Search form

UZ Gent afdeling reproductieve geneeskunde

Contact details
Traineeship proposition
Abstract Bachelor Project FBT 2018-2019: Practical approach of the MIQE guidelines: Determining of reference genes for gene expression analysis during the in vitro maturation of primordial follicles from ovarian tissue

Before carrying out a quantative gene expression analysis, it is important to determine the reference genes so that a normalization of the gene expression of the target genes can occur correctly. According to the MIQE guidelines at least 8 reference genes must be tested and the three most stable reference genes must be included for further analysis with the target genes. In this study, we investigated the stability of 12 reference genes in in vitro culture of ovarian tissues under different conditions to determine the three most stable reference genes for this application.

A total of 27 pieces’ ovarian tissue were obtained from 3 transgender patients (9 pieces per patient). The tissue pieces were thawed, washed, stretched and cultured in supplement McCoy medium with or without sphinogosine-1-phosphate (S1P) (12 µM) and Jasplakinolide (JASP) (10

µM) at 37 °C (6 % CO2 en 5 % O2). After incubation, pieces were put in culture during 4 days. At several time points of culture (day 0, 2 and 4), pieces of tissue were used for gene expression analysis. RNA extraction was performed using RNeas Mini kit (Qiagen) and subsequent cDNA preparation was performed using the cDNA iScript synthesis kit from BioRad. For determination of the 12 references genes (geNorm kit) (18S, ACTB, GAPDH, ATP5B, B2M, CYC1, EIF4A2, TOP1, RPL13A, SDHA, UBC en YWHAZ) sample maximization was done with the cDNA samples of all patients and conditions by real-time qPCR using iTaq Universal Sybr Green Supermix (BioRad) and the CFX96 PCR-system (Biorad). The geNorm algorithm was used to establish the best combination of the most stable expressed reference genes. Of these three most stable reference genes, a standard curve was made to control the PCR- efficiency.

The experiments in this study show that adding JASP or S1P to the in vitro culture resulted in different amount of cDNA during the 4 days of culture. Generally, the tissue exposed to JASP showed higher amount of cDNA compared to the control group and S1P showed lower amount of cDNA. All cDNA samples (JASP, S1P and control from day 0, 2 and 4) were used for determination of the 12 reference genes. The geNorm algorithm excluded 4 reference genes (SDHA, CYC1, YWHAZ and ATP5B) because of low or absence of signal. The stability and suitability of the other 8 reference genes were further analysed. The algorithm geNorm showed that TOP1, B2M and UBC were the three most stable reference genes in this experimental setting, with a minimum M-value of 1,42. The M-value is higher than the reference according to the MIQE guidelines (0,5-1 M-value). It is clear that the M-value of 1 is difficult to reach when heterogeneous tissue pieces are used in a dynamic experimental setup.

For this experimental setting, according the MIQE quidelines, the reference genes TOP1, B2M and UBC are the most stable reference genes. When carrying out a quantative gene expression analysis in an in vitro ovarian tissue culture with JASP and S1P in order to study the in vitro maturation of primordial follicles, these three reference genes must be included to normalize the expression from the target genes.


Abstract Bachelor Project MLT 2018-2019: Optimization of seed sample processing by gradient centrifugation in order to reduce DNA fragmentation after processing

One out of six couples with a child-wish copes with an infertility problem. In the earlier days it was most likely that a female factor was considered as cause for the absence of a pregnancy. In the 20th century the possibility of a male factor, more precisely a semen quality aspect was also taken into account. From then on the semen analysis became a part of the infertility assessment. During this semen analysis, quality aspects like volume, concentration and motility of the semen is performed.

The aim of this research is to compare a two-layer and a three-layer gradient centrifugation for the processing of a semen sample. Based on the yield and DNA fragmentation it will be decided which is the better method. If the three-layer gradient has a better yield and reduced DNA fragmentation, it may be implemented.

A total of 20 semen samples were analysed with the double-layered (90/45) and three-layered (90/70/45) gradient centrifugation. The native sample was also analysed. Concentration, motility, volume and DNA fragmentation were determined on these samples. The DNA fragmentation was assessed microscopically by a junior and senior laboratory technician, each counting 200 sperm cells. The senior lab technician was an experienced evaluator, the junior lab technician was new to this microscopic assessment and received a short introduction and training. A Blant and Altman and a One Sample T test were performed on the counts of both lab technicians to determine if there was a statistically significant difference between the two analysts. The yield was calculated based on volume, concentration and motility. The Fisher Exact method was used to determine if there was a significant difference between the two gradients.

When the average yield was calculated, it was obvious that there was a higher yield for the three-layer gradient ( 83,8 %) than for the two-layer gradient (60,3 %) . A statistically significant difference (p=0.0003) was visible between the two-layered and three-layered gradients in selecting a population motile sperm.

DNA fragmentation testing showed that there was only a minimal difference between the two- layered and three-layered gradients. In the three-layer gradient there is a minimal increase, but the three-layer gradient remains an improvement compared to the native sample. Despite this minimal increase, there was no statistically significant difference (p=0.8686) between the gradients.

It can be concluded that the three-layer gradient has a statistical higher yield, namely 83.8%, compared to the two-layer gradient which only has a yield of 60.3%. Although DNA- fragmentation was higher in the three-layer gradient than in the two-layer gradient, this difference did not reach significance. Based on these two parameters, it can be concluded that the three-layer gradient is better than the two-layer gradient. As a result, the three-layer gradient will be implemented.

Abstract Bachelor project 2017-2018: Insight into the in vitro maturation of follicles from ovarian tissue by influencing the Hippo pathway


In vitro maturation of follicles starting from primordial follicles in ovarian tissue is currently not possible in humans. Telfer et al. (Telfer, McLaughlin, Ding, & Thong, 2008) has developed an in vitro system that seems promising, but is still far from implementing in a clinical setting. Female cancer patients can cryopreserve an ovary, keeping their fertility safe from destructive harm resulting from aggressive cancer therapies. The purpose of this tissue is to transplant it back after the patient has been cured. However, for certain types of cancer, mainly blood cancers, there is a possibility for reintroducing malignant cells. For these patients, in vitro maturation of primordial follicles starting from ovarian tissue could be the solution for using their stored fertility potential.

The Hippo pathway is a highly conserved signaling pathway that regulates the organ size, also in the ovary. The actin polymerization and the molecules Macrophage stimulating 1 (Mst-1) and Yes associated protein (YAP) play a very important role in this pathway and thus in the maturation of follicles. It is the aim of this thesis, to manipulate the in vitro maturation of follicles starting from ovarian tissue by adding certain components in the culture.

Material and methods

Ovarian tissue (36 pieces in total) obtained from 3 patients (12 pieces per patient), was used for in vitro maturation according to the Telfer system. In short, pieces were thawed, stretched and cultured in supplemented McCoy medium with or without sphingosine-1-phosphate (S1P) (12 µM) and jasplakinolide (JASP) (10 µM) at 37°C (6% CO2, 5% O2). After incubation, pieces were put in culture during 6 days. At several time points (d0, d2, d4 and d6), pieces of tissues were fixed and embedded and paraffin coupes were made (5 µm). Haematoxylin staining was performed for follicle count and classification (according to Gougeon (Gougeon, 1986)). Immunofluorescent staining was performed for Mst-1, YAP and phosphorylated YAP (p-YAP) in order to analyse the localization of these proteins in the developing follicle.


The experiments in this study show that adding JASP or S1P to the in vitro follicle culture results in stimulation follicle growth. A clear shift in primary follicles was observed on d2 and on d4 for S1P and JASP, respectively. From the immunofluorescence staining it can be shown that during the maturation of the follicles there a clear change in the localization of the staining per type of follicle: YAP and p-YAP are located primarily in the oocyte of primordial follicles whereas in intermediate and primary follicles, the proteins are both present in the granulosa cells and the oocyte. For Mst-1, the localization pattern was more diffuse and no clear conclusion could be reached. It was observed that in the in vitro cultures where S1P and JASP were added, more primary follicles showed a positive staining for YAP and p-YAP in both the granulosa and the oocyte.


The manipulation of the in vitro culture with S1P and JASP resulted in a shift in follicles showing an enhanced activation of primordial follicles. The shift by S1P was already observed at d2, whereas the cultures manipulated by JASP, showed a maturation shift towards primary follicles on d4. In growing follicles it is clear that important molecules from the Hippo pathway find their way from the oocyte to the granulosa cells indicating the proliferative response in the granulosa cells of the growing follicle.

Abstract bachelorproef 1 2016-2017Quantification of gene expression in human cumulus cells in function of the maturity of the oocyte after in vitro maturation

More and more people and often at young age struggle with a gender disorder. A surgical intervention where the individual also changes its gender physically can only occur when the person is an adult and has followed cross hormone therapy for at least one year. When female to male transgenders, also called trans men have a future childwish, they often choose to freeze ovarian tissue at the moment of this transition surgery. During the preparation of the ovarian tissue, immature oocytes can also be obtained and they can be further matured in vitro. The cumulus-oocyte-complex is very important in the maturation of the oocytes. It is known from literature that the cumulus mass (CM) is positively correlated with the maturation stage of the residing oocyte. The more layer of cumulus cells (CM0 -> CM2) there are around the oocyte, the more likely the oocyte is mature. The aim of this study was to evaluate the expression of certain genes in cumulus cells according to maturation status of the oocyte after in vitro maturation in samples derived from different thickness of cumulus mass (CM). 

Cumulus cells were obtained mechanically from in vitro matured oocytes and scored according to the characteristics of the CM: where CM1 was characterized as having >3 and <10 layers of cumulus cells, and where CM2 as having >10 layers of cumulus cells. The maturation stage (GV, MI, MII) of the oocytes inside the cumulus complex was determined after removal of the cumulus cells. For gene expression analysis, RNA extraction was performed using the RNeasy Mini Kit (Qiagen) and subsequent cDNA preparation was performed using the cDNA iScript synthesis kit from Invitrogen. For the determination of the reference genes (geNorm kit) (RPL19, ACTB, 18S, YWHAZ, CYC1, RPL13A, EIF4A2, TOP1, ATP5B, B2M, GAPDH, UBC and SDHA) and the primer testing of the target genes (Serpine2, GJA1, CX43, STAR and RGS2) cDNA of eighteen cumulus cell samples from individual cumulus-oocyte-complexes were used. The geNorm algorithm was used to establish the best combination of the most stable expressed reference genes. For target gene analysis correlating to the maturation status of the oocyte, nine CM1 samples (three GV, three MI, three MII) and twelve CM2 (four GV, four MI, four MII) were analyzed by real-time qPCR using the iTaq Universal Sybr Green® Supermix (Biorad) and the CFX96 PCR-system (Biorad). 

The algorithm geNorm showed that RPL19, 18S and ACTB were the three most stable reference genes with an M-value of 1,4. Gene expression levels of Serpine2 in cumulus cells were significantly higher in GV than in MII in both the CM1 and the CM2 group. This difference in gene expression was statistically relevant (Mann Whitney U test p≤0.05). CX43 levels has the highest expression in MI of both CM1 and CM2 and GJA1 has the highest expression in both groups in the MII samples. STAR has the highest expression in MI in CM1 samples while there is no difference in gene expression in the CM2 group. The gene expression levels for RGS2 in CM1 are the highest in GV, lower in MI and lowest in MII but by CM1 is the expression in GV lowest.

There is a difference in gene expression of Serpine2 where the expression is highest in the GV group and drops in relation to the maturation stage. This result is in accordance with what is found in the literature. Increase of the sample size and inclusion of a control group (in vitro matured oocytes obtained at the time of oophorectomy in oncology patients) will undoubtedly aid in the robustness of the data and the possibility to make conclusions for the expression pattern of the other target genes.

Abstract bachelorproef 2 2016-2017Qualitative and quantitative comparison of six staining methods used in the morphological assessment of semen

One out of six couples with a child wish copes with an infertility problem. In the earlier days it was most likely that a female factor was considered as cause for the absence of a pregnancy. In the 20th century the possibility of a male factor, more precisely a semen quality aspect was also taken into account. From then on the semen analysis became a part of the infertility assessment. During this semen analysis, quality aspects like volume, concentration and morphological evaluation of the semen is performed. Morphological assessment of sperm is generally performed by staining of a fixated semen smear and subsequent microscopic evaluation.

The aim of this project is to compare six staining methods used in the morphological assessment of semen. Based on this comparison the SpermBlue technique, used in the laboratory for diagnostic semen analysis at the Ghent University Hospital, will be evaluated.

In total 50 semen samples were analyzed with 6 different staining methods; 4 methods starting from a semen smear (Sperm Stain, SpermBlue, Spermac Stain, Shorr Stain), 2 methods using a prestained slide (Goldcyto SB prestained slides, Testsimplets ® prestained slides). The morphological evaluation was performed by a senior and junior lab technician, each counting 200 spermcells. The senior lab technician was an experienced evaluator, the junior lab technician was new to this morphological evaluation and received a short introduction and training. A qualitative and quantitative comparison between the 6 methods was made. The comparison of the morphological assessment from the 6 different tests was analyzed by using both continuous data (% of normal morphology) and the categorical data (normal or abnormal). Statistical analysis was performed using SPSS (v23): a reliability analysis, Friedman test, Mc Nemar test and related samples Cochran’s Q test where p-value ≤0.05 was considered statistically relevant.

When the 6 methods were compared from a qualitative point of view it was shown that vacuoles could be clearly visualized by all the staining methods apart from the Spermac Stain. The postacrosomal region was visualized by all six techniques and was excellent when the Shorr Stain and the Spermac Stain were used. The tail and midpiece of the sperm were nicely stained by all staining methods. The cytoplasmic droplets were only visible using the two prestained slides.

When the 6 methods were statistically analyzed using the continuous data it was shown that the reliability between the two raters was highest using the SpermBlue method. The results of the morphological scoring demonstrated a statistical difference (p= 0,001) between the Goldcyto SB prestained slides and the Shorr Stain when the evaluation is conducted by senior lab technicians.

There was no statistical difference between the 6 methods when the categorical data was used. The latter was confirmed by using the Mc Nemar test, a test which only takes into account how many times the same diagnosis (abnormal versus normal) is determined by the two raters. The related samples Cochran’s Q test also concludes there is no statistical significant difference between the 6 methods when evaluated by the senior lab technician or the junior lab technician. Although not statistically significant, it was noted that using the Goldcyto SB prestained slides for microscopic morphological sperm evaluation resulted in a higher % abnormal spermatozoa whereas the Testsimplets in a higher % normal spermatozoa when assessed by a junior lab technician.

Although there are some qualitative differences between the 6 staining methods, it is clear that they can all be used for morphological semen scoring. It is important to note that cytoplasmic droplets can only be clearly visualized using the prestained slides. The fixation steps and subsequent drying of the smears in the other staining methods are most probably the cause for the disappearance of these abnormalities.

There was only 1 statistical difference in the qualitative morphological assessment of the sperm; the scoring of the samples with the GoldCyto SB prestained slides was different than the result using the Shorr Stain. 

This study concludes that the morphological assessment between the 4 smear staining methods leads to similar results and care should be taken when using prestained slides since the loading of the spermatozoa might could lead to suboptimal positioning of the sperm and this could have an effect on the morphological assessment. Based on the current results, there is no need to change the currently used staining technique (SpermBlue) at the Ghent University Hospital.


Implantation is a complex process that requires a synchrony between the embryo and the endometrium. Even with the transfer of high quality embryos, implantation can fail.
Repeated implantation failure is defined as a condition in which no implantation and thus no pregnancy was achieved after three IVF/ICSI cycles, with the transfer of several embryos being of good quality and of appropriate developmental stage. There are many requirements for successful implantation, occurring during a limited time frame referred to as the window of implantation. Unfortunately the chance of a successful implantation is impaired when not all these conditions are fulfilled. The underlying causes and molecular basis for a non-receptive endometrium are not yet elucidated.

It is important that the endometrium displays an adequate decidual response during the window of implantation. During decidualization endometrial stromal cells undergo a profound transformation which enables these cells to serve as a biosensor and to interact with an embryo. It is likely that in specific groups of patients in which implantation fails repeatedly, there is a deficient decidual response.

This study intended to mimick the decidual response in vitro in an immortalized cell line (St T1b), serving as a control cell line, as well as in primary cell lines derived from endometrial biopsies from patients with repeated implantation failure. Phenotypical and biochemical changes were evaluated in these cell lines, as differences in the decidualization process could explain why these patients do not obtain pregnancy despite multiple transfers of embryos of good quality.

Primary endometrial stromal cells and the St T1b cell line was induced to decidualize for 5 days by adding 1µM cAMP and 0.5 mM MPA or was left undifferentiated. During this decidualization process, the cells were evaluated by light microscopy. After 5 days of decidualization or being left undifferentiated, the cells were collected in Trizol. After the separation of phases, proteins were separated from RNA and DNA with an optimized extraction. Proteins were then separated by SDS-PAGE, and consecutively two decidualization markers (Interleukine-11 (IL-11) and prolactin (PRL)) were detected by semi-dry western blot.

As a result, this study shows that all the used cell lines can be decidualized in vitro. Morphological decidualization was evident by light microscopy showing the reorganization from elongated to polygonal shaped cells with prominent nucleoli. This decidual phenotype seemed more pronounced in the primary cell lines as compared to the St T1b cell line.

After decidualization or being left undifferentiated, the cells were collected and proteins were extracted. Despite the fact that the total number of cells collected after decidualization, the protein content extracted from these was lower than the protein content from non decidualized cell lysates. Markers of decidualization IL-11 and PRL could not be detected by western blot, not in the decidualized cells, nor in the non decidualized cells. These markers however, could be visualized through immunocytochemistry. A potential explanation could be that the protein concentration is too low for the chosen technique (SDS-PAGE and western blot), further protein enrichment could be necessary. Furthermore, it is possible that a large part of proteins are secreted into the supernatant, which could explain the lower protein content after decidualization.

In conclusion we state that it is possible to decidualize in vitro, as can be examined based on morphological characteristics. With the chosen techniques we were not able to detect biochemical markers of decidualization. Further optimalization of the detection technique, as well as the analysis of the supernatant enable the detection of potentially secreted markers. 

Samenvatting eindwerk 1 2014-2015: Analysis of critical parameters (temp., pH) from different technical processes during IVF treatment
Objective To evaluate temperature and pH during different steps of the IVF treatment.
Background oocytes and embryos are very sensitive to changes in their surroundings. Especially temperature and pH have optimal ranges. The temperature upon which oocytes and embryos should be manipulated is between 35 and 38 °C and the pH optimum of the media used have an optimal range from 7.3 to 7.5.
Design The analysis consists out of four experiments. The first experiment uses a pH meter to monitor the behavior of different media used during the treatment. The media are first incubated at 37 °C in an atmosphere of 6% CO2, they are then placed twice onto the heated surface of the IVF workstation. The first time they are gassed for ten minutes and the second time they are exposed to the normal atmosphere. The second and third experiment have a similar design, except they are used to measure the temperature on different appliances, the IVF workstation and the ICSI stations. First the temperature of the surface is measured. Afterwards culture plates are placed on different zones of the work area. Temperature is measured inside the culture drop at the beginning and after ten minutes. The measurement of the ICSI station has two additional measuring points. One after twenty minutes and another after thirty minutes. The fourth experiment is used for temperature measurements of different manipulations during the procedure of  Ovum Pick-Up (OPU). The OPU is performed with different non-heated or pre-heated recipients and after every step the temperature is measured.
Results and conclusion In general the pH-control during IVF procedures was good. The pH fluctuations observed during manipulations are mostly within the specific pH. The gassing of the culturedishes during manipulation seems to be unnecessary because of the rapid changes in the pH, especially in situations where no embryo oil is used to cover the medium. It’s therefore recommended to leave the dishes as long as possible inside the incubator. The temperature control in the IVF lab stays mostly within the range of 35 °C to 38 °C and just to be sure it’s better to work in the front zones of the workstation. However, there should be mentioned that the temperature of one IVF workstation is set too high and one ICSI station is set too low. This definitely needs adjusting. To guarantee optimal results of fertilization it’s advised to preheat all recipients during the OPU procedure in order to minimize temperature fluctuations. In conclusion, optimization of certain steps in specific IVF procedures at the IVF lab UZ Ghent coud minimize changes in pH an temperature and keep oocytes and embryos at their optimal culture conditions.
Samenvatting eindwerk 2 2014-2015: Laser capture microdissection (LCM) as a method to isolate follicles from ovarian tissue in which by the gene expression in the granulosa cells and the oocyte can be analysed separately
Chemotherapy, radiotherapy and surgery have led to an improved survival in cancer. These treatments are gonadotoxic and can therefore induce premature ovarian failure. Currently ovarian transposition and cryopreservation of embryos, oocytes or ovarian tissues are the main options for female fertility preservation to avoid these negative effects. Transplantation of this cryopreserved and thawed ovarian tissue is possible and successful, but the risk of reimplantation of malignant cells is real, and should be avoided. In vitro culture and maturation of primordial follicles from this tissue could overcome this problem. However, the development of the optimal culture conditions for primordial follicles remains a challenge in the reproductive technology. Therefore more insight into the underlying mechanisms of primordial follicle survival, activation and development needed.
This project aims at isolating oocytes and granulosa cells from follicles in order to analyze the gene expression in these cells. Results of this study will lead to a better understanding of the mechanisms of primordial follicle survival, -development and -activation. Better understanding generates new insights into underlying pathways of follicle maturation and gives input setting up an in vitro maturation technique.
Laser capture microdissection (LCM) is the technique used to isolate the oocytes and granulosa cells from different follicles. First, the optimized settings of the LCM method for separating the granulosa cells, stromal cells and oocyte in paraffin embedded ovarian tissue slices were determined. Optimization of the RNA-extraction and subsequent RT-qPCR was performed using an ovarian granulosa carcinoma cell line as model where the detection of the reference gene was verified. Using this cell culture model, the PCR efficiency, influences of paraffinization, deparaffinization and LCM on the references gene detection could be examined.
Results of this study showed that it is possible to detect a reference gene using RT-qPCR starting from as little as 50 cells dissected from a paraffin embedded slide using LCM, that Based on this optimized protocol, further gene expression analysis on LCM isolated follicles should be possible as 50 ovarian granulosa carcinoma cells are the equivalent of 5 primordial follicles.
In conclusion, LCM is an able to separate the granulosa cells from the oocytes in primordial follicles and subsequent gene expression analysis is possible on low amount of cells. The current optimization of this procedure will undoubtedly yield important insights into the molecular pathways important in follicle maturation and will have a direct input into, the development of the optimal culture conditions for setting up an follicle in vitro maturation technique.
Samenvatting eindwerk 2013-2014:Optimisation of treatment and selection of sperm cells from hepatitis C positive patients
Couples where the male partner is infected with hepatitis C (HCV) are encouraged to use proper prophylactic measures when having sexual intercourse in order to prevent the infection of the female partner. This forms a problem when the couples desire to have children. Practising safe sex eliminates the chance of achieving pregnancy, while having unprotected sexual intercourse holds a risk of infection for the female partner.
To help these couples achieve pregnancy without risking the health of the female partner, the couples are encouraged to seek assisted reproductive treatment. If the sperm of the partner is of very poor quality, the couple relies on intracytoplasmatic sperm injection (ICSI) treatments to have a successful pregnancy. This treatment is currently unavailable to the aforementioned HCV infected couples in the department of reproductive medicine of the Ghent University Hospital. In the process of injecting the sperm cell into the oocyte there’s a risk of also injecting a viral particle present in the surrounding seminal plasma.
Therefore the objective of this thesis was to create a semen washing protocol to clear the seminal plasma of viral particles.
Artificial mixtures of semen and HCV positive plasma were used to test several protocols. Subsequent washes of seminal plasma and semen samples were tested for the presence of HCV. This validation process lead to an efficacious procedure that was able to eliminate viral particles from the sample. Moreover, after finishing the procedure, there was enough semen left for a subsequent fertility treatment.
The protocol was verified in real semen samples obtained from a HCV positive patient, where it was confirmed that the semen had an undetectable viral load after the semen washing procedure. The semen washing protocol for HCV positive patients will be implemented in the department of reproductive medicine to be used for couples for ICSI treatment where the partner is HCV positive.


De pintelaan 185
9000 Gent
De pintelaan 185
9000 Gent


Traineeship supervisor
Kelly Tilleman
Via Map