If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Low intensity repetitive magnetic stimulation reduces expression of genes related to inflammation and calcium signalling in cultured mouse cortical astrocytes
Corresponding author. Experimental and Regenerative Neuroscience, School of Biological Sciences, The University of Western Australia, Nedlands, WA, 6009, Australia.
Experimental and Regenerative Neuroscience, School of Biological Sciences, The University of Western Australia, Nedlands, WA, 6009, AustraliaPerron Institute for Neurological and Translational Science, Nedlands, WA, 6009, Australia
Experimental and Regenerative Neuroscience, School of Biological Sciences, The University of Western Australia, Nedlands, WA, 6009, AustraliaPerron Institute for Neurological and Translational Science, Nedlands, WA, 6009, Australia
Perron Institute for Neurological and Translational Science, Nedlands, WA, 6009, AustraliaSchool of Human Sciences, The University of Western Australia, Nedlands, WA, 6009, Australia
Experimental and Regenerative Neuroscience, School of Biological Sciences, The University of Western Australia, Nedlands, WA, 6009, AustraliaPerron Institute for Neurological and Translational Science, Nedlands, WA, 6009, Australia
Experimental and Regenerative Neuroscience, School of Biological Sciences, The University of Western Australia, Nedlands, WA, 6009, AustraliaPerron Institute for Neurological and Translational Science, Nedlands, WA, 6009, Australia
1 Hz and 10 Hz low intensity rMS induced reductions in astrocyte gene expression in vitro, 5 h post-stimulation.
•
Modulated genes were primarily related to inflammation and calcium signalling pathways.
•
In a subset of genes modulated by LI-rMS, protein levels were also modulated at 5 and 24 h post-stimulation.
Abstract
Repetitive transcranial magnetic stimulation (rTMS) is a form of non-invasive brain stimulation frequently used to induce neuroplasticity in the brain. Even at low intensities, rTMS has been shown to modulate aspects of neuronal plasticity such as motor learning and structural reorganisation of neural tissue. However, the impact of low intensity rTMS on glial cells such as astrocytes remains largely unknown. This study investigated changes in RNA (qPCR array: 125 selected genes) and protein levels (immunofluorescence) in cultured mouse astrocytes following a single session of low intensity repetitive magnetic stimulation (LI-rMS – 18 mT). Purified neonatal cortical astrocyte cultures were stimulated with either 1Hz (600 pulses), 10Hz (600 or 6000 pulses) or sham (0 pulses) LI-rMS, followed by RNA extraction at 5 h post-stimulation, or fixation at either 5 or 24-h post-stimulation. LI-rMS resulted in a two-to-four-fold downregulation of mRNA transcripts related to calcium signalling (Stim1 and Orai3), inflammatory molecules (Icam1) and neural plasticity (Ncam1). 10Hz reduced expression of Stim1, Orai3, Kcnmb4, and Ncam1 mRNA, whereas 1Hz reduced expression of Icam1 mRNA and signalling-related genes. Protein levels followed a similar pattern for 10Hz rMS, with a significant reduction of STIM1, ORAI3, KCNMB4, and NCAM1 protein compared to sham, but 1Hz increased STIM1 and ORAI3 protein levels relative to sham. These findings demonstrate the ability of 1Hz and 10Hz LI-rMS to modulate specific aspects of astrocytic phenotype, potentially contributing to the known effects of low intensity rTMS on excitability and neuroplasticity.
With the increasing use of repetitive transcranial magnetic stimulation (rTMS) for treating a wide range of neurological and neuropsychiatric conditions, there is a growing need to better understand the effects of electromagnetic stimulation on the brain. The forms of rTMS most commonly used in the clinic generate action potentials in neurons and elicit synaptic plasticity in the form of long-term potentiation (LTP) and long-term depression (LTD) [
Transcranial pulsed magnetic field stimulation facilitates reorganization of abnormal neural circuits and corrects behavioral deficits without disrupting normal connectivity.
], suggesting that supra-threshold stimulation (i.e. induced action potential firing) may not be necessary to induce plasticity. LI-rTMS (<120 mT) alters many types of neural function, from molecular to network-level effects, that are similar to those observed at high intensity. In vitro, LI-repetitive magnetic stimulation (rMS, no cranium) at high frequency facilitates excitability of pyramidal neurons in mouse cortical slices [
Reversed effects of intermittent theta burst stimulation following motor training that vary as a function of training-induced changes in corticospinal excitability.
]. Because LI-rTMS effects are elicited independently of action potential induction and involve universal cellular machinery such as calcium signalling, these effects may be mediated by glia as well as neurons.
Astrocytes are important modulators of neuroplasticity, primarily through the release of gliotransmitters at the synapse [
]. However, the contributions of astrocytes and other glial cells to the effects of non-invasive brain stimulation have only recently been considered [
]. In vitro, 1Hz (but not 10Hz, cTBS or BHFS) LI-rMS (18 mT) increases intracellular calcium release in the cytoplasm of purified cultures of mouse cortical astrocytes [
]. Similarly, various frequencies of HI-rTMS and LI-rTMS alter the expression of GFAP (a commonly used cytoskeletal marker of astrocytic reactivity) and the number of GFAP-positive astrocytes in numerous in vivo rodent neurotrauma and neurodegeneration models [
Intermittent theta-burst stimulation rescues dopamine-dependent corticostriatal synaptic plasticity and motor behavior in experimental parkinsonism: possible role of glial activity.
High-frequency repetitive transcranial magnetic stimulation improves functional recovery by inhibiting neurotoxic polarization of astrocytes in ischemic rats.
Repetitive transcranial magnetic stimulation inhibits Sirt1/MAO-A signaling in the prefrontal cortex in a rat model of depression and cortex-derived astrocytes.
]. However, almost nothing is known about the mRNA and protein expression changes that occur following LI-rTMS and how they might influence astrocytic phenotype.
Here, the effect of LI-rMS on gene expression was examined in purified cortical astrocyte cultures generated from neonatal mouse pups. Astrocyte cultures were stimulated with low (1Hz) and high (10Hz) frequencies at a low intensity (18 mT) and RNA was extracted 5 h following stimulation to examine 125 genes of interest relating broadly to four main astrocytic biological functions: calcium signalling, inflammation and injury, plasticity, and morphological/cytoskeletal composition. We further studied if gene expression changes resulted in protein expression changes in a selection of regulated genes at 5 and 24 h post-stimulation.
Methods
Tissue culture
Experimental procedures were conducted with ethics approval from The University of Western Australia Animal Ethics Committee (RA/3/100/1462) under the Australian Code of Practice for the Use of Animals in Research. Primary cortical cultures were prepared from euthanased (>160 mg/kg sodium pentobarbitone, i.p.) 1 day postnatal male and female C57Bl6J mice (n = 31 over 6 culture runs), obtained from the Animal Resources Centre, Western Australia. Cultures were prepared as previously shown [
] and are described in detail in the supplementary materials. Cells were grown in T75 flasks until confluent (day in vitro 11, DIV11), and then detached (0.25% trypsin, Gibco) and replated at a density of 0.04 × 106 onto 13 mm diameter glass coverslips in 24-well plates. To ensure an enriched astrocyte culture, coverslips were assessed for the proportions of various cell types. Cells at DIV14 were consistently >97% positive for GFAP across runs, with the remainder of cells immunohistochemically and morphologically identified as neurons, microglia, or fibroblasts (data not shown).
Repetitive magnetic stimulation
On DIV14, cells were stimulated with LI-rMS or sham within the cell culture incubator [
]. Custom-built round coils (inside diameter 8 mm, outside diameter 17 mm, thickness 9.5 mm) were attached to the bottom of the 24-well plates to stimulate the cells. The coils were powered by a 12V magnetic pulse generator (Global Energy Medicine). The applied frequencies were 1Hz (600 pulses, 10 min, ‘10m1Hz’), 10Hz (600 pulses, 1 min, ‘1m10Hz’; or 6000 pulses 10 min, ‘10m10Hz’), or sham (10 min, generator switched off, no pulses delivered). 1Hz LI-rMS was the primary focus of this study because of its previous effects on calcium levels in astrocytes in vitro [
]. Although 10Hz did not elicit calcium changes in our previous work, applying 10Hz offered us the opportunity to pulse match with 1Hz. In addition, 10 Hz delivered at high intensity has been recently shown to alter astrocyte gene expression in an in vitro ischemic model [
High-frequency repetitive transcranial magnetic stimulation improves functional recovery by inhibiting neurotoxic polarization of astrocytes in ischemic rats.
Repetitive transcranial magnetic stimulation inhibits Sirt1/MAO-A signaling in the prefrontal cortex in a rat model of depression and cortex-derived astrocytes.
]. The field intensity at the surface of the coverslip above the centre of the coil was measured with a Hall probe to be 18 mT, and each frequency delivered a monophasic pulse with a rise time of 320μs, equating to a dB/dt of approximately 56T/s. The coils do not generate heat or vibration above background levels during stimulation [
Changes in gene expression were examined via RT-qPCR for each frequency to investigate downstream changes relating to calcium signalling, inflammation and injury, plasticity, and morphological/cytoskeletal composition. RNA was collected from 3 biological replicates, each pooled from 45 coverslips (total 135), for each stimulation group, 5 h after the end of stimulation. The 5 h timepoint was chosen for methodological consistency with a previous experiment conducted in our laboratory investigating mRNA changes in primary neuronal cultures following LI-rMS [
]. RNA was extracted and purified from the astrocytes on coverslips using a miRNeasy kit (Qiagen), with DNase digestion (RNase-free DNase Set, Qiagen). RNA purity and quantity were measured (NanoDrop) prior to cDNA synthesis (RT2 First Strand Kit, Qiagen) from 1 μg of RNA for each biological replicate within each stimulation group. cDNA for each sample was added to custom RT2 Profiler PCR Arrays and amplified on a Bio-Rad CFX384. The custom array consisted of 125 genes known to be influential in astrocyte biology and function and focused on 4 main biological processes (see supplementary information). The amplification protocol was 10 min at 95 °C followed by 60 cycles of 95 °C for 15s and 60 °C for 1 min. Following this, a melt curve was obtained by increasing the temperature from 65 °C to 95 °C in 1 °C increments. An identical fluorescence threshold value for the exponential phase of amplification was used for each run, and genes were excluded if they passed the threshold after cycle 40 or had abnormal melt curves. The genes Brca1, Cga, Crh, FoxA2, Grm1, and Hnf1a were excluded from analysis as they reached threshold after cycle 40.
Protein expression
Changes in protein expression for selected genes were examined via immunofluorescent staining. Coverslips were stained for protein products of genes that were significantly downregulated compared to sham stimulation: antibodies to STIM1 (cat PA5-82455, ThermoFisher), ORAI3 (cat. MA5-15778, ThermoFisher), KCNMB4 (cat. PA5-77611) and CD56 (additional protein name of NCAM1, cat. MA5-11563, ThermoFisher). Five coverslips per stimulation group replicate were analysed for protein intensity at 5h and 24 h following LI-rMS. See supplementary material for further detail on staining procedure and image analysis. For consistency, the coverslips for all groups were stained in one session for each labelling combination.
Statistical analysis
mRNA and protein data were examined for outliers and normality with SPSS (v21, IBM). RT-qPCR results were normalized on Microsoft Excel and analysed with SPSS. The geometric mean of genes Gapdh, Sdha, and Sod2 were used to normalize genes for analysis and the normalized mean expression levels (log2(2-ΔΔCt)) for each stimulation group were used to determine expression differences between stimulation groups [
]. For each gene, a fold change value was used to compare expression levels between groups with a student’s t-test. Fold changes (log2) greater than ±1 with a p value < 0.05 were considered statistically significant. All significantly regulated genes above the fold change cut-off were further examined using gene set enrichment analysis via Webgestalt [
]. Genes were ranked by ± fold change ∗ –log10(p value), and a false discovery rate (FDR) threshold of 0.05 was applied. Protein expression was analysed with a generalised linear mixed model (log inverse gaussian distribution) with the number of replicates, coverslips and cell density applied as covariates. Stimulation group, time, and coverslip location were treated as fixed effects. P values were considered significant <0.05, with Sidak corrections for post-hoc analysis and multiple comparisons. Astrocyte protein level changes were graphed with Prism (GraphPad). Correlation (two-tailed) was assessed using Spearman’s rho, with p value significance determined as < 0.05.
Results
rMS stimulation decreased expression of select genes
Among the 125 genes examined, a total of 21 genes were identified with significantly regulated expression levels compared to sham 5 h following a single stimulation session (Fig. 1, summarised in Table 1). All statistically significant genes above the fold change cut-off were downregulated compared to sham stimulation; there were no significantly upregulated transcripts following LI-rMS. All frequencies induced significant changes in gene expression, with the most consistently affected pathway across frequencies related to inflammation and injury; all stimulation protocols resulted in reduced expression of at least 3 inflammatory-related genes. There was considerable overlap of genes that were modulated by both 10Hz groups (i.e. 1m10Hz and 10m10Hz), but there were also genes that were modulated by only one of the 10Hz groups (Table 1 and Fig. 1). The two 10 min protocols (1Hz and 10Hz) shared one downregulated gene in common, the intermediate filament keratin (Krt19), and there were no genes in common that were significantly regulated following 600 pulses at either 1Hz or 10Hz. These findings indicate the effects of LI-rMS on astrocyte mRNA expression are different for 1Hz and 10Hz, and are influenced by the number of pulses delivered. The expression changes for all 125 genes are indicated in Supplementary Table 1. The analysis of enriched biological processes and molecular functions across the significantly regulated genes for each stimulation protocol is detailed in the supplementary material. The main protein activity and cellular compartmentalisation of the significantly regulated genes is indicated in Fig. 1.
Fig. 1Downregulation of gene expression in astrocytes 5 h following different LI-rMS protocols. A Heatmap of the 21 significantly downregulated genes compared to sham stimulation. Colour values represent log2 fold change, asterisk indicates p value < 0.05. B Venn diagram illustrating the frequencyspecific changes in gene expression, with each gene coloured to a major biofunction of its protein product in astrocytes. C Representation of the protein compartmentalisation of the significantly regulated genes following rMS. Product is colour coded by the frequency protocol(s) that influenced it, and protein shape denotes functional role. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Table 1Significantly regulated genes in astrocyte cultures 5 h following a single session of LI-rMS. Each row is the expression of a gene significantly regulated by at least one LI-rMS protocol. mRNA details and gene function are provided. The function denotes prominent functional classification of gene product in astrocytes: C cytoskeletal, I inflammation and injury, P plasticity, S signalling. The frequency of pulses and duration of stimulation are displayed, and the log2 fold change and associated p value in comparison to sham-stimulated coverslips are presented for each gene. Significant log2 fold change in noted in bold.
1m 10Hz
10m 10Hz
10m 1Hz
REFSEQ MRNA
Gene symbol and name
Function
FC
p
FC
p
FC
p
NM_007498
Atf3, Activating transcription factor 3
I
−0.17
0.91
−1.14
0.07
−1.12
0.02
NM_007529
Bcan, Brevican
I, P
0.57
0.63
−2.06
0.01
0.73
0.21
NM_019952
Clcf1, Cardiotrophin-like cytokine
I
0.60
0.26
−1.47
0.01
−0.10
0.91
factor 1
NM_009140
Cxcl2, C-X-C motif chemokine ligand 2
I
−2.21
0.01
0.78
0.57
−0.89
0.09
NM_010516
Ccn1, Cellular communication network factor 1
I, P
−2.47
0.04
−0.89
0.38
−0.45
0.68
NM_013642
Dusp1, Dual specificity phosphatase 1
I
−1.70
0.04
−0.42
0.27
−0.23
0.75
NM_010332
Ednra, Endothelin receptor type A
I, S
0.25
0.69
−1.12
0.27
−1.74
0.04
NM_007913
Egr1, Early growth response 1
I
−1.43
0.01
−0.87
0.11
−0.89
0.30
NM_010442
Hmox1, Heme oxygenase 1
I
−0.78
0.55
−1.22
0.04
−0.65
0.18
NM_010493
Icam1, Intercellular adhesion molecule 1
I
−0.47
0.40
−0.92
0.11
−1.43
0.01
NM_010517
Igfbp4, Insulin like growth factor-binding protein 4
S, C
−1.40
0.03
−0.90
0.12
−0.99
0.14
NM_021452
Kcnmb4, Potassium calcium-activated channel subfamily M regulatory beta subunit 4
To determine whether changes in gene expression were reflected in changes in protein levels in astrocytes, 4 of the 21 downregulated genes were selected for validation by immunohistochemistry on cultured astrocytes (Figs. 2 and 3). These 4 genes were chosen based on the relevance of their functions to known changes in astrocyte biology following LI-rMS in vitro [
]. The genes Stim1, Orai3, Kcnmb4, and Ncam1 had significantly downregulated gene transcripts in both 10Hz protocols but no significant regulation in the 1Hz group compared to sham. In the astrocytic store-operated calcium entry system relating to endoplasmic reticulum calcium stores, STIM1 protein functions as a detector of depleted calcium stores, and ORAI3 is a calcium re-entry channel [
]. Preliminary studies confirmed that these proteins were expressed in astrocyte cultures, as shown in Supplementary Figure 1 for expression of KCNMB4 and NCAM1 in a GFAP-positive astrocyte. In addition, the intensity levels of the transcription factor EGR1 and the cell adhesion protein ICAM1 were explored but not pursued further as they exhibited a weak signal to noise ratio in immunofluorescent staining.
Fig. 2Changes in protein intensity of KCNMB4 and NCAM1 at 5 h and 24 h after LI-rMS. A Representative images of KCNMB4 (green) and NCAM1 (red) co-labelled with Hoechst (blue) at 24 h after stimulation for each protocol. B,C Violin plots displaying the effect of stimulation on mean pixel intensity of KCNMB4 (B) and NCAM1 (C). Dotted lines indicate median and interquartile ranges and black bars between groups represent statistically significant differences between groups (p < 0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Fig. 3Changes in protein intensity of STIM1 and ORAI3 at 5 h and 24 h after LI-rMS. A Representative images of STIM1 (green) and ORAI3 (red) co-labelled with Hoechst (blue) at 24 h after stimulation for each protocol. B,C Violin plots displaying the effect of stimulation on mean pixel intensity of STIM1 (B) and ORAI3 (C).Dotted lines indicate median and interquartile ranges and black bars between groups represent statistically significant differences between groups (p < 0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
For KCNMB4 protein immunolabel intensity (Fig. 2A), statistical analysis indicated significant differences dependent on astrocyte density (F(1,268) = 44.65, p < 0.01), location on the coverslip (F(1,268) = 18.30, p < 0.01), stimulation group (F(3,268) = 6.04, p < 0.01), and time post-stimulation (F(1,268) = 6.12, p = 0.014), but interactions between frequency and other independent variables were not significant (p > 0.05). Post-hoc analysis revealed no significant difference between stimulated cells and sham at 5h post-stimulation, but a reduced intensity of 10m10 Hz at 24h post-stimulation compared to sham (−39.4%, p < 0.001, Fig. 2B).
For NCAM1 protein immunolabel intensity (Fig. 2A), there were significant differences dependent on astrocyte density (F(1,274) = 58.19, p < 0.01), stimulation group (F(3,260) = 12.30, p < 0.01), time post-stimulation (F(1,274) = 49.07, p < 0.01), and LI-rMS frequency∗time post-stimulation (F(3,274) = 7.05, p < 0.01), but not location on the coverslip (F(1, 274) = 0.53, p = 0.47). Specifically, mean NCAM1 immunolabel intensity was significantly lower following 10m1Hz compared to sham at 5h following stimulation (−42.1%, p < 0.001). By 24h, there were significantly reduced mean intensities for NCAM1 protein compared to sham for 1m10Hz (−38.3%, p < 0.001), 10m10Hz (−46.8%, p < 0.001), and 10m1Hz (−38.2%, p < 0.001, Fig. 2C).
STIM1 immunolabel intensity (Fig. 3A) showed significant differences dependent on astrocyte density (F(1,253) = 20.11, p < 0.01), location of cells on the coverslip (F(1,253) = 4.88, p = 0.028), stimulation group (F(3,240) = 38.56, p < 0.01), time post-stimulation (F(1,257) = 29.05, p < 0.01), and frequency∗time (F(3,256) = 21.11, p < 0.01). Within the 5h timepoint, both 10m10Hz and 10m1Hz protocols led to significantly increased protein levels compared to sham (10m10Hz + 29.8%, p < 0.001; 10m1Hz +102.7%, p < 0.001). However by 24h, immunolabel intensity was significantly reduced after 1m10Hz (−35.3%, p < 0.001) and 10m10Hz protocols (−45.5%, p < 0.001) but significantly increased in the 1Hz group (+36.3%, p < 0.001, Fig. 3B).
Similar effects were found for ORAI3 immunolabel intensity (Fig. 3A), where significant differences were dependent on location of cells on the coverslip (F(1,287) = 4.20, p = 0.041), stimulation group (F(3,231) = 30.25, p < 0.01), time post-stimulation (F(1,293) = 4.55, p = 0.034), and frequency∗time (F(3,290) = 7.27, p < 0.01), but not on astrocyte density (F(1,287) = 0.29, p = 0.86). Specifically, immunolabel intensity following 10m1Hz was significantly higher compared to sham at both time points (5h + 79.8%, 24h + 55.4%, p < 0.05 for both timepoints), and the two 10Hz protocols showed reduced intensity compared to sham at 24h (1m10Hz −41.6%, 1m10Hz −48.7%, p < 0.001 for both protocols, Fig. 3C).
Overall, the density of astrocytes and/or their location on the coverslip were significant factors in the statistical analysis for each of the four proteins. Correlation analysis was performed to examine the relationship between protein intensity and astrocyte density, and protein intensity and magnetic field strength at the imaging locations within the coverslip (Supplementary Figure 2). Protein labelling intensity was positively correlated with astrocyte density. Although the correlations did reach significance for magnetic field and protein labelling intensity, the weak correlation coefficient suggest that differences in magnetic field intensity at the imaging locations had minimal effect on protein levels (Supplementary Tables 2–4).
Discussion
Many clinical and experimental rTMS studies have found changes to neuroplasticity [
Repetitive magnetic stimulation induces functional and structural plasticity of excitatory postsynapses in mouse organotypic hippocampal slice cultures.
Transcranial pulsed magnetic field stimulation facilitates reorganization of abnormal neural circuits and corrects behavioral deficits without disrupting normal connectivity.
], suggesting classically non-excitable cells such as astrocytes may play a role in mediating LI-rTMS effects. This study shows that LI-rMS at 1 and 10 Hz results in significant reduction in the expression of a subset of genes in purified astrocyte cultures. The regulated genes are involved in various functional processes of astrocyte biology, primarily calcium signalling and inflammation. Importantly, the reduction in gene expression was associated with reduced expression of protein related to calcium signalling and cell adhesion. Taken together, these results suggest that LI-rMS has multiple and significant effects on astrocyte biology.
Calcium signalling a key target of rMS in astrocytes
Calcium signalling is emerging as a key mediator of rTMS effects [
] and our results confirm downstream changes in the expression of genes related to regulation of intracellular calcium levels in astrocytes. In particular, 1Hz increased expression of STIM1 and ORAI3, proteins that are involved in sensing and restoring astrocytic endoplasmic reticulum calcium stores following depletion [
]. The finding is consistent with our previous work showing that 1Hz LI-rMS increased astrocytic cytoplasmic calcium levels during stimulation in vitro [
]. This suggests that the changes to the expression of calcium-related genes and proteins may be a consequence of acutely increased calcium levels during and/or after stimulation. However, we cannot rule out that LI-rMS might have affected gene transcriptional or translational machinery, perhaps by upregulating immediate early genes such as the transcription factor c-Fos [
Interestingly, the changes in gene expression reported here do not suggest changes to the genes directly involved in calcium release; our previous work showed that the increase in calcium during rMS is from intracellular stores because the increase persisted in calcium free media, but not when calcium stores in the endoplasmic reticulum were depleted by thapsigargin [
]. However, in the present study, there was no significant transcriptional changes in Itpr2, the gene coding for IP3R2, suggesting that a single session of rMS may not cause long-term changes in calcium release properties, but rather impacts primarily on signalling downstream of calcium release.
The timing of rTMS-induced gene expression
In this study, changes to mRNA expression were examined at 5 h post LI-rMS, consistent with our previous study examining LI-rMS effects on mRNA expression of cultured neurons [
]. However, there was no overlap of significantly regulated genes (primarily survival and apoptotic-related functions in neurons). Other studies investigating the effect of HI-rTMS on gene expression in vivo have reported changes as acutely as 45 min post-stimulation for the immediate early gene c-fos in multiple brain regions [
], and changes to excitatory and inhibitory neurotransmitter transporter and endoplasmic reticulum stress-related genes at 1 h following stimulation, which, for some genes, persisted at 4, 12, and 24 h post stimulation [
Effects of repetitive transcranial magnetic stimulation on ER stress-related genes and glutamate, γ-aminobutyric acid and glycine transporter genes in mouse brain.
Biochemistry and Biophysics Reports.2019; 17: 10-16
]. Interestingly, Ikeda et al. found that the expression of certain genes varied above and below baseline in vivo and in PC12 cells in vitro during the 24 h following the end of HI-rTMS [
Effects of repetitive transcranial magnetic stimulation on ER stress-related genes and glutamate, γ-aminobutyric acid and glycine transporter genes in mouse brain.
Biochemistry and Biophysics Reports.2019; 17: 10-16
]. These fluctuations over time likely reflect a dynamic response to the immediate gene expression changes induced by stimulation, as has been shown in cultured inhibitory and excitatory neurons following KCl depolarisation [
]. Recently, it was reported that the plasticity-related genes Ntrk2 and Mapk9 were significantly upregulated at 24 but not 6 h following HI-rMS in vitro in human SH-SY5Y cells with a neural phenotype [
]. These studies highlight the complex temporal relationship of neuronal gene expression in response to rTMS/rMS, which the current experiment now extends to astrocyte gene expression and protein levels. For example, there were changes to Kcnmb4 gene expression for 1m10Hz, but no change to protein levels, and significantly different protein levels of STIM1 and ORAI3 despite no significant changes in mRNA levels. This inconsistency between gene expression and protein levels after rMS is interesting and may reflect differences in protein stability and turn-over [
] and/or indirect, downstream changes to protein synthesis. An extended time course of both mRNA and protein would be interesting to conduct to elucidate chronic effects of rMS on astrocyte gene expression and protein levels.
Inflammation and rMS
Amongst the four functional domains, injury and inflammatory genes were the most abundantly downregulated (10/21 genes). Changes to inflammatory pathways are likely to be associated with calcium signalling. Exposure to pro-inflammatory molecules decreases calcium signalling and the production of G protein-coupled receptors that regulate calcium signalling in astrocytes [
Dual regulation of calcium oscillation in astrocytes by growth factors and pro-inflammatory cytokines via the mitogen-activated protein kinase cascade.
]. Our finding of concomitantly altered inflammatory and calcium related genes following LI-rMS strengthens the relationship between these functions, but causality remains unclear. rTMS has previously been shown to reduce inflammation in stroke and injury animal models [
High-frequency repetitive transcranial magnetic stimulation improves functional recovery by inhibiting neurotoxic polarization of astrocytes in ischemic rats.
], however modulation of inflammation in intact or normal models has not been observed. Although we did not aim to induce inflammation, the preparation of astrocyte cultures is a significant insult to the cells and is reported to stimulate inflammatory pathways [
]. Consistent with this, growing astrocytes in media-containing serum induces a gene expression profile that overlaps with profiles described for models of astrocyte reactivity [
]. Our results show LI-rMS downregulates inflammatory-related genes, but given the artificial context of these cells in vitro, it is difficult to draw conclusions about possible anti-inflammatory effects of LI-rTMS in vivo. Nonetheless, this study supports a growing body of evidence that rTMS may reduce inflammation [
High-frequency repetitive transcranial magnetic stimulation improves functional recovery by inhibiting neurotoxic polarization of astrocytes in ischemic rats.
In the rTMS literature, protocols are generally designated a binary category such as “facilitatory” and “inhibitory”, which refer to their expected effects on neuronal excitability. However, astrocytes do not communicate through action potential propagation, but instead use calcium signalling to transmit and process information that modulates neuronal activity and/or synaptic plasticity [
]. Furthermore, astrocytes are involved in other modulatory activities that do not directly alter neuron excitability but rather influence their inflammatory and injury response. Accordingly, even though pathway analysis did not yield any significant results, most likely due to the very wide range of disparate targets included in the array, the downregulated genes suggest reductions in pro-inflammatory activity, calcium signalling, cytoskeletal structure, plasticity, and cell adhesion properties. While there is clear delineation between the regulation of genes by each protocol, the main biological functions of the regulated genes are similarly spread across the protocols, complicating the distinction of potentially altered functional processes. We have previously demonstrated that rMS altered the expression of genes relating to apoptosis and cell survival in neuronal cultures [
]. Therefore, the frequency-specific effects of rMS on neurons and astrocytes are likely to be broader than just facilitatory or inhibitory to brain activity and may include anti-inflammatory and pro-survival effects. Bidirectional neuron-glia interactions are firmly established [
], and future in vitro studies using mixed cultures, potentially including a significant proportion of other cell types such as endothelial cells and immune cells, will contribute to understanding and optimising therapeutic effects of rTMS in conditions that involve inflammation such as stroke, depression and neurodegenerative conditions.
The influence of astrocyte density and magnetic field on protein changes
There was a non-uniform spatial distribution of astrocytes on coverslips, and immunohistochemical analysis revealed that astrocytes in regions of greater density were more intensely stained for KCNMB4, NCAM1, and STIM1 proteins. While such altered expression may reflect increased adhesion, gap junction coupling and intercellular communication via ion flow [
], there was no evidence that astrocytic density significantly influenced the observed LI-rMS effects on protein levels for 1Hz and 10Hz frequencies. An additional variable is the distribution of induced magnetic and electric fields is not uniform across the coverslip. The magnetic field is strongest under the centre of the coil, but the electric field is greatest underneath the coil windings [
]. In this study, the greatest magnetic field was near the centre of coverslip, whereas the greatest induced electric field was approximately halfway between the centre and edges of the coverslip (Supplementary Figure 2). There was a weak correlation between the location of astrocytes on the coverslip and the intensity of KCNMB4 and ORAI3. However, LI-rMS did not affect this relationship. This is unexpected, because of the differential intensity of the induced electric field across the coverslip. The reductions in protein levels at the centre of the coverslip might be a consequence of intercellular signalling propagating through the astrocyte network [
] from regions that were directly affected by the stronger electric field. In future studies, shielding part of the coverslip with mu-metal or applying gap junction blockers could address the role of astrocyte communication in propagating LI-rMS effects [
The effects of different repetitive transcranial magnetic stimulation (rTMS) protocols on cortical gene expression in a rat model of cerebral ischemic-reperfusion injury.
]. In this study, there was an overlap of 6 genes that were regulated by stimulation delivered at 10Hz for 600 and 6000 pulses. However, there was no overlap in gene expression between pulse matched stimulation at 1Hz and 10Hz, and the majority (4/6) of significantly regulated calcium-related genes were following 10Hz LI-rMS regardless of duration. Moreover, differences in gene expression between 1m10Hz and 10m10Hz suggests an influence of duration (dose), which mainly affected inflammatory-related genes, with less pulses counterintuitively resulting in more genes being downregulated. These results suggest that frequency is a significant determinant of LI-rMS effects on astrocytes. However, further characterisation of common frequencies with duration and pulse numbers controls is required to determine the frequency- and duration-specific (dose) effects of stimulation on astrocytes. It is important to note that dose-response in brain stimulation and pharmacological contexts differ, with the latter considering concepts such as threshold, slope, and maximal asymptote of the treatment [
LI-rTMS has the ability to induce neuroplasticity, with a wide range of effects on neuronal plasticity. However, the effect of rTMS on astrocytes is not well understood. Our data shows that a single stimulation session of 1Hz or 10Hz can reduce the expression of genes and proteins in purified astrocyte cell cultures. The functions of the affected genes and proteins primarily relate to calcium signalling and inflammation, and suggest that LI-rTMS may have therapeutic effects that contribute to LI-rTMS induced neuroplasticity.
Author declaration form
We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.
We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.
We confirm that we have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions concerning intellectual property.
We further confirm that any aspect of the work covered in this manuscript that involved experimental animals has been conducted with the ethical approval of all relevant bodies and that such approvals are acknowledged within the manuscript.
CRediT authorship contribution statement
Darren Clarke: Conceptualization, Methodology, Formal analysis, Investigation, Writing - original draft, Writing - review & editing, Visualization. Jamie Beros: Investigation. Kristyn A. Bates: Conceptualization, Methodology, Resources, Writing - review & editing, Supervision, Funding acquisition. Alan R. Harvey: Conceptualization, Methodology, Resources, Writing - review & editing, Supervision, Funding acquisition. Alexander D. Tang: Methodology, Software, Formal analysis, Writing - review & editing, Supervision. Jennifer Rodger: Conceptualization, Methodology, Resources, Writing - review & editing, Supervision, Project administration, Funding acquisition.
Declaration of competing interest
The authors declare no competing interest.
Acknowledgements
This work was supported by the Neurotrauma Research Program.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Transcranial pulsed magnetic field stimulation facilitates reorganization of abnormal neural circuits and corrects behavioral deficits without disrupting normal connectivity.
Reversed effects of intermittent theta burst stimulation following motor training that vary as a function of training-induced changes in corticospinal excitability.
Intermittent theta-burst stimulation rescues dopamine-dependent corticostriatal synaptic plasticity and motor behavior in experimental parkinsonism: possible role of glial activity.
High-frequency repetitive transcranial magnetic stimulation improves functional recovery by inhibiting neurotoxic polarization of astrocytes in ischemic rats.
Repetitive transcranial magnetic stimulation inhibits Sirt1/MAO-A signaling in the prefrontal cortex in a rat model of depression and cortex-derived astrocytes.
Repetitive magnetic stimulation induces functional and structural plasticity of excitatory postsynapses in mouse organotypic hippocampal slice cultures.
Effects of repetitive transcranial magnetic stimulation on ER stress-related genes and glutamate, γ-aminobutyric acid and glycine transporter genes in mouse brain.
Biochemistry and Biophysics Reports.2019; 17: 10-16
Dual regulation of calcium oscillation in astrocytes by growth factors and pro-inflammatory cytokines via the mitogen-activated protein kinase cascade.
The effects of different repetitive transcranial magnetic stimulation (rTMS) protocols on cortical gene expression in a rat model of cerebral ischemic-reperfusion injury.
30310-7&id=gr1.jpg){kind=link}
30310-7&id=gr2.jpg){kind=link}
30310-7&id=gr3.jpg){kind=link}
30310-7&id=figs1.jpg){kind=link}
30310-7&id=figs2.jpg){kind=link}