Journal of Neuroscience Research 92:761–771 (2014)

Pulsed Electromagnetic Fields Potentiate Neurite Outgrowth in the Dopaminergic MN9D Cell Line Rukmani Lekhraj,1* Deborah E. Cynamon,1 Stephanie E. DeLuca,1 Eric S. Taub,1 Arthur A. Pilla,2,3 and Diana Casper1 1

Department of Neurological Surgery, Montefiore Medical Center and the Albert Einstein College of Medicine, Bronx, New York Department of Biomedical Engineering, Columbia University, New York, New York 3 Department of Orthopedics, Mount Sinai School of Medicine, New York, New York 2

Pulsed electromagnetic fields (PEMF) exert biological effects and are in clinical use to facilitate bone repair and wound healing. Research has demonstrated that PEMF can induce signaling molecules and growth factors, molecules that play important roles in neuronal differentiation. Here, we tested the effects of a low-amplitude, nonthermal, pulsed radiofrequency signal on morphological neuronal differentiation in MN9D, a dopaminergic cell line. Cells were plated in medium with 10% fetal calf serum. After 1 day, medium was replaced with serum-containing medium, serum-free medium, or medium supplemented with dibutyryl cyclic adenosine monophosphate (Bt2cAMP), a cAMP analog known to induce neurite outgrowth. Cultures were divided into groups and treated with PEMF signals for either 30 min per day or continuously for 15 min every hour for 3 days. Both serum withdrawal and Bt2cAMP significantly increased neurite length. PEMF treatment similarly increased neurite length under both serum-free and serum-supplemented conditions, although to a lesser degree in the presence of serum, when continuous treatments had greater effects. PEMF signals also increased cell body width, indicating neuronal maturation, and decreased protein content, suggesting that this treatment was antimitotic, an effect reversed by the inhibitor of cAMP formation dideoxyadenosine. Bt2cAMP and PEMF effects were not additive, suggesting that neurite elongation was achieved through a common pathway. PEMF signals increased cAMP levels from 3 to 5 hr after treatment, supporting this mechanism of action. Although neuritogenesis is considered a developmental process, it may also represent the plasticity required to form and maintain synaptic connections throughout life. VC 2014 Wiley Periodicals, Inc. Key words: pulsed electromagnetic field; neurite outgrowth; cAMP; neuronal differentiation; plasticity

The biological effects of electromagnetic fields have been explored for centuries (Beck, 2004) and, for the last 50 years, both from the safety standpoint (i.e., electrical power lines and cell phones) and for their therapeutic C 2014 Wiley Periodicals, Inc. V

potential to treat recalcitrant fractures (Bassett et al., 1974; Gossling et al., 1992; Bodamyali et al., 1998), soft tissue injuries (Itoh et al., 1991; Salzberg et al., 1995; Patino et al., 1996; Baker et al., 1997; Kloth et al., 1999; Jasti et al., 2001; Strauch et al., 2007; Callaghan et al., 2008), pain (Foley-Nolan et al., 1990; Thuile and Walzl, 2002; Shupak et al., 2004; Rohde et al., 2010), Alzheimer’s disease (Arendash et al., 2010), and depression (Loo and Mitchell, 2005). Experimental and clinical studies suggest that pulsed electromagnetic field (PEMF) signals can produce a variety of effects that share some common mechanisms, such as growth factor induction, cytokine modulation (McLeod et al., 1983; Tepper et al., 2004; Callaghan et al., 2008), and second messenger signaling (Hogan and Wieraszko, 2004; Cheng et al., 2011; Pilla et al., 2011; Pilla, 2012). Neuronal development can be divided into stages that include proliferation, migration, determination, differentiation, and maturation (Hamburger, 1952). Significant numbers of factors that influence these processes have been identified (Purves and Lichtman, 1985; Henderson, 1996), including electromagnetic fields (Yao et al., 2011). Mature neurons also display dynamic properties that reiterate developmental processes in a new context, specifically focused on the formation, maintenance, pruning, and strengthening of synapses, a property known as plasticity. These processes depend on a group of molecular second messengers such as calcium and cyclic nucleotides, growth factors, and the electrochemical potential in pre- and postsynaptic neurons (Kotaleski and Blackwell, 2010). Neuronal development has been modeled in cell Contract grant sponsor: NIH; Contract grant number: NS052576 (to D.C.). *Correspondence to: Rukmani Lekhraj, Neurosurgery Laboratory, Moses 3, Montefiore Medical Center, Bronx, NY 10467. E-mail: [email protected] Received 15 August 2013; Revised 7 November 2013; Accepted 6 December 2013 Published online 12 February 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jnr.23361

762

Lekhraj et al.

cultures, in which it can more easily be studied. Importantly, even though the processes involved in establishing neuronal cultures disrupts their relationship to neighboring cells, the extracellular matrix, and their potential targets, these cells will still differentiate given an appropriate inductive or permissive environment. PEMF signals have been shown to influence neuronal morphology (Greenebaum et al., 1996; Macias et al., 2000; McFarlane et al., 2000; Kim et al., 2008), including the differentiation of the pheochromocytoma PC12 cell line (Blackman et al., 1993; McFarlane et al., 2000; Schimmelpfeng et al., 2005; Zhang et al., 2005; Kim et al., 2008), known to transform into sympathetic neurons in the presence of nerve growth factor (Greene and Tischler, 1976). In these studies, the strength of the PEMF signals ranged from 0.004 to 12 mT, generating effects from decreased neurite lengths to enhanced and directional growth. Some of these signals required large magnets capable of generating significant thermal energy and altering neuronal membrane potentials that could influence voltage-dependent, ion-permeable channels. This study utilizes a pulse-modified radiofrequency signal configuration to deliver a very weak magnetic field (5 lT) through a single-turn, circular wire coil (Pilla, 2007). This signal delivers a uniform magnetic field perpendicular to the culture dish, inducing electric fields parallel to the plane of the coil (Bassett et al., 1974; McLeod et al., 1983; Hristov and Perez, 2011) that weakens with distance from the perimeter (Faraday’s Law) and that are thought to mediate the biological effects. A similar PEMF signal has been shown to decrease pain produced by surgery (Rohde et al., 2010) and osteoarthritis of the knee (Nelson et al., 2012) and to attenuate the induction of interleukin-1b (IL-1b) in wound exudates and following traumatic brain injury in an animal model (Rasouli et al., 2012), but no direct effects on neuronal differentiation have been reported. With the dopaminergic MN9D cell line, we investigated the effects of this signal on neurite outgrowth, cell body diameter, cell numbers, and involvement of cAMP in producing some of these effects. PEMF signals significantly increased neurite length and cell body diameter. Results also suggest that PEMF treatment was acting through the induction of cAMP, a molecule known to be involved in neuronal differentiation. MATERIALS AND METHODS MN9D Cell Culture The murine MN9D dopaminergic cell line, generated from the fusion of embryonic mouse mesencephalic neurons with mouse neuroblastoma N18TG2 cells (Choi et al., 1991), was a gift from Alfred Heller at the University of Chicago. Cultures were established at 100,000 cells per 35-mm dish coated with poly-L-ornithine in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal calf serum (Invitrogen, Carlsbad, CA) in the presence or absence of 1 mM dibutyryl cyclic adenosine monophosphate (Bt2cAMP; Sigma, St. Louis, MO). Cultures were maintained in a humidified incubator at 37 C with 8%

CO2. Treatments, addition of inhibitors, and serum withdrawal were initiated 1 day after plating. PEMF Treatment Cultures were subjected to different regimens of a nonthermal radiofrequency PEMF signal generated from a singleturn 19-cm diameter coil (Ivivi Technologies, Montvale, NJ), placed on the horizontal plane of a plastic shelf so that the coil was parallel to the surface in a dedicated incubator with CO2 and temperature controlled to 61% (model 3158; Forma, Marietta, OH). Culture dishes were placed in the central portion of the coil at 2–6 cm from the perimeter. The PEMF device transmitted a sinusoidal 27.12 MHz radiofrequency signal delivered in 5-msec bursts repeating at five bursts/sec, creating a uniform perpendicular magnetic field and inducing a 13 6 2 V/m peak gradient electric field parallel to the cell layer. PEMF parameters were assessed and verified using a calibrated shielded loop probe 1 cm in diameter (model 100A; Beehive Electronics, Sebastopol, CA) connected to a calibrated 100-MHz oscilloscope (model 2012B; Tektronix, Beaverton, OR). Control cultures were maintained in a second CO2 incubator under the same conditions, and cultures in the null group were moved to the PEMF incubator for 30 min for acute treatments or maintained in the second incubator under the same conditions. Cell Numbers Cultures were fixed with 4% paraformaldehyde, washed with PBS, and one to four consecutive microscopic fields at the fixed location of 1.4 cm from the edge of the dish (field strength 2.8 V/m) to control for electrical gradients were photographed under phase optics at 3100. Total cells and cells with processes greater than 10 lm were quantified manually on electronic images in ImageJ. In serum-free medium, the average number of cells analyzed was 308 6 138 (mean 6 standard deviation) per dish, with four dishes per group. In medium that contained serum, 183 6 83 cells per dish were analyzed (n 5 4). Protein content was determined on whole-cell lysates by the Bradford method (Bio-Rad, Hercules, CA). Cyclic AMP Cyclic AMP was quantified in cell lysates harvested in 0.1 N HCl by a commercial enzyme-linked cAMP competition immunoassay (Sigma). Morphological Analysis Neurite lengths and cell body widths were measured in the digital images described above for cell counts with the calibrated line tool of ImageJ. Process lengths less than 10 lm were excluded. Statistical Analysis Results, expressed as mean 6 standard error of the mean, were analyzed and tested for significant differences between treatment groups (n 5 4–6 dishes per group). All data sets passed the Kolmogorov-Smirnov test for normality. Student’s t-test was used for comparisons of two groups, and analysis of variance (ANOVA), followed by Fisher’s PLSD post hoc test, was Journal of Neuroscience Research

PEMF Signals Enhance Neuronal Differentiation

763

Fig. 1. Effects of Bt2cAMP on neurite length and cell numbers in MN9D cells. Cultures were plated in medium containing 10% serum in the presence or absence of 1 mM Bt2cAMP for 2 days. A: Morphology: control cells (left) were distributed in clusters, with small, phase-dark cell bodies and small neuritic processes. With Bt2cAMP

treatment, cells extended neurites (right). B: Neurite length was quantified by image analysis. Bt2cAMP increased neurite length by 58% (*P < 0.01). C: Cell numbers: total cells, cells with processes, and percentage of cells with processes. No significant differences were found between control and treated groups. Scale bar 5 50 lm.

used for comparisons of more than two groups. P < 0.05 was considered significant.

presence or absence of 1 mM Bt2cAMP for 3 days. Neurite length, total cells, and cells with processes were quantified in predesignated regions of the culture dishes, and the percentages of cells with processes were calculated. Results demonstrate that 1 mM Bt2cAMP significantly increased process length (Fig. 1B) from 30 6 4 lm in control cultures to 47.5 6 2 lm (58%; P < 0.01). In contrast, there were no significant effects of this cAMP analog on total cell numbers, cells with processes, or percentage of cells with processes (Fig. 1C). Several studies have tested the effects of electromagnetic fields on neuronal process outgrowth and nerve regeneration, with mixed results (Blackman et al., 1993; Gona et al., 1993; Drucker-Colin et al., 1994; Greenebaum et al., 1996; Macias et al., 2000; Shah et al., 2001; Schimmelpfeng et al., 2005; Zhang et al., 2005; Kim et al., 2008; Baptista et al., 2009; Koppes et al., 2011), perhaps as a result of the variety of signal configurations,

RESULTS MN9D cells are known to differentiate in culture both spontaneously and in response to treatment with differentiating agents, such as cAMP (Zhou et al., 2011). When maintained in 10% serum, the majority of cells were arranged in clusters, with minimal neuritic processes (Fig. 1A, left), and were still dividing after 3 days. In contrast, the majority of cells in cultures treated with 1 mM Bt2cAMP began to differentiate; specifically, single cells that elaborated long neuritic processes were found more frequently, with a lower tendency to form clusters (Fig. 1 A, right). The characteristics of differentiating MN9D cells after Bt2cAMP treatment can be quantified. Cultures were maintained in medium containing 10% serum in the Journal of Neuroscience Research

764

Lekhraj et al.

Fig. 2. PEMF treatment. A: The PEMF signal consisted of a 27.12 MHz sinusoidal wave with an amplitude of 5 lT (labeled B), delivered in bursts of 5 msec (T1) at 5 Hz (T2). B: Cultures were placed within a coil (arrow) delivering PEMF signals on a plastic shelf inside a cell culture incubator.

culture conditions, or injury paradigm. To investigate the effects of this PEMF signal (Fig. 2A) on neurite outgrowth, cell numbers, and cell size, we used two sets of conditions (10% serum and serum-free medium) and two PEMF treatment regimens (PEMF1 and PEMF2). Treatments were administered inside a cell culture incubator (Fig. 2B), and cultures in the null group were treated in the same manner except that no signal was delivered. A group of cultures remained in medium with serum for the duration of the experiment (control group) for comparison. Neurite length and cell numbers were quantified by the same methods as established for Bt2cAMP treatment. As before, most cultured MN9D cells were found in clusters, suggestive of clonal proliferation. Some cells at the edges of the clusters had short processes, measuring less than the widest diameter of cell bodies (Fig. 3A, arrows). When serum was withdrawn for 3 days, the numbers of cells in clusters were significantly reduced, and cell bodies assumed a teardrop shape. The majority of these cells exhibited processes that were longer than the cell body itself (Fig. 3B, arrows). When the PEMF1 regimen was applied to cells in serum-free medium for

Fig. 3. Effects of PEMF signals on neuronal morphology. Cells were plated in medium containing 10% serum, allowed to attach for 1 day, and treated for 3 days. Arrows identify neuritic processes. A: Cultures were maintained in the original plating medium with serum. B: Null (unexposed) group in serum-free medium. C: PEMF-treated cultures in serum-free medium. Scale bar 5 50 lm.

30 min once per day for 3 days, cells with neuronal morphology appeared to have longer processes (Fig. 3C, arrows). Journal of Neuroscience Research

PEMF Signals Enhance Neuronal Differentiation

765

Fig. 4. Effects of serum and two PEMF treatment regimens on neurite outgrowth and cell body width. Cultures were established in medium containing 10% serum and medium either was left in dishes (control) or was changed to either medium with or without serum (null group) and treated with PEMF signals. Neurite lengths (A–C) and cell body diameters (D) were measured after 3 days. *Significant differences from controls; **significant differences from the null group. A: Medium was maintained (control) or replaced with serum-free medium (null), and one group was exposed to PEMF1 (once per day for 3 days). Control vs. null; P < 0.002, PEMF1 vs. null group

(P < 0.004), and PEMF vs. control (P < 0.0001). B: Medium was changed to medium with 10% serum, and one group was treated with PEMF1 for 3 days. Null and PEMF vs. control (P < 0.0002 and P < 0.01 respectively). PEMF1 vs. null (P < 0.03). C: Medium was changed to serum containing medium and one group received PEMF2 treatment (15 min/hr for 3 days). PEMF2 vs. null (P < 0.02) and PEMF2 vs. PEMF1 (P < 0.01). D: Both PEMF2 (vs. null; P < 0.03) and 1 mM Bt2cAMP (v. null; P < 0.006) increased cell body diameter.

Neurite length was quantified in the presence and absence of serum (Fig. 4). Serum withdrawal alone increased the length of neuritic processes from 28.5 6 0.6 lm to 38.7 6 1.8 lm, a 36% increase compared with the control group that remained in 10% serum (Fig. 4A; P < 0.002). PEMF1 treatment further enhanced this effect to 67% compared with serum controls (P < 0.0001) and 23% compared with the null group (P < 0.004). Effects of PEMF signals on neurite length were also tested in the presence of serum. Medium with 10% serum was removed from cultures on the day after plating and

replaced with new medium containing 10% serum. Designated dishes were then exposed to PEMF1. Interestingly, medium change significantly reduced neurite length in both null and PEMF treatment groups, by 18% (23.2 6 0.9 lm; P < 0.0002) and 10% (25.6 6 0.6 lm; P < 0.01), respectively, compared with mean neurite length in the control group (28.5 6 0.6 lm; Fig. 4A,B). However, neurites were 10% longer in the group that received PEMF1 treatment compared with those of the null group (P < 0.03), albeit by a percentage lower than that achieved in serum-free medium. To determine

Journal of Neuroscience Research

766

Lekhraj et al.

Fig. 5. Effects of PEMF and Bt2cAMP on protein content. Cultures were established in medium with 10% serum. After 1 day, medium was changed to serum-free medium (A,C) or serum-containing medium (B). Groups of cultures were exposed to PEMF1 treatment or 1 mM Bt2cAMP for 3 days. A: PEMF1 reduced protein content compared with the null group (*P < 0.004), an effect similar in magnitude to treatment with 1 mM Bt2cAMP (*P < 0.007) or the combination of PEMF1 and Bt2cAMP (*P < 0.0001). B: In 10% serum, there were no significant effects with PEMF1 treatment, although Bt2cAMP (*P < 0.03) and the combination of Bt2cAMP and PEMF signals (*P < 0.02) decreased protein content. C: No significant effects of PEMF were found on the percentage of cells with processes.

whether additional PEMF treatment would increase the magnitude of this effect in the presence of serum, culture medium was replaced, and cultures were exposed to the PEMF signals administered continuously for 15 min every hour (PEMF2 regimen; Fig. 4C). Results show that this regimen increased neurite length by 15% over the null group (P < 0.02) as well as the PEMF1 group (P < 0.03), which was exposed less frequently to PEMF signals. Neuronal differentiation is characterized by neurite outgrowth and increased cell size, reflected in the size of

the cell body. Therefore, we measured the effects of PEMF signals on cell body diameter. Cultures were established and medium with serum was changed after 1 day to equivalent medium. One group of cultures was exposed to the PEMF2 treatment regimen and another group was maintained in the presence of 1 mM Bt2cAMP. Both of these treatments increased cell body size (Fig. 4D) from 26 6 3.1 lm to 33 6 0.7 lm (P < 0.03) and 36 6 0.7 lm (P < 0.006), respective increases of 27% and 38% compared with the null group. These changes were in the same direction as effects on neurite length but were greater in magnitude. We demonstrated that, in both serum-enriched and serum-depleted cultures, PEMF treatment increased neurite length. In development, neuronal stem cells cease proliferation before they extend processes toward their anatomical targets. Therefore, we investigated whether this signal would also affect cell numbers. Using protein content as an index of cell numbers, we measured protein levels in cultures exposed to serum and PEMF treatment (Fig. 5A,B). PEMF1 reduced protein content by 28% compared with the null group (P < 0.004), similar to cultures treated with Bt2cAMP (26% decrease; P < 0.007 vs. null group). The combination of PEMF and Bt2cAMP resulted in an average reduction in protein content of 41% with respect to the null group (P < 0.0001); however, this reduction was no different from reductions produced by either the cyclic nucleotide or PEMF signals alone. In the presence of 10% serum (Fig. 5B), there were no significant effects of the PEMF1 treatment regimen on protein content; levels in cultures exposed to PEMF1 were similar to the control group levels without medium change. Only Bt2cAMP and the combination of Bt2cAMP and PEMF signals significantly decreased protein content by 16% (P < 0.03) and 18% (P < 0.02), respectively. It should be noted that these reductions, though reflecting cell numbers, were not attributed to cell death, because there was no visual evidence of dying cells under any of these conditions. The percentage of cells with processes appeared to be higher than that in the null group with PEMF1 treatment, but the difference was not statistically significant (Fig. 5C). Cyclic AMP and its analogs have been shown by others to induce the differentiation of neurons (Rydel and Greene, 1988; Dugan et al., 1999; Sanchez et al., 2004), including the MN9D cell line (Zhou et al., 2011). Our experiments confirmed that Bt2cAMP increased neurite length and decreased protein content, an index of cell numbers, and showed that PEMF had similar effects. Therefore, it was possible that PEMF signals were acting through the formation of cAMP. We tested this possibility using several approaches. Because the PEMF-mediated decrease in protein content was the most robust effect with the lowest variability within groups, measuring protein content would be expected to have the highest sensitivity to detect treatment effects. Protein content was therefore used as an endpoint to examine the contribution of cAMP. Cultures were treated with 213 lM dideoxyadenosine (DDA), an Journal of Neuroscience Research

PEMF Signals Enhance Neuronal Differentiation

767

protein achieved by once-daily PEMF treatment (PEMF1 vs. PEMF1 1 DDA; P < 0.02). In another experiment, cells were treated with 1 mM Bt2cAMP, a concentration producing optimal effects on neurite length, in the presence or absence of PEMF signals. Neurite length was then assessed (Fig. 6B). If PEMF induced differentiation through a cAMPindependent pathway, the effects of each type of treatment should be additive. Results demonstrate that Bt2cAMP alone and the combination of this agent with PEMF resulted in mean neurite lengths of 55 6 7 lm and 57 6 6 lm, respectively, indicating that PEMF treatment did not further enhance neurite growth. Taken together, our results were consistent with the possibility that PEMF signals mediated neurite elongation through cAMP. Therefore, we investigated whether endogenous cAMP levels were increased by PEMF. MN9D cells were plated, and one day after plating the medium was changed to serum-free medium and cultures were treated with PEMF (null) signals. Cell lysates were collected at specified times, from 15 min to 5 hr after the beginning of treatment, where all groups, except for the group that was treated for 15 min, received 30 min of PEMF treatment, and cAMP was quantified by ELISA. Both the PEMF and the null groups demonstrated an increase in cAMP levels over time (Fig. 6C). Importantly, cultures treated with PEMF exhibited a twofold increase in cAMP production at 2 hr after treatment from 121 to 248 pmol/ml (P < 0.05), which increased to a 2.5-fold increase at 5 hr (P < 0.01).

Fig. 6. The actions of PEMF on neurite outgrowth may be mediated by cAMP. Cultures were established in medium with 10% serum and changed to serum-free medium upon treatment. A: Inhibition of cAMP: cultures were treated with PEMF1 signals in the presence or absence of dideoxyadenosine (DDA), a competitive inhibitor of adenylyl cyclase. **PEMF1 vs. DDA 1 PEMF1 (P < 0.02). B: Cultures were treated with 1 mM Bt2cAMP 6 PEMF1 treatment. C: Time course of cAMP accumulation: Cultures were treated with PEMF1 or null signals for 15 (first time point) to 30 min (remaining time points), and lysates were harvested at specified times after treatment. cAMP levels were quantified by ELISA. *PEMF vs. null at 2 hr (P < 0.05); **PEMF vs. null at 5 hr (P < 0.01).

adenylate cyclase inhibitor (Holgate et al., 1980), in the presence or absence of the PEMF1 treatment regimen for 3 days. Results demonstrate that, in the absence of serum (Fig. 6A), DDA, which had no effects on the protein content of the null group, entirely blocked the decrease in Journal of Neuroscience Research

DISCUSSION Neuronal development is a stepwise process that includes proliferation and determination of phenotypic fate, followed by differentiation, migration, and maturation (Hamburger, 1952). A similar scheme is exhibited by MN9D cells, a hybridoma cell line derived from neural crest and primary mesencephalic dopaminergic neurons. These cells proliferate in medium with serum, and a small percentage of these cells can differentiate spontaneously, whereby they cease proliferation and extend long neuritic processes and the size of the cell body increases. Importantly, changing culture conditions, such as removing serum (Figs. 3–5), the addition of signaling molecules such as Bt2cAMP (Figs. 1, 4, 5), or treatment with retinoic acid, will enhance this process (Rydel and Greene, 1988; Eom et al., 2005; Zhou et al., 2011). Although the cessation of proliferation, somal enlargement, and neuritic outgrowth describe a developmental process, neurite extension itself may also represent synaptic plasticity, the ability to form new synapses by collateralization and remodeling. Therefore, cultured MN9D cells will also serve to test strategies to increase synaptic plasticity in aging and disease. Here we demonstrate that a low-amplitude radiofrequency PEMF signal delivered through an external source (Fig. 2) enhanced the morphological differentiation of MN9D cells as demonstrated by increased neurite length

768

Lekhraj et al.

and cell body size over the course of 3 days (Fig. 3). In contrast to previous reports of effects on sympathetic neurons, this cell line originated from mesencephalic dopaminergic neurons from the central nervous system that were fused to neuroblastoma cells. In addition, this is one of the few reports in which neurite extension was achieved without the addition of an exogenous neurotrophic factor, such as NGF. In the presence of serum, in which most MN9D cells were in the proliferative state, continuous PEMF exposure produced effects superior to oncedaily treatment on neurite outgrowth (Fig. 4). PEMF signals were more effective after serum withdrawal, which also stimulated neurite growth and decreased protein content. Serum withdrawal has been shown to induce the differentiation of neuroblastoma and pheochromocytoma (PC12) cells through extracellular signal-regulated kinase (ERK1/2) signaling, known to be active in development when cells differentiate, and stress-activated protein kinases, such as c-Jun N-terminal kinase (JNK; Leppa et al., 1998; Evangelopoulos et al., 2005). Others have demonstrated that these signaling molecules were also required for retinoic acid-induced MN9D cell differentiation, and when these molecules were blocked neurite growth was eliminated (Eom et al., 2005). It is therefore not surprising that PEMF treatment produced superior results in the absence of serum; its actions have been described as “enhancement” of ongoing processes, which in this culture system can be interpreted as either “development” or “stress.” These results are similar to results with PC12 cells, in which a PEMF signal enhanced neurite outgrowth in medium with 4% serum but inhibited outgrowth when cells were cultured in 15% serum (McFarlane et al., 2000). As with cAMP, PEMF also increased the mean cell body diameter (Fig. 4), although it is not known whether this represents an actual increase in the volume of the cell body, because increased cell diameter may also reflect increased adhesion to polyornithine-coated culture dishes. However, both possibilities are associated with differentiation, as cells increase in size when they mature as well as forming a stronger attachment to the substrate as they cease division to initiate neurite outgrowth, a process that requires substrate adhesion (Purves and Lichtman, 1985). Numerous studies have demonstrated that PEMF signals could facilitate neurite outgrowth, nerve regeneration, and behavioral recovery after nerve injuries in vivo (Wilson and Jagadeesh, 1976; Ito and Bassett, 1983; Raji and Bowden, 1983; Kanje et al., 1993; Drucker-Colin et al., 1994; Walker et al., 1994, 2007; Longo et al., 1999), including Parkinson’s disease patients (Sandyk, 1999), as well as increase the length, number, and branches of neuritic processes in vitro (Blackman et al., 1993; Drucker-Colin et al., 1994; Greenebaum et al., 1996; Macias et al., 2000; McFarlane et al., 2000; Schimmelpfeng et al., 2005; Zhang et al., 2005), although others have demonstrated negative effects or failed to show similar effects (Shah et al., 2001; Baptista et al., 2009). Unfortunately, multiple PEMF signal configurations are

rarely tested in identical paradigms, and the two treatment regimens tested here do not formally address dosimetry, so it is not possible to determine why different investigators obtained discordant results. However, it may be possible to draw some broad conclusions. For example, in our study, it appears that more frequent treatment with the same PEMF signal produced a greater effect in cells maintained in medium with 10% serum. In a study by another group, a stronger field (1.36 mT peak amplitude) was applied to PC12 cells, and effects depended on the duty cycle (pulse duration 3 repetition rate); low duty cycles (10%) decreased the number of neurite-bearing cells, but their length was greater than in controls, and high duty cycles increased the percentage of neuritebearing cells, but they had shorter neurites (Zhang et al., 2005). With a significantly lower duty cycle of 2.5% (0.005-sec burst width 3 5/sec repetition rate 3 100), our results are consistent with the notion that low duty cycles tend to increase neurite length (Fig. 4). This decrease was not reflected in increased percentages of differentiated cells by direct counting (Fig. 5C), possibly because of the variability inherent in the method, which sampled a small region of each culture dish (compare error bars in Figs. 1C, 5B). However, the significant decrease in protein content after PEMF treatment in the absence of cell death suggests that this PEMF signal also caused cells to cease mitosis and differentiate. Unlike the study discussed above, changes in protein were similar in magnitude and in the same direction as the observed increases in process length, suggesting that this particular PEMF signal configuration can both induce differentiation and facilitate neurite elongation. Although others have demonstrated that PEMF enhanced neuronal differentiation, the pathway by which this is achieved is largely unknown. It has been shown that PEMF-induced osteoblast proliferation and differentiation were mediated through the induction of the nitric oxide/cyclic guanosine monophosphate/protein kinase G (NO/cGMP/PKG) pathway (Fitzsimmons et al., 2008; Cheng et al., 2011). Cerebellar NO and cGMP could be elevated by PEMF (Miura et al., 1993; Noda et al., 2000), and apoptosis was attenuated by PEMF in granule neurons in the cerebellum (Oda and Koike, 2004), where levels of nitric oxide synthase are relatively high (Bredt et al., 1990), but we found no evidence that NO or cGMP were involved in neuronal differentiation in MN9D cells. Instead, it is possible that the NO/cGMP/ PKG pathway is more important in preventing neuronal death (Farinelli et al., 1996; Kim et al., 1999; Fiscus, 2002), and, even though our serum-withdrawal paradigm may kill neurons eventually, we saw no evidence of cell death within the 3-day period of PEMF treatment. Alternatively, retinoic acid and cAMP appear to be major determinants of differentiation in neurons (Rydel and Greene, 1988; Dugan et al., 1999; Johnston et al., 2004; Sanchez et al., 2004; Eom et al., 2005; Evangelopoulos et al., 2005; Kim et al., 2008), including MN9D cells (Zhou et al., 2011). A previous study demonstrated that cAMP was induced by PEMF in hippocampal slices Journal of Neuroscience Research

PEMF Signals Enhance Neuronal Differentiation

(Hogan and Wieraszko, 2004), which the authors associated with evoked potentials, but cell specificity and morphological changes were not evaluated. We previously reported that the PEMF signals increased cAMP levels in MN9D cells in response to temperature change (Pilla et al., 2011). Although Bt2cAMP had no significant effects on either numbers or proportions of differentiating cells (Figs. 1C, 5C), both PEMF and Bt2cAMP decreased protein content (Fig. 5), perhaps a more sensitive index of cell numbers, because protein was quantified in lysates from all cells in each dish. Moreover, the decrease in protein content could be blocked by DDA, an inhibitor of adenylyl cyclase (Fig. 6A), suggesting that PEMF signals, acting through cAMP, inhibited cell proliferation. Effects of PEMF treatment and Bt2cAMP together on neurite outgrowth or protein content were not additive, supporting the notion that PEMF and cAMP were inducing neurite elongation along the same pathway (Fig. 6B). Finally, the demonstration that PEMF exposure increased endogenous cAMP levels after a single treatment (Fig. 6C) provides additional support for the idea that PEMF produced the biological effects reported here by increasing cAMP production. Further studies are required to elucidate the effects of PEMF on signal transduction molecules both upstream and downstream of cAMP. Developmental processes can be governed or modulated by endogenous and exogenous electromagnetic fields (Metcalf and Borgens, 1994; Yao et al., 2011), but little is known about their role in healthy adult tissue. In contrast, neuropathological evidence and results from studies using experimental models of neurodegenerative diseases led to the proposition that developmental processes are reiterated when neurons are challenged with toxic stress. Examples include challenge of cultured neurons with neurotoxic glutamatergic agonists, which will induce a “trophic” sprouting response (Kwon and Sabatini, 2011). Dopaminergic neurons appear to compensate for injury to the striatum in humans with Parkinson’s disease who received neural transplants through the sprouting of surviving neurons (Kordower et al., 1991) and by the collateralization of surviving axons in animal models (Shults et al., 1995). From these examples, it could be argued that these responses are an attempt at synaptic plasticity that may be insufficient to compensate for neuronal loss, resulting in disease progression. Therefore, the application of a nontoxic PEMF signal to increase outgrowth and sprouting of existing axons to increase the efficacy of neurotransmission in surviving neurons may be of clinical interest for traumatic and pathological neurodegenerative conditions. ACKNOWLEDGMENTS The authors thank Andre DiMino and Matthew Drummer, Ivivi Technologies (Montvale, NJ), for providing and calibrating the PEMF device; Shahla Powell for participating in image analysis studies; and Parviz Lalezari for his scientific guidance. A.A.P. is a scientific consultant to Ivivi Technologies. Journal of Neuroscience Research

769

REFERENCES Arendash GW, Sanchez-Ramos J, Mori T, Mamcarz M, Lin X, Runfeldt M, Wang L, Zhang G, Sava V, Tan J, Cao C. 2010. Electromagnetic field treatment protects against and reverses cognitive impairment in Alzheimer’s disease mice. J Alzheimers Dis 19:191–210. Baker LL, Chambers R, DeMuth SK, Villar F. 1997. Effects of electrical stimulation on wound healing in patients with diabetic ulcers. Diabetes Care 20:405–412. Baptista AF, Goes BT, Menezes D, Gomes FC, Zugaib J, Stipursky J, Gomes JR, Oliveira JT, Vannier-Santos MA, Martinez AM. 2009. PEMF fails to enhance nerve regeneration after sciatic nerve crush lesion. J Peripher Nerv Syst 14:285–293. Bassett CA, Pawluk RJ, Pilla AA. 1974. Augmentation of bone repair by inductively coupled electromagnetic fields. Science 184:575–577. Beck W. 2004. The Bioelectromagnetics Society: history of the first 25 years. Frederick, MD: The Bioelectromagnetics Society. Blackman CF, Benane SG, House DE, Pollock MM. 1993. Action of 50 Hz magnetic fields on neurite outgrowth in pheochromocytoma cells. Bioelectromagnetics 14:273–286. Bodamyali T, Bhatt B, Hughes FJ, Winrow VR, Kanczler JM, Simon B, Abbott J, Blake DR, Stevens CR. 1998. Pulsed electromagnetic fields simultaneously induce osteogenesis and upregulate transcription of bone morphogenetic proteins 2 and 4 in rat osteoblasts in vitro. Biochem Biophys Res Commun 250:458–461. Bredt DS, Hwang PM, Snyder SH. 1990. Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347:768–770. Callaghan MJ, Chang EI, Seiser N, Aarabi S, Ghali S, Kinnucan ER, Simon BJ, Gurtner GC. 2008. Pulsed electromagnetic fields accelerate normal and diabetic wound healing by increasing endogenous FGF-2 release. Plast Reconstr Surg 121:130–141. Cheng G, Zhai Y, Chen K, Zhou J, Han G, Zhu R, Ming L, Song P, Wang J. 2011. Sinusoidal electromagnetic field stimulates rat osteoblast differentiation and maturation via activation of NO-cGMP-PKG pathway. Nitric Oxide 25:316–325. Choi HK, Won LA, Kontur PJ, Hammond DN, Fox AP, Wainer BH, Hoffmann PC, Heller A. 1991. Immortalization of embryonic mesencephalic dopaminergic neurons by somatic cell fusion. Brain Res 552:67– 76. Drucker-Colin R, Verdugo-Diaz L, Mendez M, Carrillo-Ruiz J, Morgado-Valle C, Hernandez-Cruz A, Corkidi G. 1994. Comparison between low frequency magnetic field stimulation and nerve growth factor treatment of cultured chromaffin cells, on neurite growth, noradrenaline release, excitable properties, and grafting in nigrostriatal lesioned rats. Mol Cell Neurosci 5:485–498. Dugan LL, Kim JS, Zhang Y, Bart RD, Sun Y, Holtzman DM, Gutmann DH. 1999. Differential effects of cAMP in neurons and astrocytes. Role of B-raf. J Biol Chem 274:25842–25848. Eom DS, Choi WS, Ji S, Cho JW, Oh YJ. 2005. Activation of c-Jun N-terminal kinase is required for neurite outgrowth of dopaminergic neuronal cells. Neuroreport 16:823–828. Evangelopoulos ME, Weis J, Kruttgen A. 2005. Signalling pathways leading to neuroblastoma differentiation after serum withdrawal: HDL blocks neuroblastoma differentiation by inhibition of EGFR. Oncogene 24:3309–3318. Farinelli SE, Park DS, Greene LA. 1996. Nitric oxide delays the death of trophic factor-deprived PC12 cells and sympathetic neurons by a cGMP-mediated mechanism. J Neurosci 16:2325–2334. Fiscus RR. 2002. Involvement of cyclic GMP and protein kinase G in the regulation of apoptosis and survival in neural cells. Neurosignals 11: 175–190. Fitzsimmons RJ, Gordon SL, Kronberg J, Ganey T, Pilla AA. 2008. A pulsing electric field (PEF) increases human chondrocyte proliferation through a transduction pathway involving nitric oxide signaling. J Orthop Res 26:854–859.

770

Lekhraj et al.

Foley-Nolan D, Barry C, Coughlan RJ, O’Connor P, Roden D. 1990. Pulsed high frequency (27MHz) electromagnetic therapy for persistent neck pain. A double blind, placebo-controlled study of 20 patients. Orthopedics 13:445–451. Gona AG, Yu MC, Gona O, al-Rabiai S, Von Hagen S, Cohen E. 1993. Effects of 60 Hz electric and magnetic fields on the development of the rat cerebellum. Bioelectromagnetics 14:433–447. Gossling HR, Bernstein RA, Abbott J. 1992. Treatment of ununited tibial fractures: a comparison of surgery and pulsed electromagnetic fields (PEMF). Orthopedics 15:711–719. Greene LA, Tischler AS. 1976. Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. Proc Natl Acad Sci U S A 73:2424–2428. Greenebaum B, Sutton CH, Vadula MS, Battocletti JH, Swiontek T, DeKeyser J, Sisken BF. 1996. Effects of pulsed magnetic fields on neurite outgrowth from chick embryo dorsal root ganglia. Bioelectromagnetics 17:293–302. Hamburger V. 1952. Development of the nervous system. Ann N Y Acad Sci 55:117–132. Henderson CE. 1996. Role of neurotrophic factors in neuronal development. Curr Opin Neurobiol 6:64–70. Hogan MV, Wieraszko A. 2004. An increase in cAMP concentration in mouse hippocampal slices exposed to low-frequency and pulsed magnetic fields. Neurosci Lett 366:43–47. Holgate ST, Lewis RA, Austen KF. 1980. Role of adenylate cyclase in immunologic release of mediators from rat mast cells: agonist and antagonist effects of purine- and ribose-modified adenosine analogs. Proc Natl Acad Sci U S A 77:6800–6804. Hristov J, Perez VH. 2011. Critical analysis of data concerning Saccharomyces cerevisiae free-cell proliferations and fermentations assisted by magnetic and electromagnetic fields. Int Rev Chem Eng 3:3–20. Ito H, Bassett CA. 1983. Effect of weak, pulsing electromagnetic fields on neural regeneration in the rat. Clin Orthop 181:283–290. Itoh M, Montemayor JS Jr, Matsumoto E, Eason A, Lee MH, Folk FS. 1991. Accelerated wound healing of pressure ulcers by pulsed high peak power electromagnetic energy (diapulse). Decubitus 4:24–25, 29– 34. Jasti AC, Wetzel BJ, Aviles H, Vesper DN, Nindl G, Johnson MT. 2001. Effect of a wound healing electromagnetic field on inflammatory cytokine gene expression in rats. Biomed Sci Instrum 37:209–214. Johnston CA, Beazely MA, Bilodeau ML, Andrisani O, Watts VJ. 2004. Differentiation-induced alterations in cyclic AMP signaling in the Cath.a differentiated (CAD) neuronal cell line. J Neurochem 88:1497– 1508. Kanje M, Rusovan A, Sisken B, Lundborg G. 1993. Pretreatment of rats with pulsed electromagnetic fields enhances regeneration of the sciatic nerve. Bioelectromagnetics 14:353–359. Kim S, Im WS, Kang L, Lee ST, Chu K, Kim BI. 2008. The application of magnets directs the orientation of neurite outgrowth in cultured human neuronal cells. J Neurosci Methods 174:91–96. Kim YM, Chung HT, Kim SS, Han JA, Yoo YM, Kim KM, Lee GH, Yun HY, Green A, Li J, Simmons RL, Billiar TR. 1999. Nitric oxide protects PC12 cells from serum deprivation-induced apoptosis by cGMP-dependent inhibition of caspase signaling. J Neurosci 19:6740– 6747. Kloth LC, Berman JE, Sutton CH, Jeutter DC, Pilla AA, Epner ME. 1999. Effect of pulsed radio frequency stimulation on wound healing: a double-blind pilot clinical study. In: Bersani F, editor. Electricity and magnetism in biology and medicine. New York: Plenum. p 875–878. Koppes AN, Seggio AM, Thompson DM. 2011. Neurite outgrowth is significantly increased by the simultaneous presentation of Schwann cells and moderate exogenous electric fields. J Neural Eng 8:046023. Kordower JH, Cochran E, Penn RD, Goetz CG. 1991. Putative chromaffin cell survival and enhanced host-derived TH-fiber innervation

following a functional adrenal medulla autograft for Parkinson’s disease. Ann Neurol 29:405–412. Kotaleski JH, Blackwell KT. 2010. Modelling the molecular mechanisms of synaptic plasticity using systems biology approaches. Nat Rev Neurosci 11:239–251. Kwon HB, Sabatini BL. 2011. Glutamate induces de novo growth of functional spines in developing cortex. Nature 474:100–104. Leppa S, Saffrich R, Ansorge W, Bohmann D. 1998. Differential regulation of c-Jun by ERK and JNK during PC12 cell differentiation. EMBO J 17:4404–4413. Longo FM, Yang T, Hamilton S, Hyde JF, Walker J, Jennes L, Stach R, Sisken BF. 1999. Electromagnetic fields influence NGF activity and levels following sciatic nerve transection. J Neurosci Res 55:230–237. Loo CK, Mitchell PB. 2005. A review of the efficacy of transcranial magnetic stimulation (TMS) treatment for depression, and current and future strategies to optimize efficacy. J Affect Disord 88:255–267. Macias MY, Battocletti JH, Sutton CH, Pintar FA, Maiman DJ. 2000. Directed and enhanced neurite growth with pulsed magnetic field stimulation. Bioelectromagnetics 21:272–286. McFarlane EH, Dawe GS, Marks M, Campbell IC. 2000. Changes in neurite outgrowth but not in cell division induced by low EMF exposure: influence of field strength and culture conditions on responses in rat PC12 pheochromocytoma cells. Bioelectrochemistry 52:23–28. McLeod BR, Pilla AA, Sampsel MW. 1983. Electromagnetic fields induced by Helmholtz aiding coils inside saline-filled boundaries. Bioelectromagnetics 4:357–370. Metcalf ME, Borgens RB. 1994. Weak applied voltages interfere with amphibian morphogenesis and pattern. J Exp Zool 268:322–338. Miura M, Takayama K, Okada J. 1993. Increase in nitric oxide and cyclic GMP of rat cerebellum by radio frequency burst-type electromagnetic field radiation. J Physiol 461:513–524. Nelson FR, Zvirbulis R, Pilla AA. 2012. Non-invasive electromagnetic field therapy produces rapid and substantial pain reduction in early knee osteoarthritis: a randomized double-blind pilot study. Rheumatol Int 33:2169–2173. Noda Y, Mori A, Liburdy RP, Packer L. 2000. Pulsed magnetic fields enhance nitric oxide synthase activity in rat cerebellum. Pathophysiology 7:127–130. Oda T, Koike T. 2004. Magnetic field exposure saves rat cerebellar granule neurons from apoptosis in vitro. Neurosci Lett 365:83–86. Patino O, Grana D, Bolgiani A, Prezzavento G, Mino J, Merlo A, Benaim F. 1996. Pulsed electromagnetic fields in experimental cutaneous wound healing in rats. J Burn Care Rehabil 17:528–531. Pilla AA. 2007. Biological and medical aspects of electromagnetic fields. Boca Raton, FL: CRC Press. p 351–411. Pilla AA. 2012. Electromagnetic fields instantaneously modulate nitric oxide signaling in challenged biological systems. Biochem Biophys Res Commun 426:330–333. Pilla AA, Fitzsimmons RJ, Muehsam DJ, Rohde C, Wu JK, Casper D. 2011. Electromagnetic fields as first messenger in biological signaling: application to calmodulin-dependent signaling in tissue repair. Biochim Biophys Acta 1810:1236–1245. Purves D, Lichtman J. 1985. Principles of neural development. Sunderland, MA: Sinauer Associates. Raji AR, Bowden RE. 1983. Effects of high-peak pulsed electromagnetic field on the degeneration and regeneration of the common peroneal nerve in rats. J Bone Joint Surg Br 65:478–492. Rasouli J, Lekhraj R, White NM, Flamm ES, Pilla AA, Strauch B, Casper D. 2012. Attenuation of interleukin-1beta by pulsed electromagnetic fields after traumatic brain injury. Neurosci Lett 519:4–8. Rohde C, Chiang A, Adipoju O, Casper D, Pilla AA. 2010. Effects of pulsed electromagnetic fields on interleukin-1 beta and postoperative pain: a double-blind, placebo-controlled, pilot study in breast reduction patients. Plast Reconstr Surg 125:1620–1629. Journal of Neuroscience Research

PEMF Signals Enhance Neuronal Differentiation Rydel RE, Greene LA. 1988. cAMP analogs promote survival and neurite outgrowth in cultures of rat sympathetic and sensory neurons independently of nerve growth factor. Proc Natl Acad Sci U S A 85:1257–1261. Salzberg CA, Cooper-Vastola SA, Perez F, Viehbeck MG, Byrne DW. 1995. The effects of non-thermal pulsed electromagnetic energy on wound healing of pressure ulcers in spinal cord-injured patients: a randomized, double-blind study. Ostomy Wound Manag 41:42–44, 46, 48 passim. Sanchez S, Jimenez C, Carrera AC, Diaz-Nido J, Avila J, Wandosell F. 2004. A cAMP-activated pathway, including PKA and PI3K, regulates neuronal differentiation. Neurochem Int 44:231–242. Sandyk R. 1999. Treatment with AC pulsed electromagnetic fields improves olfactory function in Parkinson’s disease. Int J Neurosci 97: 225–233. Schimmelpfeng J, Weibezahn KF, Dertinger H. 2005. Neuronal outgrowth of PC-12 cells after combined treatment with nerve growth factor and a magnetic field: influence of the induced electric field strength. Bioelectromagnetics 26:74–77. Shah JP, Midkiff P, Brandt PC, Sisken BF. 2001. Growth and differentiation of PC6 cells: the effects of pulsed electromagnetic fields (PEMF). Bioelectromagnetics 22:267–271. Shults CW, Kimber T, Altar CA. 1995. BDNF attenuates the effects of intrastriatal injection of 6-hydroxydopamine. Neuroreport 6:1109–1112. Shupak NM, Prato FS, Thomas AW. 2004. Human exposure to a specific pulsed magnetic field: effects on thermal sensory and pain thresholds. Neurosci Lett 363:157–162. Strauch B, Patel MK, Navarro JA, Berdichevsky M, Yu HL, Pilla AA. 2007. Pulsed magnetic fields accelerate cutaneous wound healing in rats. Plast Reconstr Surg 120:425–430.

Journal of Neuroscience Research

771

Tepper OM, Callaghan MJ, Chang EI, Galiano RD, Bhatt KA, Baharestani S, Gan J, Simon B, Hopper RA, Levine JP, Gurtner GC. 2004. Electromagnetic fields increase in vitro and in vivo angiogenesis through endothelial release of FGF-2. FASEB J 18:1231–1233. Thuile C, Walzl M. 2002. Evaluation of electromagnetic fields in the treatment of pain in patients with lumbar radiculopathy or the whiplash syndrome. Neurorehabilitation 17:63–67. Walker JL, Evans JM, Resig P, Guarnieri S, Meade P, Sisken BS. 1994. Enhancement of functional recovery following a crush lesion to the rat sciatic nerve by exposure to pulsed electromagnetic fields. Exp Neurol 125:302–305. Walker JL, Kryscio R, Smith J, Pilla A, Sisken BF. 2007. Electromagnetic field treatment of nerve crush injury in a rat model: effect of signal configuration on functional recovery. Bioelectromagnetics 28:256– 263. Wilson DH, Jagadeesh P. 1976. Experimental regeneration in peripheral nerves and the spinal cord in laboratory animals exposed to a pulsed electromagnetic field. Paraplegia 14:12–20. Yao L, Pandit A, Yao S, McCaig CD. 2011. Electric field-guided neuron migration: a novel approach in neurogenesis. Tissue Eng Part B Rev 17:143–153. Zhang Y, Ding J, Duan W, Fan W. 2005. Influence of pulsed electromagnetic field with different pulse duty cycles on neurite outgrowth in PC12 rat pheochromocytoma cells. Bioelectromagnetics 26:406–411. Zhou Z, Kim J, Insolera R, Peng X, Fink DJ, Mata M. 2011. Rho GTPase regulation of alpha-synuclein and VMAT2: implications for pathogenesis of Parkinson’s disease. Mol Cell Neurosci 48:29–37.