Biofabricating murine and human myo-substitutes for rapid volumetric muscle loss restoration

The importance of skeletal muscle tissue is undoubted being the controller of several vital functions including respiration and all voluntary locomotion activities. However, its regenerative capability is limited and significant tissue loss often leads to a chronic pathologic condition known as volumetric muscle loss. Here, we propose a biofabrication approach to rapidly restore skeletal muscle mass, 3D histoarchitecture and functionality. By recapitulating muscle anisotropic organization at the microscale level, we demonstrate to efficiently guide cell differentiation and myobundle formation both in vitro and in vivo. Of note, upon implantation, the biofabricated myo-substitutes support the formation of new blood vessels and neuromuscular junctions – pivotal aspects for cell survival and muscle contractile functionalities – together with an advanced along with muscle mass and force recovery. Together, these data represent a solid base for further testing the myo-substitutes in large animal size and a promising platform to be eventually translated into clinical scenarios.


Introduction
Skeletal muscle (SM) tissue accounts for over 40% of the total body mass, controlling all voluntary locomotion activities and regulating vital functions such as respiration and metabolic homeostasis.
The clinical scenarios of SM-related disorders are complex as the functionality of this tissue can be impaired by several conditions that can either be acquired, as a consequence of contusions, lacerations or surgical interventions (sarcoma removal), or be congenital/genetic as in muscular dystrophies or metabolic diseases. Regardless the disorder origin, treating any type of SM loss or disorder remains an unmet therapeutic need [5,6].
In this context, skeletal muscle tissue engineering (often referred as SMTE) holds great promises for addressing the two key challenges in the field: i) the development of a reliable 3D in vitro models and ii) the biofabrication of a macroscopic engineered SM tissue (construct volume > 1 cm 3 ) [7][8][9].
As far as the first challenge is concerned, SMTE is expected (likely) to yield in the near future functional in vitro models for developing and testing approaches that may help addressing muscle dysfunctions caused by genetic or metabolic diseases. The success in the development of dependable in vitro models of human SM, aside from contributing to sparing animal experimentation, it is expected to overcome some of the problems caused by the failure of some animal models and to reproduce some important features of the human disease. In addition, such in vitro models will therefore be of great help in better understanding the etiology of SM specific diseases -including dystrophies such as Sarcoglycanopathy or Facio-Scapulo-Humeral-Dystrophy and metabolic disorders such as Pompe disease and Mitochondrial Metabolic Myopathy [10][11][12].
Finally, in the context of drug discovery and development, SMTE-based in vitro models will be especially valuable for the characterization of drug molecular mechanisms, rapid prioritization of lead candidates, toxicity testing and new biomarker identification [13].
The development of in vitro SM tissue models has been at the core of intensive industrial and academic research over the past ten years and an ever-increasing number of innovative approaches have been proposed including 3D bioprinting and organ-on-a chip solutions [14][15][16][17].
Despite the introduction of new technologies and refined protocols for culturing engineered muscles -derived from either murine or human muscle progenitors -so far, SM maturation and maintenance have been only partially recapitulated in vitro and, most importantly, to a small scale [18,19]. These approaches, while promising, are still far from satisfactorily addressing the second key challenge of SMTE -i.e. the development of macroscopic tissue-equivalents of a size suitable for treating volumetric muscle loss (VML): a condition resulting from large traumatic injury or massive surgical ablation upon tumor removal [20,21]. VML, which is characterized by muscular mass loss, functional deficit and permanent disability, is currently treated by surgical tissue transfer (autologous transplantation). This invasive approach is often associated with unsatisfactory outcomes due to poor tissue engraftment and donor site morbidity [22][23][24].
In this study, we present a microfluidic wet-spinning system [25] that allows rapid fabrication (fiber extrusion velocity > 2 m/min) of macroscopic yarns of cell-laden hydrogel microfibersloaded either with murine (mesoangioblasts, Mabs) or human (primary myoblasts) muscle precursors -that closely mimic the structure of the SM tissue. Our findings demonstrate that by optimizing key parameters such as i) hydrogel precursor composition (i.e. bioink), ii) cell seeding density, iii) fiber architecture and iv) culturing protocol, we are able to successfully fabricate advanced artificial myo-substitutes that can be used to restore SM mass-loss and functionality in vivo.

Materials
All chemicals were purchased from Sigma-Aldrich and used without further purification unless otherwise stated. Sodium alginate (ALG, Mw 33 kDa) was a kind gift from FMC Biopolymers.

Design and fabrication of wet-spinning set-up
Cell-laden hydrogel fibers were fabricated using a custom wet-spinning approach [25].

In vivo construct implantation
Two-month-old male SCID/Beige mice were anesthetized with an intramuscular injection of physiologic saline (10 ml/kg) containing ketamine (5 mg/ml) and xylazine (1 mg/ml) and then the 3D constructs were implanted in Tibialis Anterior muscle (TA), according to following surgical procedure [29]: a) limited incision on the medial side of the leg has been performed in order to reach the TA; b) utilizing a cautery to avoid bleeding, the muscle fibers were deeply removed to create a venue for the implant; c) the construct was placed in the removed TA fibers lodge and the incision was sutured. As control, the contralateral TA was surgically ablated but no construct was

Force measurements
Contractile performance of treated and control muscles was measured in vivo using a 305B muscle lever system (Aurora Scientific, Aurora, ON, Canada) in animals anesthetized with a mixture of tiletamine, zolazepam, and xilazine. Mice were placed on a thermostatically controlled table, with knee kept stationery and foot firmly fixed to a footplate, which, in turn, was connected to the shaft of the motor. Contraction was elicited by electrical stimulation of the peroneal nerve. Tefloncovered stainless-steel electrodes were implanted near the branch of the peroneal nerve as it emerges distally from the popliteal fossa. The two thin electrodes were sewn on both sides of the nerve, and the skin above was sutured. The electrodes were connected to an AMP Master-8 stimulator (AMP Instruments, Jerusalem, Israel). Isometric contractions induced at different frequencies were performed to determine the force-frequency relationship.

Histology and immunostaining
In vitro 3D constructs after 20 days of culture, were fixed in PFA 2%, while in vivo grafts were processed for histology as previously described [7]. Briefly, 3D constructs were surgically

Image analysis
The alignment of the differentiated muscle fibers was assessed using OrientationJ plugin for ImageJ/FIJI [30]. The local orientation and isotropic properties, such as coherency, are evaluated for every pixel of the image based on structure tensors [31]. In order to analyze the images, we converted them in highly contrasted 8-bit images and processed them using OrientationJ dominant direction and OrientationJ distribution tools. OrientationJ dominant direction computes the coherency value, an index ranging between 0 and 1, with 1 indicating highly oriented structures and 0 indicating isotropic areas. OrientationJ distribution processes a weighted histogram of orientations, where the weight is the coherency itself.

Vessel density evaluation
To obtain a quantification of the vascularization levels in the normal muscle tissue and in the engrafted tissue, images from immunofluorescence analysis against SMA (Smooth Muscle Actine) and vW (von Willebrand factor), well-known vascularity markers, were analyzed using ImageJ

Statistical analysis
All experiments were performed in biological triplicate. Data were analyzed using GraphPad Prism using Student's t-test. A probability of less than 5% (P < 0 05) was considered to be statistically significant.

Assembly and characterization of a wet spinning set-up.
Nowadays, biofabrication technologies hold great promises for the in vitro production of functional tissue equivalents thanks to their flexibility and unprecedent precision in cell and biomaterial deposition. Such features are of paramount importance for the recapitulation of organ/tissue complexity and functionality; especially true for the skeletal muscle, a tissue characterized by a hierarchical, uniaxially aligned architecture [32].
In a recent study, we presented an extrusion-based 3D bioprinting approach that enables to biofabricate hydrogel scaffolds composed of aligned myoblast-laden fibers [7]. Despite the remarkable results in terms of muscle architectural guidance, as shown by the accomplishment of oriented, mature myofibers that spontaneously contract in vitro, the developed approach suffers from several limitations. Firstly, ectopic implantation in a mouse model revealed some limitations in terms of scaffold remodeling, as large areas were actually devoid of myotubes (likely because of the large diameter of printed fibers ≈ 250 µm). Furthermore, in order to carefully pack adjacent fibers as close as possible, printing speed was relatively low (≈ 4 mm•s -1 ), thus limiting the possibility of biofabricating large constructs in an acceptable time.
In order to overcome these constraints, we have developed a custom-made wet-spinning system to generate well-organized myo-substitute with higher resolution (fiber diameter down to approx. 80 µm). Our novel wet-spinning system is notably simpler and easier to use then conventional 3D printers, as it comprises just a co-axial nozzle system and a Teflon drum connected to a stepper motor (Figure 1a-c). The highly aligned skeletal muscle architecture can be achieved by collecting the hydrogel fibers on the rotating drum, without any need of ad-hoc printing code. Additionally, the hydrogel fibers that are collected onto the drum form a highly packed bundle with almost no distance between adjacent fibers. This is hardly achievable with conventional extrusion bioprinting approaches (Figure 1d,e).
Another advantageous feature of this system consists in the possibility to fine tune the hydrogel fiber size by controlling the flow rate of the bioink and the rotating speed of the drum. After a thorough characterization of the system, we found that hydrogel fibers can be produced in a reproducible manner using a motor speed in the 15÷70 rpm range (i.e. linear speed ≈ 20÷92 mm•s -1 ), which is 10-to 100-fold faster in comparison to conventional extrusion-based bioprinting approaches [33], and a bioink flow rate of 50 µL/min. As for the calcium chloride flow rate, we found that values in the range 15÷25 µL/min were equally suitable for fiber production, with no major impact on fiber size. As anticipated, the diameter of wet-spun fibers and the motor speed are inversely related (Figure 1f). The theoretical values of fiber diameters (dfiber) could be calculated by assuming a volumetric equivalence with the flow rate of the bioink (Qbioink) using the formula [27]: where V is the motor speed. Interestingly, such values were actually higher than the measured ones, for all motor speeds. Such effect was most likely a consequence of alginate shrinking upon exposure to calcium ions and was estimated to be of approximately 30%.
As bioink, we used a formulation already well established by our group containing PEG-Fibrinogen (PF), responsible for myogenic differentiation, and alginate (ALG), enabling immediate gelation of spun hydrogel fiber upon exposure to divalent calcium ions. This latter feature was exploited by using a co-axial nozzle extrusion system that allows to supply simultaneously the bioink and the primary crosslinking solution (CaCl2).
After system optimization, we exploited the setup for the production of aligned yarns of cell-laden microfibers (Figure 2). In SMTE, next to matrix composition and scaffold architecture, the major role is played by the biological component, i.e. myogenic precursors. In this study, we initially fabricated hydrogel fiber yarns loaded with mouse-derived perivascular progenitor cells, mesoangioblasts (Mabs), and tested them for generation of skeletal muscle both in vitro and in vivo.
Mabs represent a primary myogenic cell population extensively characterized for its capability to robustly differentiate into skeletal muscle in vitro and in vivo [34]; human Mabs, have been employed in a clinical trial for Duchenne Muscular Dystrophy (Eudract Number: 2011-000176-33). As a preliminary test, we encapsulated Mabs within the fibers at different cell densitiesranging from 1 to 5x10 7 cells/mL. As a result, we found that the best myogenic differentiation was obtained for concentration around 2÷2.5x10 7 cells/mL (data not shown), a value in agreement with previously published studies [7]. As shown in Figure 2, Mabs had a typical round morphology upon encapsulation (day 0); cell elongation and spreading took place since the very first days of culture and was clearly appreciable at day 3 in vast areas of the engineered constructs. This was followed by an abundant myogenesis, with the formation of long-range, multi-nucleated myotubes within two weeks of culture (day 15). These myotubes were characterized by a remarkable aligned architectural organization.
To further validate our approach and especially its cell-compatibility towards primary cells, a live/dead staining based on calcein-AM (live cells in green) and propidium iodide (dead cells in red) was performed on the scaffolds at day 1 and 3. In both cases, the majority of Mabs appeared alive, with notable cell elongation at day 3 (see Figure S1), thus confirming the suitability of our approach for the biofabrication of myo-substitutes.

3.3
In vitro characterization of mouse-derived myo-substitutes.
In order to assess the effectiveness and advantages of our wet-spinning system in the biofabrication of functional myo-substitute, we compared the in vitro myogenesis potential of Mabs in two hydrogel systems: a bulk gel and a wet-spun fiber yarn. The two systems are characterized by the same matrix composition and cell density but significantly different 3D architecture. In vitro myogenesis in the engineered myo-substitutes was evaluated by immunofluorescence staining against myosin heavy chain (MHC, a muscle-specific marker highly expressed in fully differentiated myotubes) after 20 days of culture (Figure 3). Mabs-laden yarns revealed a substantial myotube parallel organization and positivity for MHC (Figure 3 upper panel), while Mabs embedded in the bulk gels showed a similar MHC expression but an entangled, disordered myotube layout (Figure 3 bottom panel). Being a critical parameter for myo-substitutes functionality, myotube organization was further analyzed quantitatively through fluorescent image analysis. As shown in the polar plots in Figure 3, in Mabs-laden yarns all myotubes were distributed within ± 30° of average fiber orientation, whereas, in the case of bulk gels, myotube orientation was random with no preferential directions.

Tibialis Anterior (TA) regeneration by means of Mabs-derived myo-substitutes.
Based on the promising in vitro results, Mabs-laden myo-substitutes were further tested in vivo in a volumetric muscle loss (VML) model following a surgical protocol previously established by our group [29]. The approach consists of surgically dislodging around 90% of the tibialis anterior muscle (TA) while leaving the surrounding tendons intact and in place (Figure 4a). In order to precisely distinguish Mabs-driven regeneration from that originating from host's cells, the former were labelled by lentiviral transduction with nuclear b-Galactosidase (nLacZ) before grafting [29,34].
The results in the control animals (untreated) after 20 days post-implantation showed that, their leg presented a void where the TA was resected. Most of the missing muscular mass, was unable to self-regenerate after VML injury (Figure 4b). Conversely, wet-spun myo-substitute containing nLacZ-expressing Mabs implanted into the contralateral leg yielded a significant muscle regeneration whereby the TA defect was replenished by muscle-like tissue (Figure 4b).
Furthermore, the recovered TA from the treated legs (implanted with wet-spun myo-substitutes) demonstrated larger mass compared to untreated control legs, as evidenced by the measured wet weights of both groups (Figure 4c, d).  Figure   5 bottom panels). Accordingly, the disclosed sarcomeres were detected in close proximity to black X-Gal labeled nuclei, indicating the nLacZ/Mabs based myo-substitute were able to recapitulate all the morphological features of functional skeletal muscle tissue (Arrows in Figure 5 bottom panels). As widely demonstrated in our previous publication, acellular PF is completely inefficacious in VML TA mouse model, hence we prevent this further control avoiding unnecessary waste of animals [29].
In addition to the muscle-like architectures there were also uncharacteristic gaps within the muscle tissue, these being noticeable as black voids in the cross-sections of the grafted TA. We speculate that these voids are due to the presence of some alginate residues from wet-spun fibers within the engineered muscle, which in physiological conditions degrades slowly and exclusively through the gradual exchange of calcium with monovalent ions [35]. Although this would appear as imperfections in the muscle repair, we hypothesize that such alginate could also exert a beneficial effect in the grafted muscle by acting as pillars, thus rapidly guiding myotube organization around them and providing additional mechanical support to the neo-tissue. Moreover, because of the small fiber size, such gaps do not significantly impair the functionality of the muscle as demonstrated by absolute force measurements of the muscle contraction (see Figure 7). Regenerated TA derived from nLacZ/Mabs-based myo-substitutes showed neo-vascularizationa key feature for artificial tissue subsistence and engraftment. By immunolabelling the explanted grafts for von Willebrand factor (vW, specifically marking endothelial cells) and dystrophin (Dys, a key protein located in proximity of myotube membranes essential for proper muscle functionality, Figure 6a), it was possible to demonstrate that myotubes containing LacZ positive nuclei were surrounded by small vessel and capillaries (Figure 6b-e). In order to quantify vessel density in the reconstructed TA side, normal muscle tissue (native tissue) and reconstructed TA portion (graft) has been compared for vW fluorescent signal area and intensity by ImageJ analysis (Supplementary Figure 2). The evaluation revealed no statistically significant differences among the analyzed muscle moiety, further confirming the good vascularization of the reconstructed TA (Figure 6f). According to, vessel caliber and numbers were also evaluated by smooth muscle actin (SMA) vessel labelling marker. Despite an apparent larger area of the vessels in the normal muscle tissue (native) due to the greater caliber, in the reconstructed TA portion the number of the scored vessels is remarkably higher, confirming vW results and revealing a more than satisfactory vascularization of the reconstructed TA (Supplementary Figure 3).
In addition to the morphological analysis, we also carried out a functional assessment of the harvested engineered muscle to evaluate TA muscular activity following massive ablation. Muscle  [36]. The testing consisted of a direct electrical stimulation of the common peroneal nerve and the simultaneous measurement of the force generated by the TA. The results demonstrated a higher force production in the engineered TA compared to the control, reaching almost a two-fold increase of tetanic force for frequencies around 100 Hz (Figure 7d).
Besides all the above analyzed parameters demonstrating the completeness of the reconstructed TA in terms of histoarchitecture and functionality, we have not leaved out the replenishment of stem cell niche. Hence, we investigated by means of immunofluorescence the presence of Pax-7 positive muscle resident stem cells, namely satellite cells, in the engineered TA muscle. Notably, we identified satellite cells in between LacZ positive myotubes (arrowheads in Supplementary   Figure 3), revealing also the provision of myogenic stem cells in the reconstructed mouse TA.
Taken together, the morphological and functional characterization of the engineered TA reveal an unprecedented reconstructive capability of our system, resulting in a swift recovery of the VML defect. The fact that the engineered TA recovered partial functionality only 20 days after implantation further underscores the remarkable speed of the muscle recovery after such a traumatic injury.

In vitro characterization of human-derived myo-substitutes.
Mouse derived myo-structures showed a remarkable capacity in generating functional myosubstitutes in vitro and in vivo. However, in order to eventually translate such technology to a clinical scenario, the performance of the proposed biofabrication approach in combination with human myogenic progenitors needs to be assessed. Stem cells stability and functionality represent a key task for fabrication of large 3D structures [37]. Mechanical forces acting on cells during bioink extrusion -such as high shear-stresses -have been demonstrated to directly impact cell proliferation and cell lineage commitment [38,39].
Moreover, different cell lineages, namely mouse Mabs and human primary myoblast, could react differently to the bioink mechano-chemical properties as extensively demonstrated for other myogenic stem/progenitor cells [40,41]. Hence, we performed a preliminary set of in vitro experiments with the sole intent of demonstrating the compatibility of the proposed approach with primary human myoblasts (hMyob) and verifying cell stability in terms of myogenic differentiation capacity. To this aim, we benchmarked -as in the case of Mabs -the myosubstitutes fabricated through the proposed wet-spinning technology with standard bulk gel preparations (Figure 8).
The hMyob confined into the tiny fibers underwent myogenic differentiation following a timeline similar to murine Mabs, with a pronounced cell elongation appreciable as early as day 3 (see black arrows in Figure 8

Discussion
The in vitro fabrication of human tissue equivalents and the in vivo restoration of tissue/organ functionalities is one of the biggest biomedical challenges of our times. Thanks to the biotechnological progress of the last two decades, substantial advances in the tissue engineering field have been achieved with the fabrication of liver tissue with sinusoids [42,43], lung functions [44], spleen [45], skeletal muscle model functionality [46,47] in vitro models that are capable of mimicking native organ morphology and functionalities.
In the specific context of skeletal muscle tissue engineering (SMTE), researchers have greatly improved myo-substitute 3D morphological organization and functionality by i) developing specific dynamic culturing protocols, which generally include an electro-mechanical stimulation of the engineered samples, ii) implementing advanced technologies such as 3D bioprinting or organs-on-a-chip, and iii) supplying bioactive molecules in combination with tailored matrices to promote myogenic differentiation. Over the past decade, a condition clinically referred to as volumetric muscle loss (VML), has been the subject of a number of studies. A variety of approaches including acellular gels [48], cell sheetderived [49], minced muscle [50] or tissue engineering grafts [29] have been proposed for VML treatment. Among these strategies, as described in a recent excellent review on this topic [20], the use of grafts delivered in combination with a myogenic cell source should be preferred to the other methods. In fact, acellular gels, as ultimately demonstrated in a recent study by Corona [51] do not offer any beneficial effect for VML recovery. Other strategies suffer from major drawbacks as well, such as scalability, difficulty of handling/surgical implantation (especially in the case of cell sheet) and limited volume of muscle tissue harvestable from the patient. All of these drawbacks complicate the clinical translation for such approaches. On the other hand, tissue engineered muscle grafts (TEMGs) offer the advantages of enhanced construct durability, ease of surgical manipulation and high scalability potential. However, due to the lack of a standard reference VML model, comparison between studies is not trivial due to the broad range of myogenic precursors and scaffold materials as well as differences in in vitro pre-culturing time, construct size, and type of VML model -i.e. animal used, size of VML defect, muscle anatomy and pennation. Therefore, in order to compare more effectively our findings with the state-of-the-art, we have selected only those studies in which a VML defect was created in the tibialis anterior of small size animals -i.e. mice and rats. A brief overview of the results described in these studies is reported in Table 1.
As can be observed in the  Table 1. State-of-the-art of TA-VML reconstruction in small size animals (mice and rats).
Despite the promising outcomes in force and muscle mass recovery, the obtained results are still unsatisfactory due to the recurring problems related to cell engraftment, differentiation efficacy, muscle fibrosis, long regeneration time (up to 6 months) and limited system scalability [8,9,48].
Scalability -being a pivotal requirement for translating animal model approaches into human clinical therapy -is a critical feature that has so far often limited the possibility to translate the aforementioned approaches into VML models of large size animals [51].
In this context, our work proposes an easily scalable SMTE approach, which is completely independent from muscle defect geometry, based on a simple yet robust biofabrication wetspinning strategy that enables the assembly of large myo-substitutes capable of efficiently guiding and speeding up cell differentiation and 3D muscle tissue organization both in vitro and in vivo.
These features are successfully achieved by assembling highly anisotropic bundles composed of tiny hydrogel fibers where myogenic precursors are geometrically confined. In spite of their simplicity, such architectures cannot be fabricated using other fiber-based technique -e.g. 3D bioprinting -due to technical limitations such as printing speed and resolution [7].
In our previous work, we have reported an advanced reconstruction of a functional mouse TA [29]; nevertheless, a long implantation time (6 months) was needed to reestablish TA mass, morphology and functionality. Moreover, the proposed strategy relied on the polymerization in situ of a liquid hydrogel-based precursor solution -a strategy suitable for the TA lodge thanks to its concave shape but hardly transferable to other muscle locations.
Here, we have demonstrated that the aforementioned limits can be overcome by employing a biofabricated wet-spun fiber yarn loaded with primary muscle precursors. Such constructs, in fact, can be easily manipulated, thus being totally independent from the transplantation muscle shape and size. Moreover, these constructs have shown an extraordinary capacity in regenerating muscle mass, vasculature and innervation thus allowing recovery of muscle functions within only 20 days after implantation upon a massive TA ablation (~ 90% muscle mass removed). Last, but definitely not least, the preliminary evidences of compatibility of our approach with human primary myoblast open a practical possibility to translate this approach into a clinical scenario.
Currently, we are investing a great deal of work to improve construct matrix remodeling in vivo by precisely tuning bioink composition and to incorporate vasculature and innervation features during the biofabrication process. The latter aspects, in fact, are undoubtedly needed to successfully translate our approach into large animal VML models. In addition, we are planning to test our approach for VML recovery in a swine model using autologous cells.
In conclusion, in light of the presented results, we think that the proposed strategy represents a breakthrough in the field and, with further refinements, may be successfully translated to the clinic as a reconstructive therapy for VML patients.

The paper explained
Problem Skeletal muscle regenerative capability is related to damage size, in other words small injuries are able to self-repair while larger lesions need therapeutic treatment and/or surgical intervention.
Muscular dystrophies and large trauma due to both accidental tearing or cancer ablation surgery lead to extensive loss of tissue, requiring a reconstructive approach to recover the muscle damage.
Within this scenario skeletal muscle tissue engineering represent a real opportunity to replace damaged or lost muscular tissue with complete and functional implantable artificial muscle. But in spite of everything there is still an open challenge related to the architectural muscular organization replica and overall the functionality in terms of vessel and nerves supply.

Results
Exploiting homemade 3D printing technology to assemble in orderly way myogenic precursor cells and supporting matrix, we were able to replace large damaged muscle mass and rapidly restore (within 3 weeks) muscular histoarchitecture and functionality in a mouse model of extensive muscle damage. Forcing parallel alignment by 3D printing innovative approach, macroscopic artificial muscular structures were generated in vitro. Upon implantation, the printed muscle substitutes supported the formation of new blood vessels and neuromuscular junctions, fundamental for artificial muscle survival and functional recovery.

Impact
These data represent a solid base for further testing the 3D printing technology here proposed in large animal size, before to eventually be translated into clinical scenarios for the treatment of a wide range of pathology eliciting muscle wasting.

Acknowledgments
We would like to thanks Giulio Cossu for critical reading. This study was supported by National CZ and RB provided surgical support for TA muscle ablation; BB performed electrophysiological analysis together with critical reading; DS provided PEG-fibrinogen characterization and production next to critical reading. MC, ST, EF and CG wrote the manuscript. All co-authors contributed to discussion and analysis of the data.     Table S1. Absolute force raw data obtained from electrophysiological set-up.