Cell Physiology

Physiological and histological changes in skeletal muscle following in vivo gene transfer by electroporation

Joseph A. Roche, Diana L. Ford-Speelman, Lisa W. Ru, Allison L. Densmore, Renuka Roche, Patrick W. Reed, Robert J. Bloch


Electroporation (EP) is used to transfect skeletal muscle fibers in vivo, but its effects on the structure and function of skeletal muscle tissue have not yet been documented in detail. We studied the changes in contractile function and histology after EP and the influence of the individual steps involved to determine the mechanism of recovery, the extent of myofiber damage, and the efficiency of expression of a green fluorescent protein (GFP) transgene in the tibialis anterior (TA) muscle of adult male C57Bl/6J mice. Immediately after EP, contractile torque decreased by ∼80% from pre-EP levels. Within 3 h, torque recovered to ∼50% but stayed low until day 3. Functional recovery progressed slowly and was complete at day 28. In muscles that were depleted of satellite cells by X-irradiation, torque remained low after day 3, suggesting that myogenesis is necessary for complete recovery. In unirradiated muscle, myogenic activity after EP was confirmed by an increase in fibers with central nuclei or developmental myosin. Damage after EP was confirmed by the presence of necrotic myofibers infiltrated by CD68+ macrophages, which persisted in electroporated muscle for 42 days. Expression of GFP was detected at day 3 after EP and peaked on day 7, with ∼25% of fibers transfected. The number of fibers expressing green fluorescent protein (GFP), the distribution of GFP+ fibers, and the intensity of fluorescence in GFP+ fibers were highly variable. After intramuscular injection alone, or application of the electroporating current without injection, torque decreased by ∼20% and ∼70%, respectively, but secondary damage at D3 and later was minimal. We conclude that EP of murine TA muscles produces variable and modest levels of transgene expression, causes myofiber damage due to the interaction of intramuscular injection with the permeabilizing current, and that full recovery requires myogenesis.

  • intramuscular injection
  • muscle transfection
  • muscle damage
  • myogenesis
  • green fluorescent protein

electroporation (EP) of genetic material such as DNA or small interfering RNA is commonly used to alter gene expression or protein levels in skeletal muscle (4, 31, 36). EP is more effective than intramuscular or intravenous injection alone (2, 22, 24), but its effects on the overall health of muscle have not been studied in detail. Several groups have suggested that myofiber damage is an undesired consequence of EP irrespective of the transgene that is studied and have explored modifications of the technique to minimize damage (14, 19, 22, 33). Here we investigate the loss and recovery of contractile function and normal muscle morphology over several hours through several weeks following EP of the tibialis anterior (TA) muscle of young adult mice. To study the efficiency of transgene expression, we injected DNA encoding green fluorescent protein (GFP) before EP. We chose GFP for this study, as it has been extensively used as a reporter in skeletal muscle (22, 25), and there are no reports of GFP expression by itself being harmful to myofibers. We also examined the role of myogenesis in the recovery of skeletal muscle from damage resulting from EP by ablating satellite cells with a single dose of X-irradiation. Our results show that EP damages large muscles like the TA both functionally and morphologically, and that recovery from this damage involves inflammation and requires myogenesis.


Animal subjects.

Male C57Bl/6J mice (N = 74) were obtained from the Jackson Laboratories (Bar Harbor, ME). Animals were 12–16 wk of age at the time of EP. With the exceptions of X-irradiation and perfusion fixation, all experimental protocols involving live animals were performed under general anesthesia induced and maintained by a commercial table-top anesthesia system (VetEquip, Pleasanton, CA). X-irradiation and perfusion fixation were performed under general anesthesia induced by intraperitoneal injection of ketamine and xylazine (40 and 10 mg/kg, respectively). All protocols were approved by the Institutional Animal Care and Use Committee of the University of Maryland School of Medicine.

Assessment of contractile torque.

Measurements of contractile torque were performed in vivo as described (21, 28, 29) using the Small Animal Unit for Muscle Injury, Muscle Testing and Muscle Training (SUMITT, patent pending). An anesthetized animal was placed supine with its foot strapped onto a footplate and the tibia stabilized with a 27-gauge needle through the head of the tibia. Isolated contractions of the ankle dorsiflexor muscles (DFs) were elicited by depolarizing the peroneal nerve transcutaneously with a bipolar electrode (BS4 50–6824, Harvard Apparatus, Holliston, MA) placed over the head of the fibula. Twitches were used to optimize electrode placement and current amplitude. As the TA muscle accounts for ∼90% of the torque generated by the DFs (16), maximum isometric tetanic torque of the DFs was considered to be representative of the maximal force-generating capacity of the TA.

In vivo gene transfer by EP.

In an anesthetized animal, the left TA was exposed by an anterior incision of the skin and epimysium along the length of the muscle. The TA muscle was kept moist with periodic drops of sterile phosphate-buffered saline (PBS, pH 7.2) on the muscle. DNA encoding GFP (1 μg/μl, in sterile 0.9% saline) was injected into the TA muscle (1 μl/g body wt; ∼0.5 μl/mg of TA muscle) with a 30-gauge needle and 1-ml syringe. After injection, the TA was electroporated by applying five electrical pulses at an intensity of 180 V/cm, at 20-ms duration per pulse, separated by a 200-ms interpulse interval (ECM830, Electro Square Porator, Harvard Apparatus). We used this voltage based on pilot data that showed it to be the lowest voltage at which we consistently achieved transfection of ∼20% of the fibers in the TA studied 28 days after EP. Lower voltages, such as those used by others for effective transfection of the rat soleus muscle (see, e.g., Ref. 17), produced low levels of transfection in the mouse TA. After EP, the incision was closed under sterile conditions with nylon suture (Ethilon C3, 668G, Ethicon, Somerville, NJ).

Eight animals were also subjected to a sham procedure that involved exposure of the TA as described above and injecting the muscle with only the vehicle (0.9% sterile saline). Of these animals, four were exposed to a single dose of hindlimb X-irradiation before the sham procedure to blunt myogenesis.

Additional controls.

We elucidated the changes induced by specific components of our experimental procedure to identify parameters that contributed to the damage sustained by electroporated muscle. To do so, we studied animals (N = 4 per group) subjected to the following experimental conditions: 1) EP of a vector encoding a sequence of FLAG, 3XFLAG, a linker, and c-myc (FLAG EP) to test the effect of the molecular weight of the gene product (p3XFLAG-Myc-CMV-26 Expression Vector, E7283, Sigma-Aldrich, St. Louis, MO) (15). The molecular mass of the polypeptide is ∼5.6 kDa compared with GFP, which is ∼27 kDa; 2) EP of 20% of the volume of GFP DNA, at five times the concentration, to assess the effect of the volume of injected material (5XGFP EP); 3) EP after injecting only saline, without plasmid DNA, to assess the effect of saline alone (saline EP, Sal EP); 4) EP without injection of either saline or saline + DNA to assess the effect of applied voltage in the absence of injected material (no injection EP, NI EP); 5) injection of saline alone without surgical exposure of muscle to assess the effect of injection without the confounding effects of surgery and EP (closed saline, CSal); 6) surgical procedure used for EP without injection or EP to assess the effect of the surgical procedure alone (surgery, Surg); and 7) electrophysiological measurements only, to assess changes induced by our in vivo torque measurement protocol (electrophysiology, Ephys).


We irradiated the left hindlimbs of mice with a single localized dose of X-irradiation to ablate satellite cells and therefore inhibit myogenesis, as described (20). A single 25 Gy (2,500 rad) dose was delivered with a Pantak-Seifert 250KpV X-Irradiator (HF 320, North Branford, CT). This dosage inhibits activation of satellite cells (SCs) and subsequent SC-mediated myogenesis, with no significant systemic side effects or effects on mature muscle (1, 13, 30). We performed ion chamber dosimetry (PTW model 31006, Freiburg, Germany) in the path of the beam and within the lead shielding to ensure proper dosage and confirm the absence of significant backscatter of X-rays, respectively.

Tissue collection.

We collected both fixed and unfixed TA muscles to study transgene expression and histological changes after EP. Fixed tissue, collected after animals were perfused through the left ventricle with ice-cold 2% paraformaledhyde (PFA) in PBS, was snap frozen in a slurry of liquid nitrogen. Unfixed tissue was rinsed in PBS containing protease inhibitors (Roche Complete, Roche Applied Science, Indianapolis, IN) and then snap frozen in isopentane cooled by liquid nitrogen. Frozen samples were stored at −80°C.

Hematoxylin and eosin staining.

We performed hematoxylin and eosin (H&E) staining to assess the general morphology of the TA and to visualize cellular infiltrates, areas of necrosis, and centrally nucleated muscle fibers (CNFs; a marker of myofiber regeneration). We stained 8-μm cross sections of TA muscles with Harris modified hematoxylin (HHS 16, Sigma) and aqueous eosin-Y (HT110216, Sigma). Sections were hydrated in tap water for 5 min. After hydration, sections were immersed in hematoxylin (stains nuclei dark blue) for 3 min, rinsed in tap water briefly, treated for 30 s with alkaline tap water (pH 8.0) to intensify nuclear staining, and then rinsed thoroughly in tap water to remove excess hematoxylin. Sections were dipped in eosin (stains myoplasm pink), washed thoroughly in tap water to remove excess eosin, dried with a suction pipette, and mounted in mineral oil under a coverslip to prevent dehydration and preserve morphology and staining. The sections were observed under a light microscope within a week (Zeiss Axioskop, ×20 objective, Carl Zeiss), and digital images of overlapping fields were taken covering the entire cross section of the TA. Images were combined in Photoshop (Adobe, San Jose, CA) to generate tiled images of the whole TA

Analysis of GFP expression.

We studied expression of GFP in cross sections of TA muscles by counting the number of fibers labeled with GFP in images obtained under confocal fluorescence optics (LSM 510 Duo, Carl Zeiss, Poughkeepsie, NY). The tiling function on the microscope was used to take multiple images at ×10 across the cross section robotically and then combine them to generate images of the whole TA (e.g., see Figs. 2 and 4). To enable counting of all myofibers and CNFs, sections were additionally stained with Alexa 555-conjugated wheat germ agglutinin (WGA, W32464, Invitrogen, Carlsbad, CA) to label plasma membranes and 4′,6′-diamidino-2-phenylindole (DAPI, 71-03-01, KPL, Gaithersburg, MD) to label nuclei, based on methods described earlier with slight modifications (5). Cryostat sections, 16-μm thick, were fixed for 5 min with 2% paraformaldehyde (PFA) in PBS, washed with 0.05 M glycine for 5 min, and then washed with PBS for 5 min. Sections were then incubated for 10 min with WGA conjugated to Alexa 555 at a dilution of 1:400, washed four times (5 min per wash) with PBS, after which they were fixed again to immobilize WGA. Finally, sections were washed once with PBS, incubated with DAPI for 3 min, washed twice with PBS (5 min per wash), and mounted in Vectashield (Vector, Burlingame, CA) under a coverslip. We counted the total number of fibers, the number of GFP+ fibers, and the number of CNFs from tiled images of TA muscles from four animals at each time point (e.g., see Figs. 2, 4, 5). WGA also labels nuclei (5), which accounts for overlapping labeling of nuclei with WGA and DAPI.

Labeling of developmental myosin heavy chain.

To assess myogenic activity, we immunolabeled cryosections of unfixed muscle with antibodies to developmental isoforms of myosin heavy chain (dMHC; VP-M664, Vector) (28, 29). As dMHC is expressed during regeneration at ∼3–7 days following myofiber damage, its expression is commonly used to quantify myogenesis (18, 32). Sections were counterstained with DAPI to label nuclei. Tiled images were obtained with an LSM 510 Duo confocal microscope, and dMHC+ cells were counted from 10 unique 300 μm × 300 μm areas per TA muscle from two animals per time point (Fig. 6).

Labeling of macrophages.

To assess the identity of infiltrating mononuclear cells in necrotic areas of electroporated muscle, we labeled 16-μm cryosections of unfixed tissue with antibodies to CD68 (FA-11, Macrosialin), which is a pan-macrophage marker in mice (12, 26). Cryosections were fixed and permeabilized with ice-cold acetone and allowed to air dry. Sections were rehydrated with PBS (10 min), incubated with 3% BSA-PBS (30 min) to reduce nonspecific labeling, and incubated overnight at 4°C with rat anti-mouse CD68 (MCA 1957, AbD Serotec, Raleigh, NC) at a dilution of 1:100 in 3% BSA-PBS. Sections were washed once with 0.1% Triton X-100 in PBS (10 min), washed three times with PBS (5 min per wash), and incubated at room temperature for 60 min with goat anti-rat secondary antibodies conjugated to biotin (A10517, Invitrogen, Carlsbad, CA), diluted 1:200 in PBS. After incubation with anti-rat secondary antibodies, sections were washed once with 0.1% Triton X-100 in PBS (10 min), washed three times with PBS (5 min per wash), and incubated for 10 min with streptavidin conjugated to Alexa 568 (S11226, Invitrogen), diluted 1:400 in PBS. Sections were washed again, as above, incubated with DAPI for 3 min, washed twice with PBS (5 min per wash), and mounted in Vectashield (Vector Labs) under a coverslip. The solutions for incubations routinely contained 0.01% Triton X-100, to improve penetration of antibodies. Images were obtained at ×63 using an LSM Duo confocal microscope, with the pinholes set for illumination at 1 Airy unit. Counts of macrophages from 8 unique optical fields from TA muscles of 2 different animals per time point were used for quantitations (Fig. 7).

Research design and statistical methods.

Our standard protocol involved measurement of torque before (baseline torque) and after our EP (or sham) protocol (see above), with subsequent measurements and harvesting of tissue at desired time points.

The data for mice assayed at D0, in Fig. 1, are from 32 animals. Of the 32 animals, contractile torque was also measured at 3, 6, 12, and 24 h after EP (3H, 6H, 12H, and 24H, respectively) from a single set of four animals. After torque measurement at 24 h, these animals were returned to the animal facility for study 1 yr later. The remaining 28 animals were divided into seven groups of 4 and designated for torque measurements and tissue collection immediately after EP (∼10 min), or 3, 7, 14, 21, 28, or 42 days after EP (D3, D7, D14, D21, D28, and D42, respectively). Torque data for the EP group that received X-irradiation (X-IRR + EP, N = 4), the sham group (Sham, N = 4), and the sham group that received irradiation (X-IRR + Sham, N = 4) are from animals that were studied before their respective procedure and then studied longitudinally at ∼10 min, D3, D7, D14, D21, D28, and D42 post-procedure. For the X-IRR + EP, Sham and X-IRR + Sham groups, tissue was harvested at D42, and histological studies were performed only at that time point. For all the additional control groups (groups 1–7), torque data were collected before and after the respective procedure, and then 3 days later, following which unfixed TA muscles were collected as described.

Fig. 1.

Changes in contractile torque after electroporation (EP). Contractile torque drops by ∼80% after EP and recovers completely by D28. Partial recovery occurs within 3 h after EP. Irradiated (X-IRR) muscles recover partially by D7 but do not recover further thereafter. Sham injected muscles that were not electroporated lose a small amount of torque within ∼10 min and recover completely by D28. †Significant difference compared with ∼10 min post-EP; *Significant difference compared with EP at a given time point; ‡Significant difference compared with a preceding time point other than D0; NS, no significant difference.

Statistical analyses were performed with SigmaStat 3.5 (Aspire Software International, Ashburn, VA). Contractile data in Figs. 1 and 8 were analyzed by one-way ANOVA and repeated measures two-way ANOVA, respectively. Quantitations on GFP+ fibers, CNFs, dMHC+ cells, and CD68+ cells were analyzed by Kruskal-Wallis ANOVA on ranks. Pair-wise post hoc comparisons were made based on the Student-Newman-Keuls method. Alpha was set at 0.05. P values are reported only for parametric tests. All data are reported as means ± SE.


Contractile torque.

Before EP, the DF muscles of mice gave a maximal tetanic torque of 2.49 ± 0.001 Nmm (N = 32). We set the time for complete recovery from EP as the time at which the mean torque reached ≥100% of the value before EP (pre-EP). Contractile data normalized to pre-EP data are shown in Fig. 1.

Immediately after our EP protocol (∼10 min post-EP; see methods), torque decreased by ∼80% from the pre-EP value. Torque recovered partially, but significantly, between ∼10 min and 3 h post-EP and stayed at that level until D3. Recovery then progressed slowly and returned to pre-EP levels by D28 (P = 0.0169, D3 vs. D28).

In the group that was exposed to X-irradiation before EP (X-IRR + EP), the initial loss of torque following EP was not altered compared with the group that was exposed to EP alone. Torque levels recovered significantly between ∼10 min and D7 (P = 0.0333) but not at later times. As a result, the contractile torque of irradiated muscles remained at ∼55% of control levels for at least 6 wk following EP.

In the sham group, torque decreased by ∼25%, and complete recovery from this small deficit took 28 days. Prior irradiation did not have a significant effect on contractile function following the sham procedure. At D42, although the X-IRR + Sham group showed a deficit below baseline of ∼5%, it was not statistically different from the unirradiated sham group.

GFP expression.

GFP expression was detected at D3 and peaked at ∼25% of fibers at D7 (Fig. 2, and Fig. 3, top). Even a year after EP a small but significant number of fibers expressed GFP.

Fig. 2.

Green fluorescent protein (GFP) expression and centrally nucleated fibers (CNFs) after EP. Representative fluorescent micrographs of tiled images of the whole tibialis anterior (TA) sectioned at the midbelly are shown at different time points after EP. GFP+ fibers (green) can be seen as early as D3 and increase through D7, after which they begin to decrease in number. The variability in the location and numbers of fibers expressing the GFP and differences in GFP intensity among transfected fiber, are apparent within the same muscle. The area of each muscle defined by a blue rectangle is shown at higher magnification as an inset. The sections have been costained with Alexa 555-WGA to label plasma membranes (red) and DAPI to label nuclei (blue). Many CNFs (for example, fibers indicated with white arrows) can be seen at D7 and many more appear over the next several weeks.

Fig. 3.

Quantitation of GFP expression and CNFs after EP. Top: quantitations of GFP+ fibers; bottom: quantitations of CNFs. GFP expression was detected at D3 and peaked at D7. Even a year after EP, a small but significant number of fibers express GFP. The number of CNFs increased gradually over time and peaked at D28, with a significant number of CNFs present even 1 yr after EP. At D42, CNFs were significantly less in the X-IRR + EP and Injection groups than after EP alone. †Significant difference compared with control; other symbols as in Fig. 1.

We observed a high level of variability among mice in the number and location of fibers expressing GFP. This is illustrated in A in the supplementary data, which shows data from individual animals at all time points. The variability of EP and transfection efficiency are also illustrated qualitatively in Fig. 4, which shows images from individual animals at D28, the time when recovery of contractile torque was complete.

Fig. 4.

GFP expression and CNFs 28 days after EP. Images of all 4 TAs that were studied at D28, the time at which functional recovery was complete, are shown for comparison. The variable patterns of GFP fluorescence (green fibers) and the large number of CNFs (for example, fibers indicated with white arrows) are apparent in several regions of the muscle. As in Fig. 2, sections are costained with Alexa 555-WGA (red) and DAPI (blue).

Presence of CNFs.

The number of CNFs increased gradually over time and peaked at D28. Even a year after EP, ∼15% of the fibers counted were CNFs. Representative fluorescent micrographs of TA muscles showing CNFs as a function of time are shown in Fig. 2. Quantitations are in Fig. 3, bottom. The presence of CNFs and the high level of variability in their number and distribution can be seen in Fig. 4, which shows images of all TAs from all animals studied at D28. CNF data from individual animals studied at all time points are presented in A in supplementary data.

Our data also show that X-IRR before EP significantly blunts the increase in CNFs, and that sham injected muscles show <5% CNFs at D42 (supplementary data A and D; quantitations of CNFs for X-IRR + Sham not shown).

The H&E images in Fig. 5 also show the presence of CNFs as a function of time, as well as the presence of areas of necrosis and cellular infiltration from D3-D21.

Fig. 5.

Histological changes after EP. Representative images of hematoxylin and eosin stained sections are shown. These images show the presence of CNFs (for example, fibers indicated with blue arrows), areas of necrosis (for example, regions indicated with asterisks), and cellular infiltration (for example, regions indicated with green circles) after EP.

dMHC expression after EP.

In addition to counting CNFs, we confirmed the presence of myogenic activity after EP by studying the expression of developmental isoforms of myosin heavy chain (dMHC, Fig. 6). The number of dMHC+ cells increased significantly from control levels by D3 and stayed elevated until D42, with a broad peak from D7–21.

Fig. 6.

Expression of developmental myosin heavy chain (dMHC) after EP. Top: representative images of dMHC labeling; bottom: complete quantitations. Expression of dMHC (red) peaks at D7 but stays elevated even up to D42. Symbols for quantitations are as in Fig. 3. Sections have been colabeled with DAPI to mark nuclei (blue). GFP+ fibers appear green.

Macrophage infiltration after EP.

As seen in Figs. 2, 4, and 5, many areas of electroporated muscle appeared necrotic and contained cellular infiltrates. We identified the majority of cells in these infiltrates as macrophages by labeling with antibodies to CD68, a macrophage marker, and counting labeled cells (Fig. 7). The counts of CD68+ cells showed a large peak at D3 and a smaller peak at D14. A significant number of CD68+ cells were present even at D42.

Fig. 7.

Macrophage infiltration after EP. Representative images and quantitations of macrophage (CD68+ cells; red) infiltration after EP are shown. The number of CD68+ cells shows a bimodal distribution with a large peak at D3 and smaller peak at D14. A significant number of CD68+ cells are seen even up to D42. Symbols for quantitations are as in Fig. 3. Sections have been colabeled with DAPI (blue) to mark nuclei. GFP+ fibers appear green.

Additional controls.

We tested the various steps and components of our EP procedure to determine whether the injection or the composition or volume of the material injected, surgery, or measurement of torque had a significant effect on the outcome. The losses of torque ∼10 min after the standard EP procedure (EP), EP with FLAG tag DNA (FLAG EP), EP with 20% of the volume of 5× concentrated GFP DNA (5XGFP EP), EP after saline injection (Sal EP), and EP without any injection (NI EP) were statistically indistinguishable (19–37% of baseline torque). Among these groups, only the NI+EP group showed significantly higher levels of torque than the EP group 3 days later (Fig. 8).

Fig. 8.

Changes in contractile torque after other experimental conditions. EP following injection of GFP DNA, FLAG DNA, 20% volume of 5× concentrated GFP DNA, or saline alone, and EP without prior injection of saline or DNA solutions, induce similar, high levels of torque loss ∼10 min after the procedure. Other conditions show lower levels of torque loss ∼10 min after the procedure. Uninjected, electroporated muscle recover significantly at D3, suggesting that torque loss at ∼10 min is transient. Symbols for statistical differences are as in Fig. 1.

The losses of torque ∼10 min after injection of saline without surgery (CSal), surgery alone (Surg), and electrophysiology alone (Ephys) were significantly lower than those seen under the other conditions we tested and were not significantly different from each other (87–99% of baseline torque). None of these groups differed significantly in torque 3 days later (Fig. 8).

Three days after the respective procedure, macrophage infiltration in muscles electroporated with FLAG DNA was not significantly different from standard EP with GFP DNA. However, macrophage infiltration was significantly lower in all other conditions compared with standard EP. Injecting 20% of the volume of 5× concentrated GFP DNA or injecting saline before EP produced the same moderate level of macrophage infiltration. EP without injection, saline injection without surgery, and surgery alone produced similar low levels of macrophage infiltration. Finally, torque measurements alone also produced a very low, but significant, increase in macrophage infiltration (Fig. 9).

Fig. 9.

Macrophage infiltration at D3 after other experimental conditions. Representative images and quantitations of macrophage (CD68+ cells; red) infiltration after other experimental conditions are shown. The highest levels of macrophage infiltration occur 3 days after EP with FLAG DNA (compare with D3 in Fig. 7). Moderate levels of infiltration occur after the low volume GFP DNA and saline EP conditions. Low levels of infiltration occur after the other conditions. Sections have been colabeled with DAPI (blue) to mark nuclei. Symbols for quantitations are as in Fig. 3.

Complementary to our data on torque and macrophage infiltration, histological changes studied from H&E-stained sections also showed higher levels of tissue damage in the experimental conditions that involved injection and EP (FLAG, 5XGFP EP, and Sal EP) than in those that did not (Fig. 10).

Fig. 10.

Histological changes at D3 after other experimental conditions. Necrotic areas infiltrated by mononuclear cells (indicated by green circles) are more widespread in muscles that were injected with genetic material or saline before electroporation than in the other groups. EP with FLAG DNA show the greatest amount of damage and is similar to damage induced by EP with GFP DNA (compare with D3 in Fig. 5).

Our results suggest that current and a conductive solution such as saline both contribute significantly to the damage experienced by mouse TA muscles during EP, and that saline injection by itself induces much lower levels of damage.


The goal of our study was to describe the physiological and histological changes in murine TA muscle following in vivo gene transfer by EP of the TA muscle. We were particularly interested in assessing transgene expression following EP and in determining whether EP by itself was injurious to the muscle. Our results suggest that EP under conditions that transfect ∼25% of the myofibers in TA muscle is indeed injurious, that recovery from the injury it induces requires myogenesis, and that evidence of the damage it causes can last many weeks and perhaps as long as 1 yr.

To our knowledge, changes in contractile function as a function of time after EP have not been reported. One report documented an ∼85% loss of contractile torque in the extensor digitorum longus (EDL) measured in vitro soon after EP (11), which is consistent with our observation of an ∼80% decrease. This loss of function is much larger than what we and others have observed after contraction-induced injuries by lengthening contractions (27, 28). Our data further indicate that electroporated muscles recovered ∼40% of their contractile function, lost immediately after EP, within a period of 3 h. This rapid recovery is likely to be due to local repair of fibers that were transiently damaged by EP.

Consistent with this, TA muscles that were exposed to X-irradiation to blunt myogenesis also recovered partially and were indistinguishable from unirradiated muscles 3 days after EP, suggesting that this early phase of recovery did not require myogenesis. Further recovery of function was blocked by irradiation, however. These results suggest that myogenesis is required for complete recovery from the damage caused by EP. In agreement with this, CNFs and dMHC+ fibers also increased in the days following EP, and the increase in CNFs was suppressed by irradiation (Fig. 3, bottom, and in supplementary data A). Our data therefore indicate that recovery, beyond what is rapidly achieved within 3 h, is mediated by activation of SCs and formation of new muscle fibers. Sham-injected TA muscles also lost a small percentage of their original contractile torque and recovered slowly by D28, suggesting that local repair of myofibers was not sufficient to facilitate earlier recovery from the surgery and the intramuscular injection associated with EP. However, irradiation did not have a significant effect on recovery, suggesting that myogenesis plays a less important role in enabling recovery from the sham procedure than after our standard EP protocol. The modest levels of damage seen in our additional control experiments, which involved injection without surgical exposure of the muscle or surgery alone, suggest that the sham procedure induces damage that is very different from damage caused by injecting and then electrically permeabilizing the muscle.

In addition to counting CNFs, which are often used to assess regeneration in skeletal muscle, we also counted the number of TA myofibers that expressed dMHC after EP. These data, too, indicate that myogenic replacement of damaged fibers is prominent following EP. The presence of a large number of dMHC+ cells 7–21 days after EP, and the persistence of a significant number of cells for at least 42 days after EP, differ from the pattern of dMHC expression that has been reported after injury from lengthening contractions or myotoxin. Typically, following myofiber damage, dMHC expression increases between 3 and 7 days and then drops (6, 7, 28). The presence of a significant number of dMHC+ cells over several weeks suggests that myogenic activity after EP continues much longer compared with myogenesis after concerted muscle damage induced by myotoxins or injurious contractions. As many myofibers remain centrally nucleated even 1 year after EP, low levels of damage are likely to continue to occur over a prolonged period of time after EP.

These results strongly suggest that the formation of new myofibers is required for the recovery of TA muscle from EP. This is because the fibers that undergo necrosis, especially in the week following EP (Fig. 5), are being replaced by the generation of new muscle fibers through the activation of satellite cells. This explanation is consistent with the persistent deficit in torque in muscles irradiated before EP and with an ∼40% decrease in the total number of fibers at the midbelly of the TA after EP (control = 1,935 ± 184 and D42 X-IRR+EP = 1,136 ± 189; supplementary data A).

Given that many fibers undergo necrosis following EP, we were not surprised to observe a significant increase in the number of macrophages in recovering muscle. Counts of CD68+ cells indicate that the TA muscle is heavily infiltrated by macrophages after EP; their most likely role is clearing dead fibers initially (34) and supporting myogenesis later (3, 35). The large peak in macrophage counts at D3, followed by a second smaller peak at D14, further suggests that inflammatory infiltration of muscle occurs in more than one wave.

Our results indicate not only that EP causes significant damage to TA muscle but also that its effects are quite variable from mouse to mouse. This was most apparent in our tiled images of entire cross sections of the muscle midbelly (e.g., Figs. 2 and 4), which revealed major differences in the number and distribution of fibers expressing GFP after EP in different mice. These images also illustrate the variability in the intensity of the GFP signal, which can be interpreted as the extent of expression in a fiber that has been successfully transfected. The results indicate that the pattern of expression of GFP across the muscle, as well as the extent of expression from fiber to fiber, varied considerably from mouse to mouse. The distribution of necrotic fibers also varied. Both observations are consistent with the idea that the path of the voltage surge across the muscle during EP in itself is highly variable.

A caveat to our current study is that we only studied the TA muscle. The TA is significantly larger in diameter than muscles such as the soleus, EDL, and the flexor digitorum brevis, which have also been studied following EP by various investigators (8, 11, 17). It is possible that the extensive damage and high variability in transgene expression that we observe after EP might be due to the high electrical impedance encountered in the tissue between the electrodes (10), which is likely to be less of a factor in EP of muscles that are smaller than the TA or in EP of other tissues [e.g., Lefesvre et al. (19)]. However, the findings of Gissel and Clausen (11) suggest that even a smaller muscle, like the rat EDL, sustains significant damage, reflected in altered calcium signaling and force production measured in vitro soon after EP.

Our systematic study of the effects of specific aspects of our EP protocol gives us several useful insights into the mechanism of damage from EP, which might be useful in future refinements of the protocol. For example, we find the size of the encoded peptide or protein does not affect the extent of damage, at least with relatively small molecules (GFP vs. FLAG). Surprisingly, we find that application of the electroporating voltage without prior injection creates minimal damage, suggesting that the injected material and applied voltage interact to induce extensive damage. This is likely to be due to the fact that the injected material, with or without DNA, is a saline solution, which conducts the current required for effective EP. Consistent with this, reducing the volume injected fivefold leads to lower levels of muscle damage; it also reduces the transfection efficiency considerably ( in supplementary data B; quantitations not shown). Because of the interaction that injection and EP have on damage, we injected GFP DNA intraperitoneally before EP but could not identify any transfected myofibers (supplementary data C; quantitations not shown). Injection of saline without surgery or EP causes a reduction in contractile function, albeit without high levels of myofiber damage 3 days after injection and EP. The very small but significant increase in macrophages seen 3 days after torque measurements alone is consistent with earlier evidence of increased inflammation following electrically stimulated isometric contractions (23).

Our data suggest that the damage produced by EP must be considered in experiments using these techniques. The adverse consequences of EP and the variability in the extent and pattern of transfection make it challenging to quantify the effects of transgene expression versus the effects of EP alone, at least in murine TA muscle. However, EP may still be suitable to alter gene and protein expression in studies in which variable levels of muscle damage and transgene expression are not confounding factors. For example, EP is effective and useful for soleus muscles (17) and for flexor digitorum brevis myofibers, which can be isolated and studied in culture for the location or activity of transgene-encoded proteins (9). It can also be used successfully for studies of individual fibers of TA muscles, if appropriate quantitative controls are performed to ensure that the results are not simply due to nonspecific effects of the EP procedure itself.


This research was supported by fellowships to J. A. Roche from the Jain Foundation, Development Awards to D. L. Ford-Speelman and P. W. Reed from the Muscular Dystrophy Association, by grants to P. W. Reed and R. J. Bloch from the National Institutes of Health (1R21 AR057519-01 to P. W. Reed; 1RO1 AR056330, 1RO1 AR055928 to R. J. Bloch) and to R. J. Bloch from the Jain Foundation (no. 71245) and the Muscular Dystrophy Association (no. 157601), and by a contract to R. J. Bloch from the Wellstone Muscular Dystrophy Cooperative Research Center (5U54 HD060848, Dr. C. Emerson, PI).


No conflicts of interest, financial or otherwise, are declared by the author(s).


We thank Dr. S. C. Kandarian, Dept. of Health Sciences, Boston University, Boston, MA, for kindly providing training and valuable insights on EP. Our methods were adapted from Dr. Kandarian's methods, but they differ significantly in several key aspects.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
View Abstract