Effect of hyperbaric oxygen on BDNF-release and neuroprotection: Investigations with human mesenchymal stem cells and genetically modified NIH3T3 fibroblasts as putative cell therapeutics

Abstract

Hyperbaric oxygen therapy (HBOT) is a noninvasive widely applied treatment that increases the oxygen pressure in tissues. In cochlear implant (CI) research, intracochlear application of neurotrophic factors (NTFs) is able to improve survival of spiral ganglion neurons (SGN) after deafness. Cell-based delivery of NTFs such as brain-derived neurotrophic factor (BDNF) may be realized by cell-coating of the surface of the CI electrode. Human mesenchymal stem cells (MSC) secrete a variety of different neurotrophic factors and may be used for the development of a biohybrid electrode in order to release endogenously-derived neuroprotective factors for the protection of residual SGN and for a guided outgrowth of dendrites in the direction of the CI electrode. HBOT could be used to influence cell behaviour after transplantation to the inner ear. The aim of this study was to investigate the effect of HBOT on the proliferation, BDNF-release and secretion of neuroprotective factors. Thus, model cells (an immortalized fibroblast cell line (NIH3T3)-native and genetically modified) and MSCs were repeatedly (3 x - 10 x) exposed to 100% oxygen at different pressures. The effects of HBO on cell proliferation were investigated in relation to normoxic and normobaric conditions (NOR). Moreover, the neuroprotective and neuroregenerative effects of HBO-treated cells were analysed by cultivation of SGN in conditioned medium. Both, the genetically modified NIH3T3/BDNF and native NIH3T3 fibroblasts, showed a highly significant increased proliferation after five days of HBOT in comparison to normoxic controls. By contrast, the number of MSCs was decreased in MSCs treated with 2.0 bar of HBO. Treating SGN cultures with supernatants of fibroblasts and MSCs significantly increased the survival rate of SGN. HBO treatment did not influence (increase / reduce) this effect. Secretome analysis showed that HBO treatment altered the protein expression pattern in MSCs.

Fig 1. Proliferation of HBO-treated cells.

The effect of hyperbaric oxygen (HBO) treatment on the proliferation of NIH3T3 fibroblasts (A), genetically modified NIH3T3/BDNF fibroblasts (B) and MSCs (C) is depicted. HBO was applied at three different pressures: 1.0 bar (red circle and line), 1.5 bar (blue square and line) and 2.0 bar (orange triangle and line). Both NIH3T3 and NIH3T3/BDNF fibroblasts show a highly significantly increased cell number compared to the control cells (normoxic and normobaric conditions (NOR), green inverted triangle and line) after five consecutive HBO treatments. By contrast, the proliferation of MSCs started to decrease after five consecutive HBO treatments with 2.0 bar until the end of ten treatments in comparison to NOR. Values are given as mean ± standard error of the mean (SEM); N = 3, n = 3. Asterisks indicate the significance of cell numbers of the different used pressures compared to the normoxic control. Statistical assessment was performed using one-way ANOVA with Bonferroni’s multiple comparison test (n.s. = not significant; *p < 0.05; **p < 0.01; ***p < 0.001).

Fig 2. Representative images of HBO-treated MSCs.

Microscopic brightfield images demonstrate the effect of HBO treatment on MSCs. Decreased cell numbers were observed after five (not shown), eight and ten treatments (corresponding to eight, 11 and 15 days of in vitro cultivation, respectively) in MSCs treated with HBO at 2.0 bar. By contrast, HBO treatment with 1.0 bar showed a cell density comparable with the control (NOR). All pictures are presented at a 200-fold magnification. Scale bar: 100 μm.

Fig 3. Effect of HBO treatment on BDNF-secretion of genetically modified fibroblasts.

Quantitative determination of BDNF secreted by recombinant NIH3T3/BDNF fibroblasts exposed to HBO at an oxygen pressure of 2.0 bar (intensely dotted) and their corresponding controls (NOR, without pattern) by ELISA. The release of BDNF is significantly increased after three and five consecutive HBO treatments and three to five NOR treatments (light grey and grey) when compared to their respective controls. The continuous line represents the recommended BDNF concentration (50 ng/mL) for optimal neuronal survival of SGN. All values are given as mean ± standard error of the mean (SEM). Statistical analysis was performed by one-way ANOVA with Bonferroni’s correction (n.s. = not significant; *p < 0.05; **p < 0.01; ***p < 0.001), comparing the respective single treatment value (1xNOR, 1xHBO) to the repeated treatment.

Fig 4. Survival rates of SGN cultivated in supernatants obtained from HBO-treated cells.

Mean survival rates of spiral ganglion neurons (SGN) were determined by the amount of neurons in relation to the seeding control after cultivation in supernatants obtained from HBO-treated NIH3T3 and NIH3T3/BDNF fibroblasts (A) and MSCs (B). The treatment with supernatants significantly enhanced the survival of SGN compared to the negative control (NC, white column without pattern). Values are given as mean ± standard error of the mean (SEM); N = 3, n = 3; Controls: NC (spiral ganglion medium control, serum-free), BDNF (spiral ganglion medium substituted with 50 ng/mL human recombinant BDNF, positive control), FCS (fetal calf serum containing fibroblast/MSC medium, medium control). Statistical analysis was performed using one-way ANOVA with Bonferroni’s multiple comparison test (n.s. = not significant; *p < 0.05; **p < 0.01; ***p < 0.001). Asterisks over the error bars of the columns indicate the significance compared to the NC.

Fig 5. Neurite length of SGN cultivated in supernatants obtained from HBO-treated cells.

Mean neurite length of SGN, treated with supernatants from HBO-treated and NOR-treated NIH3T3 as well as NIH3T3/BDNF fibroblasts (A) and MSCs (B), are shown. Supernatants of 1.0 and 1.5 bar HBO-treated MSCs significantly increased the neurite length in comparison to the BDNF control and negative control. The asterisks over the error bars of the columns refer to the NC. N = 3; n = 3; n.s. = not significant; *p < 0.05; ** p< 0.01; ***p < 0.001.