Transfection of mesenchymal stem cells at physiological oxygen concentrations

Aim. Numerous works showed the beneficial effect of oxygen in physiological concentrations on mesenchymal stem cells (MSC) cultures. The aim of the present work was to study the impact of physiological oxygen tensions on non-viral transfection of MSC from human Wharton jelly (WJ-MSC). Methods. WJ-MSC at passage 2 were cultivated for 48 hours in different gas mixtures: nitrogen-based (oxygen – 3 %, CO 2 – 5 %, nitrogen – 92 %) and argon-based (oxygen – 3 %, CO 2 – 5 %, argon – 92 %), while the control group was cultivated under standard conditions of CO 2 -incubator (ambient oxygen concentration, CO 2 – 5 %). After 48 hours of cultivation the cells were transfected with polyplexes pEGFP-C1/PEI/ and pEGFP-C1/TurboFect, containing 3 μg of plasmid DNA. The cultures were in contact with com­ plexes for 1 hour, at standard conditions of CO 2 -incubator. Next, the media were changed, and WJ-MSC were cultivated for 48 hours in described gas mixtures, and CO 2 – incubator. Transfection effectiveness was estimated by flow cytometry as the number of eGFP+ fluores­ cent cells, (BD FACSAria). Results. In the groups cultivated in gas mixtures containing 3 % of oxygen, the number of cells, synthesizing eGFP+ before and after the transfection procedure, was on average 2.58 times higher in the nitrogen-based mixture and 1.37 times higher in the argon-based mixture than in control groups from the CO 2 –incubator. Conclusions. The cul­ tivation of human WJ-MSC under physiological oxygen tensions allowed an increase in the percentage of transfected cells, and is promising to be used as a method for optimization of transfection.


Introduction
Mesenchymal stem cells (MSC), a population of adult stem cells, are considered to be per spective tool for regenerative medicine [1]. At present, their therapeutic potential is attributed N. S. Shuvalova, E. K. Toporova, V. A. Kordium ISSN 1993-6842 (on- to their unique immunomodulatory and para crine activity. The ability to produce a wide range of bioactive molecules is viewed as the underlying mechanism of their therapeutical effects.

Methods
Considering, that the original number of MSC in different tissues is comparatively low [2], their ex-vivo multiplication is required for obtaining the cells in amount, sufficient for scientific research and clinical application [3]. Since the widely acknowledged condi tions of cultivation differ from those of MSCs natural location sites, the process of in vitro cultivation is often accompanied by the cell damage, degradation of culture and, as the result, loss of the therapeutic potential [4][5][6]. Thus, one of the major tasks of regenerative medicine is the development of cultivation protocols for MSC (and other populations of stem cells) to provide the maximal preserva tion of their original therapeutically signifi cant properties [4].
At present, one of the novel methods, that allows an increase in the amount of therapeu tically significant molecules and paracrine factors, and enhancement of the therapeutic potential of MSC from various sources, is the transfection of MSC with target genes, using various viral and non-viral constructions [7][8][9]. The important emerging biotechnological task is to optimize the existing systems for the delivery of genetic material into cells, and to develop the new ones. Main disadvantages of the existing viral transfection methods include toxicity, immunogenicity, and potential ge netic instability in cells, leading to onco gene sis. All of above drawbacks restrict the pos sibility of practical application of these me thods [10]. Nonviral vectors, comparing to viral vec tors, are considered to be safer, noninfectious, non-immunogenic, and less toxic. They can be obtained by simple methods in large amounts and have the ability to transfer the genes with large size [11]. Non-viral methods are thought to be optimal for the cases, when the constitu tive transgene expression is not necessary. But their major disadvantage is a low transduction efficiency [12]. Therefore, the development of methods, that allow the improvement in the non-viral transfection efficiency and the en hancement of the following expression of tar get product is considered to be a perspective trend in biotechnology. One of the most im portant tasks is the optimization of above methods for transfection of MSC.
It was shown, that one of the MSC distin guishing features is their localization in the sites of organism with low oxygen concentra tion, irrespectively of their tissue origin [13]. At present, the MSC cultivation at physiolo gical oxygen concentrations (approximately 1.5-8 %) is considered to be the promising method, that allows preventing pretime senes cence [14], enhancement of migratory poten tial and increasing the number of survived cells after transplantation to ischemic sites [15,16]. Additionally, the enhancement of MSC para crine activity in conditions of mild hypoxia is detected [17]. The studies showed the effec tiveness of various strategies of using lower oxygen concentrations: from long-term culti vation [14,18] to shortterm preconditioning [19,20]. Considering the present literature, we hypothesized that conducting the transfection in conditions of mild hypoxia would have beneficial effects on the cell survival and the expression of target product.
The aim of the present work was to study the impact of low oxygen concentrations on the effectiveness of transfection and to deter mine whether the cultivation in mild hypoxia conditions can be used as a method for opti mization of non-viral polyplex-gene delivery to stem cells in vitro.

Materials and Methods
Obtaining the primary MSC cultures from Wharton jelly. MSC were isolated from human Wharton jelly (WJ) by the explant method, as described in our previous works [21,22]. Umbilical cords (UC) were obtained from three healthy donors (39-40 weeks of gesta tion, normal delivery), after informed consent. The UC fragment (5-10 cm) was washed twice with PBS, the blood vessels were mechani cally removed. WJ was minced, and the frag ments were placed in the cultural flacks, 75 cm 2 , with complete expansion medium: aMEM(BioWest, supplemented with 10 % fetal bovine serum (HyClone), penicillin 100 U/ml (Arterium, Ukraine), streptomycin 100 μg/ml (Arterium, Ukraine). The attached cells were observed on the 5-7th day. After approx. 14 days, the clones reached the suf ficient number (>10) and size, and 70-80 % confluence. After that, the cells were passed with the trypsin-EDTA (0.1 % trypsin and 0.02 % EDTA) solution.
At passage 1 the cells were studied for the expression of surface marker proteins CD90, CD73, CD105 (over 90 % positive), CD 34 and CD45 (negative) using flow cytometry (BD FACS Aria) with fluorescein -and rho damineconjugated antibodies (UsBiological, USA), as descirbed in our previous works [23,24] . Moreover, the cell cultures were checked for their differentiation ability, as described in our previous works [25].
Gas mixture preparation. For the present study, a system for preparation of nitrogen and argon-based gas mixtures was developed, and described in our previous works [26]. The system consists of 3 gas cylinders with carbon dioxide, nitrogen and argon, air pump, the gas mixing device and the terminal container. The terminal container is connected to the gas an alyzer PGA 200 (JSC "Electronstandartpribor", Russian Federation). The gas mixture of required concentration was prepared by regulating the gases outflow from cylinders with air pump and monitoring the indications of gas analyzer.
The flacks and Petri dishes with WJ-MSC cultures were placed into the polyethylene bags with hermetic clasp (1.5 l). Additionally, the Petri dishes with sterile water were placed, to maintain the humidity in the bags. The bags were washed twice with the gas mixture, con taining 5 % of CO 2 and 95 % of nitrogen or argon (depending on the group), after that they were filled with the prepared gas mixture and placed in the vacuum containers (Scarlet). The containers with bags were kept at 37° C.
All liquids (cultivation media and PBS), used in the transfection procedure, were pre conditioned in required gas mixture, in Petri dishes, for 2 hours. According to the literature, the "liquid media/gas" ratio must be 1:100 [27], to provide normal gas exchange, and the gas concentration in liquid phase becomes the same as in gas phase after maintenance for 1 hour.
After 48 hours of cultivation the cells were transfected with polyplexes pEGFP-C1/PEI/ and pEGFP-C1/TurboFect, containing 3 μg of plasmid DNA and 6μg of PEI. Cultures were in contact with complexes for 1 hour, in stan dard CO 2 -incubator conditions [29]. After that, the transfection mixture was removed. The cells were washed with PBS twice, and the complete growth media was added.
Fluorescence of transgenic eGFP in trans fected cells was observed with an inverted microscope Leica DMIL, at excitation range 470/40 nm and emission wavelength ~530 nm (suppression filter 520/530 nm).
After that, transfection effectiveness was estimated as the number of eGFP-positive cells (cells expressing fluorescence) by flow cytom etry (BD FACSAria), using the FACS Diva software.
Statistics. Seven independent experiments were performed. Statistical significance of the difference was determined using Mann Whitney U-test at P < 0.05.

Results and Discussion
The major task of the work was to study the impact of conditions of mild hypoxia on the transfection efficiency of human MSC of Wharton jelly. The WJ-MSC cultures were used at passage 2, because our previous works showed that at this passage WJ-MSC preserve high level of proliferative activity, can be ob tained in sufficient amounts [21], and have the most homogeneous morphology, without the presence of senescent cells [28].
The characterization of obtained cultures at passage 1 showed, that they fit the generally accepted MSC criteria, being over 95 % posi tive by CD90, CD73 and CD105, and >0.3 % negative for CD34 and 45 (Fig 1.) MSC cultures, maintained in two types of gas mixtures, containing 3 % oxygen (nitro genbased and argonbased ones), and in stan dard CO 2 -incubator conditions (ambient oxygen concentration), were transfected with pEGFP-C1 DNA plasmid. Five experiments were conducted using PEI, and two -with TurboFect. The fluorescence of transgenic eGFP in transfected cells became detectable after 24h posttransfection. However, the most intensive fluorescence could be observed after 48 hours (Fig 2). Thus, the general fluores cence intensity, and the percentage of trans fected cells were measured by flow cytometry (BD FACSAria) 48 hours after transfection. We also examined the morphology of cultures, and estimated the" live to dead" cells ratio.
Observation by light microscopy did not show any visual difference in morphology between the cells in control and experimental groups. In general, the cells preserved normal spindleshaped morphology, despite the slight "darkening" of cytoplasm and appearance of small vacuoles in some cells, that can be viewed as a mark of probably slight toxicity of transfection mixture (Fig. 2).
However, the task of present pilot work was to compare the effects of mild hypoxia conditions. Because of the detected difference between the variants, to evaluate the results regarding the effect of low oxygen concentra tions, we estimated and compared the ratio between the percentage of eGFP+ cells in experimental groups (from nitrogen-based and argon-based gas mixtures), to the groups from standard CO 2 incubator conditions ( Table 1). In all experiments, the number of eGFP+ cells was greater in the groups maintained in conditions of mild hypoxia. The most promi nent positive effect was detected in nitrogen based mixture. On average, the number of fluorescent eGFP+ cells in nitrogen-based mixture, was significantly, 2.58-fold higher, than in control groups, whereas only 1.37fold in argon-based mixture (Fig. 5). Interestingly, in our previous works, the positive impact on proliferative potential and MSC morphology in nitrogenbased and ar gon-based mixtures was nearly at the same level, only slightly stronger in nitrogenbased mixture [21,28]. However, in the groups, cul tivated in argon-based mixtures, the enhance ment of transfection effectiveness was much less pronounced, despite the same concentra tion of oxygen in mixture.
The " Live\dead" assay performed with BD FACSAria showed that the percentage differ ences between control and experimental groups were not significant (on average, 89.6±5.29 % in control group, 85.85±7.4 % in nitrogenbased mixture, and 88.4±5 % in argon-based mixture, data shown as mean±SD). However noteworthy, the number of live cells appeared to be slightly lower in both gas mixtures.
We realize, that in this pilot study, the ob tained data are not sufficient for full statistical treatment. The individual features of the donor (like, potentially, the peculiarities of protein expression) may be the cause of marked dif ference between the cultures. Interestingly, other researchers have also described the dif ference in transfection efficiency between the cultures, obtained from different donors. For example, Madeira et al. described, that in the set of experiments, the percentage of GFP+ cells was 4fold higher for one of the donors, though the number of plasmid copy per cell was approximately similar [30].
Despite the current interest in developing safe methods of genetic modification of MSCs, we should note that rather small share of all transfection research involves nonviral trans fection prodecures [31]. Besides, the difference in experiment conduction hampers the com parison and analysis of the data, known from literature. Moreover, it is important to note that MSC properties may vary in the populations of different tissue origins and different donors [1,3]. As far as we know, at present, our work is the only one, which focuses on using the cultivation under physiologic oxygen concen trations as a transfection optimization method, so we are not able to compare our results with the data, obtained by other investigators. However, the enhancement of the efficiency of transfection shown in this work may be explained by other findings concerning the At present, many researchers described, that the conditions of mild hypoxia can increase the number of proliferating cells in MSC cul ture, and in WJ-MSC cultures as well [32,33]. Recently, Zhang et al. [18] showed, that MSC from bone marrow, cultivated under lower oxygen tensions (1 % and 5 %), had higher proliferation rates and greater percentage of cells in S + G2/M-phase, comparing to nor moxia. Possibly, enhancement of proliferation in MSC cultures under low oxygen tensions is often accompanied by the shift in cell cycle propagation, comparing to standard cultivation conditions.
There are several works which show that the effectiveness of transfection, especially for the methods using cationic carriers, depends on the cell cycle phase [34,35] and the prolif erative characteristics of cultures. For example, some works demonstrated that the cultivation conditions, inhibiting the division of human MSC, can reduce transfection efficiency [36].
Usually, in research works on mammalian cell transfection, large number of initially plated cells is used to facilitate further analysis (for example, flow cytometry) [37]. In this connection, it is important to mention the data of Boura et al. [38], who described the impact of plating density of MSC on transfection ef ficiency, and showed the possibility to enhance the efficiency of MSC transfection by perform ing the procedure on the actively proliferating MSC cultures. The authors [38] used the initial plating density (1000 and 3000) that allowed the culture to proliferate. The transfection pro cedure was conducted in 72 hours to confirm the division of cell in the cultures. This ap proach provided an increase in the effective ness of lipofection by two -to three-fold, comparing to their previous works.
We hypothesize, that in our work the en hancement of transfection efficiency in the cultures, maintained under physiological oxy gen tension, can possibly have similar underly ing mechanism. In our previous works, we showed that MSC cultures had higher prolif eration rates under 3 % oxygen, comparing to the standard conditions of CO 2 incubator [21]. Though, it is important to note that in the pres ent work we used a higher plating density (nearly 9400 per cm 2 ) than in the experiment with MSC multiplication (3000 cells per cm 2 ). However, the monitoring of the cultures mor phology showed the presence of dividing cells, which indicates that during the transfection procedures, at least a part of the culture was in the state of active proliferation (Fig. 3).
The enhancement of transfection efficiency was more prominent in the nitrogenbased gas mixture. This finding is in-line with our results on the WJ-MSC proliferation. Also notewor thy, the statistical evaluation did not show significant difference between the results from argon-based gas mixture and control group, though in 5 of 7 experiments the number of eGFP+cells was larger. In the present work, the period of WJ-MSC cultivation before trans fection was much shorter, than the routine maintenance at one passage during the multi plication in vitro. Interestingly, in our previous works, the beneficial effects of argon-based gas mixtures manifested themselves only in longterm cultivation [22]. This fact is in ac cordance with the findings in the present work, but needs future investigation.
The shift in the cell cycle propagation can probably be a possible reason for a higher transfection efficiency under lower oxygen tensions. In this connection, we hypothesize, that the experiments, conduct ed on the synchronized MSC cultures would be useful in the further studies on this phe nomenon. But still, the use of physiological oxygen concentrations seems to be perspec tive approach to increase the effectiveness of transfection.

Conclusions
The developed method of using the cultivation under physiological oxygen tensions as the method of transfection optimization for human WJ-MSC allowed the increasing of the per centage of transfected cells. The impact of nitrogen-based and argon-based mixtures, con taining oxygen in the same concentrations, differed: the effect of argon-based mixture was significantly less pronounced.