• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Biol. Author manuscript; available in PMC Nov 1, 2007.
Published in final edited form as:
PMCID: PMC1781256

Cellular Electroporation Induces Dedifferentiation in Intact Newt Limbs


Newts have the remarkable ability to regenerate lost appendages including their forelimbs, hindlimbs, and tails. Following amputation of an appendage, the wound is rapidly closed by the migration of epithelial cells from the proximal epidermis. Internal cells just proximal to the amputation plane begin to dedifferentiate to form a pool of proliferating progenitor cells known as the regeneration blastema. We show that dedifferentiation of internal appendage cells can be initiated in the absence of amputation by applying an electric field sufficient to induce cellular electroporation, but not necrosis or apoptosis. The time course for dedifferentiation following electroporation is similar to that observed following amputation with evidence of dedifferentiation beginning at about 5 days postelectroporation and continuing for 2 to 3 weeks. Microarray analyses, real-time RT-PCR, and in situ hybridization show that changes in early gene expression are similar following amputation or electroporation. We conclude that the application of an electric field sufficient to induce transient electroporation of cell membranes induces a dedifferentiation response that is virtually indistinguishable from the response that occurs following amputation of newt appendages. This discovery allows dedifferentiation to be studied in the absence of wound healing and may aid in identifying genes required for cellular plasticity.

Keywords: Dedifferentiation, Cellular plasticity, Regeneration, Electric field, Electroporation, Newt, Notophthalmus viridescens


Newts have the ability to regenerate lost appendages and injured organs, including their limbs, tails, spinal cords, retinas, lenses, optic nerves, jaws, heart ventricles, and intestines (Brockes and Kumar, 2002; Butler and Ward, 1967; O’Steen, 1958; Turner and Singer, 1974a; Turner and Singer, 1974b). These regenerative events are dependent upon an unusual degree of cellular plasticity near the site of injury. For example, following limb or tail amputation, the wound is closed within 1 day by the migration of epithelial cells from the proximal epidermis. The internal cells underlying this newly-formed wound epithelium begin to dedifferentiate to form a pool of proliferating progenitor cells known as the regeneration blastema. This dedifferentiation process is characterized in vivo by a general histolysis of the internal tissues, cell cycle reentry in normally quiescent cells, downregulation of cell differentiation markers, and upregulation of blastemal markers (Bodemer and Everett, 1959; Chalkley, 1954; Hay and Fischman, 1961; Kintner and Brockes, 1984; Thornton, 1938a; Thornton, 1938b). Later in the regenerative process, the blastemal cells will redifferentiate to form all of the internal tissues of the regenerated structure, except the nerve axons.

Loss of an appendage or injury of an organ initiates a regenerative response involving the dedifferentiation of cells near the wound. Several studies have suggested that severe injury is the main requirement for inducing regeneration or the related phenomenon of supernumerary limb formation. Supernumerary limbs can form when a deep incision is made through the limb followed by the placement of a tight ligature through the incision and around the remaining uncut portion of the limb (Della Valle, 1913; Tsonis, 1996; Wallace, 1981). Application of carcinogens or inflammatory substances to a urodele limb can induce dedifferentiation of internal tissues and supernumerary limb formation (Breedis, 1952; Tsonis and Eguchi, 1981). Crushing injuries can also produce a regenerative response that perfectly repairs the tissues of the crushed region (Mescher, 1982).

We show here that application of an electric field sufficient to cause electroporation of internal limb cells, but insufficient to cause necrosis or apoptosis, can initiate a dedifferentiation process characterized by cell cycle reentry of appendage cells, histolysis of internal tissues, and appropriate regulation of differentiation and blastemal markers. There is a direct correlation between pore formation in cell membranes and dedifferentiation of internal limb cells, suggesting that widespread, quickly reversible cell membrane damage is sufficient to initiate the dedifferentiation process. Microarray and real-time RT-PCR analyses reveal that electroporated and amputated newt limbs exhibit similar temporal gene expression patterns, while in situ hybridization experiments suggest that upregulated genes are expressed in the same tissues following both types of injuries. These results indicate that at the histological, cellular, and molecular levels, amputation- and electroporation-induced dedifferentiation are virtually indistinguishable. This discovery allows dedifferentiation to be studied in the absence of the wound healing process that normally follows appendage amputation and may aid researchers in identifying genes required for the cellular plasticity response.

Materials and methods

Care of animals

Adult newts, Notophthalmus viridescens, were obtained from Charles Sullivan and housed in large tanks containing slow-flowing dechlorinated water. Newts were fed live California blackworms. All animal protocols were approved by the University of Utah Institutional Animal Care and Use Committee.

Amputations and collection of regenerating tissues

Newts were anesthetized by submersion in Tris buffered 0.1% tricaine (pH 7.3) for 10 minutes and then placed on ice. Limb amputations were performed through the stylopodium of the forelimb or hindlimb midway between the proximal and distal epiphyses of the humerus or femur. The bone was not trimmed following amputation. Tail amputations were performed 1 cm from the distal tip of the tail. Newts were allowed to recover on ice for 1 hour and then placed in shallow water with their heads above the surface until they recovered from anesthesia. Once conscious, the newts were housed in dechlorinated water until the regenerating tissues were collected for analyses.

Newts that were to be used to assess cell cycle reentry and histolysis were injected intraperitoneally with 0.1 ml of 10 mM BrdU 12 hours before sacrifice. Forelimb tissues were collected for this analysis at 1, 2, 3, 4, 5, 6, 7, 14, 21, 28, and 35 days postamputation. Hindlimb and tail tissues were collected at 14 days postamputation. Newts that were to be used for immunofluorescence or apoptosis assays were not injected with BrdU. To collect the regenerating tissues, the newts were anesthetized as described above and the limbs were reamputated at either the shoulder or hip. Tails were reamputated 1 cm proximal to the regenerating tip. The regenerating tissues from BrdU-injected newts were fixed overnight in Carnoy’s fixative (60% ethanol, 30% chloroform, 10% glacial acetic acid), while those tissues that were to be used for apoptosis assays were fixed in 4% paraformaldehyde in PBS overnight. Tissues fixed in Carnoy’s were then decalcified for 1 hour in 2 M HCl/PBS/0.5% Triton X-100, rinsed twice in PBS, dehydrated through a series of ethanol/PBS washes with increasing concentration of ethanol, treated with Hemo-De, and embedded in paraffin blocks. Tissues fixed in paraformaldehyde were embedded in paraffin blocks using the same procedure, except that the decalcification step was omitted. Tissues to be used for immunofluorescence assays were fixed in a paraformaldehyde-lysine-periodate fixative (0.05% paraformaldehyde, 100 mM lysine-HCl, 10 mM sodium periodate, 120 mM NaCl, 3 mM KCl, 10 μM CaCl2, 5 mM HEPES, pH 7.4) for 45 minutes, rinsed in wash solution (120 mM NaCl, 3 mM KCl, 10 μM CaCl2, 5 mM HEPES, pH 7.4) for 30 minutes, and frozen in O.C.T. for cryosectioning.

Application of electric fields to newt limbs and tails

Newts were anesthetized as described above and then gently strapped to an IC-Spacing perfboard (RadioShack) with Stretch Magic elastic cord (Helby Import Co.) to prevent movement of the limbs or tail during the application of the electric field. The newts were submerged in PBS and 3 mm Genetrodes (BTX) were placed parallel to the proximal-distal axis, one on the anterior side of the stylopodium and the other on the posterior side (Fig. 1). For tails, the electrodes straddled the tail and ran parallel to its proximal-distal axis. The gap between the electrodes was 3 mm for newt forelimbs and tails and 4 mm for newt hindlimbs. This allowed the electrodes to be placed such that they were not contacting the newt skin on either side of the limb or tail. Electric fields ranging from 33 to 167 V/cm were applied in five 100 msec pulses with 1 second between each pulse. Pilot studies had indicated that this was an effective range of electric field strengths for producing cellular electroporation, while not overtly damaging the limb tissue. Following the application of the electric field, the newts were placed on ice for 10 minutes and then allowed to recover from anesthesia as described above.

Fig. 1
Method used to apply electric fields to newt forelimbs. (A) Anesthetized newts were gently strapped to an IC-Spacing perfboard with Stretch Magic elastic cord (restraints). Restraints were applied to the neck, trunk, and both forelimbs. Newts were submerged ...

For ectopic EGFP expression, 3 μg of the expression construct pCMV-SPORT6-EGFP was injected in a 1 μl volume into the dorsal muscles of the stylopodium using a Drummond II Nanoject injector and a glass needle with a bore size of at least 60 μm. Electroporation was accomplished by pulsing using electric fields ranging from 33 to 167 V/cm electric field as described above. Limbs were monitored for EGFP expression over several weeks using a Zeiss M2Bio fluorescence Stemi SV 11 stereomicroscope and photographs were taken using a MicroMax cooled, high-performance digital camera (Princeton Instruments).

Collection of electroporated limbs and tails

Newts were injected with BrdU as described above when the collected tissues were to be used for assessing cell cycle reentry or histolysis. Time points for the collection of electroporated tissues were the same as those used for amputated tissues (see above). Limbs and tails were collected and either embedded in paraffin after fixing the tissues overnight in Carnoy’s fixative or 4% paraformaldehyde in PBS or embedded in O.C.T. after briefly fixing in the paraformaldehyde-lysine-periodate solution as described above.

Cell cycle reentry and histolysis assays

Decalcified tissues were sectioned at 10 μm and the paraffin was removed by washing the slides twice in Hemo-De for 10 minutes. The tissues were rehydrated in a series of solutions containing increasing amounts of PBS-0.5% Triton X-100 (PBSTx) and decreasing amounts of ethanol. Tissues were treated with 2 M HCl-PBSTx for 1 hour to denature the DNA and then neutralized for 1 minute in 100 mM sodium borate (pH 8.4). Following two rinses in PBS, endogenous peroxidases were inactivated by incubating the tissues in 1% hydrogen peroxide for 30 minutes and immunohistochemistry was performed according to the manufacturer’s instructions using anti-BrdU as the primary antibody (Chemicon; 1/350 dilution in PBS) and the Elite Mouse IgG Vectastain ABC Kit (Vector Laboratories). Tissues were incubated with the DAB substrate for 10 minutes, rinsed for 5 minutes in tap water and counter-stained for 10 seconds with a 1/3 dilution of Gill #3 hematoxylin. The tissues were rinsed for 5 minutes in running tap water, dipped 10 times in acid rinse solution (4 ml glacial acetic acid in 196 ml ddH2O), dipped 10 times in tap water, and incubated in bluing solution (3 ml NH4OH in 197 ml of 70% ethanol). The slides were dipped 10 times in tap water and then dehydrated through a series of solutions containing increasing amounts of ethanol. The tissues were placed in Hemo-De and mounted with Cytoseal.

Immunofluorescence assays

Frozen tissues were cryosectioned at 10 μm and immunofluorescence assays were performed as described by Kintner et al. (Kintner and Brockes, 1984) using the anti-12/101 (Kintner and Brockes, 1984) and anti-MT1 (Onda et al., 1991) mouse monoclonal antibodies (antibodies were obtained from the Developmental Studies Hybridoma Bank at the University of Iowa). The primary anti-12/101 (IgG) and anti-MT1 (IgM) antibodies were diluted 1/100 and 1/50, respectively, in 100 mM phosphate buffer (81 mM Na2HPO4 + 19 mM NaH2PO4) and a secondary Alexa 594-conjugated goat anti-mouse IgG or goat anti-mouse IgM antibody (Molecular Probes) was diluted 1/200. Primary antibodies were allowed to react with tissue sections for 45 minutes at room temperature and the secondary antibodies were incubated with the tissues for 30 or 45 minutes. The slides were mounted using the Slow Fade Light Anti-fade Kit with DAPI according to manufacturer’s instructions (Molecular Probes).

TUNEL assay for apoptosis and histological examination for necrosis

Paraformaldehyde-fixed and paraffin-embedded tissues were sectioned at 10 μm, deparaffinized by Hemo-De, and gradually rehydrated in PBS. Given that these tissues were not decalcified, only regions containing soft tissues (e.g., epidermis, dermis, and muscle) were sectioned. TUNEL assays were performed according to manufacturer’s instructions using the In Situ Cell Death Detection Kit, POD (Roche). Briefly, tissue sections were incubated in 3% H2O2 in methanol for 10 minutes to inactivate endogenous peroxidases, washed in PBS, and permeabilized with ~20 μg/ml recombinant PCR grade proteinase K (Roche) for 15 minutes at room temperature. The tissues were washed twice with PBS and incubated in TUNEL reaction mixture for 1 hour at 37°C in a humidified chamber. Tissues were washed in PBS and labeling was detected using the Converter-POD (30 minutes) and DAB (5–10 minutes) solutions. Tissue sections were counter stained with diluted Gill #3 hematoxylin as described above.

Microarray analysis

Microarray slides containing quadruplicate spots of 521 cDNA fragments representing 366 regeneration-enriched newt genes were prepared at the Huntsman Cancer Institute Microarray Core Facility at the University of Utah. Newt limbs were either amputated or electroporated as described above and the tissues were collected at 1, 3, and 5 days postinjury and flash frozen in liquid nitrogen. Intact, nonelectroporated limbs served as controls. Limbs were pooled for each time point and ground to fine powder with a mortar and pestle while cooled in liquid nitrogen. The limb powder was transferred to Trizol (Invitrogen) and total RNA was extracted according to the manufacturer’s instructions. The integrity of the total RNA was assessed by formaldehyde agarose gel electrophoresis. The RNA was amplified and purified using a RiboAmp RNA amplification procedure from Arcturus Engineering. Probes were prepared by incorporation of either Cy-3 or Cy-5 fluorescent dyes during reverse transcription of the amplified RNA templates. Hybridization was performed at 42°C for 24 hours in a solution containing 50% formamide, 5× SSC, 5× Denhardt’s, 0.1% SDS, and labeled probe/Cot-1 DNA (final concentration, 0.13 mg/ml) using lifter cover slips. Following hybridization, the slides were washed as follows in a series of solutions with increasing stringency: 1) 1× SSC, 0.2% SDS for 5 minutes at 42°C; 2) 0.1× SSC, 0.2% SDS for 5 minutes at 52°C; 3) 0.1× SSC, 0.2% SDS for 10 minutes at 52°C; and 4) four 2 minute room temperature washes in 0.02× SSC. The Cy-3 and Cy-5 signals were detected by scanning the slides using a GenePix 4000B Microarray Scanner (Axon Instruments) with analysis by ImaGene 5.5 (BioDiscover, Inc.). After removing intensity data from low-quality spots, the values of remaining spots for each cDNA fragment were averaged, and log (base 2) ratios were calculated between the treated and control samples. No significant spatial variation was observed on any of the microarrays. The median intensity value for each microarray was nearly identical, so no further normalization was performed. Spots that had a signal less than 20% greater than background signal were eliminated from further analysis. Only those cDNAs that produced an appropriate signal at all three time points were included in the final analyses. The data were analyzed using Spotfire DecisionSite for Functional Genomics (Spotfire, Inc.), which allows for statistical analyses and comparison of data from multiple experiments.

Real-time RT-PCR

Reverse transcription of total RNA was performed according to manufacturer’s instructions using the Bio-Rad iScript cDNA synthesis Kit and a mixture of poly-dT and random primers. Specific gene primers for real-time PCR were designed using the Beacon Designer 3.01 program (Premier Biosoft International). Sequences and amplification conditions are shown in Table 1. Real-time PCR was performed on an ABI Prism 7700 Sequence Detection System using the iTaq SYBR Green Supermix with ROX (Bio-Rad) for detection of amplified DNA. Data were analyzed using the standard curve method (Applied Biosystems, Inc.) and normalized to the control gene histone acetyltransferase 1. Histone acetyltransferase 1 was selected from a group of 10 genes that had exhibited the least variation between time points based on microarray and/or northern blot analyses. Normalized values were then converted to relative values by using the intact nonelectroporated newt limb controls as the calibrator.

Table 1
Sequences of primers used for real-time RT-PCR

RNA in situ hybridization

RNA In situ hybridization was performed as previously described using antisense and sense digoxigenin-labeled riboprobes for nCol, MMP3/10a, MMP3/10b, and MMP9 (Vinarsky et al., 2005).


Electric fields induce dedifferentiation in intact newt limbs and tails

When newt forelimbs were subjected to five 100 msec pulses of 50 V (Electric field strength = Applied voltage/distance between electrodes in cm = 50V/0.3 cm = 167 V/cm), internal limb tissues began to dedifferentiate in a manner resembling the dedifferentiation that occurs following limb amputation (Fig. 2). Starting at about 5 days postelectropulse, cells within the muscle tissue, periosteum, and dermal layer of the skin began to reenter the cell cycle, while myofibers appeared to cleave to form either smaller myofibers or mononucleated cells (Fig. 2C, E, G, I). The timing and sequence of these cellular events closely follows those events that occur during the dedifferentiation phase of limb regeneration (Fig. 2B, D, F, H). This suggests that the application of an electric field to the newt forelimb induced a dedifferentiation response in the intact limb that was histologically indistinguishable from amputation-induced dedifferentiation. The dedifferentiation response continued through at least day 14 (Fig. 2B–G) and by day 21, the myofibers were beginning to re-form in the electropulsed limbs. There was still evidence of DNA synthesis in many internal cells, especially in the periosteal cells that lined the bone and in cells residing between the myofibers (Fig. 2I).

Fig. 2
Cell cycle reentry and general histolysis in newt forelimbs following application of an electric field. Newts were injected intraperitoneally with BrdU approximately 12 hours before collecting the limbs for examination. (A) An intact nonelectroporated ...

To determine whether the application of electric fields could induce a dedifferentiation response in other regeneration-competent appendages, we performed similar experiments on the newt hindlimb and tail. At day 14 postinjury, both the amputated and electrically-stimulated hindlimbs and tails contained internal cells that had reentered the cell cycle as evidenced by BrdU labeling (Fig. 3). In each case, the dedifferentiation responses following amputation and electropulsing were histologically indistinguishable, suggesting that the electrically-induced dedifferentiation response is universal in the regeneration-competent appendages of the newt.

Fig. 3
Application of an electric field induces cell cycle reentry and general histolysis in newt hindlimbs and tails. (A and D) Intact, nonelectroporated hindlimb or tail control, respectively. (B and E) Hindlimb and tail, respectively, 14 days postamputation. ...

Another hallmark of dedifferentiation is that markers of differentiation are downregulated during the process (Kintner and Brockes, 1984; McGann et al., 2001; Odelberg et al., 2000), while blastemal markers are upregulated (Kintner and Brockes, 1984; Onda et al., 1991). We used the myogenic differentiation marker 12/101 (Griffin et al., 1987; Kintner and Brockes, 1984) to determine whether the application of an electric field reduced the expression of this protein in newt forelimbs. Fig. 4 compares the reduction of 12/101 in amputated newt forelimbs (Fig. 4B) to the reduction observed following the application of electrical pulses (Fig. 4C–E). By day 7 postelectropulsing, there is a marked reduction in 12/101 antigen and this reduction is even more noticeable at days 10 and 14. By day 21, the 12/101 marker is starting to return as the myofibers are beginning to re-form (Fig. 4F), a result that is consistent with the BrdU/histolysis studies shown in Fig. 2. In contrast, the blastemal marker MT1, which is an epitope on the extracellular matrix protein tenascin (Onda et al., 1990; Onda et al., 1991), is upregulated in the internal tissues within 10 days of limb amputation or electropulsing and remains upregulated through at least day 14 (Fig. 4G–I). These results are consistent with the BrdU incorporation studies in demonstrating that the application of electric fields to newt appendages induces a dedifferentiation response.

Fig. 4
Regulation of muscle differentiation and blastemal markers following the application of an electric field to newt forelimbs. (A) Immunofluorescence assay on intact nonelectroporated control limbs demonstrated high expression of 12/101 in myofibers (red ...

A time course experiment using hematoxylin and eosin stained tissue sections (Fig. 5) also demonstrated that histolysis reached a peak at 7–14 days after the application of an electric field. Breakdown of the muscle tissue is quite evident by day 7 and the number of nuclei have dramatically increased by day 14. Dermal tissues are also affected during the period of histolysis and appear to exhibit both a decrease in the number skin glands and a disorganization of the remaining glands (Fig 5A, B). By day 21, the myofibers are starting to re-form and by day 35, the electrically pulsed limbs are often indistinguishable from intact control limbs (Fig. 5E, F).

Fig. 5
A time course for complete regeneration of tissue structure following application of an electric field. Hematoxylin and eosin-stained tissue sections from newt forelimbs were taken at weekly intervals following the application of an electric field. (A ...

We performed TUNEL assays and searched for signs of necrotic myofibers to determine whether the apparent dedifferentiation events following the application of an electric field were associated with increases in cell death due to apoptosis or necrosis (Fig. 6). The TUNEL assays revealed the presence of only a few apoptotic cells following either amputation or application of an electric field (Fig. 6A–D) and histological examination showed no evidence of necrosis, such as weak cytoplasmic staining with eosin or the infiltration of macrophages or other leukocytes (Fig. 6E). During the breakdown of the myofibers, many of the resulting cells contained a single nucleus surrounded by cytoplasm that stained brightly with eosin. Mitotic figures were also observed during myofiber breakdown. As shown above, myofiber nuclei reentered the cell cycle (Figs. 2 and and3)3) and some myofibers continued to express the 12/101 marker during the early stages of cleavage (Fig. 4C, D). In addition, the blastemal marker MT1 was upregulated during the dedifferentiation process (Fig. 4G–I). These indicators of cell survival and growth during the early stages of electrically-induced histolysis are inconsistent with marked cell necrosis. Instead, our results strongly suggest that electrical pulses delivered to intact newt limbs initiate a dedifferentiation response that is histologically and cellularly indistinguishable from the response elicited by limb amputation.

Fig. 6
Application of low level electric fields causes little cell death in newt forelimbs. TUNEL assays were performed to assess level of apoptosis following application of an electric field. Brown nuclei represent cells positive for the TUNEL assay, while ...

Correlation between electroporation and dedifferentiation

If the strength of an electric field reaches a certain threshold, it can produce the formation of transient pores in the cell membrane, a process known as electroporation. Electroporation is often used to deliver molecules such as DNA into cells. During the electrical pulse, the negatively charged DNA molecules will be driven towards the anode and through pores that have formed in the cell membranes. This electrophoretic effect has been shown to be essential for efficient DNA transfection, given that DNA plasmids added after the administration of the electrical pulse but before closure of the transient pores do not transfect cells (Golzio et al., 2002; Mir et al., 1999; Sukharev et al., 1992). The existence of the transient pores can be measured in minutes. For example, in mouse and rat myofibers, most of these pores have resealed in about 9 minutes (Bier et al., 1999; Gehl et al., 2002).

To determine whether electroporation of limb cells was required for the dedifferentiation process, we injected newt limbs with an EGFP expression construct and applied five electrical pulses at varying electrical field strengths ranging from 33 V/cm to 167 V/cm. We were able to observe very low levels of EGFP expression when electric field strengths reached 67 V/cm and the expression levels increased as the electric fields increased to 167 V/cm, indicating that electroporation of the limb cells begins at electric field strengths of about 67 V/cm and increases with the application of higher electric field strengths (Fig. 7A, C, and E). When we performed BrdU incorporation assays on either the same limbs or other limbs pulsed with equivalent voltages, we observed cell cycle reentry in the periosteal and muscle cells starting at electric fields of 67 V/cm and increasing in number as the strength of the electric fields increased to 167 V/cm (Fig. 7B, D, and F). Myofiber breakdown was apparent at the higher electric field strengths, but not at 67 V/cm. The correlation between electric field strengths required to produce electroporation as measured by transgene expression and those required to induce a dedifferentiation response suggests that the minor and transient injury that occurs in cells following application of a series of electrical pulses is sufficient to induce the genetic programs that regulate dedifferentiation in the newt.

Fig. 7
Dedifferentiation correlates with electric field strengths sufficient to cause electroporation of cell membranes. Newt limbs were injected with an EGFP expression construct followed by the application of electrical pulses at varying electric field strengths. ...

Analysis of temporal gene expression during amputation- and electroporation-induced dedifferentiation

Given that the dedifferentiation responses following amputation and electroporation were indistinguishable at the histological and cellular levels, we next examined whether the two processes were defined by similar gene expression profiles. Similar profiles would strongly suggest that both types of dedifferentiation are controlled by the same molecular signals and pathways. For these studies, we examined gene expression at three different time points following newt forelimb amputation or electroporation. The time points chosen were 1, 3, and 5 days post-injury. By performing microarray analyses on amputated and electroporated newt forelimbs using an in-house regeneration-enriched cDNA chip, we found that 153 of 203 genes (75.4%) exhibited similar expression patterns (less than 2-fold difference) at all three time points following either amputation or electroporation, while only 7 genes (3.4%) exhibited greater than 2-fold difference in expression at all three time points. Of these 153 genes, 35 (22.9%) were up- or down-regulated at least a 2-fold when compared to intact limbs (Table 2). Using the less stringent criterion where expression data between the two types of injury models only had to match (less than 2-fold difference) at two of three time points, 175 of 203 (86.2%) genes were deemed to have similar expression patterns.

Table 2
Differentially-expressed genes that exhibited similar or markedly different differential expression patterns following amputation and electroporation as determined by microarray analysis

In some cases, genes that were highly upregulated following both amputation and electroporation did not meet our stringent criteria for inclusion, because either electroporation or amputation induced a greater than 2-fold difference in the level of expression between the two types of injuries. For example, MMP9, which was highly upregulated following both types of injuries, exhibited a much higher level of expression following electroporation than amputation at two of the three time points, while MMP3/10a exhibited higher expression at one of the three time points. Therefore, these genes were not included in Table 2, although their expression patterns suggest that they play important roles in the response to both types of injuries. All 7 genes that showed markedly different expression patterns were expressed at much higher levels following limb amputation than following electroporation, suggesting that they either might be involved in wound closure or be regulated by physiological conditions that normally follow amputation, such as hypoxia.

Of the 35 genes that exhibited similar expression patterns between the two types of injuries, 19 of them could be placed into known gene or EST families based on sequence analysis. The remaining 16 genes fell into one of three categories: 1) genes that were novel; 2) genes of unknown function; or 3) cDNAs that were too short to make any assignments. Of the 19 known genes, 3 encoded proteases with 2 of these, nCol (AY857753) and MMP3/10b (AY857754), belonging to the matrix metalloproteinase (MMP) family and the remaining gene being the cysteine protease cathepsin L. As noted above, two additional MMP genes, MMP3/10a (AY857751) and MMP9 (AY857752), were also highly upregulated following both amputation and electroporation, but did not meet our criteria for inclusion. Some of the other known genes encode protease inhibitors, receptors, antimicrobial proteins, and putative anti-inflammatory agents.

The seven genes exhibiting markedly different expression patterns between the two injury models make up a mixed group encoding the muscle proteins parvalbumin, cardiac α-actin, troponin C, and genes that can be upregulated by hypoxia, including those encoding the glycolytic enzymes enolase 3 or 1 and fructose-bisphosphate aldolase I (Discher et al., 1998; Ferry et al., 1983; Kouno et al., 2000; Semenza et al., 1996). The upregulation of hypoxia-induced genes following amputation is consistent with previous studies that have shown a decrease in the number of blood vessels just proximal to the amputation plane during the first week following limb amputation (Peadon and Singer, 1966; Rageh et al., 2002; Smith and Wolpert, 1975). We have observed no such decrease in the number of blood vessels following electroporation and therefore would not expect to observe hypoxia-induced gene upregulation in electroporated limbs.

Real-time RT-PCR on a selected group of 24 genes (histone acetyltransferase 1 was used as the control gene--see Materials and methods) confirmed the results of the microarray analyses, while giving us a more dynamic assessment of the degree of differential expression for these selected genes (Yuen et al., 2002). For this analysis, we chose both genes that exhibited marked upregulation following injury as well as genes that showed very little variation in expression levels. The 11 genes that exhibited at least a 2-fold upregulation following both electroporation and amputation when analyzed by microarray analysis were also shown to be upregulated more than 2-fold by real-time RT-PCR (first 11 genes in Table 3), while the 10 genes that exhibited little change in expression by microarray analysis also showed little variability when analyzed by real-time RT-PCR (last 10 genes in Table 3). There were small discrepancies between the microarray and real-time RT-PCR results for 3 of the 24 genes. For 174a, microarray analysis detected a small increase (<2-fold) in mRNA levels following amputation and a >2-fold increase following electroporation. Real-time RT-PCR, however, showed that 174a is upregulated >2-fold following both types of injuries, a result that is consistent with northern blot analysis (data not shown). Microarray analysis indicated that 193c and profilin 2 were upregulated >2-fold following amputation but did not reach this threshold following electroporation. However, real-time RT-PCR demonstrated that both of these genes were upregulated >2-fold following both types of injuries. These results suggest our microarray data provide conservative estimates for the degree of differential expression following limb injury. Therefore, any gene we identify as being up- or down-regulated by at least 2-fold by microarray analysis is likely to be differentially expressed.

Table 3
Relative differential expression of selected genes following amputation and electroporation as determined by real-time RT-PCR

Real-time RT-PCR also revealed underlying differences in the differential expression levels of some genes following amputation or electroporation. However, when these differences occurred, they were often due to a slight shift in the timing of gene expression. For example, MMP3/10a exhibited very high levels of gene expression within 1 day of limb amputation or electroporation. By day 3, the levels had dropped considerably in the amputated limbs but remained very high following electroporation. By day 5, however, expression levels were greatly reduced in both injury types. Whether these differences in temporal expression have biological significance has yet to be determined.

Analysis of spatial gene expression during amputation- and electroporation-induced dedifferentiation

To further compare the dedifferentiation responses between amputated and electroporated limbs, we examined spatial expression patterns of selected genes in the two injury model systems. We have previously demonstrated that matrix metalloproteinases (MMPs) are required for normal newt limb regeneration and that at least three different MMPs, nCol, MMP3/10b, and MMP9 are upregulated in the internal limb tissues that undergo dedifferentiation during the early stages of limb regeneration (Vinarsky et al., 2005). To determine whether these genes are upregulated in the same tissues following electroporation, we performed RNA in situ hybridization on limb tissue sections at days 1, 3, and 5 postelectroporation and compared the spatial expression patterns to those observed in amputated limbs (Fig. 8). nCol and MMP9 were expressed in the periosteum and epithelia (skin and/or apical epithelial cap) following both types of injuries, while MMP3/10b was expressed in the epithelium and in the muscle tissues that will be undergoing dedifferentiation. MMP9 was also expressed in the endosteal cells following either amputation or electroporation. The common spatial expression patterns of these important regeneration genes further strengthens the argument that amputation- and electroporation-induced dedifferentiation are nearly identical processes controlled by the same genetic program.

Fig. 8
Spatial expression patterns of upregulated genes are similar following amputation or electroporation. RNA in situ hybridization of amputated and electroporated forelimb using riboprobes directed towards three MMP genes revealed that these genes were expressed ...


We present evidence demonstrating that electric fields sufficient to cause electroporation of newt limb cells can induce a dedifferentiation response in limbs and tails that is virtually indistinguishable at the histological and molecular levels from the dedifferentiation response following appendage amputation. The data suggest that dedifferentiation is most likely a result of the widespread, but transient opening of pores in the cell membrane following the application of the electric field. This minor, transient injury is apparently sufficient to initiate the genetic program that leads to the dedifferentiation of cells in the injured appendage. We reach this conclusion based on the correlation between the electric field strength required to induce both dedifferentiation and electroporation. An alternative explanation would be that the electric fields themselves might induce this response either by activating ion channels directly or by releasing growth factors and cytokines from the extracellular matrix (ECM). These signals could then lead to the activation or repression of downstream genes that regulate the dedifferentiation response. The ion channel hypothesis would be consistent with previous work demonstrating the importance of electrical current and sodium ion channels during the initial stages of limb regeneration (Borgens et al., 1979; Jenkins et al., 1996), while the release of cytokines from the ECM following electrical stimulation has been demonstrated for other systems (Braunhut et al., 2004; Zhou et al., 1998). These latter explanations would be satisfying, because they could help bridge the gap between two historically different ways of viewing the initiation of salamander limb regeneration, i.e., through the regulation of genes or through currents controlled by ion channels.

Although the dedifferentiation processes following amputation and electroporation are nearly indistinguishable, only amputation leads to the regrowth of a new appendage. This suggests that the signals required for appendage outgrowth and/or patterning of a new limb are not present following electroporation. Instead, the minor and transient injury caused by the electrical pulses induce a robust dedifferentiation response that the newt efficiently resolves by regenerating the normal internal tissues of the original appendage. This regenerative process is similar to the response that is observed following a mild crush injury in larval axolotls (Mescher, 1982) and appears to be an efficient method for regenerating tissues following an injury that does not involve the complete loss of an appendage. In contrast, a different result can occur when the injury is more severe and involves the severing and deviation of the brachial nerves, removal of large patches of skin, and the destruction of muscle tissue. When such injuries are located near the shoulder, newts often respond by growing supernumerary limbs (Bodemer, 1958; Bodemer, 1959). Supernumerary limbs can also be produced at the site of limb wounds by combining nerve transection and deviation with the juxtaposition of two pieces of skin with opposite axial limb polarities (Endo et al., 2004; Lheureux, 1977; Maden and Mustafa, 1984; Reynolds et al., 1983). The insertion of carcinogenic microcrystals into connective tissues beneath the surface of the skin on the newt forelimb can occasionally induce supernumerary limb formation (Tsonis and Eguchi, 1981). Presumably, the induction of the new limb is a result of the injury created by the insertion of the microcrystal through the skin coupled with the extended effects of the carcinogen.

Another study has recently presented evidence suggesting that the combination of plasmid injection and application of an electric field to axolotl tail myofibers can induce dedifferentiation in 5–10% of the electroporated axolotl myofibers (Schnapp and Tanaka, 2005). In the present study, we demonstrate that the application of an electric field to newt appendages in the absence of any other injury, such as injection, can induce widespread dedifferentiation that is indistinguishable from the response observed following limb amputation. Our results also suggest that limb tissues may be more susceptible to the effects of electroporation than tail tissues (compare Figs. 2 and and3)3) and this difference might explain, at least in part, the low percentage of tail myofibers that dedifferentiated in the axolotl study. Other explanations might include species differences in the response to electrical stimulation and differences in methodology.

Our gene expression studies revealed several potential dedifferentiation/cellular plasticity genes, including four members of the MMP family. Two of these genes, nCol and MMP3/10b, exhibited comparable differential expression patterns between amputated and electroporated forelimbs at all time points examined, while the other two genes, MMP9 and MMP3/10a, were highly upregulated following both types of injuries but exhibited significantly higher expression levels at one or more time points in intact electroporated limbs. We have previously demonstrated that MMP function is required for normal newt limb regeneration and that nCol, MMP3/10b, and MMP9 are expressed in the early stages of regeneration in tissues that will undergo dedifferentiation (Vinarsky et al., 2005). In this study, we show that following the application of an electric field these same MMP genes are also expressed in tissues that will be undergoing dedifferentiation. These results further support the hypothesis that the MMP genes are involved in the dedifferentiation process. Whether MMPs play an active role in this process by either activating signaling proteins or releasing dedifferentiation-initiating cytokines from the extracellular matrix (ECM) or whether they play a more permissive role by remodeling the ECM has not yet been determined.

Finally, this study indicates that caution must be exercised when interpreting results where electroporation has been used in vivo for transfecting expression constructs. This is especially true if the experiments are designed to examine various aspects of cellular plasticity, e.g., dedifferentiation, cell cycle reentry, or multipotency. Such responses could be a consequence of gene activation by cellular electroporation, rather than the effects of transgene expression. Caution is also warranted in cases where the transgene acts only as a marker for following cell fate, given that electroporation might affect the potency of the transfected cell(s). In future in vivo studies involving electroporation, it will be important to implement controls that assess the effects of electroporation alone.

In conclusion, our data demonstrate that application of an electric field sufficient to induce transient electroporation of cell membranes induces a dedifferentiation response that is virtually indistinguishable from the response that occurs following amputation of newt appendages. This discovery allows us to predict whether a gene that is differentially expressed following amputation will likely function in the dedifferentiation process. It also provides a possible method for determining whether a gene is required for dedifferentiation by knocking down its function in intact newt appendages using electroporation-delivered morpholinos or RNA interference.


We would like to thank George Eisenhoffer, Alejandro Sanchez Alvarado, Nestor Oviedo, and Kevin Flanigan for helpful suggestions and Brian Dalley and the Huntsman Microarray Core Facility for their technical assistance with the microarray analyses. We also thank David Kent and Katherine Zukor for technical assistance. This work was funded by grants from the National Institute of Neurological Disorders and Stroke (Grant Numbers R01 NS043878 and R01 NS043878S) (SJO).


  • Bier M, Hammer SM, Canaday DJ, Lee RC. Kinetics of sealing for transient electropores in isolated mammalian skeletal muscle cells. Bioelectromagnetics. 1999;20:194–201. [PubMed]
  • Bodemer CW. The development of nerve-induced supernumerary limbs in the adult newt, Triturus viridescens. J Morphol. 1958;102:555–582.
  • Bodemer CW. Observations on the mechanism of induction of supernumerary limbs in adult Triturus viridescens. J Exp Zool. 1959;140:79–99. [PubMed]
  • Bodemer CW, Everett NB. Localization of newly synthesized proteins in regenerating newt limbs as determined by radioautographic localization of injected methionine-S35. Dev Biol. 1959;1:327–342.
  • Borgens RB, Vanable JW, Jr, Jaffe LF. Reduction of sodium dependent stump currents disturbs urodele limb regeneration. J Exp Zool. 1979;209:377–386. [PubMed]
  • Braunhut SJ, McIntosh D, Vorotnikova E, Zhou T, Marx KA. Development of a smart bandage: applying electrical potential to selectively release wound healing growth factors from cell free extracellular matrices. Material Research Society Journal. 2004;711:1–3.
  • Breedis C. Induction of accessory limbs and of sarcoma in the Newt (Triturus viridescens) with carcinogenic substances. Cancer Res. 1952;12:861–866. [PubMed]
  • Brockes JP, Kumar A. Plasticity and reprogramming of differentiated cells in amphibian regeneration. Nat Rev Mol Cell Biol. 2002;3:566–574. [PubMed]
  • Butler EG, Ward MB. Reconstitution of the spinal cord after ablation in adult Triturus. Dev Biol. 1967;15:464–486. [PubMed]
  • Chalkley DT. A quantitative histological analysis of forelimb regeneration in Triturus viridescens. J Morphol. 1954;94:21–70.
  • Della Valle P. La doppia rigenerazione inversa nella fratture della zampe di Triton. Boll. Soc. Nat. Napoli. 1913;25:95–160.
  • Discher DJ, Bishopric NH, Wu X, Peterson CA, Webster KA. Hypoxia regulates beta-enolase and pyruvate kinase-M promoters by modulating Sp1/Sp3 binding to a conserved GC element. J Biol Chem. 1998;273:26087–26093. [PubMed]
  • Endo T, Bryant SV, Gardiner DM. A stepwise model system for limb regeneration. Dev Biol. 2004;270:135–145. [PubMed]
  • Ferry JA, Nichols RC, Condon SJ, Stubbs JD, Bowen ST. Artemia hemoglobins. Increase in net synthesis of the beta-polypeptide (relative to the alpha-polypeptide) in hypoxia. Biochim Biophys Acta. 1983;739:249–257. [PubMed]
  • Gehl J, Skovsgaard T, Mir LM. Vascular reactions to in vivo electroporation: characterization and consequences for drug and gene delivery. Biochim Biophys Acta. 2002;1569:51–58. [PubMed]
  • Golzio M, Teissie J, Rols MP. Direct visualization at the single-cell level of electrically mediated gene delivery. Proc Natl Acad Sci U S A. 2002;99:1292–1297. [PMC free article] [PubMed]
  • Griffin KJ, Fekete DM, Carlson BM. A monoclonal antibody stains myogenic cells in regenerating newt muscle. Development. 1987;101:267–277. [PubMed]
  • Hay ED, Fischman DA. Origin of the blastema in regenerating limbs of the newt Triturus viridescens. An autoradiographic study using tritiated thymidine to follow cell proliferation and migration. Dev Biol. 1961;3:26–59. [PubMed]
  • Jenkins LS, Duerstock BS, Borgens RB. Reduction of the current of injury leaving the amputation inhibits limb regeneration in the red spotted newt. Dev Biol. 1996;178:251–262. [PubMed]
  • Kintner CR, Brockes JP. Monoclonal antibodies identify blastemal cells derived from dedifferentiating limb regeneration. Nature. 1984;308:67–69. [PubMed]
  • Kouno A, Inoue H, Bajanowski T, Maeno Y, Iwasa M, Nakayama M, Nishi K, Brinkmann B, Matoba R. Development of haemoglobin subtypes and extramedullary haematopoiesis in young rats. Effects of hypercapnic and hypoxic environment. Int J Legal Med. 2000;114:66–70. [PubMed]
  • Lheureux E. [Importance of limb tissue associations in the development of nerve-induced supernumerary limbs in the newt Pleurodeles waltlii Michah (author’s transl)] J Embryol Exp Morphol. 1977;38:151–173. [PubMed]
  • Maden M, Mustafa K. The cellular contributions of blastema and stump to 180 degrees supernumerary limbs in the axolotl. J Embryol Exp Morphol. 1984;84:233–253. [PubMed]
  • McGann CJ, Odelberg SJ, Keating MT. Mammalian myotube dedifferentiation induced by newt regeneration extract. Proc Natl Acad Sci U S A. 2001;98:13699–13704. [PMC free article] [PubMed]
  • Mescher AL. Neurotrophic control of events in injured forelimbs of larval urodeles. J Embryol Exp Morphol. 1982;69:183–192. [PubMed]
  • Mir LM, Bureau MF, Gehl J, Rangara R, Rouy D, Caillaud JM, Delaere P, Branellec D, Schwartz B, Scherman D. High-efficiency gene transfer into skeletal muscle mediated by electric pulses. Proc Natl Acad Sci U S A. 1999;96:4262–4267. [PMC free article] [PubMed]
  • O’Steen WK. Regeneration of the intestine in adult urodeles. J Morphol. 1958;103:435–477.
  • Odelberg SJ, Kollhoff A, Keating MT. Dedifferentiation of mammalian myotubes induced by msx1. Cell. 2000;103:1099–1109. [PubMed]
  • Onda H, Goldhamer DJ, Tassava RA. An extracellular matrix molecule of newt and axolotl regenerating limb blastemas and embryonic limb buds: immunological relationship of MT1 antigen with tenascin. Development. 1990;108:657–668. [PubMed]
  • Onda HLPM, Tassava RA, Chiu IM. Characterization of a newt tenascin cDNA and localization of tenascin mRNA during newt limb regeneration by in situ hybridization. Developmental Biology. 1991;148:219–232. [PubMed]
  • Peadon AM, Singer M. The blood vessels of the regenerating limb of the adult newt, Triturus. J Morphol. 1966;118:79–89. [PubMed]
  • Rageh MA, Mendenhall L, Moussad EE, Abbey SE, Mescher AL, Tassava RA. Vasculature in pre-blastema and nerve-dependent blastema stages of regenerating forelimbs of the adult newt, Notophthalmus viridescens. J Exp Zool. 2002;292:255–266. [PubMed]
  • Reynolds S, Holder N, Fernandes M. The form and structure of supernumerary hindlimbs formed following skin grafting and nerve deviation in the newt Triturus cristatus. J Embryol Exp Morphol. 1983;77:221–241. [PubMed]
  • Schnapp E, Tanaka EM. Quantitative evaluation of morpholino-mediated protein knockdown of GFP, MSX1, and PAX7 during tail regeneration in Ambystoma mexicanum. Dev Dyn. 2005;232:162–170. [PubMed]
  • Semenza GL, Jiang BH, Leung SW, Passantino R, Concordet JP, Maire P, Giallongo A. Hypoxia response elements in the aldolase A, enolase 1, and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor 1. J Biol Chem. 1996;271:32529–32537. [PubMed]
  • Smith AR, Wolpert L. Nerves and angiogenesis in amphibian limb regeneration. Nature. 1975;257:224–225. [PubMed]
  • Sukharev SI, Klenchin VA, Serov SM, Chernomordik LV, Chizmadzhev Yu A. Electroporation and electrophoretic DNA transfer into cells. The effect of DNA interaction with electropores. Biophys J. 1992;63:1320–1327. [PMC free article] [PubMed]
  • Thornton CS. The histogenesis of muscle in the regenerating fore limb of larval amblystoma punctatum. J Morphol. 1938a;62:17–47.
  • Thornton CS. The histogenesis of the regenerating forelimb of larval Amblystoma after exarticulation of the humerus. J Morphol. 1938b;62:219–235.
  • Tsonis PA. Limb Regeneration. Cambridge University Press; New York: 1996.
  • Tsonis PA, Eguchi G. Carcinogens on regeneration. Effects of N-methyl-N′-nitro-N-nitrosoguanidine and 4-nitroquinoline-1-oxide on limb regeneration in adult newts. Differentiation. 1981;20:52–60. [PubMed]
  • Turner JE, Singer M. An electron microscopic study of the newt (Triturus viridescens) optic nerve. J Comp Neurol. 1974a;156:1–18. [PubMed]
  • Turner JE, Singer M. The ultrastructure of regeneration in the severed newt optic nerve. J Exp Zool. 1974b;190:249–268. [PubMed]
  • Vinarsky V, Atkinson DL, Stevenson TJ, Keating MT, Odelberg SJ. Normal newt limb regeneration requires matrix metalloproteinase function. Dev Biol. 2005;279:86–98. [PubMed]
  • Wallace H. Vertebrate Limb Regeneration. John Wiley & Sons; New York: 1981.
  • Yuen T, Wurmbach E, Pfeffer RL, Ebersole BJ, Sealfon SC. Accuracy and calibration of commercial oligonucleotide and custom cDNA microarrays. Nucleic Acids Res. 2002;30:e48. [PMC free article] [PubMed]
  • Zhou T, Braunhut SJ, Medeiros D, Marx KA. Potential dependent endothelial cell adhesion, growth and cytoskeletal rearrangement. Material Research Society Journal. 1998;489:211–216.
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...