Neural cell integration into 3D bioprinted skeletal muscle constructs accelerates restoration of muscle function

A bioengineered skeletal muscle construct that mimics structural and functional characteristics of native skeletal muscle is a promising therapeutic option to treat extensive muscle defect injuries. We previously showed that bioprinted human skeletal muscle constructs were able to form multi-layered bundles with aligned myofibers. In this study, we investigate the effects of neural cell integration into the bioprinted skeletal muscle construct to accelerate functional muscle regeneration in vivo. Neural input into this bioprinted skeletal muscle construct shows the improvement of myofiber formation, long-term survival, and neuromuscular junction formation in vitro. More importantly, the bioprinted constructs with neural cell integration facilitate rapid innervation and mature into organized muscle tissue that restores normal muscle weight and function in a rodent model of muscle defect injury. These results suggest that the 3D bioprinted human neural-skeletal muscle constructs can be rapidly integrated with the host neural network, resulting in accelerated muscle function restoration.

The paper titled "Neural Cell Integration into 3D Bioprinted Skeletal Muscle Constructs Accelerates Restoration of Muscle Function" is a highly topical study. It reports outstanding quality results on inclusion of neuronal cells in bioprinting of skeletal muscle to rapidly integrate the bioprinted muscle in the host neuronal system and to accelerate the muscle regeneration. The authors report on the optimisation of the bioprinting of the muscle with an integrated neuronal component and implantation of the bioprinted muscle in an athymic nude (RNU) rat model for 8 weeks. The study reports some impressive results, including the accelerated formation of neuromuscular junctions to integrate the muscle with the host neuronal system, and evidence of rapid vascularization of the bioprinted nerve. The study also reports a functional recovery of the muscle by measuring muscle weight and muscle tetanic force measurement. The authors set their results well within the framework of recent literature and highlight a future direction of this research. The paper is expertly written and referenced and the figures are of high publication standard. The results of the study are of excellent quality and merit publication in Nature Communications. The general subject area of this study is highly topical and this study definitely sets a high benchmark for future work in this field. I can recommend publication of this study after the authors have considered a few small suggestions: 1) The authors could add a scale bar to Figure  2) As I understand from the manuscript the PCL is used as a support material for the in vitro culture pre-implantation, but it was not part of the implant. Where exactly were the pillars positioned, were the pillars on the outside of the bioprinted construct or were they also inside, as prongs, to retain the bioprinted construct in place. A slightly more detailed description of the bioprinting process and the role of the PCL pillars would help.
3) Figure 5 shows clearly a striated structure for the bioengineered muscles, but there are also clearly ~10 µm gaps observable, which I assume are produced by the remnants of the sacrificial material to make the bioprinted muscle. There seems to be a distinct difference in density of the muscle tissue definitely in between week 4 and 8 MP+NSC and MPC only ( Figure 5). Maybe the authors would like to comment on these observations. 4) The authors correctly state that "Host responses, including inflammatory response and foreign body reactions, in the regeneration process, need further investigation" but provide little detail. Do the authors expect any hostile host response to the implant, in particular if the suggested cellular components are progenitors (hMPCs) or stem cells (hNSCs). A very short discussion could be added. Figure 6C x-axis label "myofiebrs" should be "myofibers".

5)
Reviewer #2 (Remarks to the Author): The authors present a bioengineered skeletal muscle construct embedding human muscle progenitor cells and human neural stem cells. Muscle progenitor cells differentiate into aligned myotubes while neural stem cells contribute to the generation of neuronal and glial populations. The presence of neural stem cells is shown to increase the in vitro maturation and long term survival of the bioprinted skeletal muscle constructs. Moreover, the integration of neural cells promoted early signs of innervation and restored the muscle weight and contractility in a rodent model of tibialis anterior defect.
The authors clearly emphasize in the introduction the importance of engineering muscle tissue constructs that could restore the functionality of irreversibly damaged muscles. Several recent advancements in the field are presented. However, the authors should also consider other milestones, including but not limited to: -Levenberg, Rouwkema et al. Nature Biotech, 2005-Shandalov, Egozi et al. PNAS, 2014 In addition, the authors do not mention important contributions in the field of bioengineered The result section presents several characterizations of the bioprinted construct, both in vitro and in vivo. However, many aspects are not clear and they would need more data/explanations to go beyond a simple incremental study.
Here are some specific comments/questions (in order of appearance in the manuscript): -In the discussion of the effect of neural cells on myotube formation, it is not clear if the same number of human muscle progenitor cells was used in all conditions where the cell ratio was tuned. Is the number of muscle progenitor cells the same comparing 1:0 and 300:1 conditions? -The MHC staining is not always convincing. Additional staining of markers of muscle differentiation and maturation are strongly recommended -The authors present several images of the bioengineered construct, however these images should always be coupled with quantifications (e.g. images presented in Fig. 1 (quantification of co-localizations), Fig.5 (quantification of MTS)) -In Figure 1, are the authors considering single slices or projected 3D stacks? Images are not exhaustive and electron microscopy images are strongly recommended to prove the presence of neuromuscular junctions -In the quantification of Live&Dead assay, muscle+neural cell samples showed an absolute value of 94.99% viability (and not a 94.99% increase in cell viability compared to constructs embedding muscle cells only) - Figure 3D: it is not clear why the fiber is much thicker in the co-culture condition - Figure 3J: quantification is encouraged. Moreover, images are not clear and the color choice is questionable because it does not allow to correctly discriminate each marker. Also, which is the difference between co-culture and monoculture constructs? Quantification is required. Finally, electron microscopy images would provide a clearer picture of the differences between the two conditions - Figure 4: statistical differences are not well explained (also true for Figure 6). -Do bioengineered muscles contract in vitro? Is there any difference in the co-culture vs. monoculture conditions in response to electrical stimulation? This aspect would be really important when discussing the formation of neuromuscular junctions - Figure 5: it is not clear if the enlarged images come from the same anatomical region within the defect created in the tibialis anterior. Also, it is not clear if the tissue sections are collected from muscle samples with the same orientation. Finally, the authors should clearly highlight with dashed lines the defect in the muscle samples -The section "Host nerve integration of the 3D bioprinted skeletal muscle constructs" presents conclusions which are not fully supported by the data provided in Fig. 7 Some additional general questions are: -How do the authors explain that co-culture constructs have more neuromuscular junctions and AChR clusters than the Sham group? Is it physiological? -The bioengineered constructs are contributing to muscle regeneration, but are the integrating nerves functional? Is the presence of functional nerves correlating with the improved muscle function (i.e. force generation)?
-The authors discuss that factors secreted by differentiating neural stem cells might contribute to muscle maturation. Which factors? Which is the biological mechanism? -An useful control for all the experiments would be an acellular construct with an empty ECM The authors conclude the manuscript with a long discussion emphasizing that additional validations and refinements are required to completely characterize the bioengineered muscle constructs. Most of these refinements would be necessary to go beyond an incremental study for the field of muscle tissue engineering.
There are minor grammar errors and sentences which are not correctly formulated, including but not limited to: "In this study, we investigated the feasibility using the bioprinted neural cell-integrated..." "We evaluated the effect of neural cells in the bioprinted constructs with the ratio of hMPCs and NSCs for aspects of viability..."

RESPONSE TO REFEREES LETTER
Reviewer #1: The paper titled "Neural Cell Integration into 3D Bioprinted Skeletal Muscle Constructs Accelerates Restoration of Muscle Function" is a highly topical study. It reports outstanding quality results on inclusion of neuronal cells in bioprinting of skeletal muscle to rapidly integrate the bioprinted muscle in the host neuronal system and to accelerate the muscle regeneration. The authors report on the optimisation of the bioprinting of the muscle with an integrated neuronal component and implantation of the bioprinted muscle in an athymic nude (RNU) rat model for 8 weeks. The study reports some impressive results, including the accelerated formation of neuromuscular junctions to integrate the muscle with the host neuronal system, and evidence of rapid vascularization of the bioprinted nerve. The study also reports a functional recovery of the muscle by measuring muscle weight and muscle tetanic force measurement. The authors set their results well within the framework of recent literature and highlight a future direction of this research. The paper is expertly written and referenced and the figures are of high publication standard. The results of the study are of excellent quality and merit publication in Nature Communications. The general subject area of this study is highly topical and this study definitely sets a high benchmark for future work in this field. I can recommend publication of this study after the authors have considered a few small suggestions: 1) The authors could add a scale bar to Figure 3A, or dimensions to Figure 2B/C to give an indication of the size of the implant. Figure 3.

Response: We appreciate the reviewer's comment. We have added a scale in
2) As I understand from the manuscript the PCL is used as a support material for the in vitro culture pre-implantation, but it was not part of the implant. Where exactly were the pillars positioned, were the pillars on the outside of the bioprinted construct or were they also inside, as prongs, to retain the bioprinted construct in place. A slightly more detailed description of the bioprinting process and the role of the PCL pillars would help.
Response: We appreciate the reviewer's comment. The detailed information, including design concept, printing path, and bioprinting process, has been added in the Methods section. We have also added the printing path as Figure 3A for a better understanding. The PCL pillar only presented on the outside of the construct to support the cell-laden construct. For the implantation, the PCL structure was removed from the construct and only cellular structure was implanted.
3) Figure 5 shows clearly a striated structure for the bioengineered muscles, but there are also clearly ~10 µm gaps observable, which I assume are produced by the remnants of the sacrificial material to make the bioprinted muscle. There seems to be a distinct difference in density of the muscle tissue definitely in between week 4 and 8 MPC+NSC and MPC only ( Figure 5). Maybe the authors would like to comment on these observations.

Response:
We appreciate the reviewer's comment. We assumed that the gaps between newly formed myofibers might be the undifferentiated muscle cells in the construct. In Figure 6B, the area of myofibers/HFP increased over time. This could be evidence that the bioengineered muscle with NSCs accelerated its maturation and development.
4) The authors correctly state that "Host responses, including inflammatory response and foreign body reactions, in the regeneration process, need further investigation" but provide little detail. Do the authors expect any hostile host response to the implant, in particular if the suggested cellular components are progenitors (hMPCs) or stem cells (hNSCs). A very short discussion could be added.
Response: We appreciate the reviewer's comment. According to the reviewer's comment, a short discussion has been added to the Discussion. Figure 6C x-axis label "myofiebrs" should be "myofibers".

Response: The typo has been corrected. Thank you for the correction.
Reviewer #2: The authors present a bioengineered skeletal muscle construct embedding human muscle progenitor cells and human neural stem cells. Muscle progenitor cells differentiate into aligned myotubes while neural stem cells contribute to the generation of neuronal and glial populations. The presence of neural stem cells is shown to increase the in vitro maturation and long term survival of the bioprinted skeletal muscle constructs. Moreover, the integration of neural cells promoted early signs of innervation and restored the muscle weight and contractility in a rodent model of tibialis anterior defect.
1. The authors clearly emphasize in the introduction the importance of engineering muscle tissue constructs that could restore the functionality of irreversibly damaged muscles. Several recent advancements in the field are presented. However, the authors should also consider other milestones, including but not limited to: Response: We appreciate the reviewer's thorough literature review in the field of skeletal muscle tissue engineering. We have added the studies mentioned above to the Introduction.
The result section presents several characterizations of the bioprinted construct, both in vitro and in vivo. However, many aspects are not clear and they would need more data/explanations to go beyond a simple incremental study. Here are some specific comments/questions (in order of appearance in the manuscript): 4. The authors present several images of the bioengineered construct, however these images should always be coupled with quantifications (e.g. images presented in Fig. 1 (quantification of co-localizations), Fig.5 (quantification of MTS)). Figure 1 and Figure 5. Figure 1, are the authors considering single slices or projected 3D stacks? Images are not exhaustive and electron microscopy images are strongly recommended to prove the presence of neuromuscular junctions.

In
Response: We appreciate the reviewer's thorough review and comment. Unfortunately, we had a technical difficulty to prepare 2D culture sample for the electron microscopic images. Alternatively, we have added 3D stacked confocal microscopic images to support the colocalization of βIII tubulin + neuron and AChR (NMJ formation) as shown in Supplementary Figure 4.
6. In the quantification of Live&Dead assay, muscle+neural cell samples showed an absolute value of 94.99% viability (and not a 94.99% increase in cell viability compared to constructs embedding muscle cells only).
Response: We appreciate the reviewer's comment. This has been corrected.
7. Figure 3D: it is not clear why the fiber is much thicker in the co-culture condition.
Response: We appreciate the reviewer's comment. The image of the MPC+NSC shows two printed cell-laden struts overlapped. In order to avoid any misinterpretation, the image has been modified.
8. Figure 3J: quantification is encouraged. Moreover, images are not clear and the color choice is questionable because it does not allow to correctly discriminate each marker. Also, which is the difference between co-culture and monoculture constructs? Quantification is required. Finally, electron microscopy images would provide a clearer picture of the differences between the two conditions.
Response: We appreciate the reviewer's comment. As the reviewer's recommendation, we have added the number of AChRs/HFP ( Figure 3M). The quantification result showed that a higher number of AChRs were expressed in the co-culture than the monoculture constructs. In addition, we quantified the number of NMJs (βIIIT + AChR + )/HFP ( Figure  3N). To clearly show the βIIIT + AChR + NMJs, the construct was stained with only βIIIT and AChRs using two colors (Green, βIIIT; Red, AChR + ) (Supplementary Figure 4).
9. Figure 4: statistical differences are not well explained (also true for Figure 6).

Response:
We appreciate the reviewer's comment. We have improved the explanation of statistical differences in Figure 4 and Figure 6.
10. Do bioengineered muscles contract in vitro? Is there any difference in the co-culture vs. monoculture conditions in response to electrical stimulation? This aspect would be really important when discussing the formation of neuromuscular junctions.
Response: We appreciate the reviewer's comment. We agree that the muscle contractility in vitro in response to electrical stimulation can be an indicator of the functionality of NMJs. Even though we tried to measure the in vitro contractility of the bioprinted muscle constructs, the contractile force generated by the differentiated MPCs in the bioprinted muscle constructs was not sufficient to be measured. Since our approach aimed to develop the implantable muscle constructs to treat extensive muscle defect injuries, the bioprinted constructs were cultured in vitro only for 4-5 days in the differentiation medium before implantation. This could be a reason that the contractile force was not sufficiently generated by the bioprinted constructs. Alternatively, we performed calcium uptake imaging in vitro to validate the functionality of the NMJs in terms of synaptic transmission and calcium channels opening which results in muscle contraction. We observed the increased number of cells with a high level of intracellular calcium in the MPC+NSC constructs compared with the MPC only constructs (Supplementary Figure 5).
11. Figure 5: it is not clear if the enlarged images come from the same anatomical region within the defect created in the tibialis anterior. Also, it is not clear if the tissue sections are collected from muscle samples with the same orientation. Finally, the authors should clearly highlight with dashed lines the defect in the muscle samples.
Response: We appreciate the reviewer's comment. All the sections were collected from tibialis anterior muscle samples with the same orientation (longitudinal-section of the tibialis anterior muscle of each group). The original defect area of each group was highlighted with a dashed line, and the area where the enlarged images were obtained was indicated with a solid line.