Direct Effect of Electrical Stimulation on Induction of Brain-Derived Neurotrophic Factor from Cultured Retinal Müller Cells

Abstract

purpose. To investigate the direct effect of electrical stimulation (ES) on the induction of brain-derived neurotrophic factor (BDNF) from cultured retinal Müller cells.

methods. Müller cells were isolated from rat retinas. ES was applied to passage 1 Müller cells with biphasic pulses (duration, 1 ms; frequency, 20 Hz; current, 10 mA) for 30 minutes. The changes in gene expression after ES were analyzed with microarrays. The mRNA and protein levels of BDNF were determined at each time point after ES by RT-PCR and ELISA, respectively. RT-PCR was also performed at 3 hours after ES of Müller cells that had been exposed to 1 μM nifedipine, a blocker of L-type voltage-dependent calcium channels (L-VDCCs).

results. Microarray analyses showed an upregulation of 245 genes, including BDNF. The mRNA level of BDNF increased significantly (P < 0.05; by ∼1.2-fold over that of the control) at 2 and 3 hours after ES. The intracellular protein level was upregulated significantly (by ∼1.4-fold) at 6 hours after ES, whereas the extracellular level did not change at any time point. The total protein level of BDNF increased significantly (∼1.3-fold) at 6 hours after ES. The increase in the mRNA level of BDNF was fully suppressed by exposure of the Müller cells to nifedipine.

conclusions. These results demonstrate that ES directly upregulates the transcriptional induction of BDNF through L-VDCCs in cultured Müller cells. The ES of Müller cells may be used to supply endogenous BDNF to the retina.

Müller cells, the predominant glial element in the retina, are believed to play important roles in maintaining its integrity and function. For example, Müller cells have been shown to express glutamate transporters, ¹ which are postulated to contribute to the clearance of glutamate and protect retinal ganglion cells (RGCs) from glutamate neurotoxicity. ^{2 3 4 5} Müller cells also synthesize glutamine synthetase ⁶ which amidates glutamate to form the non-neuroactive compound glutamine. ⁷ In addition, Müller cells produce various neurotrophic factors, including brain-derived neurotrophic factor (BDNF), ^{8 9} basic fibroblast growth factor (bFGF or FGF-2), ^{10 11} and insulin-like growth factor (IGF)-1, ¹² in reaction to different in vivo and in vitro conditions. 

We have demonstrated that electrical stimulation (ES) induces the transcription of IGF-1 from cultured Müller cells and that the induction was dependent on calcium (Ca²⁺) influx through L-type voltage-dependent calcium channels (L-VDCCs). ¹³ Ca²⁺ ions are the most widely used second messengers, and many of the processes that occur in the central nervous system (CNS; e.g., gene transcription), are controlled by Ca²⁺ influx. ¹⁴ For example, Ou and Gean ¹⁵ showed that the Ca²⁺ influx through N-methyl-d-aspartate (NMDA) receptors and L-VDCCs upregulated the transcription of BDNF in the amygdala. Sasaki et al. ¹⁶ demonstrated that Ca²⁺ entry into cortical neurons through L-VDCCs induced neuronal nitric oxide synthase (nNOS) message as well as protein synthesis. On the other hand, ES has been reported to induce different gene expressions (e.g., BDNF, ^{17 18} bFGF, ¹⁸ and growth-associated protein [GAP]-43 ¹⁹ in both central and peripheral neurons. In addition, ES has been demonstrated to induce expression of various genes by Ca²⁺ influx through VDCCs in neurons other than those of the retina. ^{20 21}  

Thus, we hypothesized that the Ca²⁺ influx into Müller cells via L-VDCCs that is induced by ES increases the gene expression of neurotrophic factors. To test this hypothesis, we initially performed microarray analyses to evaluate the changes in gene expression with and without ES in cultured Müller cells. Our results showed that the mRNA of BDNF was upregulated, and the time course of BDNF expression at both the message and protein levels was determined. 

Materials and Methods

All experimental procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Research Committee, Osaka University Medical School. 

Müller Cell Cultures

Müller cells were obtained by a method that isolates cells that were >95% pure Müller cells. ^{13 22} Briefly, eye cups from Long-Evans rats at postnatal days 12 to 14 were soaked in Dulbecco’s modified Eagle’s medium (DMEM; Nikken Biomedical Laboratory, Kyoto, Japan) supplemented with 1:1000 penicillin/streptomycin (Invitrogen Japan, Tokyo, Japan) overnight at room temperature in the dark. The eye cups were then incubated in DMEM containing 0.05% trypsin/EDTA (Invitrogen Japan) and 70 U/mL collagenase (Sigma-Aldrich, Tokyo, Japan) for 30 minutes at 37°C. The retinas were dissociated into small aggregates with a narrow-bore Pasteur pipette in culture medium. The culture medium was low-glucose DMEM supplemented with 10% fetal bovine serum (Invitrogen Japan) and 1:1000 penicillin-streptomycin. The cells were seeded into culture dishes and maintained at 37°C in a humidified atmosphere of 5% CO₂ and 95% air. When the primary cells proliferated to 80% to 90% confluence, the cells were passaged with Dulbecco’s phosphate-buffered saline (DPBS; Nikken Biomedical Laboratory) supplemented with 0.05% trypsin/EDTA. The cells were seeded at the same concentration into 35-mm dishes in 2 mL fresh culture medium. 

Cells of passage 1, which were cultured strictly for the same period to subconfluent condition, were used in all experiments for the statistical analyses. 

Conditions of ES

ES was applied to Müller cells on a clean bench. ¹³ Briefly, silver/silver chloride needle-type electrodes (φ = 0.2 mm) were inserted into culture medium without touching the Müller cells. The distance between the two electrodes was 20 mm. A rectangular biphasic train of pulses (pulse duration, 1 ms; pulse frequency, 20 Hz; current intensity, 10 mA) was delivered between the electrodes continuously for 30 minutes through a stimulus isolation unit for delivery of a constant current (Stimulator SEN-7203; Nihon Kohden, Tokyo, Japan; Isolator A395R; World Precision Instruments, Sarasota, FL), as in our previous study. ¹³ The voltage between the electrodes was monitored on a 40-MHz oscilloscope (CS-4135A; Kenwood, Tokyo, Japan). 

In control experiments, cells without ES were maintained on the bench for 30 minutes. 

Microarray Analyses

To minimize the variations among the cell samples, the Müller cells from three dishes with or without ES were combined as the ES or control groups, respectively. Total RNA was isolated immediately after ES with an extraction reagent (RNeasy Mini Kit; Qiagen Japan, Tokyo, Japan) according to the manufacturer’s protocols. The RNA samples were hybridized to a rat array containing 31,099 target probe sets (GeneChip Rat Genome 230 2.0 Array; Affymetrix Japan, Tokyo, Japan) and the following experiments were performed according to the manufacturer’s recommendations. 

The signal intensities were calculated with the microarray software (GeneChip Operating Software; Affymetrix Japan). If the raw signal intensities were <0.01, they were set to 0.01 according to the manufacturer’s protocols. The raw intensity values in each chip were normalized to the 50th percentile of the measurements. Each gene was normalized to the median of that gene in the respective ES or control groups. Further analyses were performed with a second software program (GeneSpring 7.2 software; Agilent Technologies Japan, Tokyo, Japan) to compare the transcriptional inductions in the ES group with the control. 

Quantitative Real-Time PCR

The mRNA levels of BDNF were determined by quantitative real-time PCR (RT-PCR) at 0, 1, 2, 3, and 6 hours after ES and calculated by the comparative C_T method. ²³ β-Actin was used as the endogenous control. Total RNA was isolated by the method just described and was reverse transcribed to synthesize cDNA (First-Strand cDNA Synthesis System for Quantitative RT-PCR; Marligen Bioscience, Ijamsville, MD). RT-PCR was then performed (Prism 7900HT; Applied Biosystems [ABI] Japan, Tokyo, Japan) in a 384-well format by selecting primer and probe sets (TaqMan) from an online catalog (BDNF: TaqMan [ABI] Gene Expression Assay ID, Rn02531967_s1, Celera Annotation GenBank accession No. NM_178866.2; β-actin: ID Rn00667869_ml, No. NM_031144.2; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). The sequences of the primers and probes are not disclosed. The thermal cycling conditions were: 50°C for 2 minutes, 95°C for 10 minutes, and 40 cycles at 95°C for 15 seconds and 60°C for 1 minute. Each measurement was performed in duplicate in six independent runs. 

The expression of BDNF gene in the ES group was compared to that in the control group. The relative data are presented as the mean multiple of change (x-fold) ± SD calculated from the six separate experiments. The Wilcoxon signed rank test was used to analyze the statistical significance (P < 0.05) of differences between the median values. 

Enzyme-Linked Immunosorbent Assay

Both extracellular and intracellular BDNF protein levels were measured at 3, 6, 12, 24, and 48 hours after ES by quantitative sandwich enzyme immunoassay (Chemicon International, Temecula CA) according to the manufacturer’s instructions. This technique can measure both human and rat BDNF. The culture medium and the cell lysate were collected as the samples for extracellular and intracellular BDNF, respectively. To lyse the Müller cells, the cells were soaked in 200 μL of lysis buffer (RIPA buffer; Sigma-Aldrich) supplemented with a protease inhibitor cocktail (Protease Inhibitors Mixture for Protease and Esterase; Wako, Osaka, Japan). The lysate was centrifuged at 15,000 rpm for 10 minutes at 4°C, and the BDNF in the supernatant was measured as intracellular BDNF. 

The optical density of each sample was measured at 450 nm with the correction wavelength set at 540 nm (ARVO_MX; PerkinElmer Japan, Kanagawa, Japan). Each measurement was performed in duplicate in six independent runs. The absolute protein amount of BDNF was calculated by the following equations:  

\[\mathrm{Extracellular\ BDNF}\ (\mathrm{pg}){=}\mathrm{medium\ concentration}\ (\mathrm{pg}/\mathrm{mL}){\times}2\ (\mathrm{mL},\ \mathrm{medium\ volume})\]

,  

\[\mathrm{Intracellular\ BDNF}\ (\mathrm{pg}){=}\mathrm{lysate\ concentration}\ (\mathrm{pg}/\mathrm{mL}){\times}0.2\ (\mathrm{mL},\ \mathrm{lysate\ volume})\]

, and  

\[\mathrm{Total\ BDNF}\ (\mathrm{pg}){=}\mathrm{sum\ of\ extra-\ and\ intracellular\ BDNF}\ (\mathrm{pg}).\]

 

The expression of the protein BDNF in the ES group was compared with that in the control group. The relative data are presented as the mean multiple of change ± SD calculated from six separate experiments. The Wilcoxon signed rank test was used to analyze the statistical significance (P < 0.05) of differences between the medians. 

Quantitative Real-Time PCR with L-VDCC Blocker

The mRNA level of BDNF in Müller cells exposed to 1 μM nifedipine was determined by RT-PCR at 3 hours after ES and compared with that in Müller cells without nifedipine. Briefly, 1 μM nifedipine was added to the culture medium immediately before the start of ES. The dose of nifedipine was enough to block the Ca²⁺ influx via L-VDCCs. ¹³ ES was applied to Müller cells, and thereafter the medium of the dishes with and without ES was replaced by 2 mL fresh culture medium. 

The mRNA level of BDNF in Müller cells without ES and nifedipine was normalized to 1, and the relative data are presented as the mean multiple of change ± SD calculated from twelve separate experiments. The Kruskal-Wallis one-way ANOVA on ranks was used to analyze the statistical significance (P < 0.05) of the differences among the median values, followed by Dunn’s method to determine significant differences between each group. 

Results

These experiments were conducted on passage 1 Müller cells because the primary-culture cells were contaminated by many other cell types, and preliminary assays by RT-PCR showed marked variations in the expression levels of several genes among the culture dishes (data not shown). Examination of the cells with an optical microscope showed that the morphologic appearance of the Müller cells did not change after the different experimental procedures. 

Microarray Analyses

In our earlier study, ¹³ a Ca²⁺ influx into Müller cells was detected immediately after the beginning of ES, and the mRNA of IGF-1 was upregulated at the cessation of ES. Thus, to evaluate the direct effect of ES on the changes in gene expression, we performed microarray analyses on cells isolated immediately after ES. 

The results showed that 7.3% (2268) of the probes expressed more than a twofold higher signal intensity in the ES group than in the control (Fig. 1A) . The probes whose signals were absent in both the ES and control groups were not analyzed, according to the manufacturer’s recommendations. The number of probes showing sufficient intensities that were more than two times higher in the ES than in the control was 245 (Fig. 1B) . The genes associated with nervous system development were selected and are listed in Table 1 . The most well-known neurotrophic factor was BDNF, with a gene expression that changed from absent in the control to a 2.6-fold increase in the ES group. Thus, the following experiments were focused on BDNF. 

Upregulation of BDNF mRNA by ES in Cultured Müller Cells

The time course of transcriptional induction of BDNF by ES was determined by RT-PCR (Fig. 2) . The mRNA level was significantly increased by approximately 1.2-fold at 2 and 3 hours after ES over that of the control, and the level was decreased at 6 hours. 

Upregulation of BDNF Protein by ES in Cultured Müller Cells

Because the relative value of the mRNA of BDNF increased at 2 and 3 hours after ES, the protein level of BDNF was determined from 3 hours after ES. The level in the ES group was compared to that in the control group at each time point (Fig. 3) . The absolute values of the extracellular proteins of BDNF ranged from 10.2 to 129.6 pg and the intracellular values from 25.5 to 151.3 pg. The intracellular level of BDNF protein was often higher than that of extracellular protein. These relatively large variations were probably due to the different number of cultured Müller cells between the experimental runs. 

The intracellular protein level of BDNF in the experimental cells was significantly increased (>1.4-fold) at 6 hours after ES compared with that in the control cells, whereas the extracellular level did not change significantly at any time point. The total protein was significantly upregulated (by ∼1.3-fold) at 6 hours after ES. 

Suppression of BDNF mRNA by Blocking L-VDCCs in Cultured Müller Cells

In earlier work, we showed that ES induced the transcription of IGF-1 by Ca²⁺ influx via L-VDCCs in cultured Müller cells. ¹³ Because Ca²⁺ ions are one of the common second messengers, ¹⁴ we hypothesized that the transcriptional induction of BDNF also depends largely on the Ca²⁺ influx through L-VDCCs. A common pharmacologic hallmark of L-VDCCs is their sensitivity to dihydropyridines. ²⁴ Thus, the gene expression of BDNF was determined at 3 hours after ES by RT-PCR in Müller cells loaded with 1 μM nifedipine, a dihydropyridine calcium channel blocker (Fig. 4) . 

The mRNA level of BDNF was approximately 0.5-fold at 0 mA with nifedipine and 0.4-fold at 10 mA with nifedipine compared with that of 0 mA (control). Statistical significance was detected among these four groups (P = < 0.001; Kruskal-Wallis one-way ANOVA on ranks), and the levels after exposure to nifedipine were significantly lower than that with 0 mA (P < 0.05; Dunn’s method). These results indicated that the transcription of BDNF in cultured Müller cells was upregulated by ES, was most likely due to Ca²⁺ influx into the cells via L-VDCCs, and was downregulated by the application of nifedipine, which prevented Ca²⁺ influx via L-VDCCs. 

Discussion

To investigate the direct effect of ES on the production of neurotrophic factors from Müller cells, ES was applied to cultured Müller cells. The major findings were that both the message and protein levels of BDNF in cultured Müller cells were upregulated by ES, and that the transcriptional induction of BDNF is fully depressed by the application of an L-VDCC blocker. These findings suggest that the transcriptional induction by ES depends largely on Ca²⁺ influx through L-VDCCs. 

Although BDNF was first purified from brain cells, ²⁵ it has been shown to play crucial roles, not only in CNS ^{26 27 28} but also in the retina. Recent studies have demonstrated that BDNF promoted the survival of RGCs in different models. ^{29 30 31 32 33} Furthermore, Paskowitz et al. ³⁴ demonstrated that BDNF increased photoreceptor survival after verteporfin photodynamic therapy, indicating that BDNF had a neuroprotective effect on the photoreceptors. These findings support the conclusion that BDNF acts as a neurotrophic agent in the retina. 

However, one crucial question regarding the clinical application of BDNF is how to deliver it to the retinal neurons. Previous studies ^{29 30 31 32 33 34} have demonstrated that exogenous BDNF, delivered by intravitreous injections, had a protective effect on retinal neurons. However, exogenous BDNF had a transient and limited protective effect, and repeated applications were necessary, which can have adverse effects on the retina. On this point, we have already developed a system of transcorneal ES to the retina and showed the protective effect on the eyes with optic nerve diseases without serious complications. ³⁵ These results indicate the possibility that ES may safely induce the production of endogenous BDNF from Müller cells to promote the survival of retinal neurons. 

In this study, the extracellular protein level of BDNF did not increase at any time point. The reason for this is still not clear, but one possibility is that BDNF behaves as an autocrine factor in cultures of pure Müller cells. An earlier study demonstrated that Müller cells not only synthesize BDNF but also express TrkB, the specific receptor of BDNF. ³⁶ These results indicate that BDNF secreted from Müller cells may be expanded by an autocrine mechanism and that the transcriptional induction of BDNF may be suppressed at 6 hours by negative feedback of the increased protein of BDNF. 

The mechanism of how ES stimulates L-VDCCs in cultured Müller cells is not fully understood. It is well known that Ca²⁺-regulated gene expression plays a critical role in diverse neural functions and that the activation of L-VDCCs is required for depolarization-mediated gene induction. ²⁰ Most studies that have examined mechanisms of activity-dependent gene expression have used chronic membrane depolarization to raise intracellular calcium levels and stimulate gene expression. ²¹ Further studies are needed to determine the mechanisms involved in the gene expression of BDNF by ES in Müller cells. 

One of the limitations of this study is that the microarray analyses were performed at only one time point, immediately after ES. In fact, the mRNA level of BDNF determined by RT-PCR was higher at 2 and 3 hours than at 0 hours after ES, although the transcriptional induction depended mainly on Ca²⁺ influx through L-VDCCs, similar to that for IGF-1. Thus, we cannot exclude the possibility that other neurotrophic factors may have been induced by the ES. 

In summary, we have demonstrated the direct effect of ES on BDNF production in cultured rat Müller cells. Our findings raise the possibility that ES may deliver endogenous BDNF from Müller cells without serious complications and promote the survival of retinal neurons. 

 Supported by Health Sciences Research Grant H16-sensory-001 from the Ministry of Health, Labor, and Welfare, Japan, and by Grant 18591918 from the Ministry of Education, Culture, Science, and Technology.

 Submitted for publication March 19, 2008; revised April 23 and May 20, 2008; accepted August 8, 2008.

 Disclosure: T. Sato, None; T. Fujikado, None; T. - S. Lee, None; Y. Tano, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Results of microarray analyses. The horizontal and vertical axes represent the signal intensity of the genes expressed in the control and ES groups, respectively. (A) Signal intensity image plot of all gene expressions (31,099 probes). For each probe, the ratio of the signal intensity in the ES group to that in control is represented by a color, according to the vertical strip. (B) The circle of Flag contains the genes with signals that were “present” or “marginal” in the ES or control groups or both. The circle of ×2.0 holds the genes showing a more than twofold signal intensity in ES compared with the control. Left: signal intensity image of the probes, which corresponds to the pie chart.

Table 1.

 

Results of Microarray Analyses Immediately after ES

Description	GenBank Accession No.	Normalized Signal Intensity in Control	Flags of Control	Normalized Signal Intensity in ES	Flags of ES
Dihydropyrimidinase-like 3	AI059953	1.00	A	2.02	P
Zinc finger and BTB domain containing 16	BG371725	1.00	A	2.15	P
Ectodermal-neural cortex 1	AA997271	1.00	M	2.19	A
Brain derived neurotrophic factor	X67108	1.00	A	2.58	P
Stathmin-like 3	NM_024346	1.00	A	3.31	M
UDP galactosyltransferase 8	L21698	1.00	A	13.77	P

 A, absent; M, marginal; P, present.

Figure 2.

 

Relative gene expressions of BDNF after ES at 10 mA determined by RT-PCR. The relative values are presented as the mean multiple of change (x-fold) ± SD calculated from six separate experiments. Statistical analysis was performed by the Wilcoxon signed rank test (*P < 0.05).

Figure 3.

 

Relative protein expressions of BDNF after ES at 10 mA determined by ELISA. The relative values are presented as the mean multiple of change (x-fold) ± SD calculated from six separate experiments. Statistical analysis was performed by Wilcoxon signed rank test (*P < 0.05). (A) Intracellular protein level of BDNF. The absolute levels of intracellular BDNF ranged from 25.5 to 151.3 pg. (B) Extracellular level of BDNF. The absolute extracellular level of BDNF ranged from 10.2 to 129.6 pg. (C) Total level of BDNF. The absolute level of total BDNF ranged from 60.0 to 249.3 pg.

Figure 4.

 

Relative gene expressions of BDNF with or without 1 μM nifedipine at 3 hours after 10 mA ES, determined by RT-PCR. The relative data are presented as the mean multiples of change (x-fold) ± SD in 12 separate experiments. Statistical analysis was performed by Kruskal-Wallis one-way ANOVA on ranks, followed by the Dunn test (*P < 0.05).

The authors thank Takayuki Harada for crucial comments on this study.