount of RNA from five patients.


Protein interactions and pathways were analyzed using the IPA software and database. Ingenuity's knowledge base is compiled from the scientific literature, supported by experimental results and structured into ontology in the knowledge database. We analyzed 426 proteins from the total protein list and 235 proteins from the gene-chip matched protein list separately, as well as data published by others for genes transcribed in granulosa cells. For network analysis, both direct and indirect interactions between the molecules were analyzed and displayed as a solid and dotted line, respectively. To find the most relevant canonical pathways, we picked pathways that were statistically significant, with P-values of less than 0.05 by Fisher's exact test.


Through LC/MS/MS, we were able to identify most of the components of both common and alternative pathways of the complement system (C1qA, C1qB, C1qC, C1r, C1s, C2, C3, C4BP, C4A, C4B, C7, C8 alpha, C8 beta, C9, complement factor D, complement factor H, complement factor H-1, complement factorH-5, complement factor I, complement properdin, Mannan-binding lectin serine protease-1) (Fig. 1) in hFF. Many of these complement factors are detected by our mass spectroscopy proteomic approach (Table 2) as well as by Western blot analysis (data not shown) for the first time, and have not been previously described by other proteomic approaches. In addition, we found that C5 and C6 are absent from hFF; LC/MS/MS showed, however, that C5 and C6 are present in serum of these same patients.

Figure 1.

Common and alternative pathways of the complement system in human follicular fluid. Ingenuity Pathway Analysis using the ingenuity knowledge database for the total 426 proteins and 234 gene-chip matched proteins. Shape shows function: square (cytokine), horizontal diamond (peptidase), circle (others), double circles (complex/group). Gray color indicates the total proteins, orange circle indicates gene-chip matched proteins. C, complement component; C1QA,B,C, complement component 1, q subcomponent, A, B, C chain; C1R (S), complement component 1, r (s) subcomponent; C4BPA, complement component 4 binding protein, alpha; MASP1, mannan-binding lectin serine peptidase 1; MBL2, mannose-binding lectin 2; SERPING1, serpin peptidase inhibitor, clade G member 1.

Table 2. Presence of complement system proteins in granulosa cells and FF
AccessionGenemRNA presenceProteinRT-PCR
  1. Content of mRNA of complement system genes in granulosa cells was identified by gene-chip microarray. Subsequently, the presence of protein was verified by MS. The level of mRNA for selected genes was also assessed by RT-PCR as indicated above. C, complement; MS, mass spectroscopy; NA, not analyzed; RT-PCR, real-time polymerase chain reaction.

NP_057075.1C 1q A+MSNA
NP_000482.3C1q B+MSNA
NP_001107573.1C 1q C+MSNA
NP_000055.2C 3+MS+
NP_000706.1C 4 BP+MSNA
NP_009224.2C 4A+MSNA
NP_001002029.3C 4B+MSNA
NP_000553.1C8 alpha+MSNA
NP_000057.1C8 beta+MSNA
NP_000597.2C 8 gamma+MSNA
NP_001728.1C 9+MS+
NP_000177.2C H+MSNA
NP_002104.2C H-1+MSNA
NP_110414.1C H-5+MSNA
NP_000195.2C factor I+MSNA
NP_001870.3Mannan-binding lectin serine protease- 2+MSNA

Furthermore, we utilized the gene chip for granulosa cell RNA. Our results showed that mRNA of complement factors, except for C5 and C6, is present in granulosa cells (Table 2). By matching the results with data from LS/MS/MS, we found for the first time that these components of the complement system are produced by granulosa cells.


The complement system participates in diverse processes, such as immune response, synapse maturation, angiogenesis, tissue regeneration and lipid metabolism.5 However, the role of complement in oocyte maturation is still unclear. It is intriguing that C5 and C6 were found not to be produced by the granulosa cells in this study.

The complement cascade is activated by three pathways: classic, alternative or lectin pathways. All these three pathways merge into a final effector, the membrane attack complex (MAC), which is composed of C5b, C6, C7, C8 and C9.5 Several studies demonstrated that C5b-9 MAC might be involved in the pathophysiology of pregnancy-related disorders. One study found that MAC caused villous trophoblast injury in vivo.6 It has been shown that MAC co-localizes to areas of fibrin, and binds to the surface membrane of trophoblasts, increasing apoptosis of cytotrophoblasts exposed to lower-than-ambient oxygen tension.11 Another study found that patients with unexplained fetal death and those developing pre-eclampsia had a higher level of serum C5a than normal pregnant women.4 This result indicates that C5 might be implicated in the process of fetal injury. As fetal tissues are semi-allogenic, the placenta is subject to the damage by the MAC. C5a has also been found to be a key mediator of fetal injury in murine models of antiphospholipid antibody syndrome. The damaging effects of C5a have been prevented by antibodies that block C5a–C5a receptor interaction.12 Given the deleterious effects of C5 during pregnancy, we speculate that from an evolutionary perspective, humans with C5 and C6 present in the hFF were infertile, leading to the elimination over time of these complement factors from the oocyte milieu. Previously we demonstrated that the granulosa cells also produce the enzyme heme oxygenase −1,13 which could provide another protective mechanism from a complement-induced damage.14,15 The absence of C5 and C6 from the hFF seemingly protects the oocyte from injury by the MAC. It may represent a survival advantage for oocyte maturation.

The function of the complement system is usually regarded as an important system in host defense and inflammation. The absence of C5 and C6 from the hFF blocks an important step in the activation of the complement cascade. This phenomenon argues against the immunologic role of the complement system in FF and thus begs questions regarding the function of the complement system identified here.

C3 is an important factor in the complement system with several functions in repair16 and early embryogenesis.17 Complement, and specifically C3a, also has important regulatory effects on the differentiation and migration of neural progenitor cells thereby affecting neurogenesis.18

Besides its role in immunity, repair, embryogenesis, and inflammation, evidence demonstrates that it plays important roles in the reproductive process: C3 plays a bridge role in fertilization by enhancing spermatozoa-egg membrane apposition.19 The complement receptors CR1 and CR3 are expressed in human oocyte. Through cleavage of C3 to C3b by sperm acrosomal protease, C3b binds to the membrane cofactor protein (MCP) on the sperm, as well as CR1 and CR3 on the oocyte together, which facilitate the process of apposition. In the chicken, C3 is transported into oocytes by LR-8-mediated endocytosis and becomes a yolk component of the egg.20 Another group found that oviduct cells secrete inactivated C3b, which stimulates growth and development of the embryo.21 This group also found C3 deficiency impairs early pregnancy in mice.22 In their study, C3-deficient mice had a longer estrogen cycle length, a higher post-implantation loss, and lower fetal and placental weight as compared to normal mice. All of these data suggest that C3 in hFF has a role in maturation and development of the oocyte. Further studies are needed to elucidate the role of the complement factors in oocyte maturation.

Our data are the first to demonstrate that most of the complement factors are present and C5 and C6 are absent in hFF. Our experimental approach combining, gene-chip analysis and LC-MS/MS goes beyond the established proteomics approaches and has made it possible for the simultaneous identification of complement system components at both mRNA and protein levels. The RNA data from the gene-chip analysis provides some insight into the function of the granulosa cells in hFF. These findings provide insights into the molecular basis for understanding the role of granulosa cells in the follicular process and in controlling the process of oocyte maturation.


The authors would like to thank Dr Richard Reindollar, Department of Obstetrics and Gynecology, Dartmouth Medical School for advice and help with experimental work, analysis and editing; Dr Alan Penzias, Boston IVF for helpful discussions; and Dr Chong Shou and Martin Lange, Department of Endocrinology, Beth Israel Deaconess Medical Center, Boston for editing.


The authors have no relevant commercial or research-related conflict of interest to disclose.