In this study, we first used a microfluidic quantitative PCR system to elucidate the gene expression profiles of single human oocytes and small groups of cumulus cells and subsequently improved upon in vitro maturation of cumulus-free oocytes to obtain parthenogenic blastocysts. We also observed that parthenotes from in vitro matured oocytes have similar mRNA expression levels of the majority of studied genes compared to IVF/ICSI embryos at the blastocyst stage.
Analysis of gene expression in single human oocytes and small groups of cumulus cells revealed notable similarities and differences from previous reports including human and model organisms. FSH, LH, estradiol and EGF are commonly used to promote maturation of cumulus-enclosed oocytes. According to the mRNA expression analysis in our study, there is negligible expression of the receptors for FSH, LH and EGF in human oocytes. This is consistent with a recent animal study 
but contradictory to a previous report showing FSHR
mRNA expression in mouse oocytes and pre-implantation embryos 
. Furthermore, parthenote development in Sage medium with hormone supplementation did not improve late stage (further than 8-cells) embryo development compared to IVM-medium alone. Our gene expression and embryo development data suggest that FSH and LH may not affect oocyte nuclear and cytoplasmic maturation directly. In contrast, both ESR1
, the receptors for estradiol, were detected in oocytes as well as cumulus cells in our study, suggesting a role for estradiol in oocyte maturation 
. Therefore, the higher cleavage rate in Sage medium compared to IVM-medium alone is most likely due to the estradiol supplementation in Sage medium.
Both animal and human studies have demonstrated the important roles of neurotrophins (BDNF, NTF3 and NTF4) in oocyte maturation mediated by their receptors (NTRK1–3 and NGFRAP1) 
. We observed expression of BDNF
mRNA and its receptors (NTRK2
) in both oocytes and cumulus cells. The higher expression of NGFRAP1
in cumulus cells suggests that BDNF may have a role in cumulus cell functions. It is important to note that the probe for NTRK2
in our study detected both the full length as well as the truncated forms of NTRK2
. We suspect that the NTRK2
mRNA expression observed in the present study is the truncated form, since the full length form was not detected in either human oocytes or cumulus cells previously 
. Interestingly, NGFRAP1
was detected in oocytes in our study but not in the previous study 
. This may reflect the greater sensitivity of our microfluidic qPCR system over other traditional qPCR systems. Also, the previous study was conducted with cumulus-oocyte-complexes from small antral follicles without hormone stimulation, which may dramatically change the mRNA expression profiles compared to those undergoing hormone stimulation.
Expression of the mRNA encode for Glial cell line-Derived Neurotrophic Factor (GDNF) was detected in cumulus cells as shown previously 
as well as oocytes in our study. GDNF is secreted by cumulus, granulosa, and theca cells as an ovarian factor stimulated by the preovulatory LH surge 
, and induces oocyte maturation and embryo developmental competence 
by GDNFR and rearranged during transformation (RET). Based on the findings that GDNFR is only expressed in oocytes, and a low level of RET expression was detected in both oocytes and cumulus cells, our data suggest that GDNF might increase oocyte maturation by both a paracrine and autocrine manner in humans.
At the mRNA level, FGF2
was highly expressed in cumulus cells, while its receptor FGF2R
was highly expressed in oocytes in our study. FGF2 and its receptors have been shown to be expressed at the protein level by both oocytes and granulosa cells up to secondary follicles 
, but the expression patterns in antral follicles have not been studied to date. Based on our gene expression data, FGF2 could affect oocyte maturation in a paracrine manner via the cumulus cells. This notion is supported by a previous study showing a higher concentration of FGF2 in follicular fluid compared to serum 
and suggests that FGF2 production from the ovary has pivotal functions during follicle development and oocyte maturation.
Two parthenogenic blastocysts were produced from in vitro
matured GV oocytes from hormone-stimulated cycles in our study. This is a proof of principle to demonstrate that immature oocytes are able to achieve the developmental competence with selected ovarian factors in vitro
. In spite of ovarian factor supplementation, nuclear maturation of GV oocytes was not improved compared to MI stage oocytes. It may be due to the fact that the majority GV oocytes might be collected from pre-antral follicles which are not able to resume meiosis 
. However, detailed protein expression of individual ovarian factors and their receptors in oocytes need to be further studied in order to elucidate the factors that may affect oocytes in specific maturation stages.
Once we had optimized the in vitro
maturation of oocytes, we were able to obtain parthenogenic morulas and blastocysts in culture. It is noteworthy that this is the first study to show the gene expression profiles of human parthenogenic embryos at any stage. Overall, parthenotes have similar mRNA expression levels as IVF/ICSI embryos at the blastocyst stages ( and ). The altered expression patterns of Cyclin A1 (CCNA1) in parthenotes changed from high (in morula) to low (in BLs) compared to IVF/ICSI embryos, which may due to it inefficient zygote genomic activation 
. The altered expression of CCNA1 is similar to the expression of EIF1AX, another marker for zygotic genomic activation in our study as well as in a previous animal study 
. However, we cannot suspect any abnormal effect on parthenotes due to the aberrant expression of CCNA1 since the protein is not expressed at the blastocyst stage 
, and its cell cycle regulation role is limited in meiotic cells 
. Higher expression of cytoplasmic polyadenylation element-binding protein 1 (CPEB1
) in parthenotes compared to normal embryos may increase cell cycle arrest and cellular senescence by inducing p53 translation 
. This may cause lower cell number and higher apoptosis in parthenotes compared to normal embryos 
The mRNA expression pattern of growth factor receptors in parthenotes were different compared to normal embryos, suggesting that growth factors supplemented in oocyte/embryo culture media may alter the expression of their receptors 
. Interestingly, BDNF improves pregnancy rates by increasing trophoblast cell growth and survival via its receptor NTRK2 expressed by the trophectoderm of blastocysts 
. The effect of higher expression of NTRK2
mRNA in parthenotes need to be further studied.
Aberrant imprinting is the major cause of embryo lethality of parthenogenic embryos. In our study, morula stage parthenotes exhibited high expression levels of MEG3
. Maternally expressed MEG3 is regulated by a intergenic differentially methylated region in human gametes and embryos 
. A possible explanation for this finding is that residual mRNA expression from oocytes was still present in morula stage parthenotes and/or chemical activation induced the aberrant expression of certain genes in parthenotes. Paternally expressed PEG1 exhibited significantly higher levels in parthenotes at the morula stage and dramatically decreased at the blastocyst stage. The high expression levels of two genes (DNMT3A
) related to DNA methylation in our study is also consistent with a mouse study showing the elevated methylation in parthenotes compared to IVF/ICSI embryos 
. Analogous to CCNA1 and EIF1AX discussed above, the higher expression of de novo methyltransferase, DNMT3B
, in parthenotes may be due to the altered zygotic activation since DNMT3B expression has been shown to be originated from embryo 
. The regulation of epigenetic modifications in human pre-implantation embryos needs to be explored in further depth in order to understand the underlying molecular mechanisms behind early human development as well as the low efficiency of parthenogenesis and SCNT.
In our study, cleaved day 3 parthenotes were used to establish pESCs from single blastomeres using derivation methods as previously described 
. Of the blastomeres isolated, approximately 21% (8 out of 38) divided once, but no blastocele could be detected in arrested parthenotes. Only blastomeres from non-arrested 8-cell stage parthenote formed blastoceles and proliferated (data not shown). From the 8 blastomeres isolated from an 8-cell parthenote, 3 blastomeres divided, which suggested that those 3 blastomeres are the most viable from the original 8-cell parthenote. This interpretation is supported by a previous study showing that the blastomeres from arrested human embryos still divided in vitro
when isolated from the original embryos 
. The results indicate the potential for individual blastomere growth and development from parthenotes obtained from in vitro
maturation in spite of the fact that we did not derive a pESC line at this time. It would be useful to compare the potential of day 2 parthenotes which are more abundant since an ESC line was recently derived from 4-cell blastomeres 
In conclusion, cumulus-free human oocytes from hormone-stimulated cycles can be matured in vitro with the supplementation of ovarian paracrine/autocrine factors to achieve embryo developmental competence to blastocysts. Also, parthenogenic blatocysts from IVM oocytes exhibit similar gene expression profiles of selected genes compared to IVF/ICSI. Parthenogenesis is one of key steps towards successful human SCNT. Our improved IVM culture conditions may be used for supplying mature oocytes for regenerative medicine, including pESC and SCNT-ESC derivation.