Restoration of DNA integrity and the cell cycle by electric stimulation in planarian tissues damaged by ionizing radiation

ABSTRACT

Exposure to high levels of ionizing γ radiation leads to irreversible DNA damage and cell death. Here, we establish that exogenous application of electric stimulation enables cellular plasticity and the re-establishment of stem cell activity in tissues damaged by ionizing radiation. We show that subthreshold direct current stimulation (DCS) rapidly restores pluripotent stem cell populations previously eliminated by lethally γ-irradiated tissues of the planarian flatworm Schmidtea mediterranea. Our findings reveal that DCS enhances DNA repair, transcriptional activity, and cell cycle entry in post-mitotic cells. These responses involve rapid increases in cytosolic Ca²⁺ concentration through the activation of L-type Ca_v channels and intracellular Ca²⁺ stores, leading to the activation of immediate early genes and ectopic expression of stem cell markers in post-mitotic cells. Overall, we show the potential of electric current stimulation to reverse the damaging effects of high-dose γ radiation in adult tissues. Furthermore, our results provide mechanistic insights describing how electric stimulation effectively translates into molecular responses capable of regulating fundamental cellular functions without the need for genetic or pharmacological intervention.

Summary: Exogenous stimulation with electric currents enables DNA damage repair and stem cell reconstitution in planarian tissues exposed to a lethal dose of ionizing radiation.

INTRODUCTION

Since the initial observations made by Luigi Galvani in the late 1700s, scientists have been fascinated by the effects of the exogenous application of electric currents to animal tissues (Adee, 2018; Bresadola, 1998). This long-lasting interest has revealed near universal roles for electricity during embryonic development, tissue regeneration and disease (Levin, 2007, 2014; McCaig et al., 2005). The effects of direct current stimulation (DCS) vary depending on the location in the body, the tissue type that is targeted, the intensity of the electric current and the duration of treatment. Furthermore, the outcome of DCS may differ depending on the context in which it is assayed (e.g. embryonic development, adult stages, disease or homeostasis). It is important to note that currents used in DCS are subthreshold, implying that they do not induce action potentials in excitable cells or tissues (Bikson et al., 2004). All cells, including stem cells, progenitors and differentiated cells, have built-in mechanisms designed to sense electric changes, which consequently influence their innate cellular transmembrane potential (V_mem). Changes in V_mem are caused by changes in ion fluxes at the cell plasma membrane (Jaffe, 1981a,b; Levin, 2007; McLaughlin and Levin, 2018; Nuccitelli and Jaffe, 1974). V_mem acts as a potent regulator of cellular migration, proliferation, differentiation, cell cycle and cell death, and is highly sensitive to manipulation by DCS (Levin, 2014; McCaig et al., 2005).

Studies performed in vitro and in vivo have consistently found that electric stimulation has strong effects on stem cell behavior. Specifically, electric stimulation alters patterns of transcription, migration, proliferation and fate decisions through diverse cellular mechanisms. For example, electric stimulation can be used to effectively guide the differentiation of neural precursor cells (Zhao et al., 2015), mesenchymal stem cells (Chang et al., 2021; Mobini et al., 2016) and other progenitors (Hernández et al., 2016; Oliveira et al., 2020; Tang et al., 2018). These electric-guided stem cell responses are relevant to bioengineering and medical applications where biophysical properties can be exploited individually or combined with biochemical treatments to enhance cellular engraftment potential and tissue rehabilitation (Chen et al., 2019; Rando and Ambrosio, 2018). Furthermore, DCS, at the organismal level, is capable of altering axial polarity, tissue repair and organ specification (Borgens et al., 1987; Jaffe, 1981a; Marsh and Beams, 1952; McCaig et al., 2005; McLaughlin and Levin, 2018; Nuccitelli, 2003; Zhao et al., 2006). For example, galvanotactic responses have been observed in migratory cells of humans (Guo et al., 2010), fish (Graham et al., 2013), frogs (Stump and Robinson, 1983), Caenorhabditis elegans (Chrisman et al., 2016) and even Paramecium (Ogawa et al., 2006). Similarities in DCS-mediated cellular responses across vertebrate and invertebrate organisms argue for evolutionarily conserved mechanisms. However, the molecular bases of DCS in various cell types remain largely unknown.

To further investigate the effects of DCS, we developed an experimental strategy using the planarian flatworm Schmidtea mediterranea, which is known for a high rate of cellular turnover and an extraordinary capacity to regenerate tissues that relies upon adult stem cells called neoblasts (Raz et al., 2021; Reddien, 2018; Rink, 2018; Zeng et al., 2018; Zhu and Pearson, 2016). Neoblasts are the only cell type with the capacity to divide in asexually reproducing planarians. Thus, neoblast division alone provides the cellular progeny required to renew and repair all planarian tissues (Raz et al., 2021; van Wolfswinkel et al., 2014; Wagner et al., 2011; Zeng et al., 2018). This constant cellular crosstalk allows planarians to maintain a diverse cellular population through neoblast division, making them powerful models to analyze the effects of DCS on adult stem cells and differentiated cells at the cellular, subcellular and organismal levels. Here, we introduced a simplified platform to apply exogenous DCS to the whole body of planarians and analyze the resulting cellular and subcellular responses in real time. We found that brief exposure to DCS overrides cellular decisions in tissues exposed to lethal doses of ionizing radiation. Furthermore, our results reveal that DCS is a rapid and robust method with the potential to enhance DNA damage repair and activate the transcription of stem cell markers in tissues damaged by ionizing radiation. Moreover, we identified that these DCS-mediated responses are tightly regulated by the transcription of immediate early genes and rapid intracellular Ca²⁺ flux. These findings provide insights into the effects of DCS in the adult body, demonstrating the ability to influence fundamental cellular processes such as transcription, DNA repair, the cell cycle and cellular plasticity without the need for genetic or pharmacological treatments.

RESULTS

Bodywide application of steady-state direct current

We immobilized planarians by applying our recently developed method (ThermoPress immobilization), which combines the anesthetic chloretone (0.2%) with a 1% agar encasing chamber (Davidian et al., 2021). This method of agar immobilization keeps planarians alive while preserving tissue integrity and restricting body movement for hours to days with complete and rapid recovery of sensory functions and locomotion (Fig. 1A). ThermoPress immobilization was used to administer steady-state direct current stimulation to the whole planarian body – a process that we refer to as ‘pDCS’ (Davidian et al., 2021).

Fig. 1.

Schematic summary of planarian immobilization and pDCS setup. (A) Schematic representation of planarian immobilization (LMP, low melting point). (B) Illustration of current-clamp circuit created by pDCS electrodes and planarian (VDC, direct current voltage). (C) Amplified schematic representation of planarian tissues showing electrode placement and operational amplifiers used to quantify current passing through the 100 MΩ current clamp. (D) Picture of the general pDCS setup including the general arrangement of the microscope, power supply and various components as shown in B. (E,F) Images showing (E) outside and (F) inside of a customized power supply containing 9 V batteries in series used to deliver electric currents in pDCS.

Electric current was delivered through pulled borosilicate sharp microelectrodes that were filled with 3 M KCl and placed at the anterior and posterior ventral side of each planarian (i.e. the pre-pharyngeal and tail regions, respectively). The microelectrode resistance was consistently measured at 1–2 MΩ. The control group (sham-treated) consisted of animals exposed to the same procedure, including electrode penetration, with the absence of electric current. To facilitate circuit conduction and minimize electrode byproducts in planarian tissues, the microelectrodes were coupled to an electrode bath containing 3 M KCl solution using 1.0% agar bridges (Fig. 1B). The amount of current delivered was limited through a 100 MΩ resistor, and the most consistent results were obtained when the current traversing the animal was 7 µA (Fig. 1C). DCS exceeding 7 µA resulted in noticeable tissue damage and eventual animal lysis. The electric field generated with this setup has previously been calculated as ∼71.81 mV/mm (Davidian et al., 2021). All experiments were performed with an electric polarity of a positive pole in the anterior and a negative pole in the posterior, unless otherwise stated. Current delivered to the animal was differentially measured and acquired with an analog-to-digital converting board controlled using custom-made LabView-based software (Elliott et al., 2007). Animals were under constant surveillance to ensure that the electrode placement and tissue integrity were maintained. Overall, this setup was effective in keeping electrodes correctly positioned and allowed analysis of DCS for up to 6 h (Fig. 1D–F).

pDCS activates transcription of stem cell markers in tissues exposed to a lethal dose of ionizing radiation

DCS techniques are known to affect both stem cells and differentiated cells in many model organisms (Feng et al., 2017; Huang et al., 2015; McCaig et al., 2005; Zhao et al., 2006). Planarian neoblasts are generally scattered along the antero–posterior axis, except for regions in front of the eyes and pharynx, and uniquely express the gene smedwi-1 (referred to as piwi-1 henceforth; Fig. 2A) (Reddien et al., 2005; van Wolfswinkel et al., 2014; Wagner et al., 2011; Zeng et al., 2018). Expression of piwi-1 is currently used as the standard marker to recognize the presence and distribution of neoblasts (Reddien et al., 2005; Wagner et al., 2011; Zeng et al., 2018). Exposing planarians to lethal doses of ionizing radiation (60 Gy) irreversibly eliminates neoblasts and the corresponding piwi-1 expression (as shown in Fig. 2A), consequently abolishing their regenerative capabilities. As a result, planarians ultimately perish within 3–4 weeks following lethal ionizing radiation (Bardeen and Baetjer, 1904; Reddien et al., 2005). However, lethally irradiated planarians can be rescued by transplanting tissue containing neoblasts (Guedelhoefer and Sanchez Alvarado, 2012a,b). This rescue occurs as neoblasts gradually migrate from transplanted tissues to repopulate the entire irradiated host in ∼1 month (Guedelhoefer and Sanchez Alvarado, 2012a). Engrafted tissue becomes more structurally stable after 4 days post transplantation (dpt); a stage at which the majority of neoblasts are still within the transplanted tissue. Over the first 5 dpt, neoblast repopulation is uniform with no bias towards anterior, posterior or medial–lateral tissues (Guedelhoefer and Sanchez Alvarado, 2012a). Thus, this neoblast repopulation paradigm was used as a model to study the effects of pDCS on both adult stem cells and differentiated cells in their natural environment.

Fig. 2.

pDCS triggers a transcriptional response and cell cycle in γ-irradiated host tissues. (A) WISH showing piwi-1⁺ signal in both WT and γ-irradiated (60 Gy) planarians (image is representative of n=10/10 animals). (B) Schematic depicting transplantation procedure using WT donor and irradiated host planarian with subsequent exposure to DCS. (C) WISH of piwi-1 gene expression at 4 days post transplantation in both sham-treated (control, n=10/10) and animals subjected to 60 min pDCS (n=12/15). The insets in the lower portion of the image represent magnifications of the distal part of the animal tail indicated by the boxes. The piwi-1 signal is indicated with arrows in animals subjected to pDCS. (D) Gene expression levels of piwi-1 in the tails of sham-treated and 60 min pDCS-treated animals, as determined by qPCR (pool of six animals/replicate and three biological replicates). (E–G) RNA-seq data collected from the host tail tissue of sham-treated animals and animals subjected to 60 min pDCS (data was collected by pooling tails from four independently treated pDCS or sham planarians across three independent experimental trials). Gene expression heatmaps display differentially expressed transcripts (FDR<0.05) as averaged log₂ CPM Z-scores. Heatmap representations of RNA-seq data display differentially expressed genes characteristic of the indicated neoblast subclass populations defined in the following recent publications: (E) van Wolfswinkel et al. (2014) and Wagner et al. (2012); (F) Raz et al., 2021; and (G) Zeng et al., 2018 (Late prog, late progeny; Pan-neob, pan-neoblasts). (H) Gene expression levels of cyclin-B obtained by qPCR using tail fragments from sham-treated and 60 min pDCS animals (pool of six animals/replicate and three biological replicates). (I) RNA-seq heatmap displaying differentially expressed genes commonly associated with cell cycle regulation. Differential expression and Z-scores were defined as in E–G. (J) Whole-mount immunostaining with anti-phosphorylated histone H3 (H3P) antibody (green) showing H3P⁺ cells inside and outside of the transplant (indicated by the dashed yellow line). Note the H3P⁺ cells in the pDCS experimental group far away from the transplanted tissue (white arrows) following 6 h pDCS, compared to sham-treated control (n=10/15). (K) Magnified images around the transplanted tissue (outlined by the dashed yellow line) for both sham-treated and pDCS-treated animals after 6 h of treatment. Red and yellow arrowheads indicate mitotic cells in the regions anterior and posterior of the transplant, respectively. The boundary between the regions is marked by the pink dotted line. (L) Average number of H3P⁺ cells in sham-treated and experimental groups after 6 h of pDCS. pDCS experiments were executed with positive pole to the anterior and negative pole to the posterior for 60 min, except for J–L, which were treated for 6 h. Data in D,H,L are presented as mean±s.e.m. *P<0.05; ****P≤0.0001 (two-tailed unpaired Student's t-test). Scale bars: 500 µm.

Tissue containing neoblasts was transplanted into lethally irradiated hosts, followed by pDCS treatment at 4 dpt (Fig. 2B). This experimental setup led to essential differences in piwi-1 gene expression in sham- versus pDCS-treated animals. Specifically, 60 min after pDCS, piwi-1 expression was widely detected outside the transplanted tissue towards the posterior region of the animal; whereas, in the sham-treated group, piwi-1 expression was restricted to tissues adjacent to the transplant, as expected (Fig. 2C). Levels of piwi-1 expression in the tails of pDCS-treated planarians increased by more than twofold as determined using qPCR (Fig. 2D). Transcriptomic analysis using RNA sequencing of the tail fragment confirmed that pDCS treatment for 60 min was accompanied by an increase in the expression of piwi-1 [log₂ fold change (FC)=1.97, Benjamini–Hochberg false discovery rate (BH FDR)=0.0006, moderated t=9.89]. Furthermore, we also detected an increase in the expression of the transcription factor soxP-2 (Fig. 2E; log₂ FC=0.61, BH FDR=0.005, moderated t=4.82), which is a marker of the classic sigma neoblast subpopulation required for stem cell function and planarian regeneration (van Wolfswinkel et al., 2014; Wagner et al., 2012). Likewise, we observed that pDCS slightly reduced the expression of the piwi family member Smed-piwi-3 after 60 min treatment (log₂ FC=0.21, BH FDR=0.0068, moderated t=4.24) (Kim et al., 2019; Palakodeti et al., 2008), whereas there was moderate upregulation of the late progeny marker agat-1 after 60 min pDCS (Fig. 2E; log₂ FC=0.18, BH FDR=0.016, moderated t=3.74).

We also observed differential expression of other markers (Fig. 2F,G) associated with the recently expanded neoblast subclasses (Raz et al., 2021; Zeng et al., 2018). We identified a select group of neoblast and muscle markers (including soxP-2, MCM-4, piwi-1, bruli and soxP-5) that were significantly upregulated (Fig. 2F; log₂ FC>0.5, BH FDR<0.01; Raz et al., 2021). Furthermore, from the list of 189 possible neoblast markers recently identified by Zeng et al. (2018), we found that 25 transcripts were differentially expressed with a BH FDR of less than 0.05. The transcriptomic analysis tested the contrasts between the 60 min pDCS-treated animals and sham-treated controls, evidencing 1778 genes differentially expressed and with all 25 of the neoblast markers expressed below the 1% significance level, indicating a less than 1% probability that a differentially expressed gene in this list is a false positive. The transcriptomic analysis of tissues treated with 60 min pDCS showed that key markers for the neoblast subclasses Nb5, Nb8 and Nb12 were strongly downregulated compared to the control at this timepoint (Fig. 2G). The Nb5 and Nb12 subclasses contain populations of cells expressing piwi-1 at high levels (piwi-1^high) and are hypothesized to include early progenitor cells for intestinal tissue (BH FDR<0.05). Conversely, markers for the muscle cell progenitor subclasses Nb4 and Nb6 (including TDP2, TPI1 and ACTB) were significantly upregulated (Fig. 2F,G; BH FDR<0.01). Muscle progenitors are critical for providing positional information and are key drivers of tissue patterning during regeneration (Cote et al., 2019; Scimone et al., 2020, 2017). Additionally, putative markers for the pharyngeal neoblast progenitors and differentiated populations (Nb7, Nb8 and Nb10) were overall differentially expressed. However, markers of the piwi-1^high pharyngeal progenitor populations Nb7 and Nb8 (PTK7 and PPIC, respectively) had lower expression compared to markers of the piwi-1^low differentiated pharyngeal population (Nb10; BH FDR<0.05). Key markers for intestinal and pharynx progenitors (neoblast subclasses Nb5, Nb8, Nb12 and Nb9) were downregulated (Fig. 2G). The Nb10 neoblast markers PDIA6, CALU and P4HB were significantly upregulated after the 60 min application of pDCS (BH FDR<0.05). Only one marker (dd_Smed_v6_1399_0_1) for the putative neural progenitor (Nb11) was significantly upregulated after 60 min of pDCS (Fig. 2G; log₂ FC=0.34, BH FDR=40.043, moderated t=3.71). Importantly, key markers for piwi-1^high clonogenic neoblasts (cNeoblasts, Nb2) were significantly upregulated (Aats-asp and soxP-2; Fig. 2E,G; log₂ FC>0.5, BH FDR<0.01), indicating an increase in progenitors of muscle, neural and gut tissues. Our findings show that the 60 min application of pDCS triggers the expression of the pan-neoblast marker piwi-1 along with markers of neoblast subpopulations with high piwi-1 expression in host tissues where neoblasts had been permanently eliminated by exposure to a lethal dose of ionizing radiation.

Prolonged pDCS leads to a directional cell cycle entry in tissues exposed to lethal ionizing radiation

We observed that pDCS-induced transcription of piwi-1 in irradiated tissue was also accompanied by an increase in the expression of the neoblast marker Smed-cyclin B and other components associated with the regulation of cell division (e.g. cyclin-dependent kinases, MCM proteins, checkpoint kinase, polo-like kinase, and Rb-binding protein) (Fig. 2H,I). However, the presence of mitotic cells far away from the transplanted tissue was evident after extending the pDCS treatment to 6 h (Fig. 2J,K). Strikingly, the number of mitotic cells outside of the transplant was significantly increased upon pDCS (Fig. 2L). In these animals, mitotic cells within irradiated tissues were primarily distributed towards the posterior region of the host, being observed as far as the tip of the tail (Fig. 2L). Proportionately, a smaller number of mitotic events occurred in the anterior of pDCS-treated transplanted planarians, implying asymmetrical effects likely due to pDCS polarity (Fig. 2L). In the sham-treated control group, we did not observe asymmetries in dividing cells within the scarce population of mitotic cells outside the transplanted tissues (Fig. 2J–L). We also observed variability in the effects of pDCS on the number of mitotic cells. Some animals showed low or no response (26%), whereas the large majority displayed moderate effects (67%) with noticeable mitotic signal outside of the transplant (Fig. S1A).

To address the possibility of pDCS-induced mitotic asymmetry within irradiated tissues, a similar magnitude of DCS was applied to 4 dpt planarians with the opposite polarity (i.e. reversed polarity, with the positive pole located in the tail and the negative pole in pre-pharyngeal tissue in front of the transplant). These experiments led to inconsistent results, thus we decided to continue with the characterization of pDCS based on polarity with the positive pole in the anterior region and the negative pole implanted in the posterior region of the animal. We also found that pDCS had a similar effect on gene expression when the tissue graft was placed in the posterior region (Fig. S1B,C). However, tissue transplants in the anterior region were more reliable and convenient for characterization of the effects of pDCS.

pDCS-induced transcription of stem cell markers originates in lethally irradiated tissues

Exposure to a lethal dose of ionizing radiation permanently eliminates neoblasts in less than 24 h (Peiris et al., 2016a). Distinctively, pDCS led to the transcription of neoblast markers and the presence of mitotic cells in the host tissue several days after exposure to lethal ionizing radiation. This finding prompted us to investigate the potential source of the neoblast-related cells.

First, since the exogenous application of electric fields is widely known to guide the movement of cells through electrotaxis (McCaig et al., 2005), we addressed the dynamics of cell migration from the transplanted tissue as the potential mediator of pDCS effects. Previous work has determined that neoblasts migrate from transplanted tissue to gradually repopulate the lethally irradiated host at a rate of ∼3–5 µm/h, which is about 72–120 µm/day (Abnave et al., 2017; Eisenhoffer et al., 2008; Guedelhoefer and Sanchez Alvarado, 2012a,b; Newmark and Sánchez Alvarado, 2000; Reddien et al., 2005; Saló and Baguñà, 1985). Because neoblasts are the only dividing cells in planarians, the spatiotemporal path of neoblast-related gene expression and mitotic cells (i.e. mitotic wave) are commonly used to infer migration rates between two points (Abnave et al., 2017; Guedelhoefer and Sanchez Alvarado, 2012a,b; Newmark and Sánchez Alvarado, 2000; Saló and Baguñà, 1985). Our results show that following 6 h of pDCS, mitotic cells could be found in the tail of the irradiated host, which is ∼5 mm away from the transplanted tissue (Fig. 2J,K). If the transplanted tissue was the source of these dividing cells, the neoblasts must have been displaced at ∼833 µm/h (i.e. ∼200 times faster than previously reported) to arrive at the tip of the tail more than 700 h earlier than would be expected based on previous reports (Guedelhoefer and Sanchez Alvarado, 2012a; Saló and Baguñà, 1985). Furthermore, we found that shorter pDCS treatment led to robust piwi-1 expression throughout the animal (see Fig. 5A, 15 min pDCS). Were this to be the result of cellular migration from the engrafted tissue, the neoblasts must have migrated at rates exceeding 20,000 µm/h (2 cm/h), which is not only four orders of magnitude faster than previously established but also unlikely due to physical tissue-derived obstacles in their path.

Fig. 5.

pDCS induces a rapid transcriptional response. (A,B) Gene expression of piwi-1 assessed using (A) WISH and (B) qPCR at 15, 30 and 45 min of pDCS compared to the sham-treated group. (A) Lower-left corner shows the whole-body image of lethally irradiated (60 Gy) animals 4 days after transplantation of tissue from a WT animal. Boxes indicate regions shown in magnified images of the tail section (sham, n=10/10; 15 min pDCS, n=8/10; 30 min pDCS, n=9/13; 45 min pDCS, n=10/14). Asterisks signify where electrodes were placed in planaria. (B) qPCR data, represented as mean±s.e.m., obtained from three technical replicates consisting of five pooled tail samples per experiment, with PCR repeated twice. Data are normalized to the expression level in the sham-treated group. (C) Gene expression heatmap of neoblast markers from tail tissue explants, measured by qPCR. Data shown are log₂ FC with column-scaled Z-score (data is from six pooled tail fragments/replicates and three biological replicates). Nb, neoblast; cNb, clonogenic neoblast. (D–G) RNA-seq data collected from tail tissue explants of irradiated host animals that were either sham-treated or pDCS-treated for 15 min following transplantation of WT tissue (data was collected by pooling tails from four independently treated pDCS- or sham-treated planarians across three independent experimental trials). Gene expression heatmaps display differentially expressed transcripts (FDR<0.05) as averaged log₂ CPM Z-scores (key shown in G). (D–F) Differentially expressed genes that are markers for the indicated neoblast subclass populations based on the following previously published reports: (D) van Wolfswinkel et al. (2014) and Wagner et al. (2012), (E) Raz et al. (2021) and (F) Zeng et al. (2018). (G) Differentially expressed immediate-early gene putative homologs. DEG, delayed early genes; IEG, immediate early genes; ILG, immediate late genes. The polarity of the electric field for pDCS was oriented with the positive pole to the anterior and the negative pole to the posterior. ****P≤0.0001 (one-way ANOVA with Dunnett's multiple comparison test). Scale bars: 500 µm.

Second, recent findings have demonstrated that cellular migration in planarians depends on the expression of genes encoding epithelial–mesenchymal transcription factors (Snail-1, Snail-2 and Zeb-1) and β1-integrin, along with components of matrix metalloproteinase (Abnave et al., 2017; Bonar and Petersen, 2017; Isolani et al., 2013; Seebeck et al., 2017). We compared the expression of markers for cellular migration in two segments of sham- and pDCS-treated animals – the trunk fragment, which included transplanted tissue, and the tail region – at 1 h of treatment (Fig. 3A). The results show that in the trunk the expression increased for both Snail-1 and Snail-2, whereas there was no change in expression of Zeb-1 and β1-integrin. In the tail region, we found no changes in expression, except for a slight increase in expression of the β1-integrin gene (Fig. 3B,C). Next, we used bromodeoxyuridine (BrdU) to trace migratory cells, but our attempts were unsuccessful due to inconsistent tissue engraftment likely associated with the friability of tissue fragments obtained from donors treated with BrdU.

Fig. 3.

Lethally irradiated host tissue is the main source of pDCS-induced neoblast transcription. (A) Schematic representation depicting different regions of planarians subjected to tissue transplants from WT animals. (B,C) Gene expression levels of genes associated with cellular migration in planarians (Snail-1, Snail-2, Zeb-1, β1-integrin). The tissues used for each experiment were (B) trunk and (C) tail from sham-treated and 60 min pDCS-treated animals. Gene expression was measured using qPCR. IRR, irradiated; TP, transplant. (D,F,H) Schematics showing experimental design for tissue transplantations between different donors and hosts, following which tail fragments were processed to measure gene expression at 4 dpt. (D) Transplant of piwi-1(RNAi) tissue into a lethally irradiated (X-ray) host. (F) Transplant of WT tissue into a piwi-1(RNAi) host. (H) Transplant of γ-irr (X-ray) tissue into a γ-irr host. (E,G,I) Gene expression in tail region tissues obtained from sham-treated animals and planarians subjected to 60 min pDCS. Gene expression levels, normalized to levels in sham-treated animals (dashed line), were quantified for markers of diverse neoblast populations (piwi-1, soxP-2, fgfr-1, egr-1, hnf-4, nkx2.2 and inx-13). pNb, pan neoblast; cNb, clonogenic neoblast; Nb, neoblast. Data are presented as mean±s.e.m. All gene expression experiments were obtained from three biological replicates consisting of four pooled samples per replicate. The polarity of the electric field was positive pole to the anterior and negative pole to the posterior for 60 min of pDCS. *P≤0.05; **P≤0.01; ***P≤0.001; ****P≤0.0001; ns, not significant (one-way ANOVA with Dunnett’s multiple comparisons test).

Third, we did not observe the anterior-to-posterior progressive pattern of cells expressing neoblast markers or of dividing cells that is characteristic during migration-mediated neoblast repopulation of the irradiated host. For instance, stimulation with pDCS for a shorter time (i.e. 15 min) resulted in strong expression of piwi-1 at distant places from the transplanted tissue (see Fig. 5A, 15 min). These findings are in stark contrast to the previously described gradual progression of piwi-1 expression over the antero–posterior axis that takes about 40 days to reach the tip of the tail in the absence of electrical stimuli (Abnave et al., 2017; Guedelhoefer and Sanchez Alvarado, 2012b).

Fourth, we designed a series of experiments involving tissue transplantations between wild-type (WT), piwi-1(RNAi) and lethally irradiated animals to measure the expression of neoblast markers in the tail region of the host (Fig. 3D,F,H). We performed selective elimination of piwi-1 expression in either the host or donor tissue to verify the source of piwi-1-positive cells and other progenitor subtypes. It is important to note that piwi-1(RNAi) specifically silences piwi-1 expression without affecting neoblast number or function (Reddien et al., 2005). Transplanting tissue from a piwi-1(RNAi) animal into a lethally γ-irradiated host resulted in a sixfold increase in piwi-1 expression in the tail of the host subjected to pDCS compared to the piwi-1 expression in sham-treated animals (Fig. 3D,E). There was also a prominent increase in gene expression for other neoblast markers, suggesting a generalized neoblast response (Fig. 3E). Since piwi-1 expression was originally silenced in the transplanted tissue, the increased expression of piwi-1 away from the transplant, specifically in the tail, suggests that the piwi-1 expression originates in host tissues. To confirm this, tissue containing neoblasts from WT animals was transplanted into a piwi-1(RNAi) host and was subjected to identical treatment (Fig. 3F). The results show that piwi-1 expression in pDCS-treated animals was equivalent to that in sham-treated animals, as expected, but there was an important increase in the expression of other neoblast markers in the tail of animals that received pDCS (Fig. 3G). These findings confirm the specificity of the RNAi strategy and provide evidence in support of the hypothesis that lethally irradiated host tissue is the source of expression of neoblast markers upon pDCS. However, it remained unclear whether the presence of neoblasts in the graft was needed for the pDCS effects. To address this, neoblasts were eliminated from both the donor and host tissue by lethal γ-irradiation (Fig. 3H). Under these conditions, piwi-1 expression in the pDCS-treated animals remained similar to that in the sham-treated control (Fig. 3I), while there was a mixed effect on the expression of markers of neoblasts (i.e. expression of most markers was either reduced or did not change, with the exception of soxP-2; Fig. 3I). We also noted that the application of pDCS in a lethally irradiated animal without transplanted tissue did not trigger expression of neoblast markers or cell division. These results suggest that the presence of neoblasts in transplanted tissue is necessary for the pDCS-mediated expression of neoblast markers. Taken together, our findings indicate that pDCS elicits transcription of stem cell markers in lethally irradiated tissues – an effect that originates in host tissues but requires the presence of grafted neoblasts. Nevertheless, based on the spatiotemporal expression pattern of genes required for migration, the rapid presence of piwi-1-expressing cells, and cellular division upon short application of electric stimulation, we propose that pDCS-induced effects on neoblast transcription and cell division are not due to electrotactic cellular migration from the transplanted tissue but rather act through the activation of expression of neoblast progenitors and subsequent cell division originating in lethally irradiated host tissue.

pDCS enhances the DNA damage repair response in tissues exposed to a lethal dose of ionizing radiation

Exposure to a high dose of ionizing radiation induces DNA damage and subsequent cell death (Barghouth et al., 2019; Peiris et al., 2016a,b; Pellettieri et al., 2010). Nonetheless, 60 min pDCS-treatment of γ-irradiated animals activated gene transcription, a process known to require DNA integrity. Therefore, we assessed DNA integrity and repair mechanisms in dissociated cells obtained from the lethally irradiated tail region of animals in the sham-treated and pDCS groups. Ionizing radiation increases DNA double-strand breaks (DSBs), which, in planarians, are mainly repaired through homologous recombination (HR) (Barghouth et al., 2019; Peiris et al., 2016b). We used immunostaining to detect markers of the DNA damage and repair response after 60 min of either pDCS treatment or control sham treatment. The results revealed that pDCS led to a noticeable increase in phosphorylation of the histone H2AX (γ-H2AX; Fig. 4A), which is often observed in the early response to DSBs (Bonner et al., 2008; Marti et al., 2006). Likewise, pDCS increased RAD51 protein nuclear localization by 20% (Fig. 4B).

Fig. 4.

pDCS enhances the DDR, re-establishes DNA integrity and decreases cell death within γ-irradiated tissues. (A,B) Dissociated cells isolated from tail fragments were immunostained with (A) anti-γH2AX and (B) anti-RAD51 antibodies (green) to visualize nuclear versus cytoplasmic localization in sham-treated animals and animals subjected to 60 min pDCS. Nuclei were stained using DAPI (blue). (A) The classification of γH2AX signal includes four classes of nuclear signal, as displayed in the representative images on the left and as previously described (Barghouth et al., 2019; Thiruvalluvan et al., 2018). (B) The RAD51 signal was classified based on localization with respect to DAPI, as shown in the representative images on the left. The single-cell extract staining experiments in A and B consisted of five pooled tail fragments and three biological replicates. (C) DNA integrity was measured with a comet assay (mean tail length) using cells isolated from the host tail fragment in sham-treated animals and worms subjected to 60 min pDCS (n=12 animals in each group). (D,E,H) RNA-seq data obtained from tail fragments of sham-treated and 60 min pDCS tissue explants (data was collected by pooling tails from four independently treated pDCS-treated or sham-treated planarians across three independent experimental trials). Gene expression heatmaps display differentially expressed transcripts (FDR<0.05) as averaged log₂ CPM Z-scores for (D) putative DNA damage regulators, (E) DNA repair regulators, and (H) cell death regulators. (F) FACS analysis showing the distribution of live, pre-apoptotic, apoptotic and necrotic cells using annexin V staining in sham-treated and 60 min pDCS-treated animals. The data include four pooled tail fragments and three biological replicates. (G) Single-cell immunostaining using anti-caspase-3 antibody to identify three possible staining classes used for quantification (active caspase, pro-caspase and no caspase; representative images are shown on the left). Caspase immunostaining was obtained from five pooled tail fragments and three biological replicates. Data are presented as mean±s.e.m. obtained from experiments independently repeated at least three times. The polarity of the electric field was positive pole to the anterior and negative pole to the posterior for 60 min of pDCS. *P≤0.05; **P≤0.01; ****P≤0.0001; ns, not significant (two-tailed unpaired Student's t-test in A–C; one-way ANOVA with Dunnett's multiple comparisons test in F,G). Scale bars: 10 µm.

Nuclear translocation is essential for the function of RAD51 during DSB repair (Haaf et al., 1999; Peiris et al., 2016b). We further determined, using a comet assay, that pDCS-mediated activation of the DNA damage response (DDR) was accompanied by a noticeable reduction in DSBs caused by γ-irradiation (Fig. 4C). The results were expanded by performing transcriptomic analysis with a focus on genes associated with DNA damage and repair. RNA was extracted from the tail fragments from sham- and pDCS-treated animals after 60 min of treatment, and we used BLAST domains as annotations for the S. mediterranea transcriptome dd_Smed_v6 (Grohme et al., 2018). When examining orthologs of Homo sapiens DNA damage and repair pathways, the analysis confirmed that genes associated with these pathways were differentially expressed, with a majority upregulated (Fig. 4D,E).

In addition to the upregulated DNA repair-related and DNA damage-related transcripts, there was a strong upregulation of transcripts associated with DNA replication (Fig. S2, Tables S3–S22). Approximately 60% of significantly differentially expressed genes related to replication were upregulated, further supporting the notion that nucleic acid activity is enriched upon application of 60 min pDCS treatment. pDCS effects were also accompanied by improvements in cell viability as determined by flow cytometry assays of annexin V and by immunostaining using an anti-caspase-3 antibody (Peiris et al., 2016a; Thiruvalluvan et al., 2018). Reduced cytoplasmic signal of caspase-3 suggested lower levels of pre-apoptotic cells (Kamada et al., 2005) in the pDCS-treated group compared to the sham-treated group (Fig. 4F). Consistently, we also noticed a reduction in the percentage of pro-caspase-3⁺ cells, which are commonly associated with pre-apoptotic cells (Fig. 4G). These results were accompanied by differential expression of genes known to regulate apoptosis (Fig. 4H; Tables S3–S22). In summary, these findings demonstrate that 60 min of pDCS is capable of activating DNA repair, DNA damage and replication mechanisms, leading to reduced overall DNA damage in lethally γ-irradiated tissues.

pDCS activates transcription of immediate early genes in lethally irradiated tissues

Previous work has demonstrated the capacity for electric stimulation to produce rapid cellular responses, beginning at the transcriptional level (Dragunow and Robertson, 1987; Saha et al., 2011). To determine whether pDCS treatment is capable of influencing transcription of neoblast markers on a more rapid timescale, tissue from WT animals was grafted into lethally irradiated hosts and exposed to different lengths of pDCS (15, 30 and 45 min; Fig. 5A). Strikingly, expression of piwi-1 and other neoblast markers was not only detected but found maximally enriched during the first 15 min of pDCS (Fig. 5A–C). The expression of these neoblast-specific genes gradually reduced over time (Fig. 5A–C). Transcriptomic analysis using tail fragments from the 15 min timepoint showed a strong upregulation  in the expression of soxP-2, agat-1, piwi-1, piwi-2 and piwi-3 (BH FDR<0.05), which are markers of the sigma and pan-neoblast populations according to the classic classification (Fig. 5D; Tables S3–S22) (Eisenhoffer et al., 2008; Kim et al., 2019; Palakodeti et al., 2008; van Wolfswinkel et al., 2014). Moreover, we also detected significant differential expression of neoblast markers (Fig. 5E,F) based on the most recent transcriptomic analyses (Raz et al., 2021; Zeng et al., 2018). For example, markers of clonogenic neoblasts and muscle cells from Raz et al. (2021) were strongly upregulated after 15 min of applied pDCS (Fig. 5E; log₂ FC>0.5, BH FDR<0.01). Similar to the 60 min timepoint, we observed a stronger upregulation of soxP-2, MCM-4, piwi-1, bruli, and soxP-5. The analysis also revealed that 35 out of 189 markers (Fig. 5F) for neoblast subclasses were differentially expressed (Tables S3–S22; Zeng et al., 2018). In general, there was a strong upregulation in the expression of markers associated with cNeoblasts (Nb2) and progenitors of the pharynx (Nb7), epidermis (Nb1) and muscle (Nb4, Nb6) after 15 min of pDCS (BH FDR=0.05; Fig. 5F; Tables S3–S22).

This rapid upregulated transcription pattern was also observed in markers of the DDR, DNA repair and DNA replication (Fig. S3A–C). Differentially expressed Smed transcripts were sorted for annotations of Homo sapiens orthologous genes involved in DNA damage, DNA repair, and replication pathways. These Smed orthologs were plotted and show a critical upregulation in DNA repair pathways. For example, expression of Rad54 and Rad51 was upregulated with a log₂ FC greater than 1.2 (Fig. S4B, BH FDR=0.05). Checkpoint kinase 1 (CHK1) is a critical mediator of the DDR and cell cycle activation, and CHK1 gene expression was upregulated by 1.7-fold with early application (15 min pDCS treatment) of pDCS (BH FDR=0.01, moderated t=5.98). Expression of DNA polymerases, helicases and topoisomerases was upregulated twofold. Cyclin-dependent kinase 1 (CDK1) exhibited the strongest increase of expression, by nearly fourfold, further demonstrating the strong activation of cell cycle pathways (BH FDR=0.0046, moderated t=4.72). In comparison, the expression of CHK1 and CDK1 at the 60 min timepoint was slightly dampened, which supports the high enrichment of cell cycle regulators in our gene set enrichment analysis after 15 min pDCS (Tables S1–S22). The results indicate that pDCS elicits a rapid transcriptional response geared toward markers of neoblast, DNA damage and DNA repair within the host irradiated tissues.

These remarkably rapid changes in gene expression after pDCS are temporally consistent with activation of immediate early gene (IEG) transcription, generally defined as genes expressed in the absence of de novo protein synthesis (Bahrami and Drabløs, 2016; Greer and Greenberg, 2008; Herschman, 1991; Morgan and Curran, 1991; Saha and Dudek, 2013). Furthermore, the mRNAs of IEGs are detectable within minutes of exposure to a wide range of stimuli, such as stress, mitogens, immune challenge, neuronal signals and electric stimulation (Bahrami and Drabløs, 2016; Cohen and Greenberg, 2008; Greer and Greenberg, 2008; Saha and Dudek, 2013). pDCS timecourse qPCR results revealed a rapid and transient expression of a well-characterized member of the IEG family, early growth response gene-1 (egr-1; Fig. 5C), which is a well-characterized member of the IEG family (Bahrami and Drabløs, 2016; Greer and Greenberg, 2008) and an established neoblast marker required for regeneration and stem cell function in planarians (Lei et al., 2016; Sandmann et al., 2011; Tu et al., 2015; Wagner et al., 2012; Zeng et al., 2018).

IEGs are classified based on their induction profile and are separated into rapid, delayed or slow expression response groups following a stimulus (Bahrami and Drabløs, 2016; Saha and Dudek, 2013). Thus, we compiled a list of 138 known IEGs previously reported in other model organisms (Fig. S4) (Cullingford et al., 2008; Tullai et al., 2007; Uhlitz et al., 2017) and assessed the expression of their putative planarian orthologs following pDCS. Genes that were considered significantly differentially expressed at the 15 min timepoint were listed after applying P-value and FDR cut-offs of <0.05; this yielded a total of 1778 upregulated genes. Out of these 1778 genes, we found 25 planarian orthologs to previously published IEGs (Fig. 5G; Fig. S4). This analysis confirmed the upregulated expression of egr-1 and revealed that the increased expression of IEGs persists through the 60 min application of pDCS (log₂FC=1.32, BH FDR=0.0013, moderated t=7.42). Additionally, other IEGs were considerably upregulated after 15 min pDCS treatment, including genes encoding the RAS oncogene and dual-specificity phosphatase 10 (DUSP10), which is known to affect components of the signaling pathways mediated by the mitogen-activated protein kinases JNK and ERK (Fig. 5G). Within the 25 planarian putative IEGs, we found representation for all three subclasses [immediate early genes (IEG), delayed early genes (DEG), and immediate late genes (ILG)], and the number of upregulated and downregulated members of those groups were split evenly. It is not entirely clear how pDCS transcriptional regulation of IEGs is translated into cellular actions, but our observations suggest that short treatment with pDCS stimulates the rapid induction of IEGs within lethally irradiated tissues.

pDCS induces ectopic expression of neoblast markers mediated by Ca²⁺ signaling

Since exposure to a lethal dose of irradiation irreversibly eliminates neoblasts, we assessed the identity of host cells expressing neoblast markers following 15 min pDCS. To recognize the spatiotemporal distribution of cells at different stages of differentiation, we performed expression analysis using double fluorescent in situ hybridization (FISH) in tissue sections (Fig. 6A). Specific genes were chosen to reflect two distinct stages of cellular differentiation, piwi-1 and agat-1, which label early progenitors and late post-mitotic progeny, respectively (Eisenhoffer et al., 2008). Our results show that 15 min pDCS induced a 56-fold increase in piwi-1⁺ cells, and this upregulation coincided with a simultaneous 21-fold increase in agat-1⁺ cells (Fig. 6B,C). Intriguingly, 70.1% of agat-1⁺ cells co-expressed piwi-1, whereas 24.5% of piwi-1⁺ cells co-expressed agat-1 (Fig. 6B). Moreover, sham-treated control planarians exhibited minimal numbers of piwi-1⁺ and/or agat-1⁺ cells, as expected in 4 days post-irradiation tissue (Fig. 6A,C). Previous research has reported that minimal, if any, overlap between piwi-1 and agat-1 expression occurs in WT planarians (<2.0%; Eisenhoffer et al., 2008). The ectopic expression of the neoblast marker in post-mitotic cells suggests that the brief application of pDCS disrupts patterns of gene expression across cellular lineages.

Fig. 6.

Ca²⁺ signaling mediates pDCS-induced ectopic expression of piwi-1. (A) Double fluorescent in situ hybridizations (FISH) for agat-1 (red) and piwi-1 (green) were performed on sagittal cross-sections (dashed line in the diagram) of sham-treated and pDCS-treated animals (n=6). Nuclei were stained using DAPI (blue). The region shown in the micrographs is indicated by the red box in the diagram. Dashed box indicates the region shown enlarged in the top-right image. (B) Venn diagrams showing the percentage of the agat-1⁺ cell population also expressing piwi-1 (left) and the percentage of piwi-1⁺ cells also expressing agat-1 (right). (C) The average number of piwi-1⁺ and agat-1⁺ cells per section of sham-treated and 15 min pDCS-treated samples. (D,E) Double FISH for agat-1 (red) and piwi-1 (green) performed on sagittal cross-sections following (D) 24 h soak with 5 µM nicardipine or (E) 15 min incubation with 100 µM EGTA-AM (n=6). Nuclei were stained using DAPI (blue). Dashed boxes indicate regions shown in enlarged images. In A,D,E, yellow arrows point to cells expressing both piwi-1 and agat-1, green arrows point to cells only expressing piwi-1, red arrows point to cells only expressing agat-1. (F,G) Venn diagrams showing the percentage of the agat-1⁺ cell population also expressing piwi-1 (left) and the percentage of piwi-1⁺ cells also expressing agat-1 (right) in planarians exposed to (F) 5 µM nicardipine or (G) 100 µM EGTA-AM. (H) Bar graph displaying the average number of piwi-1⁺ and agat-1⁺ cells per section within each inhibition group. (I,J) Gene expression levels of neoblast markers piwi-1, soxP-2, fgfr1, egr1, hnf4, nkx2.2 and inx-13 assessed using qPCR following (I) 24 h 5 µM nicardipine soak or (J) 100 µM EGTA (data is from five pooled tail fragments/replicates and three biological replicates). All cases involve tail tissue isolated following 15 min pDCS for in the experimental group. Expression levels are normalized to the level in the sham-treated group (dashed lines). Data are presented as mean±s.e.m. pNb, pan neoblast; cNb, clonogenic neoblast; Nb, neoblast. The polarity of the electric field was orientated with the positive pole to the anterior and negative pole to the posterior for 15 min. **P≤0.01; ****P≤0.0001; ns, not significant (two-tailed unpaired Student's t-test in C; Kolmogorov–Smirnov test, KS<0.05 in H; one-way ANOVA with Dunnett’s multiple comparisons test in I,J). Scale bars: 100 µm.

Ca²⁺ signaling is among the most prominent mediators of excitation–transcription coupling and IEG activation (Greenberg et al., 1986; Greer and Greenberg, 2008; Saha and Dudek, 2013; Saha et al., 2011; Yan et al., 2014). For example, voltage-dependent Ca²⁺ channels at the plasma membrane can be electrically stimulated to allow the rapid influx of Ca²⁺ to the cytoplasm (Yan et al., 2014). Similarly, Ca²⁺ signaling mechanisms have been suggested as mediators of acute signaling events in various experimental models, including planarians (Bahrami and Drabløs, 2016; Chan et al., 2017; Cohen and Greenberg, 2008; Greenberg et al., 1986; Greer and Greenberg, 2008; Herschman, 1991; Kandel, 2012; Ma and Yan, 2014; Marchant, 2019; Morgan and Curran, 1991; Saha and Dudek, 2013; Saha et al., 2011; West and Greenberg, 2011; Yan et al., 2014). In concert with these reported findings, inhibiting Ca²⁺ flux through L-type voltage-gated Ca²⁺ (Ca_v) channels using a dihydropyridine (DHP), nicardipine, dramatically suppressed the effects of rapid (15 min) pDCS-mediated expression of neoblast markers (Fig. 6D,I). The effects of nicardipine inhibition persisted even when pDCS was extended to 60 min (Fig. S5A–C). These results were confirmed using nifedipine, another DHP that blocks L-type Ca_v channels via a different high-affinity binding site (Fig. S5D–G). Likewise, buffering of intracellular Ca²⁺ with EGTA-AM also disrupted pDCS-mediated piwi-1 and agat-1 expression (Fig. 6E,G,H,J). These results suggest that Ca²⁺ released from intracellular Ca²⁺ stores (i.e. endoplasmic reticulum) mediate pDCS effects.

DISCUSSION

Our findings underscore the overriding capacity of bioelectric signaling to rapidly affect essential cellular processes such as transcription, the cell cycle and DNA repair. This robust and effective strategy is capable of altering cellular behavior in situ, without the need for genetic or pharmacological intervention. Our results in planarians are consistent with the overriding effects obtained following DCS in mice models of the Rett syndrome (RTT). The RTT is caused by inactivation of the X-linked gene methyl-CpG-binding protein 2 (MECP2) (Amir et al., 1999) and is a complex degenerative dysfunction involving many genes and neuronal groups, in which pharmacotherapy is unlikely to succeed (Baker et al., 2013; Chahrour et al., 2008; Johnson et al., 2017; Sugino et al., 2014). However, the application of DCS with electrodes implanted in the brain of the RTT mouse model led to activation of neurogenesis and restored neural circuits and spatial memory, and the behavior of the experimental group was found to be indistinguishable from sham-treated mice (Hao et al., 2015; Lu et al., 2016; Pohodich et al., 2018). Jointly, the results in both vertebrates and invertebrates suggest that the overriding effects of DCS (pulsing or steady state) consistently overcome conditions involving dysfunctional DNA.

We introduce planarians as a simplified platform to carry out comprehensive analyses aimed at dissecting the molecular basis of electric stimulation at the organismal, cellular and subcellular levels. Chloretone was used in this study to sedate planarians before immersion in molten agar. Chloretone has known interactions with transmembrane ion channels, specifically, voltage-gated sodium channel type II (Na_V1.2) (Kracke and Alandrum, 2011). Thus, concerns over the long-term impact chloretone may have on the bioelectric state of planarian cells, including Ca²⁺-mediated intracellular Ca²⁺ release, are warranted. Nonetheless, the effective inhibition of Na_V1.2 channels caused by chloretone is known to be dose-dependent and reversible (Kracke and Alandrum, 2011). In addition, work in mammalian cells has shown that chloretone bathing does not alter Ca²⁺ dynamics of discharging afferent neurons, nor does it influence the shape of action potentials (Fischer, 2000). Altogether, given the reversible nature of chloretone and the timing imposed on immobilized planarians before experimentation (15–30 min of rest), we believe the pDCS effects are independent of any effects of chloretone exposure.

We observe extensive commonalities between DCS effects in planarians and mammals. For example, the time and strength of the currents used in our DCS are similar to the ones used in humans [e.g. transcranial direct current stimulation (tDCS), muscle, bone repair] (Gerovasili et al., 2009; Kadosh et al., 2010; Moreno-Duarte et al., 2014). Ca²⁺ signaling consistently recurs as a mediator of DCS effects in planarians and mammals. Likewise, the overall changes upon pDCS are transient, thus providing self-contained regulatory mechanisms that can be calibrated­ to gain desired cellular responses under different circumstances. Uniquely, our findings demonstrate a cost- and time-effective alternative for studying rapid activation of transcription in tissues exposed to high doses of ionizing radiation. DNA damage is central to cancer, aging and radiotherapy, but there are limited options to effectively enhance genomic repair. We present evidence demonstrating that short exposure to pDCS activates transcription of genes involved in the DDR, which together lead to the re-establishment of DNA integrity in tissues exposed to a high dose of ionizing radiation. Currently, we do not have data to address potential leakage or the anatomical positioning of electrodes. Our previous pDCS characterization has shown that alternative ionic conduits and varying ionic concentrations of salt solutions within the microelectrodes results in similar effects across them (Davidian et al., 2021). These findings suggest that pDCS effects are likely the result of electric stimulation and not salt leakage around the electrode. However, we do not discard that variation between experiments may partly be due to differences in internal current caused by leakage or inconsistencies in electrode placement. Future experiments will be designed to address less invasive strategies to overcome the undesirable effects of the electrodes. We also plan to characterize both the fidelity of pDCS-induced DNA repair and the molecular mechanism mediating this process. One candidate mechanism might involve small non-coding RNAs (sncRNAs), which recent evidence shows may facilitate the recruitment of repair components in both the HR and non-homologous end joining pathways to sites of DSBs (Gao et al., 2014; Qi et al., 2016; Wei et al., 2012).

pDCS triggers the ectopic transcription of stem cell and differentiated tissue markers followed by mitotic activity in tissues damaged by ionizing radiation. The molecular mechanism of this intriguing finding is still unclear, but it is possible that pDCS may affect cell fate regulators associated with the lineage progression of post-mitotic progenitors to increase cellular plasticity. Indeed, our finding showing overlapping expression of agat-1 and piwi-1 in post-mitotic cells is consistent with recent findings demonstrating enhanced cellular plasticity by disturbing Hippo and egr-5 signaling pathways (de Sousa et al., 2018; Tu et al., 2015). The cellular plasticity evidenced upon pDCS treatment is also consistent with a recent proposal stating that specialized neoblasts may retain pluripotency, which is not an exclusive property of any known neoblast subpopulation (Raz et al., 2021). However, to our knowledge, there is no precedent in which a short electric stimulus can robustly coordinate genetic and cellular events toward stem cell reconstitution in tissues damaged by ionizing radiation. Future studies will be needed to understand the epistatic relationship between cell fate regulators and the identity of the cells expressing neoblast markers ectopically, along with the long-term stability of the piwi-1⁺ cells.

One possible explanation for the pDCS-mediated expression of stem cell markers is that distinctive post-mitotic progenitors can sense the electric stimulus and respond. Consistent with this idea, we propose that post-mitotic lineages expressing L-type Ca_v channels (i.e. neural, epidermal, parenchymal, protonephridia and cathepsin⁺ cells; Fincher et al., 2018; Plass et al., 2018) are likely the targets of pDCS-enhanced plasticity (Fig. S6). Future experiments will address the individual contributions of post-mitotic lineages expressing L-type Ca_v channels after pDCS. Another possible scenario may involve the presence of radio-resistant neoblasts or neoblasts with low cycling activity that are sensitive to the electric stimulus. Recent evidence supports the possibility of slow-cycling neoblasts with distinctive regenerative properties (Molinaro et al., 2021). As stated above, pDCS treatment may also enable proliferative expansion of specialized neoblasts (Raz et al., 2021). The complete picture of neoblast heterogeneity and regulation is an evolving subject that is far from being understood (Fincher et al., 2018; Molinaro and Pearson, 2016; Plass et al., 2018; Raz et al., 2021; Reddien, 2018; Rink, 2018; van Wolfswinkel et al., 2014; Wagner et al., 2011; Zeng et al., 2018).

We propose a model whereby pDCS may lead to enhanced DNA repair followed by transcription. Our results suggest that pDCS effects are mediated by increases in intracellular Ca²⁺ concentration via L-type Ca_v channels and/or release from intracellular Ca²⁺ stores (Fig. 7A). The initial effects of pDCS stimulate transcription of IEGs that are tightly regulated by increases in intracellular Ca²⁺ concentration via L-type Ca_v channels and/or release from intracellular Ca²⁺ stores (Fig. 7B). The increase in cytosolic Ca²⁺ concentration may serve to boost DNA repair, improving DNA integrity, followed by the transcription of IEGs (Fig. 7B). The consequences of these Ca²⁺-mediated events significantly reduce overall levels of DNA damage, leading to transcription of stem cell-related genes and cell cycle re-entry in tissues damaged by ionizing radiation. Further experiments are needed to determine the order in which these events take place and to further define the identity of the cells expressing neoblast markers. It is unclear whether radio-resistant cells, slow-cycling neoblasts or post-mitotic cells with enhanced plasticity are associated with the pDCS-induced effects (Fig. 7C).

Fig. 7.

Schematic summary of pDCS-induced effects. (A) pDCS-mediated effects in lethally γ-irradiated host tissue (gray cells) require engrafted neoblasts (colored cells, gray arrows). (B) Proposed cellular effects responsible for observed pDCS-induced transcriptional activation. Transcription sensitive to (1) Ca²⁺ influx through L-type Ca_v and subsequent (2) Ca²⁺ release from intracellular stores (ER, endoplasmic reticulum) leading to (3) Ca²⁺-mediated transcription and (4) expression of pDCS-induced genes (Neo, neoblast marker genes). The ectopic overlapping expression of agat-1 and piwi-1 suggests enhanced plasticity of transcriptional programs within the lethally γ-irradiated host tissue. (C) pDCS activates rapid transcription of genes associated with neoblast populations in lethally γ-irradiated host tissue.

MATERIALS AND METHODS

Planarian culture and maintenance

The asexual strain of the planarian Schmidtea mediterranea, clone CIW4, was used for all experiments described in the article. For transplantation experiments, planarians were acclimatized to a final culturing temperature of 13°C. Acclimatization was gradual, beginning with an initial transfer to 16°C for ∼2 weeks and then stepping to 13°C permanently until the planarians were used for transplantation. Before each change in temperature, planarian cultures destined for transfer were fed before temperature changes. Animals transferred to 13°C incubators were cleaned once per day for the first 4 days of temperature acclimatization. All planarian maintenance was performed as previously described (Oviedo et al., 2008b). After the first 4 days, planarian maintenance was resumed as previously described (Oviedo et al., 2008b). Planarians were not used for tissue transplantation until they had been cultured for at least 2 weeks at 13°C. Reduced culturing temperatures were used to decrease the mobility of recovering planarians following tissue transplantation.

Tissue transplantations

Planarians were transplanted as previously described (Guedelhoefer and Sanchez Alvarado, 2012a,b) with minor changes to tools used for transplanting tissue. We developed a transplantation tool to facilitate consistency in the size of the graft and reduce tissue damage. Briefly, the transplantation tool was made from an 18.5–19 gauge syringe that was bored out to an inner diameter of 750 µm using a Dremel drill bit. The outer diameter was polished using 500–1000 grit wet sandpaper until the edges were paper thin and smooth to reduce drag during tissue insertion.

Transplantation schedules varied with respect to experimental condition (i.e. WT, irradiated or RNAi tissues). Specifically, in experiments using irradiated planarian host or donor tissue, all irradiation was performed 24 h before tissue transplantation. 7-day starving planaria were selected and given 6 krads of γ-irradiation followed by 24 h of rest before used as hosts for tissue transplantation. For transplantation experiments using piwi-1(RNAi) tissues, transplantation was performed 48 h post-final injection (5th injection) and subsequent pDCS experiments occurred at 4 dpt. Additionally, in transplantation experiments using piwi-1(RNAi) planarians, all piwi-1(RNAi) donor tissue was derived from piwi-1(RNAi) hosts [i.e. piwi-1 donor inserts were taken from the anterior region of intact piwi-1(RNAi) animals, which were then used as hosts for WT donor tissues].

Planarian immobilization

Transplanted or intact planarians were placed in chilled 0.2% w/v chloretone solution for 3–5 min (Guedelhoefer and Sanchez Alvarado, 2012a,b) in preparation for agar immobilization. After soaking, the planarians were rinsed in chilled planarian water (Oviedo et al., 2008b). Motionless planarians were individually placed on large 75 mm×50 mm glass slides, ∼1.0 cm apart, atop the ice. All remaining planarian water was removed, and planarians were subsequently covered in 1.0% low-melting-point agarose (1.0% w/v LMP agar in planarian water; Thermo Fisher Scientific, 16520050) at near to room temperature until the planarians were entirely submerged. Planarians were gently positioned to level during the gelation process to achieve maximum body axis symmetry. Once the agar was completely solidified, excess agar was trimmed and encapsulated planarians were placed into the center of chilled 35 mm Petri dishes (Corning, CLS430588) that were prefilled halfway with solidified 1.0% agarose (1.0% w/v agar in planarian water; Sigma, A9539). The remaining Petri dish volume was filled with 1.0% agarose until the agarose level was flush with the top of the encapsulated planarian. All agar encapsulation processes were performed on ice.

Administration of current and electric field generation for pDCS experiments

Planarians were immobilized in agarose and subjected to applied currents via current clamped microelectrodes. To deliver the current, a power supply was fashioned using thirty 9 V batteries in series, sectioned in 45 V increments; the power supply was fashioned with a 100 K rotary potentiometer to adjust output to the desired voltage (Fig. 1D–F).

To deliver the current to the planarians, borosilicate sharp microelectrodes were pulled using a P-97 Flaming/Brown pipette puller (Sutter Instruments P-97). Microelectrodes were filled with 3 M KCl and placed vertically in a 3 M KCl bath with the pulled tip in the solution. To deliver current to the planarians, microelectrodes were bridged with 6–8″-long 1/32″ internal-diameter PVC tubing filled with 3 M KCl 1.0% agar connected to 3 M KCl baths joined to the power supply via Pt electrodes. The current was clamped using an RNX 3/8 1GO 100 MΩ resistor (Mouser RNX-3/8-100M) in series with the planarian.

For all pDCS experiments, the power supply output was 70 V (7 µA delivered to the planarian) and the immobilized planarians were impaled through their ventral epithelial layer in the pre-pharyngeal region, proximal to the brain, and in the tail tip with each glass microelectrode. The duration of current administration ranged from minutes to hours depending on the needs of the experiment. The polarity of the electric field used was positive pole in the anterior and negative pole to the posterior region of the animal.

RNA isolation

RNA from tissues was extracted as previously described (Oviedo et al., 2008c). Tail fragments from sham-treated and pDCS-treated animals were placed in Trizol immediately after amputation, and RNA was extracted for each sample. Triplicate analysis was performed for each data point, consisting of pooled samples from four planarian tails per replicate.

Library preparation and RNA sequencing

The cDNA sequencing libraries were prepared using an automated system at the UC Davis Technologies Core. All samples were accompanied with quality control (QC) documentation and profiled with Bioanalyzer (Agilent) for QC before sequencing. Poly-A enrichment was used to remove ribosomal RNA contamination and maximize mRNA detected. Samples were indexed and pooled for multiplexing. Using the Illumina HiSeq 4000 platform, paired-end reads were sequenced to a length of 200 bp by the DNA Technologies Core at the UC Davis Genome Center. This generated high-quality RNA-seq data for thorough downstream bioinformatic analysis to detect delicate changes in phenotype. Paired-end sequencing was also used to resolve ambiguous differences with high repeat regions.

Read mapping and gene expression analysis

Trimmed fastq files were assessed for quality control and mapped to the recently published complete Schmidtea mediterranea genome dd_Smes_g4_1 from the PlanMine database (Grohme et al., 2018). The Bioconductor package Rsubread version 2.4.2 (Liao et al., 2013) was used to map reads to the reference genome using a robust and efficient seed-and-vote algorithm, followed by the featureCounts algorithm to assign counts. The raw count data were normalized and filtered for genes with log₂ counts-per-million (CPM) greater than 0.5 (Tables S21,S22). The sample variation was assessed for quality. A customized pipeline using the limma package and voom transformation for precision weights was developed (Law et al., 2014; Phipson et al., 2016; Ritchie et al., 2015). Limma version 3.46.0 and edgeR version 3.32.1 were used for the statistical analysis. Test statistics were produced using empirical Bayes moderation, and subsequent heatmaps were made using the ComplexHeatmap Bioconductor package. We separately tested the contrasts of gene expression in the 15-min and 60-min conditions against that in the sham-treated control using a Benjamini–Hochberg FDR of 5%. All bioinformatic analyses were coded with R version 4.0.3 on macOS Big Sur 10.16 using the x86_64-apple-darwin17.0 platform.

Gene set enrichment analysis

Gene set enrichment analysis was performed using the Bioconductor topGO package version 2.42.0 (https://bioconductor.org/packages/release/bioc/html/topGO.html; Alexa et al., 2006). Gene ontology annotations for the dd_Smed_v6 transcriptome were mined from the Planmine database and used to map pathway enrichment (https://bioconductor.org/packages/release/bioc/html/topGO.html; Grohme et al., 2018). Gene rankings calculated by the Limma–voom pipeline were used for determining the significance and enrichment (Alexa et al., 2006; Law et al., 2014; Liu et al., 2015; Phipson et al., 2016; Ritchie et al., 2015; Smyth and Speed, 2003). Enrichment was computed by ranking gene scores using the conservative Kolmogorov–Smirnov (K–S) test. Pathways were considered enriched with a K–S P-value <0.05. A complete RMarkdown-based notebook of code to reproduce the transcriptomic and gene set enrichment analysis is available upon request.

Ca²⁺ inhibition using nicardipine, nifedipine and EGTA-AM

Nicardipine and nifedipine inhibition was performed on transplanted planarians at 3 dpt such that the 24 h incubation period concluded at 4 dpt. Each DHP was dissolved in 100% DMSO at a concentration of 10 mM and then diluted in 50 ml of planarian water to a final working concentration of 5 µM, as described previously (Nogi et al., 2009). EGTA-AM administration as achieved via posterior injections 1 h prior to pDCS delivery. Firstly, planarian volume was estimated using where a, b and c represent planarian length, width and height, respectively. With estimated planarian volume, EGTA-AM was diluted in Milli-Q water and injected (37 nl per injection pulse) such that the final EGTA concentration was 100 µM.

Tissue preparation for cryosections

Planarian are fixed using N-acetylcysteine (NAC)–formaldehyde-based fixation as described previously (King and Newmark, 2013). Fixed animals were prepared for cryosectioning as previously described (Reddien et al., 2005). Briefly, planarians were immersed in increasing concentrations of sucrose diluted in 1× phosphate-buffered saline (PBS). 1× PBS was replaced with 15% sucrose solution for 1 h and then 30% sucrose solution overnight at 4°C. Planarians were then placed in tissue embedding molds (Thermo Fisher Scientific, 50-192-9392); 30% sucrose was removed and replaced with optimal cutting temperature (OCT) medium. Planarians were then situated in the desired position and orientation within the OCT, and molds were quickly placed in cooling bins containing dry ice immersed in 2-methyl-1-butanol. Once frozen, samples were stored at −80°C until needed for sectioning. The tissue was sectioned in a Leica Cryostat CM1860. Sectioned tissue was either used immediately for FISH or immunohistochemistry (IHC), or was stored long-term at −80°C until needed.

Fluorescent in situ hybridization on sectioned tissue

Cryo-section-containing slides were removed from storage at −80°C and given 30 min to reach room temperature. Clear Scotch tape was placed over the frosted label to ensure label longevity and assist in coverslip placement during hybridization. Custom-made multi-slide chambers were developed and used throughout the procedure. Multi-slide chambers held 12 slides and only required 16 ml of the solution to reach maximal coverage. Standard Coplin jars (made in house) may be used but require more solution volume per slide contained. After slides reached room temperature, slides were placed in the chamber and rehydrated in 1× PBS for 15 min twice. Following rehydration, FISH was performed as described previously (King and Newmark, 2013) with specific changes as follows: proteinase K incubation was performed for 10 min at room temperature at 1 µg/ml proteinase K concentration. 1:1 prehybridization/PBS with Triton X-100 was omitted. For pre-hybridization and hybridization solution incubation, slides were removed from the chamber and placed in repurposed slide-holding containers (converted to hybridization chambers using water-saturated tissue paper placed in each column), 150 µl of the solution was administered to slides, and then glass coverslips were placed atop slides to seal in the solution, decreasing evaporation. At the end of the FISH procedure, DAPI (1:1000; Thermo Fisher Scientific, cat# 62248) was added to slides within the multi-slide chamber for 30 min and washed twice with 1× PBS. Slides were mounted using Gelvatol solution (see https://doi.org/10.1101/pdb.rec10252).

Fixation of large planarians for in situ hybridization

All large planarians (8–12 mm) used for whole-mount in situ hybridization (WISH) were fixed using a modified NAC fixation protocol previously described (Guedelhoefer and Sanchez Alvarado, 2012a). Changes for fixation were introduced as follows: each planarian first underwent MgCl₂ tissue relaxation as described by Forsthoefel et al. (2014). Briefly, planarians were placed in 0.66 M MgCl₂ for 45–60 s until they fully relaxed. The MgCl₂ solution was replaced with 10% NAC for 10 min at room temperature. Lastly, bleaching was performed with a modified bleaching solution (0.5% formamide, 0.36% H₂O₂ and 0.05% Triton X-100 in 1× PBS) and placed under light overnight.

WISH

WISH was performed following a previous published protocol (Guedelhoefer and Sanchez Alvarado, 2012a). Minor modifications were made during proteinase K treatment, doubling the concentration to 2 mg/ml and treating samples for 10 min at room temperature.

Whole-mount IHC and analysis

Planarians were fixed for IHC following standard formaldehyde-based fixation with an added formamide-based step as described previously (Guedelhoefer and Sanchez Alvarado, 2012a). The number of cells scored as positive for phosphorylated histone H3 (H3P) signal was normalized to the planarian surface area (mm²). When counting H3P in transplanted planarian, the anterior–posterior axis margin was shifted to the center of the transplant. Total H3P⁺ cells were then counted in relation to the population of H3P⁺ cells within the transplant. Briefly, total H3P⁺ cells were counted in host tissue, specifically excluding cells within the transplanted tissue. Additionally, counting regions were restricted to the anterior or posterior regions within the sham-treated and pDCS-treated planaria, independently (% cells outside transplant). Cell counting was performed using the ImageJ cell counter plugin (NIH, Bethesda, MD), and all data analysis was performed in Graphpad Prism. The primary antibodies used were anti-H3P (1:250; Millipore, 05-817R), anti-activated caspase-3 (1:500; Abcam, ab13847), anti-RAD51 (1:500; Abcam, ab109107) and anti-phospho-histone H2AX (1:1000; Thermo Fisher Scientific, LF-PA0025). The secondary antibodies used were Alexa Fluor 488-conjugated goat anti-mouse IgG (1:800; Invitrogen, 673781) and HRP-conjugated goat anti-rabbit IgG (1:2000; Millipore, 12-348).

Fixation and IHC on dissociated cells

Planarian tail portions were separated and homogenized following pDCS. Homogenate was suspended in calcium and magnesium-free medium (CMF; Reddien et al., 2005) and placed on ice. Cell density was quantified using a hemocytometer, and cells were plated at 1 million/cm² on glass coverslips. Cells were given 1 h to adhere to the surface and were then fixed with Carnoy's solution (Peiris et al., 2016b) for 2 h on ice. IHC was performed using antibodies against human RAD51 (1:500; Abcam ab109107) and γH2AX (1:1000; Themo Fisher Scientific LF-PA0025), as previously shown (Peiris et al., 2016b; Thiruvalluvan et al., 2018).

RNAi experiments

RNAi was carried out via dsRNA micro-injection (Oviedo et al., 2008a). Application of piwi-1 RNAi consisted of three consecutive injections followed by one weekly injection until animals were utilized for experimentation (14 days post first injection). Injections were administered to the prepharyngeal regions and, due to the size of the planarian (8–12 mm), each planarian was given 6–8 pulses of dsRNA, 37 nl each. The piwi-1 gene was selected and identified using the SmedGD database (Robb et al., 2015). dsRNA was synthesized in vitro as previously described (Oviedo et al., 2008a).

Quantitative real-time PCR

Quantitative real-time PCR (qPCR) was performed as described previously (Peiris et al., 2016b). The ubiquitously expressed gene H.55.12e (Reddien et al., 2005) was used as a reference control. Experiments were conducted in triplicates for each condition. Each qPCR experiment was conducted independently at least two times. All qPCR experiments used RNA extractions from tails of pDCS- and sham-treated transplanted planarians, unless otherwise specified (i.e. qPCR for migration-related genes used tissue sections across the planarian, as described in the text). Extracted RNA was then converted to cDNA using the Verso cDNA synthesis kit (Thermo Fisher Scientific, AB1453A). Gene expression is expressed in fold change in comparison to the given control condition.

FACS analysis and comet assay

FACS and comet analyses were performed as previously described (Peiris et al., 2016a,b).

Imaging and data processing

All images were captured using a Nikon AZ-100 multi-zoom microscope equipped with NIS Elements AR 3.2 software. The brightness and contrast of digital images captured in NIS Elements software were further adjusted in Photoshop (Adobe). Area calculations and cellular foci quantification were carried out using ImageJ software. Mitotic counts were normalized to the planarian surface area using ImageJ.

Statistical analysis

Data are expressed as the mean±s.e.m. or fold change±s.e.m. Statistical analysis was performed in Prism 2015 software, Graphpad Inc.

Data availability

All raw and processed data files associated with this study have been deposited to the NCBI Sequence Read Archive (SRA) submission number SUB8831617. The bioinformatic and RNA-seq analyses pipeline with metadata files can be found on the Github repository at: https://github.com/mlegro/RNA-seq-of-pDCS.

Peer review history

The peer review history is available online at https://journals.biologists.com/jcs/article-lookup/doi/10.1242/jcs.259304.