Differential gene expression of human chondrocytes cultured under short-term altered gravity conditions during parabolic flight maneuvers

Background Chondrocytes are the main cellular component of articular cartilage. In healthy tissue, they are embedded in a strong but elastic extracelluar matrix providing resistance against mechanical forces and friction for the joints. Osteoarthritic cartilage, however, disrupted by heavy strain, has only very limited potential to heal. One future possibility to replace damaged cartilage might be the scaffold-free growth of chondrocytes in microgravity to form 3D aggregates. Results To prepare for this, we have conducted experiments during the 20th DLR parabolic flight campaign, where we fixed the cells after the first (1P) and the 31st parabola (31P). Furthermore, we subjected chondrocytes to isolated vibration and hypergravity conditions. Microarray and quantitative real time PCR analyses revealed that hypergravity regulated genes connected to cartilage integrity (BMP4, MMP3, MMP10, EDN1, WNT5A, BIRC3). Vibration was clearly detrimental to cartilage (upregulated inflammatory IL6 and IL8, downregulated growth factors EGF, VEGF, FGF17). The viability of the cells was not affected by the parabolic flight, but showed a significantly increased expression of anti-apoptotic genes after 31 parabolas. The IL-6 release of chondrocytes cultured under conditions of vibration was not changed, but hypergravity (1.8 g) induced a clear elevation of IL-6 protein in the supernatant compared with corresponding control samples. Conclusion Taken together, this study provided new insights into the growth behavior of chondrocytes under short-term microgravity. Electronic supplementary material The online version of this article (doi:10.1186/s12964-015-0095-9) contains supplementary material, which is available to authorized users.


Background
Joint friction at the extremities of long bones is reduced by articular cartilage. This kind of tissue is highly specialized, avascular, not innervated and consists mainly of a single cell type: the chondrocytes. The chondrocytes are tightly embedded in an extracellular matrix (ECM), which is composed of a network of collagens (predominantly collagen type II) and aggrecan. The collagen network contributes to the strength and mechanical resistance of cartilage tissue, whereas the aggrecan, a proteoglycan, comprising such molecules as chondroitin sulfate or keratin sulfate, is responsible for its viscoelasticity and flexibility due to its ability to absorb and retain considerable amounts of water [1][2][3]. Degenerative diseases of the cartilage like osteoarthritis are characterized by a progressive degradation of the ECM, caused by the increased secretion of matrix metalloproteinases (MMP) [4,5]. This process is triggered by pro-inflammatory cytokines, such as tumor necrosis factor-α (TNF-α) or interleukin-1β (IL-1β), which are transported into the cartilage via the synovial fluid [6,7]. The absence of vasculature and the extremely limited influx of chondrocyte progenitor cells [8] limit the tissues healing and restorative potential, so that in advanced stages of osteoarthritis a surgical replacement of the affected joint with a prosthesis is usually necessary.
Apart from the influences of age and use intensity, other factors have been found to mediate cartilage integrity. Most notably, microgravity (μg) has a strong impact on cartilage. It has been reported, that astronauts, after staying a longer time in Space, suffer from a reduction of cartilage mass [9] due to mechanical unloading. In addition, when cultured in Space, cartilage tissue showed a reduced aggrecan density and a less-organized collagen subtype 2 organization hinting towards an impaired resistance to mechanical stress and breaking [10,11]. It is therefore of high interest, to study the impact of microgravity on chondrocytes. On the one hand, this will help to understand the detrimental effects of prolonged stays in Space on the cartilage, but on the other it might also help to find ways to counteract this phenomenon in Space as well as to ameliorate cartilage problems caused by wear on Earth in the future.
Microgravity (μg) provides unique conditions for cell and tissue growth. It has been shown on various cell types, including chondrocytes, that cultivation under conditions of μg can induce the formation of 3D aggregates. These aggregates are especially interesting, as they do not require any potentially interfering scaffolding like those being generated under normal gravity conditions [12][13][14][15][16].
Cell cultivation in Space, however, is an extremely complicated and expensive venture. Therefore, pre-studies such as a parabolic flight, providing 31×22 s of short-term real μg, or simulated μg on ground-based facilities such as the rotating wall vessel bioreactor, clinostats or the random positioning machine (RPM) have been established [17][18][19]. Especially the parabolic flight is an attractive method to achieve real μg without going into Space. It should be taken into consideration, however, that every μg-phase is flanked by two 20 s-lasting hypergravity phases of 1.8 g. Moreover, during the flight vibrations occur and have to be taken into account for interpretation of the results [20].
This study aimed to investigate the influence of shortterm real μg during a parabolic flight campaign on the gene expression profiles of cultivated chondrocytes. In addition, the effects of 1.8 g hypergravity as well as of vibration in an extent comparable to those during a parabolic flight were separately investigated.

Influence of hypergravity and vibration on soluble factor release
Hypergravity induced a more than 2-fold increase in the release of IL-6 in the supernatant (Figure 2A). IL-8, EGF, VEGFD and FGF17 concentrations in the culture supernatant of hypergravity samples were below the detection limit. ELISA analysis revealved no significant change in the concentration of IL-6 protein in chondrocytes cultured under conditions of vibration ( Figure 2B), whereas IL-8, EGF, VEGFD and FGF17 concentrations in the culture supernatant were below the detection limit.

Parabolic flight maneuvers induced expression changes in chondrocytes
The influence of the parabolic flight after 1P and 31 P was investigated. Shortly, the 1 g vs. first parabola (1P) vs. 31st Parabola (31P) set was subjected to an F-test. Resulting significant differential expressed probes (5% FDR) were clustered using k = 6. Individual expression characteristics of the 6 clusters are documented in Figure 3.

Quantitative real-time PCR
In addition to the microarray analysis, we also employed the quantitative real-time PCR technique to validate selected genes of interest. CCNA2 as well as IL8 were significantly upregulated only after 31P ( Figure 5A, D). CD44 and TNFA were significantly upregulated only after 1P ( Figure 5B, G). VCAM showed a significant downregulation after 31P ( Figure 5E). No effects were observed for IL6, EDN1 and FGF9 ( Figure 5C, F, H).

Discussion
In this study we investigated the effects of short-term real μg, continuous hypergravity and vibration on human chondrocytes growing in monolayers. For these aims, we employed parabolic flight maneuvers as well as groundbased devices to expose the cells to isolated acceleration profiles and to vibrations as they occur in a combined manner during the flight conditions.

Short-term hypergravity affects chondrocytes
The microarray analysis of chondrocytes exposed for 2 h to 1.8 g revealed that only a very moderate amount of genes was affected in comparison to the parabolic flight effects. It is interesting to notice, that mainly biological processes were affected, which are involved in tissue morphogenesis or skeletal system development. This is a strong indication, that hypergravity directly affects cartilage development.
It has recently been shown, that mechanical load can induce vascular endothelial growth factor A (VEGF-A) expression [21]. VEGF-A belongs to a family of growth factors comprising VEGF-A, −B, −C, −D, −E, and placenta growth factor (PGF) [22,23] and is one of the key components to control angiogenesis, the development of new vessels from existing ones. VEGF is predominantly found in osteoarthritic cartilage/chondrocytes and contributes further to the disintegration of the cartilage by inducing matrix metalloproteinases, which are able to disrupt the extracellular matrix [24,25]. We have observed a similar tendency for VEGFA gene expression, which was significantly elevated by a factor of 1.65, but was omitted for the GO BP analysis due to our cut-off of a Fold-Change >2.
Interestingly, MMP3 and MMP10 transcripts were downregulated under hypergravity. We speculate that this might be due to the fact that the chondrocytes form a monolayer in the culture flasks and are not embedded in an ECM-like matrix as in their physiological environment. This strong MMP downregulation might therefore be an  HMGB1, SBNO2, ELF4, FOXO1, NFKB2, TRIB1, LIF, TSC22D1, HEXIM1, ZNF697 attempt to build up a thick ECM, and a stronger stimulus than the VEGFA induction. BMP4, on the other hand, coding for the bone morphogenetic protein 4 (BMP-4; Additional file 1) was found to be downregulated under hypergravity. BMP-4 stimulates the synthesis of collagen type 2 and aggrecan and thus, enhaces the production of articular cartilage [26,27]. EDN1 (endothelin 1) gene expression was observed to be enhanced (Additional file 1). It has been reported, that overexpression of endothelin 1 is associated with cartilage degeneration [28]. In contrast to this, the inhibitor of apoptosis, BIRC3 (Additional file 1) [29], was enhanced Figure 4 STRING analysis of the cluster 6 from the parabolic flight experiment. Chondrocytes were fixed after parabola 1 and 31 during a parabolic flight. In parallel, corresponding 1 g control samples were prepared. K-mean clustering of the resulting microarray data revealed 6 clusters of differentially expressed genes. Clusters 1-5 revealed mostly unspecific transcriptionally active genes, while cluster 6 showed a strong dominance by anti-apoptotic and cell-proliferative transcripts. Possible interactions of the corresponding proteins were visualized using the STRING software and genes involved in anti-apoptosis and cell proliferation were highlighted with white and black circles, respectively. hinting towards an improved cell survival. Furthermore, the gene expression of wingless-type MMTV integration site family, member 5A (WNT5A) was decreased (Additional file 1). Wnt-5a has been shown to be able to induce cartilage degradation through upregulation of MMPs [30]. All in all, it is obvious that in our setup chondrocytes are sensitive to mechanical stress by hypergravity, but at the moment, no definite answer can be given about the nature of the effect.

Short-term vibration is detrimental to chondrocytes
In contrast to hypergravity, the effects of vibration on cultured chondrocytes were clearer. In our experimental setup, the vibrations which were transmitted into the culture flask also caused the culture medium to stir to a certain dregree, which, as we speculate, resulted in additional shear forces. Shear forces have been shown to have a negative effect on chondrocytes and cartilage [31,32]. It has been reported, that cartilage, that was treated in such a way or was degenerating, produced increased amounts of proinflammatory interleukins, such as IL-6 or IL-8 [33,34]. In our qPCR analysis we found a strong increase of both IL6 and IL8 ( Figure 1B + D) gene expression, although no increase in IL-6 secretion (Figure 2), accompanied by decreases of EGF, VEGFD, and FGF17 ( Figure 1E, J, K) gene expression. The presence of these factors has been described as beneficial for cartilage development [35,36]. Taken together, our results indicate that vibration drives chondrocytes towards an inflammatory, cartilage destabilzing state.

The influence of parabolic flight maneuvers
The microarray analysis showed, that after only 1P relatively unspecific effects on the cells were observed, mainly connected to transcription. This is an indication that the cells have perceived the change in gravity and that they were preparing their transcriptional apparatus for an altered gene expression as a reaction to this stimulus. After 31P, we observed an increase in the enrichment of antiapoptotic genes. The qPCR analysis reflects the same tendency. Most of the investigated genes showed only transitional or no changes, such as CD44, IL6, EDN1, TNFA, and FGF9 ( Figure 5B, C, F, G, H). This seems to hint toward a short μg "shock" that the cells are able to overcome very quickly. Only CCNA2 ( Figure 5A), a cyclin involved in cell cycling and proliferaton [37] and the antiapoptotic IL8 ( Figure 5D) [38] are expressed in a manner that they exert a growth-promoting, cell-survival effect. It should be kept in mind, that these effects originate from only a short-term altered gravity (PFC) treatment and that longer exposure times have to be investigated in order to assess their significance. It is interesting to note that RPM exposure experiments resulted in increased expression of several genes responsible for cell motility, structure and integrity; control of cell growth, cell proliferation, cell differentiation and apoptosis [39] and that these results are also in very good accordance with earlier studies that also reported that chondrocytes are quite robust under μg stress [40].

Conclusions
We have shown that chondrocytes are very robust under conditions of parabolic flight maneuvers. They are able to adapt quickly to this new environment and actually profit from real μg by reducing their apoptotic rate. However, they are prone to damage/injury by hypergravity and especially by vibration/shear forces. All in all, these results are very promising and are a step further along the way to understand chondrocyte growth in μg, leading perhaps to new methods of scaffold-free preparation of cartilage grafts.

Parabolic flight
The parabolic flight experiments were conducted aboard the Airbus A300 ZERO-G operated by Novespace and based in Bordeaux-Merignac, France [19,[41][42][43]. Standard parabolic flights were performed, each with 31 parabolas in a row during the three to four hours flight. The flight manoeuver starts from the horizontal flight level followed by a 45°ascent for 20 s. During this time 1.5 g to 1.8 g are acting on the passengers and the experiments. Then the thrust is reduced and the aircraft follows the path of a parabola. The free fall (microgravity) phase starts and persists for 22 s. Afterwards, the engines are fully powered again and another phase of 1.8 g of 20 s terminates the parabola. Due to the aerodynamic forces and turbulences acting on the aircraft, the μg quality is in the range of about 10 −2 g.

Cell culture procedure
Cells were grown as published recently [19,44]. Briefly, the cells were cultured in 24 T75 cell culture flasks (75 cm 2 ; Sarstedt, Nümbrecht, Germany) until subconfluent monolayers were obtained. During this time, the cells were covered by 20 mL (T-75 flasks) CGM. One half of the flasks was used as ground control cells (1 g; n = 12), cultured and fixed further in the laboratory, the other half was taken on the parabolic flight (n = 12). During the parabolic flight, RNAlater (Applied Biosystems, Darmstadt, Germany) was injected via syringes containing the appropriate fixative. The syringes were connected to the T-75 flasks through a flexible tube and a 3-way-valve. One hour before each flight, the cell culture flasks were transported to the aircraft and placed into the 37°C preheated incubator on an experimental rack ( Figure 6A).

Cell fixation
The cells were fixed after the first parabola (1P) and after the 31st parabola (31P) using RNAlater (Applied Biosystems, Darmstadt, Germany) at a ratio of 4:1 (RNAlater:medium). After the flight, the fixative was discarded, the cells were briefly washed with PBS and covered with 10 ml of fresh RNAlater. Subsequently, the flasks were stored at 4°C and transported to the laboratory. For the quantitative real-time PCR, we collected n = 6 T75 cell culture flasks from both parabolic flight samples (1P and 31 P) and the 1 g control group, the remaining flasks were used for microarray analyses.

Hypergravity experiments
We performed experiments on the Short Arm Human Centrifuge (SAHC, DLR, Cologne, Germany) ( Figure 6B), with cells cultured in T75 cell culture flasks (75 cm 2 ; Sarstedt, Nümbrecht, Germany), growing in a monolayer. We installed two containers with floating mountings for the incubators on the SAHC ( Figure 6C). In this configuration the T75 cell culture flasks were always exposed to a correct vertical gravity (acceleration) vector during centrifugation. By using the power supply on the SAHC the incubators were constantly heated to 37°C.
On the SAHC, we exposed the samples to a continuous hypergravity phase of 1.8 g of about 2 hours corresponding to the total time frame of 31 parabolas.
We designed a homogenous centrifuge profile with constant spin-up and spin-down times of each 34 seconds.
We collected n = 5 static 1 g controls and n = 5 1.8 g hyper-g samples for the Microarray analysis. The 1 g controls were cultivated in parallel in a neighboring identical incubator. Immediately after the run, the culture medium was discarded and replaced with 25 mL RNAlater solution. For the measurement of cytokines released in the supernatant, an additional run of the SAHC was performed to obtain n = 12 static 1 g samples and n = 12 1.8 hyper-g samples for the ELISA technique.

Vibration experiments
The detailed method was published earlier [20]. In short, the Vibraplex vibration platform (frequency range 0.2 Hz -14 kHz) was used to create vibrations comparable to those occurring during parabolic flights ( Figure 6D). Corresponding vibrations to the three phases of pull-up (1.8 g), free fall (μg), and pull-out (1.8 g) were recorded and analysed by Schmidt [45]. These data were then used for the simulation experiments with the Vibraplex. For quantitative real-time PCR analyses, we collected n = 5 samples of each of the two groups (1 g controls and cells subjected to a vibration profile corresponding to 31 parabolas of a parabolic flight). The 1 g controls without vibration were grown separately in a similar incubator.
For the measurement of cytokines released in the supernatant, three additional vibration experiments were performed to obtain n = 12 static 1 g samples and n = 12 vibrated samples for the ELISA technique.

RNA isolation and cDNA synthesis
After arrival in the laboratory, the RNAlater solution on the fixed cells was replaced by PBS (Invitrogen, Darmstadt, Germany). The cells were scraped off using cell scrapers (Sarstedt, Nümbrecht, Germany), transferred to 50 ml tubes, and pelleted by centrifugation (2500 g for 10 min at 4°C). An RNeasy Mini Kit (Qiagen, Hilden, Germany) was used according to the manufacturer's instructions to isolate total RNA. RNA concentrations and quality were determined spectrophotometrically at 260 nm using an Ultrospec 2100 pro Spectrophotometer (Amersham Biosciences, Freiburg, Germany). The isolated RNA had an A260/280 ratio of >1.7. cDNA designated for the quantitative real-time PCR was then obtained with the First-Strand cDNA Synthesis Kit (Fermentas, St. Leon-Rot, Germany) using 1 μg of total RNA in a 20-μL reverse transcription reaction mixture.

Quantitative real-time PCR
Quantitative real-time PCR was used to determine the expression levels of the genes of interest. Primer Express software (Applied Biosystems) was applied to design appropriate primers with a T m of~60°C (Additional file 2). The primers were synthesized by TIB Molbiol (Berlin, Germany). All assays were run on a StepOnePlus Real-Time PCR System using the Power SYBR Green PCR Master Mix (both Applied Biosystems). The reaction volume was 25 μL including 1 μL of template cDNA and a final primer concentration of 500 nM. PCR conditions were as follows: 10 min at 95°C, 40 cycles of 30 s at 95°C and 1 min at 60°C, followed by a melting curve analysis step (temperature gradient from 60 to 95°C with +0.3°C/ cycle).
If all amplicons showed one single T m similar to the one predicted by Primer Express software, the PCR reactions were considered specific. Every sample was measured in triplicate, and relative quantification was effected by means of the comparative C T (ΔΔC T ) method. 18S rRNA was used as a housekeeping gene to normalize the expression data.

ELISA
ELISAs of IL-6, IL-8, EGF, VEGFD (R&D Systems), and FGF17 (USCN Life Science Inc.) in the cell culture supernatant from vibration and hypergravity experiments have been performed according to the protocols supplied by the manufacturer.

Microarray analysis
Prior to the analysis, RNA integrity (RIN) was checked with the bioanalyzer. Only samples meeting the required quality were included in the analysis. The Illumina HumanWG-6_V2_0_R3 arrays have been normalized using the BeadStudio Gene Expression Module v3.3.7 and quantile normalization without background correction. After quantile normalization and exclusion of low or not expressed genes (minimum Illumina detection p-value > 0.05; performed in both analyses separately) the quality of arrays and the general expression profile has been checked by Principal Component Analysis (PCA) using Partek Genomic Suite 6.6, correlation as a dispersion matrix and normalized Eigenvector scaling. No obvious batch effect or outlier was found for the hypegravity analysis, while the outlying general expression profile of one sample from the parabolic flight experiment was removed before test statistic (for an overview: see Table 4).
A parametric ANOVA comparing the conditions given in Table 5 was performed. The selection criteria for the significant differential expression are also given in Table 2. Differentiation of the expression profiles was performed using K-Mean clustering. The cluster analysis was done using Partek Genomic Suite 6.3 applying the Euclidean distance function on standardized log2 signal values. K was selected according to a local minimum of the Davies Bouldin K estimation procedure. Functional aspects of the differentially expressed probes were analyzed with g: Profiler using g: SCS threshold as significance criterion and the DAVID Bioinformatics Resources 6 [46,47]. Physical and functional interactions between proteins were determined using the String platform [48] at a low confidence score of 0.15.