This study showed that term amniotic fluid is a relatively rich source of MSCs having high expansion capacities and features similar to MSCs isolated from other sources. We also showed that harvesting amniotic fluid at term caesarean section deliveries by catheter resulted in much higher collection volumes compared with 10–20 ml volumes that were harvested in previous studies of term or pre-term amniotic fluid [2426]. We were able to harvest on average about 400 ml of fluid from the 20 samples. Comparing the TAF-MSCs described in this study with MSCs isolated from BM and AD, we show that they share a significantly similar transcriptional profile. Taken together, these results suggest that TAF-MSCs may be an alternative and complimentary source of MSCs with features making them highly attractive for further investigation in in-vivo pre-clinical functional assessments.

There are only a limited number of publications characterizing term human amniotic fluid cells [2426]. In line with these previous studies of term amniotic fluid and studies of second-trimester amniotic fluid [40], we have identified cells with either spindle-shaped or round-shaped morphology from term amniotic fluid, and assessed their growth characteristics. We have shown that the highly proliferative cells present in term amniotic fluid have the properties of MSCs, as characterized by cell surface markers, differentiation potential, and gene expression analysis. We also demonstrated the high proliferation capacity of TAF-MSCs. The three donor samples that were cultured for a longer period of time (10 passages, 41–46 days of cultivation including the time for initial colony formation) reached a maximum cumulative cell number of 1013 (range 109–1013, depending on the media type). The proliferation rate of TAF-MSCs is greater than adult sources of MSCs [21, 22], with the fastest doubling rate seen in this study being approximately 14 hours. We have observed differences in proliferation rates of cells growing in different media, which, as previously described for other MSC sources, could be due to different growth factors that are present in the respective media [41]. We have also observed a slow decrease in proliferation rate during the earlier passages of MSC expansion in all the media. A more rapid decrease in the proliferation rate of MSCs was seen only at higher passages (the actual passage number differs for each media; Additional file 4a,b). MSCs also demonstrated a continuous increase in fraction of cells enlarging in size over time in accordance with previous reports of MSCs from other sources [17, 41]. In addition, reduced differentiation and therapeutic potential of MSCs has been associated with prolonged in-vitro culture [1619]. Therefore, prolonged in-vitro cultures are not recommended, and in order to obtain the number of cells necessary for clinical applications, using the most abundant MSC sources that allow for short-term in-vitro expansion is advantageous. However, the MSC frequency is not the sole factor that needs to be considered when selecting the most appropriate MSC source for clinical application. Differences in in-vivo developmental potential of MSCs from distinct sources are also of major importance for consideration to ensure the efficacy and safety of the therapy [6, 7]. Thus, each MSC source needs to be evaluated for its in-vivo developmental potential and assure that these properties match the required criteria of each specific clinical application.

Reprogramming of cells from TAF-MSCs to true pluripotent states may offer an expanded utility of these cells for disease modelling, drug discovery and testing, and regenerative medicine. As a neonatal source, TAF-MSCs represent a valuable starting cell material for iPS cell generation considering their reduced exposure to mutagens compared with adult sources and their accessibility as a currently discarded material. Moreover, their adherent highly proliferative nature is also amenable to current reprogramming technologies. We demonstrated that TAF-MSCs can be reprogrammed to generate induced pluripotent stem (iPS) cells, which can then be efficiently differentiated to both haematopoietic and neural lineages. The TAF-iPS cell lines generated in this study showed significantly higher differentiation efficiency towards the erythroid lineage. It would be interesting to generate TAF-iPS cells from additional donors and assess whether efficient differentiation towards erythroid lineages is a general feature of TAF-iPS cell lines, or if it is a donor-related issue [42]. TAF-iPS cells could also differentiate into phenotypic NK (CD16/CD56-positive cells) and mature T cells (CD4/CD3 double-positive cells). The generation of mature lymphoid cells from pluripotent stem cell sources has been inefficient and remains a challenge [43]. When neural differentiation was induced, TAF-iPS cell lines demonstrated comparable neural differentiation ability in comparison with CB-iPS cell lines, and significantly less mesodermal differentiation as measured by COL3A1, thus generating a more homogeneous culture. The broad potential donor base of these cells and their neonatal origin support the idea of establishing iPS cell bank of normal and diseases patient samples from these cells. Together, our data show that TAF-MSCs have features desirable for generating iPS cells, and therefore potentially have value in pluripotent stem cell-based regenerative medicine applications.

Several studies have reported that the majority of isolated cells from second and third-trimester amniotic fluid express the stem cell marker OCT4 (POU5F1) [36, 4447]. The expression of OCT4 has been associated with pluripotency, and this was further advanced by showing the capacity of single or clonal OCT4-positive cells to differentiate towards various cell lineages [36, 45]. Comparative study of amniotic fluid-derived MSCs from second and third-trimester pregnancies revealed comparable levels of OCT4 expression in these cells regardless of their gestational age [48]. Consistent with previous reports, we have also identified that the majority of freshly isolated cells from term amniotic fluid are OCT4-positive. Given the interest in further evaluating OCT4-positive amniotic fluid cells in pre-clinical studies, here we demonstrate that the closed catheter extraction system is compatible for extraction of large number of OCT4-positive cells. We and others have shown that OCT4 expression is lost over time in culture [46], but it has been reported that hypoxia helps maintaining OCT4 expression and stem cell properties [49]. The closed catheter-based system can minimize air exposure compared with active suction devices.

The average number of caesarean section deliveries in OECD countries is approximately 3 million per year (approximately 26% of all births) (Statlink: http://dx.doi.org/10.1787/888932524887). Depending on the exclusion criteria for the procedure of donating term amniotic fluid (e.g. emergency C-section), and given that not every eligible donor would probably accept to participate, the actual numbers of term amniotic fluid samples collectable is difficult to predict. However, unlike collection of MSCs from some traditional sources where age-related concerns for cellular function and age/health restrictions for the medical extraction procedure exist, term amniotic fluid acquisition during a planned CS delivery could potentially provide relatively large numbers of MSCs without these complicating factors.