Background

Skeletal muscle is an important tissue with many fundamental functions (Wolfe, 2006; Nair, 2005). Consistently, diseases that affect skeletal muscle profoundly impact the organism’s fitness and survival (Demontis and Perrimon, 2009, 2010; Demontis et al., 2013a, 2014; Piccirillo et al., 2014; Rai and Demontis, 2016; Robles-Murguia et al., 2020; Rai et al., 2021a). Among the many muscle diseases, muscle wasting is a debilitating condition associated with aging and with diseases such as cancer, infections, kidney failure, sepsis, and neuromuscular disorders (Demontis et al., 2013b; Bonaldo and Sandri, 2013; Tsoli and Robertson, 2013; Piccirillo et al., 2014). Several studies have demonstrated that muscle wasting worsens disease outcome and decreases patient survival, whereas preserving skeletal muscle mass and function is protective (Zhou et al., 2010, Johnston et al., 2015). Skeletal muscles are composed of multinucleated syncytial cells known as fibers or myofibers. Myofiber types with distinct metabolic and contractile properties are differently abundant in muscles, and this varies in accordance with anatomical location and function (Schiaffino and Reggiani, 2011; Schiaffino et al., 2013). During muscle wasting, these myofiber types are differently susceptible and impacted by atrophic stimuli (Demontis et al., 2013b; Piccirillo et al., 2014; Bonaldo and Sandri, 2013; Ciciliot et al., 2013). Catabolic stimuli induce muscle wasting primarily via the induction of myofiber atrophy, whereas changes in the number of myofibers are not common. Mechanistically, a decrease in the size of myofibers occurs as a result of decreased synthesis and increased degradation of muscle protein, which is mediated by the autophagy/lysosome and ubiquitin-proteasome systems (Demontis et al., 2013b; Piccirillo et al., 2014; Bonaldo and Sandri, 2013). As for other muscle diseases, there are currently no therapies available in the clinic to prevent or cure age- and disease-associated muscle wasting. To address this unmet medical need, many experimental disease models and techniques for probing gene function in skeletal muscles have been developed over the years.

Electroporation has been used as a general method for gene delivery that is potentially applicable to all organisms and cell types (Ugen and Heller, 2003; Young and Dean, 2015). By using electrical fields, electroporation transiently destabilizes the plasma membrane and facilitates the electrophoretic movement and intracellular delivery of DNA plasmids (Ugen and Heller, 2003; Young and Dean, 2015). In skeletal muscle, this technique has been extensively used for the expression of plasmid-encoded transgenes and the subsequent assessment of gene function in skeletal muscle homeostasis (Aihara and Miyazaki, 1998; Rizzuto et al., 1999; Hoover and Magne Kalhovde, 2000; Sandri et al., 2003; Bertrand et al., 2003; Peng et al., 2005; Tevz et al., 2008). Moreover, this experimental approach has found translational application in improving the delivery of DNA plasmids, for mounting immune responses (Widera et al., 2000; Zucchelli et al., 2000; Khan et al., 2014; Vandermeulen et al., 2014; Haidari et al., 2019; Mpendo et al., 2020; Edupuganti et al., 2020), and for gene therapy in humans (Aihara and Miyazaki, 1998; Rizzuto et al., 1999; Mizui et al., 2004;Wong et al., 2005; Brolin et al., 2015).

Here, we report an optimized protocol for the electroporation of small interfering RNAs (siRNAs) into the tibialis anterior (TA) skeletal muscle of mice (protocol #1). We propose that this method can help to interrogate gene function in skeletal muscle, by inducing the acute knockdown of the intended target mRNAs in animal models. Moreover, we also report a protocol for siRNA-mediated knockdown in cultured myotubes (protocol #2), to complement in vivo studies with siRNA-mediated electroporation.

Knowledge gained with this siRNA electroporation technique could be similarly applied for improving the intramuscular delivery of other RNA species and RNA-based vaccines, as well as for other translational and basic science applications.

Experimental design and controls
In this protocol, the TA muscle in one hind leg is electroporated with siRNAs for targeting the intended mRNA, whereas the contralateral hind leg serves as control, and is electroporated with non-targeting (NT) siRNAs. Therefore, this experimental design provides a robust internal control unhindered by differences between individual animals. Nonetheless, to ensure that the results obtained from different animals in the same cohort can be cross-compared, it is important to utilize animals of the same age and sex, and reared in a consistent manner (siblings from the same litter should be utilized whenever possible).

A major advantage of the electroporation technique described here is that it allows probing gene function in animals of different ages. Previous studies have used 6-month-old mice as “young” controls (Sheard and Anderson, 2012) because, while postnatal muscle development mostly occurs in the first 3 months of age, growth still occurs in subsequent months (Ebert et al., 2015). However, whenever comparing electroporation experiments done at different ages, especially if these include ages at which postnatal development is not concluded, it is important to normalize the TA mass to the length of the tibia bone. This accounts for differences in whole body size when comparing muscles from different mice, because the tibia is typically static in fully-grown mice, and is not influenced by muscle atrophic and hypertrophic stimuli (Rowland, 2007; Shavlakadze et al., 2010; Puppa et al., 2014; Winbanks et al., 2016).

In addition to testing gene function in wild-type mice, siRNA electroporation can also be used to test the impact of mRNA knockdown in disease settings (e.g., models of cancer cachexia). In this scenario, the electroporation technique provides a suitable intervention for testing the requirement of a certain gene in disease progression. For example, the electroporation of siRNAs can be used to probe whether impeding the upregulation of a cancer-induced gene can preserve myofiber size and prevent cancer-induced TA mass loss. However, in this case, appropriate controls will consist of the electroporation of gene-targeting and control NT siRNAs, in contralateral legs of animals, with and without cancer (or another disease model).

Similarly, siRNA electroporation can be performed not only in wild-type animals but also concomitantly to another genetic intervention [e.g., Cre-Lox–mediated ablation of another gene in muscle (McCarthy et al., 2012)]. In this scenario, the electroporation probes the genetic interaction between the siRNA–targeted and the Cre–targeted genes. However, appropriate controls in this case will consist of the electroporation of gene–targeting and NT siRNAs in contralateral legs of animals with and without Cre–mediated ablation of the second gene tested.

Similar experimental design and controls are also used with the siRNA–mediated knockdown of target genes in cultured mouse C2C12 myotubes. Specifically, qRT-PCR and other cellular/molecular assays are used to test the outcome of siRNAs targeting the intended gene compared to control NT siRNAs. Although the percentage of mRNA knockdown needed to uncover a phenotype varies depending on the mRNA targeted, a previous estimate based on large-scale RNAi testing in Drosophila melanogaster suggests that phenotypes are typically uncovered when the reduction in mRNA levels is above 50% (Sopko et al., 2014; Graca et al., 2021).

We propose that testing siRNA efficacy of target gene knockdown in cultured myotubes may constitute an important step towards validation of siRNA reagents, before their use in vivo for TA muscle electroporation.

Limitations and comparison to other models
Because this protocol focuses on the electroporation of siRNAs, it is impacted by the known limitations associated with RNA interference, including the possibility that siRNAs target unintended mRNAs (RNAi off-target effects), and that the mRNA knockdown achieved via siRNAs is partial and insufficient to uncover a phenotype. These initial impediments of RNAi are largely surpassed by late-generation reagents (e.g., siRNA SMARTpools used here) (Setten et al., 2019; Neumeier and Meister, 2020).

Another limitation of siRNA delivery via electroporation is that it induces some, albeit minimal, muscle damage (McMahon and Wells, 2004; Skuk et al., 2013). In fact, the conditions reported in this protocol have been optimized to avoid extensive muscle damage, which has been possible because delivery of siRNAs requires remarkably lower electric fields and plasma membrane perturbation, when compared to the delivery of bulkier DNA plasmids (Aihara and Miyazaki, 1998; Rizzuto et al., 1999; Hoover and Magne Kalhovde, 2000; Sandri et al., 2003; Bertrand et al., 2003; Peng et al., 2005; Tevz et al., 2008), that have nonetheless lead to many landmark discoveries in muscle biology (Murgia et al., 2000; Sandri et al., 2004; Menzies et al., 2010). Specifically, the electrotransfer voltages used here have previously been demonstrated to be optimal (McMahon et al., 2001; Schertzer et al., 2006). Although the higher voltages may induce muscle damage when the skin is removed, and electrodes are placed directly on the muscle, the electroporation efficacy and damage are lower when the electrodes are placed on the skin without surgical incision, as reported here. Moreover, the method used here, by which the electrical field orientation is alternated (from lateral-medial to anterior-posterior), has been demonstrated to transduce more efficiently than the single orientation of the electrical field (Golzio et al., 2012). Therefore, in our opinion, the limitations deriving from the RNAi technology and electroporation are minimal due to the extensive optimization of siRNA reagents for this established technology, and because of the mild and non-surgical electroporation conditions that are used for siRNA delivery.

However, a major limitation of electroporation is that it allows for only a transient knockdown of gene function, which has been reported to be significant for up to 2–3 weeks (Tevz et al., 2008). Therefore, it is not possible to probe the long-term consequences of gene perturbation, as the electroporation only transiently affects mRNA levels. For the same reason, because some proteins are extremely long-lived (Savas et al., 2012; Toyama et al., 2013; Krishna et al., 2021), it may not be possible to impact their levels in the few-weeks timeframe of electroporation.

A further limitation of this protocol is that it allows for the analysis of gene function only in TA muscles. Therefore, it does not provide a means for testing gene function in muscles with different anatomical locations and physiological functions (Schiaffino and Reggiani, 2011; Schiaffino et al., 2013). Therefore, rather than surgery-mediated direct apposition to muscles, electroporation with electrodes apposed to the skin does not allow probing gene function in internal muscles, such as the soleus. Alternative methods for gene delivery to skeletal muscle are available, including adeno–associated viruses and sonodelivery (Decker et al., 2020; Manini et al., 2021), and they may overcome limitations of electroporation–mediated delivery if needed.

Advantages
Nonetheless, compared to other interventions, the siRNA electroporation protocol reported here offers several advantages. Some of the previously noted limitations (see paragraph above) can turn out to be advantageous under certain conditions. For example, the partial knockdown of the intended mRNA might be an issue and impede the uncovering of a phenotype. However, this might be advantageous when targeting a gene with fundamental functions, which would result in the death of the animal if completely ablated and/or impacted across all skeletal muscles.

It has been previously noted that, over generations, knockout animals accumulate compensatory background mutations that can reduce the phenotypic manifestation, and even lead to the complete disappearance of the phenotype (Rossi et al., 2015). In this respect, the electroporation of siRNAs provides an acute intervention for perturbing gene function, and it is therefore unencumbered by the compensatory adjustments that occur in classical genetic mutants over time. Similarly, the genetic background (i.e., the genetic makeup of the animal beyond the mutation in the intended gene) (Ungerer et al., 2003; Burnett et al., 2011; Lucanic et al., 2017; Hou et al., 2019; Koh et al., 2020), the cytoplasmic background (derived from the oocyte, with its set of mitochondria and cytoplasmic pathogens) (Toivonen et al., 2007; Joseph et al., 2013), and the microbiota (Ulgherait et al., 2016; Kim et al., 2017; Mamantopoulos et al., 2017; Poussin et al., 2018; Vujkovic-Cvijin et al., 2020) are examples of biological variables that are normally controlled only with a careful experimental design. For example, to ensure that all experimental animals do not differ in their cytoplasmic background, all experimental animals should be derived from the same mother, or at least from related mothers. To ensure that the animals are isogenic and do not differ in their genetic background, they should be back-crossed several times (typically 10×) against the same genetic reference strain.

The protocol reported here is based on the electroporation of siRNAs for the intended gene into the TA muscle of one hind leg, and the electroporation of control non-targeting (NT) siRNAs into the contralateral hind leg of the same animal. On this basis, any observed phenotype obtained with siRNAs for the intended gene versus NT siRNAs is not due to inter-individual differences in the genetic background, cytoplasmic background, or microbiota. Therefore, the protocol described here provides ideal settings for testing gene function in the TA muscle, without confounding biological variables.

Although skeletal muscle primarily consists of syncytial muscle cells (known as fibers or myofibers), there are many infiltrating and associated cells that are important for muscle homeostasis. These include immune cells, endothelial cells, fibroadipogenic progenitors, and satellite muscle stem cells. The electroporation likely targets all these cell types (Wong et al., 2005; Dean, 2013), and hence allows for the investigation of the function of a certain gene, not only in myofibers, but also in other muscle-associated cells.

Potential applications of this protocol and future directions
Due to its simplicity, this protocol can be applied to many different mouse strains and disease models. Moreover, it is potentially applicable to other rodent species that are generally used in research (such as rats), and other emerging rodent disease models, such as the long-lived naked-mole rat, and Octodon degus, a rodent that has been reported to develop spontaneous Alzheimer’s-like disease (Buffenstein, 2005; van Groen et al., 2011; Valenzano et al., 2017). Although these rodents radically differ from each other in regards to many features, including their life trajectories and disease predisposition, they all have TA muscles that are easily accessible for electroporation, without the need of any surgery. However, the TA muscle is bigger in rats; thus, the procedure reported here for mice would need to be scaled up. Alternatively, only a localized area of the TA could be electroporated and eventually sampled with a localized biopsy. Indeed, with appropriate adjustments, the siRNA electroporation described here could prove useful in investigating skeletal muscle biology also in larger model organisms, such as marmosets and monkeys (upon appropriate ethical approvals). In this case, the impact of siRNA electroporation into TA muscles could be tested by obtaining a biopsy of the electroporated area of the TA, without requiring sacrificing the animal (Joyce et al., 2012; Cotta et al., 2021). Beyond the TA, it may also be possible to adapt this method to test the impact of gene knockdown in other skeletal muscles. However, such additional testing would be limited to skeletal muscles located beneath the skin, and hence susceptible to electroporation without surgical incision.

In addition to RNAi–mediated knockdown, this electroporation method can also be used in conjunction with gene editing technologies such as CRISPR, to obtain target gene knockout and overexpression (Gaj et al., 2013; Doudna and Charpentier, 2014; Jiang and Doudna, 2017). For example, electroporation of short-guide RNAs (sgRNAs) into TA muscles of mouse models that express Cas9 or Cas9 fused with a transcriptional activator could lead to target gene deletion or overexpression, respectively, whereas electroporation of the contralateral leg with mock sgRNAs would provide a matched control.

Beyond its use in experimental animal models, it has been proposed that TA electroporation in humans may constitute a means for gene therapy, by improving the intramyofiber delivery of DNA vaccines (Widera et al., 2000; Zucchelli et al., 2000; Khan et al., 2014; Vandermeulen et al., 2014; Haidari et al., 2019; Mpendo et al., 2020; Edupuganti et al., 2020), and potentially also of recently-developed RNA-based vaccines, and for the production of cytokines, growth factors, and other therapeutic factors by the skeletal muscle of patients with a number of diseases (Aihara and Miyazaki, 1998; Rizzuto et al., 1999; Mizui et al., 2004; Wong et al., 2005;Brolin et al., 2015).