Multiphoton microscopy and tissue clearing for 3D characterization of the vasculature and fibrosis remodeling in rat dystrophic skeletal muscle

Multiphoton microscopy and tissue clearing for 3D characterization of the vasculature and fibrosis remodeling in rat dystrophic skeletal muscle | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Multiphoton microscopy and tissue clearing for 3D characterization of the vasculature and fibrosis remodeling in rat dystrophic skeletal muscle Ibrahim Hassani, Mireille Ledevin, Chantal Thorin, Tony Fiore, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6156479/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Oct, 2025 Read the published version in Scientific Reports → Version 1 posted 11 You are reading this latest preprint version Abstract Duchenne muscular dystrophy (DMD) is characterized by repeated cycles of muscle fiber necrosis/regeneration and their progressive replacement by fibrous and adipose tissues. Vascular abnormalities are also reported as pathophysiological hallmark. Traditional investigations of these morphological changes rely on two-dimensional (2D) assessments of thin tissue sections. However, they fail to capture the global spatial distribution and structural disorganization within the muscle. In this study, we developed a 3D approach combining multimodal microscopy and tissue-clearing methods to investigate the properties of microvascular and connective tissue networks in a dystrophic context. By using segmentation techniques based on deep learning models, we established a dedicated 3D image analysis workflow. We analyzed samples from healthy and dystrophic rats, quantifying key parameters for vascular and connective tissue compartments. We showed a profound spatial reorganization of the vascular network in dystrophic muscle, characterized by its embedding within connective tissue and a consequent reduction in physical interactions with muscle fibers. Our findings demonstrate that this novel imaging approach provides detailed insights into the extent of remodeling in the vascular system and connective tissue of dystrophic muscle. It holds significant potential as a powerful tool for monitoring disease progression and evaluating the impact of therapeutic interventions. Physical sciences/Optics and photonics/Optical techniques/Microscopy/Multiphoton microscopy Biological sciences/Computational biology and bioinformatics/Image processing Health sciences/Anatomy/Musculoskeletal system/Muscle/Skeletal muscle muscular dystrophy vasculature fibrosis clearing multiphoton microscopy image analysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Duchenne muscular dystrophy (DMD) is a severe X-linked recessive disorder caused by mutations in the dystrophin gene 1 , leading to a complete lack of the dystrophin protein 2 , 3 . This deficiency destabilizes the plasma membrane, triggering repeated cycles of muscle fiber necrosis and degeneration, followed by their replacement with adipose and fibrotic tissue 4 . DMD patients exhibit muscle atrophy and progressive muscle weakness, resulting in loss of ambulation, and develop cardiomyopathy and respiratory insufficiency ultimately leading to premature death typically between 20 and 30 years of age 5 , 6 . Fibrosis, corresponding to the excessive deposition of fibrillar collagen types I and III from extracellular matrix 7 , further impairs contractile muscle function and significantly contributes to the progression of the disease 4 , 8 . In addition to muscle fiber degeneration, vascular abnormalities including impaired blood flow and reduced capillary density have been documented 9 – 11 . Interstitial fibrosis exacerbates vascular dysfunction by disrupting the structural integrity of the vascular network, notably leading to impaired nutrient and oxygen exchange with muscle fibers 4 , 12 . These vascular changes, along with fibrosis, play a crucial role in pathophysiology of DMD. Their assessment has traditionally relied on 2D methodologies, as seen in previous studies 4 , 13 , 14 . However, such approaches inherently constrain the ability to fully understand the complex spatial relationships and interactions among vascular, interstitial, and muscular compartments. Yet, a detailed 3D analysis of these interactions is essential for gaining deeper insights into disease mechanisms and potential therapeutic targets. To address this need, nonlinear microscopy techniques such as second harmonic generation (SHG) and two-photon excited fluorescence (TPEF) offer valuable tools 15 – 17 . Indeed, these techniques provide detailed information on the degree and type of changes observed in vascular and connective tissue network within dystrophic muscles 8 , 18 . This enables a more comprehensive understanding of the structural and functional alterations associated with DMD. SHG is a second-order nonlinear optical process that occurs in non-centrosymmetric structures, such as collagen, myosin bands, and microtubules 15 , 16 , 19 . This property makes it particularly suitable for imaging of these components under both physiological and pathological conditions 18 – 22 . SHG microscopy enables the visualization of collagen structures in thick samples, offering insights into their organization based on their volume, shape factor, and orientation 18 , 21 . Unlike conventional fluorescence microscopy that relies on single-photon absorption, TPEF requires the simultaneous absorption of two lower-energy photons to excite fluorophores. It provides significant advantages, such as reduced photobleaching 23 , multiple absorption spectra peaks 24 and intrinsic optical sectioning capabilities 23 , making it suitable for imaging thick tissues. In 3D, TPEF-based imaging of the vasculature allows detailed morphological characterization of vessels within tissues, enabling their analysis as interconnected networks 25 . This is crucial for advancing understanding of the structural and functional dynamics of vascular networks. However, effective 3D imaging requires overcoming the challenges posed by the inherent opacity of biological tissues 26 – 28 . Optical clearing methods address this issue by allowing light to penetrate deeply into tissues while minimizing the strong diffusion and absorption of biological material. These techniques work on a common principle: light-scattering molecules within tissues, which contribute to opacity, are removed or replaced with solutions that have a uniform refractive index. This process significantly reduces light scattering, rendering large biological samples transparent and suitable for high-resolution imaging. Over the past decades, numerous optical clearing methods have been developed, each tailored to optimize imaging outcomes for various biological samples 26 , 29 . In the present study, we developed a methodology to assess the 3D remodeling of vascular and connective tissues in dystrophic muscle. We adapted and combined tissue-clearing protocols, specifically iDISCO + 30 and CUBIC 31 to achieve efficient muscle tissue clearing while maintaining compatibility with vascular labeling. Through this approach, we successfully imaged vasculature and fibrosis in cleared skeletal muscle tissue, using multiphoton microscopy. To assess the remodeling of dystrophic muscle, we designed and implemented an image analysis workflow that combines deep learning-based methods for the segmentation of vasculature and muscle fibers. We then applied this workflow to muscles from healthy and dystrophic rats aged 1 year, aiming to characterize 3D alterations in vascular morphology and how fibrosis contributes to impaired vascular function. Our findings demonstrate for the first time that 3D imaging analysis enables a detailed characterization of the structural interplay between vasculature, muscle fibers, and connective tissue in dystrophic muscle. This approach may serve as a valuable tool for studying the progression of muscle diseases and assessing the impact of novel treatments. . Results Microvessels are small blood vessels located at the periphery of myofibers, including capillaries, terminal arterioles, and terminal venules 32 , 33 . Standard 2D histological analysis of dystrophic muscle highlights capillarization defect as key feature of microvascular damage We first aimed to assess vascular impairment in dystrophic muscle using a 2D approach. To achieve this, we analyzed two global parameters (microvessel-to-fiber ratio and microvessel density) and two specific metrics (microvessel contact and sharing factor), which are commonly employed 34 – 36 to evaluate vasculature in skeletal muscle. These analyses were conducted on Biceps femoris muscle sections from 1-year-old WT and DMD mdx rats. TPEF observation of transverse sections stained with LEL DyLight 594 showed the distribution of microvessels within bundles of muscle fibers that were visualized by green auto-fluorescence signal (Fig. 1 a, left panel). A binary mask was applied to the entire section to delineate the muscle fiber area (Fig. 1 a, middle panel), allowing isolation of endomysial vessels. These vessels were then automatically detected using the Segment.ai module in NIS-Elements software (Fig. 1 a, right panel). Using this method, we determined a microvessel-to-fiber ratio of 1.05 ± 0.08 and 1.5 ± 0.09 in DMD mdx and WT rats, respectively (Fig. 1 b, top). Despite a slightly higher ratio in WT rats compared to DMD mdx ones, no statistically significant difference was observed. Additionally, microvessel density, defined as the number of microvessels within the endomysial area, was measured as 268 ± 17 mm − 2 and 243 ± 22 mm − 2 in DMD mdx and WT rats respectively, revealing a similar tissue distribution pattern (Fig. 1 b, bottom). To investigate the interactions between microvessels and individual muscle fibers, we automated the measurements of microvessel contact and the sharing factor using the GA3 module in NIS-Elements software. Muscle fibers and microvessels were detected on 2D cross-sections using deep learning models, i.e ., SegmentObject.ai and Segment.ai , respectively (Fig. 2 a, left panel). Details of the training methodology are provided in Materials and Methods section. To better capture the spatial relationship, the binary mask representing microvessels was slightly expanded to define a proximity area around each muscle fiber. A binary operation was then performed between the muscle fiber and microvessel masks to identify regions of proximity (Fig. 2 a, right panel). These regions were used to uniquely associate each proximity region with a corresponding fiber and vessel, enabling precise quantification of microvessel-to-fiber interactions. While DMD mdx rat muscle only shows a tendency towards a decrease in the number of microvessel contacts per fiber compared to WT rats (3.2 ± 0.24 vs. 3.7 ± 0.17, respectively; Fig. 2 b, top), it is characterized by a significant increase in the sharing factor, which represents the number of fibers served by a single microvessel: 2.6 ± 0.07 (DMD mdx rats) vs. 2.4 ± 0.03 (WT rats) (p < 0.01) (Fig. 2 b, bottom). Collectively, these findings indicate that 1-year-old DMD mdx rats exhibit an altered organization of the microvasculature. Multimodal 3D imaging reveals a newly formed vascular network and reduced interaction between muscle fiber-vessel in dystrophic muscle To assess whether 3D imaging analysis could offer a more precise and comprehensive investigation of vascularization, with a focus on vascular network connectivity and geometry, we utilized the same muscle samples as in previous experiments, but this time in thick section format. Initially, we identified the most appropriate clearing protocol for thick skeletal muscle, which was a combination of the iDISCO + and CUBIC methods, as described previously 37 . We then validated that the staining with LEL DyLight 594 is compatible with this protocol. After this fine-tuning stage, we generated Z-stack images from cleared and LEL DyLight 594-stained 1-mm-thick sections. Green muscle autofluorescence and red fluorescence of stained vessels were acquired with multiphotonic microscope up to a depth of 350 µm (Fig. 3 a, first column). This allowed us identifying bundles of muscle fibers and microvessel network in depth. We generated images exhibiting a high signal-to-noise ratio (SNR) that facilitates the segmentation of the vasculature across the entire field of view (FOV) measuring 1.052 × 1.052 × 0.350 mm. The muscle fibers segmented from their endogenous green fluorescence (Fig. 3 a, second column) and microvessels segmented from their red fluorescence staining (Fig. 3 a, third column) were color-coded to enhance their visibility. Vascular surfaces embedded within muscle fibers were detected by applying an "AND" logical function between the binary masks of muscle fibers and vasculature (Fig. 3 a; fourth column). We showed that the endomysial vascular network penetrating muscle consists of branched capillaries that appear much more organized in WT rat muscle compared to that of DMD mdx rat. It runs parallel to the muscle fibers in WT rat muscle while it displays multiple directions in DMD mdx rat muscle. For 3D quantitative analysis, we developed a pipeline within the GA3 module in NIS-Elements software and exploited the same deep learning models used for the 2D analysis to detect objects. These models were applied to individual image slices, and the data were reconstructed into 3D volumetric representations. 3D microvessel density, expressed as the volume occupied by vessels relative to fiber volume, was slightly higher in DMD mdx rat muscle (0.02 ± 0.002) compared to WT rat muscle (0.014 ± 0.001) (Fig. 3 b, left), but the difference was not statistically significant. However, the branching count per muscle fiber volume was significantly higher in DMD mdx rats (21,596 ± 1,981) than in WT rats (13,428 ± 2,094), indicating increased vascular network ramification in the dystrophic context (p = 0.017; Fig. 3 b, middle). To facilitate efficient nutrient and oxygen exchange, capillaries are wrapped around muscle fibers within grooves, forming close contacts—a well-documented phenomenon 38 . In Fig. 3 b, right, we explored this aspect by assessing the embedding surface of the vasculature within muscle fibers. To quantify this, we performed a binary operation using muscle fiber and microvessel masks to calculate the shared surface area, normalizing it to the number of fibers. Our analysis revealed a significant decrease in the contact surface area in DMD mdx rats (8.1 ± 0.6 µm²) compared to WT rats (20.4 ± 3.1 µm²), indicating impaired vasculature-muscle interactions in the dystrophic context (p < 0.001; Fig. 3 b, right). Additional morphometric analyses were conducted to assess the morphology of the vascular network. Vessel diameter and length were quantified using the filament algorithm in Imaris software (version 10.1, Oxford Instruments) (Fig. 4 a). The algorithm's 'loops' function was trained on data from four distinct regions before being applied to analyze three images (1.052 × 1.052 × 0.350 µm) of both WT and DMD mdx rat muscle. From each image, 150 microvessels were selected for analysis. The microvessel diameter was slightly reduced in DMD mdx rats (5.6 ± 0.1 µm) compared to WT ones (5.9 ± 0.1 µm) (p < 0.05; Fig. 4 b, top). We also determined that the microvessel length was 37.44 ± 1.3 µm and 54.218 ± 3 µm in DMD mdx and WT rat muscle, respectively (Fig. 4 b, bottom). It was significantly shorter in dystrophic context (p < 0.0001). Taken together, these results demonstrate the presence of a highly reorganized vasculature in dystrophic rat muscle, defined by microvessels that are more branched, but also shorter, wider and less in contact with muscle fibers. These findings also highlight the added value of 3D multimodal exploration for analyzing a spatially networked structure such as the vasculature. Multimodal 3D imaging is relevant for jointly assessing the intensity of microvascular and fibrotic remodeling in dystrophic muscle Fibrosis represents one of the main pathological features of dystrophic muscle 4 , 39 , 40 . It results in a change in the structural organization of muscle bundles, with the vascular network moving away from muscle fibers 13 . Recent evidence suggested that the disorganization of collagen network constitutes a potential biomarker of DMD 8 , 18 . Building on our findings regarding microvasculature, we next aimed to characterize the endomysial connective tissue in 3D and examine its interaction with the vascular network. This approach provides a more comprehensive view of muscle remodeling at the structural level. For that purpose, we analyzed forward 3D SHG on cleared muscle cross-sections. All images were acquired with consistent laser power, detector sensitivity, and gain settings. Image analysis of SHG collagen fibers was performed using the GA3 module of NIS-Elements Software and a false-coloring was applied to the binary mask. A marked increase in collagen deposition was observed in DMD mdx rats compared to WT ones, as expected (Fig. 5 a). Endomysial SHG density, corresponding to the volume of collagen network surrounding each muscle fiber relative to muscle fiber volume, was 2.79% ± 0.10% and 0.51% ± 0.11% in DMD mdx and WT rats, respectively (Fig. 5 b, left, top). It was significantly higher in dystrophic muscle (p < 0.0001). A detailed analysis of the collagen network was conducted by evaluating the morphology and spatial distribution of SHG collagen objects. As expected, DMD mdx rats were overrepresented across all volume classes, indicating increased collagen accumulation along the muscle fiber in depth (Fig. 5 b, top right). Additionally, SHG collagen objects were more frequently distributed in shorter distance classes in DMD mdx rats compared to WT ones, suggesting a denser collagen network (Fig. 5 b, bottom left). Moreover, these collagen structures exhibited a less elongated shape in DMD mdx rat muscle than that of WT rat one (p < 0.001; Fig. 5 b, bottom right). Next, we identified 3D regions where DyLight594-stained microvessels overlapped with SHG collagen network, using the binary masks of both structures previously generated (Fig. 6 a). The volume of these overlapping regions was measured and normalized to the total microvessel volume to provide a quantitative assessment of the interaction between microvasculature and connective tissue (Fig. 6 b, right). We determined that the volume of overlap between microvessels and SHG collagen was 10.07 ± 1.66% and 1.79 ± 0.43% in DMD mdx and WT rats, respectively (p < 0.001; Fig. 6 b). Overall, these data provide new information on the accumulation of connective tissue in dystrophic muscle, in terms of volume, concentration and shape, within which part of the microvasculature is embedded. Discussion In this study, we provide original and compelling data demonstrating the contribution of multimodal 3D imaging in providing insights into the spatial organization and interactions between vasculature, muscle fibers and connective tissue. The simultaneous 3D visualization of the myofibers, microvessels and/or connective tissue, followed by quantitative analysis on segmented objects, is a powerful approach for qualifying the nature and intensity of remodeling in dystrophic muscle, Disruption of microvasculature is a pathological feature of muscular dystrophy 12 , 41 , 42 . It has negative impacts on the oxygen, nutrients delivery to the muscle fiber and muscle development 33 . Like development of fibrosis, vasculature changes in dystrophic muscle are typically assessed using 2D analysis of tissue cross-sections that unfortunately exhibits some limitations 43 . Here, we introduced a novel 3D approach to analyze the structure of the vasculature and collagen network, aiming to gain more detailed insights into their spatial organization within the muscle bundles. For that, we decided to employ a multimodal imaging technique combining TPEF and SHG to investigate fluorescent lectin-stained vasculature and label-free SHG collagen on cleared rat dystrophic muscle, respectively. Unlike previous studies that relied on conventional histological methods to assess vascular and fibrotic remodeling 4 , 40 , our dual-modality imaging approach enabled simultaneous visualization of TPEF and SHG signals from both structures and analysis of their interactions within muscle fibers. Combined with iDISCO + and CUBIC clearing methods, it improved analysis of microvessel organization and collagen network connectivity over at least 1mm in depth. In a previous study, we have demonstrated the usefulness of tissue clearing and SHG imaging to study fibrosis in DMD rat heart using CUBIC protocol 18 . Here, we combine CUBIC with iDISCO + to firstly enhance the penetration of the fluorescent probes for the vessels thanks to the organic solvent pretreatment in iDISCO + protocol and secondly to improve the tissue clearing and to avoid the tissue shrinkage thanks to hyperhydration pretreatment in CUBIC protocol. The combination of iDISCO + and CUBIC tissue clearing methods has already been successfully described for in toto imaging of whole ovarian follicle with its extrinsic vascular and neuronal networks 37 . In addition to this original 3D imaging approach, we developed an image analysis protocol specifically designed to handle the complexity of the objects being measured. Deep learning have emerged as powerful tools alongside traditional hand-crafted analysis workflows, offering the potential for more accurate and efficient results 44 . Recent studies have demonstrated the effectiveness of deep learning U-net architecture to segment blood vessels in heart tissue sections, yielding accurate vascular morphometrics such as vessel length and density 44 . In our study, 3D features of the vasculature and collagen network were extracted using a U-Net model. We were able to measure vessel density, branching and sharing as well as to characterize the collagen network in terms of density packing and thickness/elongation, corroborating the conclusions of Lapierre-Landry and co-author's work on the applicability of these tools 44 . Recently, several studies have demonstrated major structural and functional vascular network defects in the main model of DMD, the mdx mice 12 , 14 . Alterations that increase in severity with age correspond to abnormal flow capacity, vessel diameter change or vessel density modification 12 , 39 . Based on their findings, these studies emphasize the crucial role of the vascular network damage in the pathophysiology of DMD. With this study, we present significant disruptions of the vascular network in another model of DMD, namely the DMD mdx rat 45 . Two-dimensional analysis revealed an increased sharing factor compared to that observed in WT animal, which are consistent with the previous findings in mdx mice 12 . Importantly, 3D investigations of the vasculature provided additional information, confirming the existence of highly significant morphological alterations in dystrophic rat muscle. This includes a high degree of microvessel ramifications, which could correspond to a compensatory response to the altered metabolic demands in dystrophic context 12 , 41 . In parallel, we also found a decrease in vessel-muscle fiber contact, which may indicate impaired functional interactions between the vascular network and muscle fibers. Similar vascular alterations were reported by Latroche et al. in a 1-year-old mdx mouse model, where structural abnormalities were characterized by a significant increase in branching and functional impairments 12 . Regarding fibrosis, 3D SHG imaging allowed us to show significant changes in the morphology and distribution of the collagen network, which were not detectable through 2D image analysis. Specifically, we demonstrated that dystrophic rat muscle was defined by a thicker and more densely packed connective tissue. It is important to note that the architecture of collagen fibers has been reported to be associated with increased passive stiffness in fibrotic skeletal muscles of mdx mice 4 , 8 . This is particularly interesting as the amount of fibrotic tissue in mdx mouse muscles correlates with contractile function. Given the much more pronounced tissue phenotype in the dystrophic rat model compared to the mdx mouse 45 as well as the variability observed depending on the type of muscle considered, it would be valuable to determine whether the same observation holds in the model used here. This could help to better understand potential impacts on contractile muscle function. Additionally, 3D SHG imaging revealed that the collagen network frequently encased vascular structures in dystrophic muscle, a phenomenon reported in DMD patients 4 . This encasing likely reflects a pathological interaction between fibrosis and vasculature. Excessive collagen deposition may restrict blood flow, contributing to the increased distance between capillaries and muscle fibers. These findings suggest a pathological feedback loop, where fibrosis exacerbates vascular pathology, which in turn drives further muscle degeneration. The combined analysis of fibrosis and vascular alterations using SHG and TPEF imaging underscores the complex interplay between these two processes in DMD. The findings suggest that therapeutic strategies targeting both extending the vascular network and fibrosis reduction could be effective in managing DMD. Enhancing vascular integrity could improve tissue perfusion and mitigate muscle damage, potentially alleviating some symptoms of the disease. Concurrently, reducing fibrosis could restore the structural and functional integrity of the muscle. Future studies should investigate the temporal dynamics of vascular and fibrotic remodeling in DMD to identify critical windows for intervention. Early therapeutic strategies targeting these processes may help slow or halt disease progression. Additionally, advanced imaging approaches, such as those used in this study, could be used to evaluate the efficacy of potential therapies in preclinical and clinical settings. Conclusion Combination of multimodal 3D SHG/TPEF imaging with iDISCO + and CUBIC tissue clearing and deep learning-based image analysis provided new tools to get valuable insights into the structural and spatial features of microvessels and collagen network in DMD. This approach is particularly well suited for producing an in depth characterization of the extent of deterioration observed in the vasculature and collagen network. Through this original approach, we demonstrated that dystrophic muscle features a spatially well-developed vascular network made up of small microvessels whose interactions with muscle fibers are limited by an encasing connective tissue. The new quantitative 3D imaging technique described here may greatly enhance the understanding of the natural progression of this disease. Additionally, it could serve as a relevant tool for assessing the impact of biotherapies on vascularization and fibrosis in preclinical studies on DMD. Methods Animals and tissue sample . All experiments were carried out in accordance with the ARRIVE guidelines ( https://arriveguidelines.org ). Wild-type (WT) and dystrophic (DMD mdx ) Sprague-Dawley rats were obtained from the transgenic rat immunophenomic platform (TRIP; Nantes, France). Three animals of each, all aged 1 year, were included in this study. Animals were maintained in a controlled environment (temperature 21 ± 1°C, 12-h light-dark cycle) at the Boisbonne Center for gene and cell therapy (Oniris, Nantes, France; agreement number: J44273). All efforts were made to minimize suffering. Rats were provided with environmental enrichment: provision of rolls is reported to potentially modify the behavior of housed animals and reduce chronic pain. Anesthesia was induced with a mixture of ketamine (100 mg/kg, Imalgene, Merial, Lyon, France) and xylazine (10 mg/kg, Rompun, Bayer, Leverkusen, Germany), after which the rats were euthanized by intravenous administration of sodium pentobarbital (300 mg Dolethal; Vetoquinol SA, Magny Vernois). The Biceps femoris muscles were collected and separated into two portions: one was transferred to a tube containing 4% paraformaldehyde for fixation overnight at 4°C and storage. The second was embedded in paraffin for classical histological analysis. The study was approved by the Ethics Committee for Animal Experiments of the Pays de la Loire Region, France. All experiments were carried out in accordance with the French National Research Council guidelines for the care and use of laboratory animals (Permit number: APAFIS #39967-2022122112121987 v10). iDISCO+ clearing pretreatment. Fixed muscles were cut into thick muscle sections (6 x 3 x 1 mm³). After washing in PBS, sections were dehydrated stepwise in 20%, 40%, 60%, and 80% methanol (MeOH, 20846.292, VWR) / distilled water (dH₂O), with gentle shaking at room temperature (RT) for 1 hour each. Tissue sections were then washed twice in 100% methanol (1 hour, RT), followed by incubation in a 66% dichloromethane (DCM, 5895811000, Sigma-Aldrich) / 33% methanol solution overnight (ON) at RT. Tissue sections were rehydrated once again stepwise in 80%, 60%, 40%, and 20% MeOH/dH₂O with gentle shaking (1 hour, RT) each. They were then washed twice with 0.2% Triton X-100 (X100, Sigma-Aldrich) in 1X PBS (1 hour). Afterward, they were soaked (ON, 37°C) in a permeabilization solution containing 0.2% Triton X-100, 20% dimethyl sulfoxide (DMSO, D4540, Sigma-Aldrich), 0.3 M glycine (G7126, Sigma-Aldrich), and 0.02% sodium azide (S2002, Sigma-Aldrich) in 1X PBS, with gentle shaking. Vascular staining. Following iDISCO+ clearing pretreatment, sections were incubated with Lycopersicon esculentum lectin conjugated with Dylight 594 (LEL DyLight 594, Invitrogen, L32471), diluted 1:50 in labeling solution (0.2% Tween 20, 10% DMSO, 0.02% sodium azide in 1X PBS) with gentle shaking (7 days, 37°C). Then, sections were washed with 0.1% Tween 20 (P1379, Sigma-Aldrich) in 1X PBS (1 day) before proceeding with the clearing procedures. iDISCO+ clearing protocol. Stained sections were gradually dehydrated using the same procedure previously described above. They were then incubated (ON, RT) in a solution of 66% dichloromethane (DCM) / 33% methanol (MeOH), washed twice in 100% DCM (20 min) and subsequently stored in dibenzyl ether (DBE, 108014, Sigma-Aldrich) at RT. CUBIC clearing protocol. iDISCO+ cleared sections were rehydrated stepwise in 80%, 60%, 40%, and 20% MeOH / dH₂O, with each step involving gentle shaking (RT, 1 hour). Tissue sections were then washed three times with 1X PBS for 1 hour each. Afterward, they were transferred to a 50% CUBIC-L (T3740, TCI) / 50% dH₂O solution (ON, 37°C), followed by incubation with 100% CUBIC-L with gentle shaking (2 days, 37°C). Following three washes with 1X PBS, sections were transferred to 50% CUBIC-R (T3741, TCI) (overnight, RT) and clarified a second time with CUBIC-R+ (T3741, TCI) (2 days, RT). Finally, sections were mounted with fibers oriented transversally on a cavity slide (Sigma-Aldrich, BR475565) using CUBIC-R+ (refractive index 1.52), and a 0.17 μm thick coverslip was placed on top before imaging. Multiphoton microscopy imaging. Images were acquired using a laser scanning multiphoton microscope (A1R-MP+, Nikon Europe B.V., Amstelveen), coupled with a tunable laser (Insight DeepSee, Spectra Physics, France) operating in the wavelength range of 680–1300 nm, with a pulse duration of 120 femtoseconds at a repetition frequency of 80 MHz. A motorized half-wave plate (MKS-Newport, USA) was used to control the laser polarization angle, which was adjustable between 45° and 90°. The objective lens employed for imaging was a Plan-Apo Lambda S 25X silicon objective (refractive index 1.406, MRD73250, Nikon Europe B.V.), with a numerical aperture (NA) of 1.05 and a working distance (WD) of 0.55 mm. The microscope was equipped with eight non-descanned detectors (NDDs), four for backward detection and four for forward detection. Further details on the setup are given in Figure 7. 3D image acquisition. Excitation at 820 nm coupled with resonant scanning mode was used to acquire green autofluorescence from muscle fibers, red fluorescence from LEL DyLight 594-labeled vessels, and second harmonic generation (SHG) signals. Green autofluorescence and red fluorescence signals were collected in reflection using band-pass emission filters, 525/50BP and 629/56BP for muscle fibers and vessels, respectively. SHG signals were collected in both reflection mode (bSHG) and transmission mode (fSHG) using short band-pass filters, 415/10BP. The field of view (FoV) was 297 µm x 297 µm, scanned at 1024 x 1024 pixels, resulting in a pixel size of 0.29 µm. Z-series were collected with a 1 µm step size within a 350 µm depth of the specimen. Tiles and depth scanning were performed to analyze at least 100 fibers, with the step size between tiles fixed at 1 µm. Images were generated as 12-bit ND2 files, and stitching and 3D reconstruction were performed using NIS-Elements software (version 5.20, Nikon Europe B.V). Deep learning model training. Muscle fiber green autofluorescence was segmented using the U-Net-based deep learning model “ SegmentObject.ai ” integrated into NIS-Elements software. This model was particularly suitable for segmenting muscle fibers due to its capability to separate densely packed objects. The training dataset comprised seven cropped images (400 × 400 × 10 µm), each containing annotations of 20 manually identified muscle fibers. The patch size was automatically set to 512 × 512. The training process consisted of a minimum of 2,500 epochs across three batches and was conducted on an NVIDIA RTX5000 graphics unit (Driver version 516.40, NVIDIA). The trained model was applied on another dataset comprised eight manually annotated images for validation. The 4 following metrics were measured (details are provided on Supplemental data S1 and S2): - Precision: It evaluates the fraction of correctly predicted positive regions out of all predicted positive regions - Recall: Also known as sensitivity or True Positive Rate, it measures how many of the true positive regions were correctly identified. - F1 score: Harmonic mean of Precision and Recall - Interception Over Union (IoU): Also called the Jaccard Index, it measures the overlap between the predicted segmentation and the ground truth. Precision Recall F1 IoU 0,90 0,82 0,84 0,74 Table 1 Score metrics for SegmentObject.ai model trained The LEL-DyLight594-labeled vasculature was segmented using the “ Segment.ai ” in NIS-Elements software. The training dataset included three images (200 × 200 × 300 µm) with a patch size automatically set to 256 × 256 pixels. Initial identification of vasculature structures was achieved through thresholding, followed by manual refinement to create accurate binary masks, which served as the ground truth for model training. The model underwent two separate training runs, each consisting of 500 epochs. To validate the vascular segmentation, the same approach previously employed for validating muscle fiber segmentation was applied. Precision Recall F1 IoU 0,92 0,77 0,82 0,71 Table 2 Score metrics for Segment.ai model trained Both segmentation models demonstrated high accuracy (SegmentObject.ai in table 1, Segment.ai in table 2), with minimal false positives (precision were 0.90 and 0.92 for the muscle fiber model and the microvessel model, respectively). However, some false negatives were observed, as indicated by recall values of 0.82 and 0.77 for the respective models. The Intersection over Union (IoU) score was 0.74 and 0.71 for the muscle fiber model and the microvessel model respectively, while the F1-score was 0.84 and 0.82, further supporting the robustness of the segmentation performance. Image analysis of the vasculature. Two-dimensional images: vascularization was evaluated on five regions per animal (n = 6 rats), each measuring 1052 × 1052 µm, using the General Analysis 3 (GA3) module in NIS-Elements software (version 6.10.1, Nikon Europe BV, Amstelveen). Vessels were detected using the trained " Segment.ai " model, which was imported into the General Analysis 3 (GA3) module of NIS-Elements software. The GA3 pipeline (details are provided in Supplemental data S3) was then applied to 30 regions (n = 15 from 3 WT rats and 15 from 3 DMD mdx rats) to analyze the vascular structures. For feature extraction, only endomysial vessels were considered using binary mask of muscle fibers. Vessel-to-fiber ratio was determined by dividing the total number of vessels by the total number of fibers. Vessel density was calculated by dividing the number of vessels by the muscle fiber area. Vessel count per fiber and sharing factor were determined by doing binary mask between muscle fibers and microvessels. For that, a proximity area was created by expanding the binary mask of vessels by 10 µm in order to identify fibers in close contact with vessels, defining the fiber-to-vessel association. Three-dimensional images: vascularization was evaluated on four regions per animal (n = 6, rats), each measuring 526 × 526 × 350 µm, using the GA3 module. Vascular network was detected using the “ Segment.ai ” model imported into the GA3 pipeline (details are provided in Supplemental data S4) and applied to 24 regions (n = 12 from 3 WT rats and 3 DMD mdx rats). This enabled the extraction of key features: vascular network density, vascular branching and surface area of vessels embedded within muscle fibers. The extracted data were exported for statistical analysis. Three-dimensional image analysis of the collagen network. Connective tissue remodeling was evaluated through analysis of SHG endomysial collagen network on four regions of interest (ROIs) per animal (n = 6, rats), each measuring 526 × 526 × 350 µm. 3D rendering and image analysis were performed using the General Analysis 3 (GA3) module in NIS-Elements software (version 6.10.1, Nikon Europe B.V.). A pre-trained deep learning denoising algorithm was applied to the SHG collagen fiber signal to remove shot noise. Subsequently, a spatial “Laplacian of Gaussian” (LoG) filter (Gaussian σ = 2.7, kernel size = 5 × 5) was applied to identify SHG collagen network boundaries. A manual threshold based on edge detection was used for segmenting the SHG + objects. Muscle fibers were segmented using the previously trained deep learning model specifically designed for muscle fiber segmentation. To restrict the analysis to SHG collagen network in the endomysial region, the segmented binary mask of autofluorescent muscle fibers was expanded by a 3 µm kernel diameter. A logical “AND” operator was then applied between the binary masks of muscle fibers and SHG collagen network. The resulting SHG collagen network binary masks in the endomysial region were used to extract features of interest: density, volume, elongation and inter-fiber distance of SHG + objects, overlapping surface between collagen network and vessels. The GA3 pipeline is described in detail in Supplementary Data S5. Statistical analysis . Statistical analyses were performed using R software (version 4.4.1). Data are presented as the mean ± standard error of the mean (SEM). Comparisons between groups were conducted using a linear mixed-effects model, with random effects on muscle, implemented in the lme4 package. All models presented in this study were assessed for independence and normality of residuals, as recommended. For datasets that did not meet these assumptions, a Wilcoxon test was used instead. For SHG dataset, loglinear models were used to compare the distribution of measures previously discretized into classes. A significant level of 0.05 was assessed in all statistical tests. Declarations Acknowledgements We thank the staff of the Boisbonne Center (Oniris, Nantes, France) for animal care and the APEX platform (INRAE/Oniris, Center of Excellence Nikon Nantes [CENN], Nantes, France) from UMR 0703 PAnTher (INRAE/Oniris, Nantes, France) for their valuable technological support. We also extend our gratitude to the FAIR CHARM consortium (H2020 program, in which UMR 703 PAnTher is a partner) for fostering a collaborative scientific environment through discussions and the sharing of biological materials. We gratefully acknowledge financial support from Région Pays de la Loire and NeurATRIS: A Translational Research Infrastructure for Biotherapies in Neurosciences , “Investissement d’Avenir-ANR-11-INBS-0011”. The authors also thank Biogenouest (the network of technology core facilities in Western France in life sciences and the environment, supported by the Conseil Régional des Pays de la Loire) for supporting APEX. Funding Ibrahim Hassani received a PhD grant from the Association Nationale de la Recherche et de la Technologie (ANRT, grant n° 2022/0868) in collaboration with Nikon France Healthcare, succursale Nikon Europe BV. Data availability The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request. Author contributions I.H. conceived, designed, and performed the experiments, analyzed the data, and wrote the manuscript. M.L. assisted with sample clearing. C.T. performed statistical analysis, reviewed, and edited the manuscript. T.F. supervised the study, assisted with data acquisition, and reviewed and edited the manuscript. M.A.C. supervised the study, reviewed, and edited the manuscript. K.R. supervised the study, contributed to data interpretation, and participated in manuscript writing. L.D. supervised the study, conceived and designed the experiments, contributed to data interpretation, and participated in manuscript writing. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. References Bushby, K. M. D. Genetic and clinical correlations of Xp21 muscular dystrophy. J. Inher Metab. Disea . 15 , 551–564 (1992). Hoffman, E. P., Brown, R. H. & Kunkel, L. M. Dystrophin: The Protein Product of the Duchenne Muscular Dystrophy Locus. Bonilla, E. et al. Duchenne muscular dystrophy: Deficiency of dystrophin at the muscle cell surface. Cell 54 , 447–452 (1988). Desguerre, I. et al. Endomysial Fibrosis in Duchenne Muscular Dystrophy: A Marker of Poor Outcome Associated With Macrophage Alternative Activation. J. Neuropathol. Exp. Neurol. 68 , 762–773 (2009). Kohler, M. et al. Disability and survival in Duchenne muscular dystrophy. J. Neurol. Neurosurg. Psychiatry . 80 , 320–325 (2009). Mazur, W. et al. Patterns of left ventricular remodeling in patients with Duchenne Muscular Dystrophy: a cardiac MRI study of ventricular geometry, global function, and strain. Int. J. Cardiovasc. Imaging . 28 , 99–107 (2012). Gillies, A. R. & Lieber, R. L. Structure and function of the skeletal muscle extracellular matrix. Muscle Nerve . 44 , 318–331 (2011). Brashear, S. E. et al. Collagen cross-links scale with passive stiffness in dystrophic mouse muscles, but are not altered with administration of a lysyl oxidase inhibitor. PLoS ONE . 17 , e0271776 (2022). Loufrani, L. et al. Flow (Shear Stress)–Induced Endothelium-Dependent Dilation Is Altered in Mice Lacking the Gene Encoding for Dystrophin. Circulation 103 , 864–870 (2001). Latroche, C. et al. Structural and Functional Alterations of Skeletal Muscle Microvasculature in Dystrophin-Deficient mdx Mice. Am. J. Pathol. 185 , 2482–2494 (2015). Kodippili, K., Thorne, P. K., Laughlin, M. H. & Duan, D. Dystrophin deficiency impairs vascular structure and function in the canine model of Duchenne muscular dystrophy. J. Pathol. 254 , 589–605 (2021). Latroche, C. et al. Structural and Functional Alterations of Skeletal Muscle Microvasculature in Dystrophin-Deficient mdx Mice. Am. J. Pathol. 185 , 2482–2494 (2015). Nguyen, F., Guigand, L., Goubault-Leroux, I., Wyers, M. & Cherel, Y. Microvessel density in muscles of dogs with golden retriever muscular dystrophy. Neuromuscul. Disord. 15 , 154–163 (2005). Loufrani, L. et al. Absence of Dystrophin in Mice Reduces NO-Dependent Vascular Function and Vascular Density: Total Recovery After a Treatment with the Aminoglycoside Gentamicin. ATVB 24, 671–676 (2004). Strupler, M. et al. Second harmonic imaging and scoring of collagen in fibrotic tissues. Opt. Express . 15 , 4054 (2007). Fuentes-Corona, C. G. et al. Second harmonic generation signal from type I collagen fibers grown in vitro. Biomed. Opt. Express . 10 , 6449 (2019). Aghigh, A. et al. Second harmonic generation microscopy: a powerful tool for bio-imaging. Biophys. Rev. 15 , 43–70 (2023). Pichon, J. et al. Label-free 3D characterization of cardiac fibrosis in muscular dystrophy using SHG imaging of cleared tissue. Biol. Cell . 114 , 91–103 (2022). Campagnola, P. J. et al. Three-Dimensional High-Resolution Second-Harmonic Generation Imaging of Endogenous Structural Proteins in Biological Tissues. Biophys. J. 82 , 493–508 (2002). Altendorf, H. et al. Imaging and 3D morphological analysis of collagen fibrils: Imaging and morphological analysis of collagen fibrils. J. Microsc. 247 , 161–175 (2012). Bianchini, P. & Diaspro, A. Three-dimensional (3D) backward and forward second harmonic generation (SHG) microscopy of biological tissues. J. Biophotonics . 1 , 443–450 (2008). Tilbury, K. et al. Second harmonic generation microscopy analysis of extracellular matrix changes in human idiopathic pulmonary fibrosis. J. Biomed. Opt. 19 , 086014 (2014). Denk, W., Strickler, J. H. & Webb, W. W. Two-Photon Laser Scanning Fluorescence Microscopy. Science 248 , 73–76 (1990). Lang, R. T. & Spring, B. Q. Two-photon peak molecular brightness spectra reveal long-wavelength enhancements of multiplexed imaging depth and photostability. Biomed. Opt. Express . 12 , 5909 (2021). Lagerweij, T. et al. Optical clearing and fluorescence deep-tissue imaging for 3D quantitative analysis of the brain tumor microenvironment. Angiogenesis 20 , 533–546 (2017). Richardson, D. S. & Lichtman, J. W. Clarifying Tissue Clearing Cell. 162 , 246–257 (2015). Silvestri, L., Costantini, I., Sacconi, L. & Pavone, F. S. Clearing of fixed tissue: a review from a microscopist’s perspective. J. Biomed. Opt. 21 , 081205 (2016). Costantini, I., Cicchi, R., Silvestri, L., Vanzi, F. & Pavone, F. S. In-vivo and ex-vivo optical clearing methods for biological tissues: review. Biomed. Opt. Express . 10 , 5251 (2019). Tainaka, K., Kuno, A., Kubota, S. I., Murakami, T. & Ueda, H. R. Chemical Principles in Tissue Clearing and Staining Protocols for Whole-Body Cell Profiling. Annu. Rev. Cell. Dev. Biol. 32 , 713–741 (2016). Renier, N. et al. A Simple, Rapid Method to Immunolabel Large Tissue Samples for Volume Imaging. Cell 159 , 896–910 (2014). iDISCO. Susaki, E. A. et al. Advanced CUBIC protocols for whole-brain and whole-body clearing and imaging. Nat. Protoc. 10 , 1709–1727 (2015). Granger, D. N. & Senchenkova, E. Inflammation and the Microcirculation . vol. 2 (2010). Latroche, C. et al. Skeletal Muscle Microvasculature: A Highly Dynamic Lifeline. Physiology 30 , 417–427 (2015). Hudlická, O. Growth of capillaries in skeletal and cardiac muscle. Circ. Res. 50 , 451–461 (1982). Deveci, D., Marshall, J. M. & Egginton, S. Relationship between capillary angiogenesis, fiber type, and fiber size in chronic systemic hypoxia. Am. J. Physiol. Heart Circ. Physiol. 281 , H241–H252 (2001). Nederveen, J. P., Betz, M. W., Snijders, T. & Parise, G. The Importance of Muscle Capillarization for Optimizing Satellite Cell Plasticity. Exerc. Sport Sci. Rev. 49 , 284–290 (2021). McKey, J., Cameron, L. A., Lewis, D., Batchvarov, I. S. & Capel, B. Combined iDISCO and CUBIC tissue clearing and lightsheet microscopy for in toto analysis of the adult mouse ovary†. Biol. Reprod. 102 , 1080–1089 (2020). Glancy, B. Vivo Microscopy Reveals Extensive Embedding of Capillaries within the Sarcolemma of Skeletal Muscle Fibers. Microcirculation 21 , 131–147 (2014). Kharraz, Y., Guerra, J., Pessina, P., Serrano, A. L. & Muñoz-Cánoves, P. Understanding the Process of Fibrosis in Duchenne Muscular Dystrophy. Biomed. Res. Int. 2014 , 1–11 (2014). Giovarelli, M. et al. Characterisation of Progressive Skeletal Muscle Fibrosis in the Mdx Mouse Model of Duchenne Muscular Dystrophy: An In Vivo and In Vitro Study. IJMS 23 , 8735 (2022). Podkalicka, P. et al. Age-Dependent Dysregulation of Muscle Vasculature and Blood Flow Recovery after Hindlimb Ischemia in the mdx Model of Duchenne Muscular Dystrophy. Biomedicines 9 , 481 (2021). Podkalicka, P., Mucha, O., Dulak, J. & Loboda, A. Targeting angiogenesis in Duchenne muscular dystrophy. Cell. Mol. Life Sci. 76 , 1507–1528 (2019). Lattouf, R. et al. Picrosirius Red Staining: A Useful Tool to Appraise Collagen Networks in Normal and Pathological Tissues. J. Histochem. Cytochem. 62 , 751–758 (2014). Lapierre-Landry, M., Liu, Y., Bayat, M., Wilson, D. L. & Jenkins, M. W. Digital labeling for 3D histology: segmenting blood vessels without a vascular contrast agent using deep learning. Biomed. Opt. Express . 14 , 2416 (2023). Larcher, T. et al. Characterization of Dystrophin Deficient Rats: A New Model for Duchenne Muscular Dystrophy. PLoS ONE . 9 , e110371 (2014). Additional Declarations No competing interests reported. Supplementary Files HASSANIetalSupplementarydatas.docx Cite Share Download PDF Status: Published Journal Publication published 21 Oct, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 09 Jul, 2025 Reviews received at journal 02 Jul, 2025 Reviewers agreed at journal 19 Jun, 2025 Reviews received at journal 08 Apr, 2025 Reviewers agreed at journal 31 Mar, 2025 Reviewers agreed at journal 21 Mar, 2025 Reviewers invited by journal 17 Mar, 2025 Editor assigned by journal 17 Mar, 2025 Editor invited by journal 12 Mar, 2025 Submission checks completed at journal 12 Mar, 2025 First submitted to journal 04 Mar, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6156479","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":440208074,"identity":"6e66be72-23e2-49c2-b5a6-0a4bf66990fb","order_by":0,"name":"Ibrahim Hassani","email":"","orcid":"","institution":"Oniris, INRAE","correspondingAuthor":false,"prefix":"","firstName":"Ibrahim","middleName":"","lastName":"Hassani","suffix":""},{"id":440208075,"identity":"2dbb4fa3-a515-41b2-939b-47303fee16d6","order_by":1,"name":"Mireille Ledevin","email":"","orcid":"","institution":"Oniris, INRAE","correspondingAuthor":false,"prefix":"","firstName":"Mireille","middleName":"","lastName":"Ledevin","suffix":""},{"id":440208076,"identity":"993ecea2-9aad-4a4c-a0e0-967fa464113e","order_by":2,"name":"Chantal Thorin","email":"","orcid":"","institution":"Oniris, INRAE","correspondingAuthor":false,"prefix":"","firstName":"Chantal","middleName":"","lastName":"Thorin","suffix":""},{"id":440208077,"identity":"175a7be8-223e-4f64-b495-5a90fbd69702","order_by":3,"name":"Tony Fiore","email":"","orcid":"","institution":"Nikon Healthcare France","correspondingAuthor":false,"prefix":"","firstName":"Tony","middleName":"","lastName":"Fiore","suffix":""},{"id":440208078,"identity":"457cddc4-ec22-4175-bdeb-ad7456c12603","order_by":4,"name":"Marie-Anne Colle","email":"","orcid":"","institution":"Oniris, INRAE","correspondingAuthor":false,"prefix":"","firstName":"Marie-Anne","middleName":"","lastName":"Colle","suffix":""},{"id":440208079,"identity":"d9ee97a5-c3e3-43e5-be43-994f3a1f0335","order_by":5,"name":"Karl Rouger","email":"","orcid":"","institution":"Oniris, INRAE","correspondingAuthor":false,"prefix":"","firstName":"Karl","middleName":"","lastName":"Rouger","suffix":""},{"id":440208080,"identity":"291982e2-5b49-43cb-b53e-a960680e6caa","order_by":6,"name":"Laurence Dubreil","email":"data:image/png;base64,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","orcid":"","institution":"Oniris, INRAE","correspondingAuthor":true,"prefix":"","firstName":"Laurence","middleName":"","lastName":"Dubreil","suffix":""}],"badges":[],"createdAt":"2025-03-04 18:08:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6156479/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6156479/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-20335-9","type":"published","date":"2025-10-21T16:16:44+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":84425661,"identity":"88ef7f24-c625-449b-ba58-ff53bd5fe7e4","added_by":"auto","created_at":"2025-06-11 19:55:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":537418,"visible":true,"origin":"","legend":"\u003cp\u003eTwo-dimensional histomorphometric analysis of global vasculature on rat \u003cem\u003eBiceps femoris\u003c/em\u003e muscle. Transverse sections from 1-year-old wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e\u003cstrong\u003e) \u003c/strong\u003erats were stained with \u003cem\u003eLycopersicon esculentum\u003c/em\u003e lectin-conjugated with Dylight 594 (LEL DyLight 594) and investigated with two-photon excited fluorescence (TPEF). (a) Left panel: muscle fibers and LEL DyLight 594-stained vessels (white arrows) are visualized by green autofluorescence and red fluorescence, respectively. Middle panel: binary mask (blue) highlights muscle fiber area used for isolating endomysial vessels. Right panel: LEL DyLight 594-stained vessels were detected using NIS-Elements software with the \u003cem\u003eSegment.ai \u003c/em\u003emodule (version 6.10.1, Nikon Europe BV, Amstelveen). Scale bar: 100 µm. (b) The microvessel-to-fiber ratio and vessel density per unit fiber area were determined in both rat groups considering 15 regions from 3 WT rats and 15 from 3 DMD\u003csup\u003emdx\u003c/sup\u003e rats. At least 100 fibers were analyzed per image. Statistical analysis was performed using a linear mixed-effects model in R (lme4 package). The normality of residuals was verified. No significant differences were observed between the two groups for the capillary-to-fiber ratio (p = 0.3) and microvessel density (p = 0.7).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/d01b40fe61df73bbfd6415e3.png"},{"id":84425667,"identity":"25c6a892-7f9a-41ff-ba99-f7e3ed9999b4","added_by":"auto","created_at":"2025-06-11 19:55:37","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":648323,"visible":true,"origin":"","legend":"\u003cp\u003eTwo-dimensional histomorphometric analysis of muscle fiber-associated vascular features on rat \u003cem\u003eBiceps femoris\u003c/em\u003e muscle. Transverse sections from 1-year-old wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e\u003cstrong\u003e) \u003c/strong\u003erats were treated to define binary mask of muscle fibers and microvessels. (a) Left: muscle fibers and microvessels were segmented using the \u003cem\u003eSegmentObject.ai\u003c/em\u003e module in NIS-Elements software (version 6.10.1, Nikon Europe BV, Amstelveen). Binary mask objects are false-colored to distinguish individual fibers. Right: binary mask between muscle fibers and microvessels were used to define capillary contacts. A proximity area, created by expanding the binary mask of vessels by 9 µm, identifies fibers in close contact with vessels, defining the fiber-to-vessel association. Scale bar: 50 µm. (b) The microvessel count per fiber and sharing factor (quantify the number of fiber supply by one single microvessel) of vessel-fiber interactions were determined in both rat groups considering 15 regions from 3 WT rats and 15 regions from 3 DMD\u003csup\u003emdx\u003c/sup\u003e rats. Statistical analysis was performed using a linear mixed-effects model in R (lme4 package). The normality of residuals was verified.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/a21537e827697413c0ea58a0.png"},{"id":84425663,"identity":"01a8f63a-f9b1-44a9-acc9-b6d8f08b8793","added_by":"auto","created_at":"2025-06-11 19:55:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":612171,"visible":true,"origin":"","legend":"\u003cp\u003eThree-dimensional histomorphometric analysis of microvessels on rat \u003cem\u003eBiceps femoris\u003c/em\u003e muscle. Transverse sections from 1-year-old wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e)\u003cstrong\u003e \u003c/strong\u003erats were submitted to iDISCO+ and CUBIC clearing protocols, stained with \u003cem\u003eLycopersicon esculentum\u003c/em\u003e lectin-conjugated with Dylight 594 (LEL DyLight 594) and submitted to segmentation of muscle fibers and microvessel network. (a) From left to right: First column: visualization of muscle fibers (green autofluorescence) and vasculature (red; LEL-DyLight594). Second column: segmented muscle fibers were identified using the \u003cem\u003eSegmentObject.ai\u003c/em\u003e module in NIS-Elements software (version 6.10.1, Nikon Europe BV, Amstelveen). Third column: segmented LEL-DyLight594-stained vascular network were generated using the \u003cem\u003eSegment.ai\u003c/em\u003e module in NIS-Elements software. Fourth column: vasculature surface embedded within muscle fibers was detected by applying an \"AND\" logical function between the binary masks of muscle fibers and vasculature. Individual components are false-colored to distinguish each object. Field of view 1.052 × 1.052 × 0.350 mm. (b) The microvessel volume density, branching count per fiber volume and mean vessel-to-fiber contact surface were determined in both rat groups considering 12 regions\u0026nbsp; from 3 WT rats and 12 from 3 DMD\u003csup\u003emdx\u003c/sup\u003e rats. Statistical analysis was performed using a linear mixed-effects model in R (lme4 package) for volume density (p = 0.06). Since the residuals were not normally distributed within the model, a Wilcoxon test was preferred for the branching and vessel-to-fiber contact surface datasets. Significance levels were set as follows: *0.01\u0026lt; p \u0026lt; 0.05, ****p \u0026lt; 0.0001\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/535d3b0216e5c9dc6e30b795.png"},{"id":84425662,"identity":"19067230-029f-47e7-873d-ffdc5da1f959","added_by":"auto","created_at":"2025-06-11 19:55:36","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":576662,"visible":true,"origin":"","legend":"\u003cp\u003eThree-dimensional representations of the vasculature in the \u003cem\u003eBiceps femoris\u003c/em\u003e muscle. Tissues were collected from 1-year-old wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e) rats. (a) Left: 3D reconstruction of the vascular organization stained with \u003cem\u003eLycopersicon esculentum \u003c/em\u003electin (LEL) DyLight 594 in 1.052 ×1.052×0.350 mm³ images. Right: Top view of the vascular network, generated using a filament algorithm from Imaris software (version 10.1, Oxford Instruments), with false coloration applied to the binary mask to differentiate vascular structures. (b) Quantification of mean diameter and microvessel length. Each group were composed of 3 WT rats and 3 DMD\u003csup\u003emdx \u003c/sup\u003erats. For each animal, 150 microvessels were analyzed. Statistical analysis was performed using the Wilcoxon test. *0.01\u0026lt; p \u0026lt; 0.05, ****p \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/d7cd690b82e3b8e4187b1cdc.png"},{"id":84425664,"identity":"612cb713-78c0-4593-8790-fcb3661d7214","added_by":"auto","created_at":"2025-06-11 19:55:36","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":328507,"visible":true,"origin":"","legend":"\u003cp\u003eThree-dimensional representation of SHG signal in rat \u003cem\u003eBiceps femoris\u003c/em\u003e muscle. (a) Left: three-dimensional (3D) reconstruction of SHG collagen fibers on a 350 µm-thick muscle section from 1-year-old wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e) rats (field of view 1.052 ×1.052×0.350 mm³). It was processed with a Laplacian of Gaussian filter for enhanced fiber boundaries. Right: individual SHG collagen fibers were detected using the GA3 module in NIS-Elements software (Nikon Europe BV, Amstelveen, Netherlands), with false-coloring applied to the binary mask. (b) Bar graphs illustrating features characterizing endomysial SHG\u003csup\u003e+\u003c/sup\u003e objects : density, volume, distance separating objects and elongation of objects. Data were determined in both rat groups considering 3 animals per group and 4 regions each. Statistical analysis was performed using the Wilcoxon test for SHG density. For volume, distance, and elongation datasets, data were discretized into classes, and a log-linear model was applied to compare the distribution between groups. *0.01\u0026lt; p \u0026lt; 0.05, ***p \u0026lt; 0.001, ****p \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/08d7c71d6cc48ed9cfc87a5e.png"},{"id":84426016,"identity":"2f1eea06-29e1-4cb2-a60e-6898ee82f5e7","added_by":"auto","created_at":"2025-06-11 20:03:37","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":415858,"visible":true,"origin":"","legend":"\u003cp\u003eTwo and three-dimensional representations combined to quantification of vasculature and collagen network in the rat \u003cem\u003eBiceps femoris\u003c/em\u003e muscle. Tissue sections from 1-year-old wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e\u003cstrong\u003e) \u003c/strong\u003erats were stained with \u003cem\u003eLycopersicon esculentum\u003c/em\u003e lectin-conjugated with Dylight594 (LEL DyLight594) and submitted to SHG. (a) DyLight594-stained vasculature and SHG collagen network were detected in thin tissue sections (left), thick ones (middle) and alongside a binary mask highlighting the overlapped regions (right). Arrows indicate SHG collagen network encapsulating red-stained vessels. Overlapped regions are false-colored for clarity. (b) Microvessel volume overlapped by SHG was determined in both rat groups considering 3 animals per group and 4 regions per rat. \u0026nbsp;A Wilcoxon test was used for statistical analysis. *** p \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/728b9194aa4de6f54022b617.png"},{"id":84426014,"identity":"83adfc92-d6c7-4679-bf88-afe6456cf474","added_by":"auto","created_at":"2025-06-11 20:03:37","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":443015,"visible":true,"origin":"","legend":"\u003cp\u003eExperimental workflow for muscle fiber autofluorescence, microvessel staining, SHG imaging and 3D analyses from cleared muscle samples. (a) Overview of the sample preparation steps, including iDISCO+/CUBIC clearing protocols. (b) Representative images of the \u003cem\u003eBiceps femoris\u003c/em\u003e muscle before and after the clearing protocols. (c) Multiphoton microscope setup for fluorescence and harmonic imaging. The configuration of emission filters is displayed above the GaAsP and PMT detectors, with excitation signals shown in red and emission signals in black and purple. (d) Representative images from DMD\u003csup\u003emdx\u003c/sup\u003e rat muscle. Top left: Muscle fibers visualized using green autofluorescence. Top right: \u003cem\u003eLycopersicon esculentum\u003c/em\u003e lectin (LEL) DyLight 594-stained microvessels detected as red fluorescence. Bottom left: SHG imaging in both backward and forward directions from cleared sections. Bottom right: Merged signals of autofluorescence, vascular staining, and SHG. White arrows indicate vessels, and white arrowheads highlight SHG signals. Scale bar: 100 µm. (e) Summary of the workflow applied to extract 3D features from cleared muscle samples.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/bc3bb4b74548ad515d999cd3.png"},{"id":94490606,"identity":"16e2e482-04b1-48e0-ac29-529f15880386","added_by":"auto","created_at":"2025-10-27 17:12:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4150421,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/52282f03-7cd0-43c5-b03d-8fe0a5dcac7a.pdf"},{"id":84425672,"identity":"f0abdebf-8e8a-441f-956e-3ca4a9a22b66","added_by":"auto","created_at":"2025-06-11 19:55:37","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":3606219,"visible":true,"origin":"","legend":"","description":"","filename":"HASSANIetalSupplementarydatas.docx","url":"https://assets-eu.researchsquare.com/files/rs-6156479/v1/fefb0bd8e9bc6e5ba939cdb2.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Multiphoton microscopy and tissue clearing for 3D characterization of the vasculature and fibrosis remodeling in rat dystrophic skeletal muscle","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDuchenne muscular dystrophy (DMD) is a severe X-linked recessive disorder caused by mutations in the dystrophin gene\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, leading to a complete lack of the dystrophin protein\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. This deficiency destabilizes the plasma membrane, triggering repeated cycles of muscle fiber necrosis and degeneration, followed by their replacement with adipose and fibrotic tissue\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. DMD patients exhibit muscle atrophy and progressive muscle weakness, resulting in loss of ambulation, and develop cardiomyopathy and respiratory insufficiency ultimately leading to premature death typically between 20 and 30 years of age\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFibrosis, corresponding to the excessive deposition of fibrillar collagen types I and III from extracellular matrix \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, further impairs contractile muscle function and significantly contributes to the progression of the disease\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. In addition to muscle fiber degeneration, vascular abnormalities including impaired blood flow and reduced capillary density have been documented\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Interstitial fibrosis exacerbates vascular dysfunction by disrupting the structural integrity of the vascular network, notably leading to impaired nutrient and oxygen exchange with muscle fibers\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. These vascular changes, along with fibrosis, play a crucial role in pathophysiology of DMD. Their assessment has traditionally relied on 2D methodologies, as seen in previous studies\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. However, such approaches inherently constrain the ability to fully understand the complex spatial relationships and interactions among vascular, interstitial, and muscular compartments. Yet, a detailed 3D analysis of these interactions is essential for gaining deeper insights into disease mechanisms and potential therapeutic targets. To address this need, nonlinear microscopy techniques such as second harmonic generation (SHG) and two-photon excited fluorescence (TPEF) offer valuable tools\u003csup\u003e\u003cspan additionalcitationids=\"CR16\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Indeed, these techniques provide detailed information on the degree and type of changes observed in vascular and connective tissue network within dystrophic muscles\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. This enables a more comprehensive understanding of the structural and functional alterations associated with DMD.\u003c/p\u003e \u003cp\u003eSHG is a second-order nonlinear optical process that occurs in non-centrosymmetric structures, such as collagen, myosin bands, and microtubules\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. This property makes it particularly suitable for imaging of these components under both physiological and pathological conditions\u003csup\u003e\u003cspan additionalcitationids=\"CR19 CR20 CR21\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. SHG microscopy enables the visualization of collagen structures in thick samples, offering insights into their organization based on their volume, shape factor, and orientation\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Unlike conventional fluorescence microscopy that relies on single-photon absorption, TPEF requires the simultaneous absorption of two lower-energy photons to excite fluorophores. It provides significant advantages, such as reduced photobleaching\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, multiple absorption spectra peaks\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e and intrinsic optical sectioning capabilities\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, making it suitable for imaging thick tissues. In 3D, TPEF-based imaging of the vasculature allows detailed morphological characterization of vessels within tissues, enabling their analysis as interconnected networks\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. This is crucial for advancing understanding of the structural and functional dynamics of vascular networks. However, effective 3D imaging requires overcoming the challenges posed by the inherent opacity of biological tissues\u003csup\u003e\u003cspan additionalcitationids=\"CR27\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Optical clearing methods address this issue by allowing light to penetrate deeply into tissues while minimizing the strong diffusion and absorption of biological material. These techniques work on a common principle: light-scattering molecules within tissues, which contribute to opacity, are removed or replaced with solutions that have a uniform refractive index. This process significantly reduces light scattering, rendering large biological samples transparent and suitable for high-resolution imaging. Over the past decades, numerous optical clearing methods have been developed, each tailored to optimize imaging outcomes for various biological samples\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn the present study, we developed a methodology to assess the 3D remodeling of vascular and connective tissues in dystrophic muscle. We adapted and combined tissue-clearing protocols, specifically iDISCO\u0026thinsp;+\u0026thinsp;\u003csup\u003e30\u003c/sup\u003e and CUBIC\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e to achieve efficient muscle tissue clearing while maintaining compatibility with vascular labeling. Through this approach, we successfully imaged vasculature and fibrosis in cleared skeletal muscle tissue, using multiphoton microscopy. To assess the remodeling of dystrophic muscle, we designed and implemented an image analysis workflow that combines deep learning-based methods for the segmentation of vasculature and muscle fibers. We then applied this workflow to muscles from healthy and dystrophic rats aged 1 year, aiming to characterize 3D alterations in vascular morphology and how fibrosis contributes to impaired vascular function.\u003c/p\u003e \u003cp\u003eOur findings demonstrate for the first time that 3D imaging analysis enables a detailed characterization of the structural interplay between vasculature, muscle fibers, and connective tissue in dystrophic muscle. This approach may serve as a valuable tool for studying the progression of muscle diseases and assessing the impact of novel treatments. .\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eMicrovessels are small blood vessels located at the periphery of myofibers, including capillaries, terminal arterioles, and terminal venules \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003cb\u003eStandard 2D histological analysis of dystrophic muscle highlights capillarization defect as key feature of microvascular damage\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWe first aimed to assess vascular impairment in dystrophic muscle using a 2D approach. To achieve this, we analyzed two global parameters (microvessel-to-fiber ratio and microvessel density) and two specific metrics (microvessel contact and sharing factor), which are commonly employed\u003csup\u003e\u003cspan additionalcitationids=\"CR35\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e to evaluate vasculature in skeletal muscle. These analyses were conducted on \u003cem\u003eBiceps femoris\u003c/em\u003e muscle sections from 1-year-old WT and DMD\u003csup\u003emdx\u003c/sup\u003e rats.\u003c/p\u003e \u003cp\u003eTPEF observation of transverse sections stained with LEL DyLight 594 showed the distribution of microvessels within bundles of muscle fibers that were visualized by green auto-fluorescence signal (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, left panel). A binary mask was applied to the entire section to delineate the muscle fiber area (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, middle panel), allowing isolation of endomysial vessels. These vessels were then automatically detected using the Segment.ai module in NIS-Elements software (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, right panel). Using this method, we determined a microvessel-to-fiber ratio of 1.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08 and 1.5\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09 in DMD\u003csup\u003emdx\u003c/sup\u003e and WT rats, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, top). Despite a slightly higher ratio in WT rats compared to DMD\u003csup\u003emdx\u003c/sup\u003e ones, no statistically significant difference was observed. Additionally, microvessel density, defined as the number of microvessels within the endomysial area, was measured as 268\u0026thinsp;\u0026plusmn;\u0026thinsp;17 mm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e and 243\u0026thinsp;\u0026plusmn;\u0026thinsp;22 mm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e in DMD\u003csup\u003emdx\u003c/sup\u003e and WT rats respectively, revealing a similar tissue distribution pattern (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, bottom).\u003c/p\u003e \u003cp\u003eTo investigate the interactions between microvessels and individual muscle fibers, we automated the measurements of microvessel contact and the sharing factor using the GA3 module in NIS-Elements software. Muscle fibers and microvessels were detected on 2D cross-sections using deep learning models, \u003cem\u003ei.e\u003c/em\u003e., \u003cem\u003eSegmentObject.ai\u003c/em\u003e and \u003cem\u003eSegment.ai\u003c/em\u003e, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, left panel). Details of the training methodology are provided in Materials and Methods section. To better capture the spatial relationship, the binary mask representing microvessels was slightly expanded to define a proximity area around each muscle fiber. A binary operation was then performed between the muscle fiber and microvessel masks to identify regions of proximity (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, right panel). These regions were used to uniquely associate each proximity region with a corresponding fiber and vessel, enabling precise quantification of microvessel-to-fiber interactions. While DMD\u003csup\u003emdx\u003c/sup\u003e rat muscle only shows a tendency towards a decrease in the number of microvessel contacts per fiber compared to WT rats (3.2\u0026thinsp;\u0026plusmn;\u0026thinsp;0.24 vs. 3.7\u0026thinsp;\u0026plusmn;\u0026thinsp;0.17, respectively; Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, top), it is characterized by a significant increase in the sharing factor, which represents the number of fibers served by a single microvessel: 2.6\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07 (DMD\u003csup\u003emdx\u003c/sup\u003e rats) vs. 2.4\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 (WT rats) (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, bottom). Collectively, these findings indicate that 1-year-old DMD\u003csup\u003emdx\u003c/sup\u003e rats exhibit an altered organization of the microvasculature.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMultimodal 3D imaging reveals a newly formed vascular network and reduced interaction between muscle fiber-vessel in dystrophic muscle\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo assess whether 3D imaging analysis could offer a more precise and comprehensive investigation of vascularization, with a focus on vascular network connectivity and geometry, we utilized the same muscle samples as in previous experiments, but this time in thick section format. Initially, we identified the most appropriate clearing protocol for thick skeletal muscle, which was a combination of the iDISCO\u0026thinsp;+\u0026thinsp;and CUBIC methods, as described previously\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. We then validated that the staining with LEL DyLight 594 is compatible with this protocol. After this fine-tuning stage, we generated Z-stack images from cleared and LEL DyLight 594-stained 1-mm-thick sections. Green muscle autofluorescence and red fluorescence of stained vessels were acquired with multiphotonic microscope up to a depth of 350 \u0026micro;m (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, first column). This allowed us identifying bundles of muscle fibers and microvessel network in depth. We generated images exhibiting a high signal-to-noise ratio (SNR) that facilitates the segmentation of the vasculature across the entire field of view (FOV) measuring 1.052 \u0026times; 1.052 \u0026times; 0.350 mm. The muscle fibers segmented from their endogenous green fluorescence (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, second column) and microvessels segmented from their red fluorescence staining (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, third column) were color-coded to enhance their visibility. Vascular surfaces embedded within muscle fibers were detected by applying an \"AND\" logical function between the binary masks of muscle fibers and vasculature (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003ea; fourth column). We showed that the endomysial vascular network penetrating muscle consists of branched capillaries that appear much more organized in WT rat muscle compared to that of DMD\u003csup\u003emdx\u003c/sup\u003e rat. It runs parallel to the muscle fibers in WT rat muscle while it displays multiple directions in DMD\u003csup\u003emdx\u003c/sup\u003e rat muscle.\u003c/p\u003e \u003cp\u003eFor 3D quantitative analysis, we developed a pipeline within the GA3 module in NIS-Elements software and exploited the same deep learning models used for the 2D analysis to detect objects. These models were applied to individual image slices, and the data were reconstructed into 3D volumetric representations. 3D microvessel density, expressed as the volume occupied by vessels relative to fiber volume, was slightly higher in DMD\u003csup\u003emdx\u003c/sup\u003e rat muscle (0.02\u0026thinsp;\u0026plusmn;\u0026thinsp;0.002) compared to WT rat muscle (0.014\u0026thinsp;\u0026plusmn;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, left), but the difference was not statistically significant. However, the branching count per muscle fiber volume was significantly higher in DMD\u003csup\u003emdx\u003c/sup\u003e rats (21,596\u0026thinsp;\u0026plusmn;\u0026thinsp;1,981) than in WT rats (13,428\u0026thinsp;\u0026plusmn;\u0026thinsp;2,094), indicating increased vascular network ramification in the dystrophic context (p\u0026thinsp;=\u0026thinsp;0.017; Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, middle). To facilitate efficient nutrient and oxygen exchange, capillaries are wrapped around muscle fibers within grooves, forming close contacts\u0026mdash;a well-documented phenomenon\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. In Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, right, we explored this aspect by assessing the embedding surface of the vasculature within muscle fibers. To quantify this, we performed a binary operation using muscle fiber and microvessel masks to calculate the shared surface area, normalizing it to the number of fibers. Our analysis revealed a significant decrease in the contact surface area in DMD\u003csup\u003emdx\u003c/sup\u003e rats (8.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.6 \u0026micro;m\u0026sup2;) compared to WT rats (20.4\u0026thinsp;\u0026plusmn;\u0026thinsp;3.1 \u0026micro;m\u0026sup2;), indicating impaired vasculature-muscle interactions in the dystrophic context (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, right).\u003c/p\u003e \u003cp\u003eAdditional morphometric analyses were conducted to assess the morphology of the vascular network. Vessel diameter and length were quantified using the filament algorithm in Imaris software (version 10.1, Oxford Instruments) (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). The algorithm's 'loops' function was trained on data from four distinct regions before being applied to analyze three images (1.052 \u0026times; 1.052 \u0026times; 0.350 \u0026micro;m) of both WT and DMD\u003csup\u003emdx\u003c/sup\u003e rat muscle. From each image, 150 microvessels were selected for analysis. The microvessel diameter was slightly reduced in DMD\u003csup\u003emdx\u003c/sup\u003e rats (5.6\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1 \u0026micro;m) compared to WT ones (5.9\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1 \u0026micro;m) (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, top). We also determined that the microvessel length was 37.44\u0026thinsp;\u0026plusmn;\u0026thinsp;1.3 \u0026micro;m and 54.218\u0026thinsp;\u0026plusmn;\u0026thinsp;3 \u0026micro;m in DMD\u003csup\u003emdx\u003c/sup\u003e and WT rat muscle, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, bottom). It was significantly shorter in dystrophic context (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). Taken together, these results demonstrate the presence of a highly reorganized vasculature in dystrophic rat muscle, defined by microvessels that are more branched, but also shorter, wider and less in contact with muscle fibers. These findings also highlight the added value of 3D multimodal exploration for analyzing a spatially networked structure such as the vasculature.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMultimodal 3D imaging is relevant for jointly assessing the intensity of microvascular and fibrotic remodeling in dystrophic muscle\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFibrosis represents one of the main pathological features of dystrophic muscle\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. It results in a change in the structural organization of muscle bundles, with the vascular network moving away from muscle fibers\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Recent evidence suggested that the disorganization of collagen network constitutes a potential biomarker of DMD\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Building on our findings regarding microvasculature, we next aimed to characterize the endomysial connective tissue in 3D and examine its interaction with the vascular network. This approach provides a more comprehensive view of muscle remodeling at the structural level.\u003c/p\u003e \u003cp\u003eFor that purpose, we analyzed forward 3D SHG on cleared muscle cross-sections. All images were acquired with consistent laser power, detector sensitivity, and gain settings. Image analysis of SHG collagen fibers was performed using the GA3 module of NIS-Elements Software and a false-coloring was applied to the binary mask. A marked increase in collagen deposition was observed in DMD\u003csup\u003emdx\u003c/sup\u003e rats compared to WT ones, as expected (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). Endomysial SHG density, corresponding to the volume of collagen network surrounding each muscle fiber relative to muscle fiber volume, was 2.79% \u0026plusmn; 0.10% and 0.51% \u0026plusmn; 0.11% in DMD\u003csup\u003emdx\u003c/sup\u003e and WT rats, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, left, top). It was significantly higher in dystrophic muscle (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001).\u003c/p\u003e \u003cp\u003eA detailed analysis of the collagen network was conducted by evaluating the morphology and spatial distribution of SHG collagen objects. As expected, DMD\u003csup\u003emdx\u003c/sup\u003e rats were overrepresented across all volume classes, indicating increased collagen accumulation along the muscle fiber in depth (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, top right). Additionally, SHG collagen objects were more frequently distributed in shorter distance classes in DMD\u003csup\u003emdx\u003c/sup\u003e rats compared to WT ones, suggesting a denser collagen network (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, bottom left). Moreover, these collagen structures exhibited a less elongated shape in DMD\u003csup\u003emdx\u003c/sup\u003e rat muscle than that of WT rat one (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, bottom right). Next, we identified 3D regions where DyLight594-stained microvessels overlapped with SHG collagen network, using the binary masks of both structures previously generated (Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). The volume of these overlapping regions was measured and normalized to the total microvessel volume to provide a quantitative assessment of the interaction between microvasculature and connective tissue (Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e6\u003c/span\u003eb, right). We determined that the volume of overlap between microvessels and SHG collagen was 10.07\u0026thinsp;\u0026plusmn;\u0026thinsp;1.66% and 1.79\u0026thinsp;\u0026plusmn;\u0026thinsp;0.43% in DMD\u003csup\u003emdx\u003c/sup\u003e and WT rats, respectively (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). Overall, these data provide new information on the accumulation of connective tissue in dystrophic muscle, in terms of volume, concentration and shape, within which part of the microvasculature is embedded.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we provide original and compelling data demonstrating the contribution of multimodal 3D imaging in providing insights into the spatial organization and interactions between vasculature, muscle fibers and connective tissue. The simultaneous 3D visualization of the myofibers, microvessels and/or connective tissue, followed by quantitative analysis on segmented objects, is a powerful approach for qualifying the nature and intensity of remodeling in dystrophic muscle,\u003c/p\u003e \u003cp\u003eDisruption of microvasculature is a pathological feature of muscular dystrophy\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. It has negative impacts on the oxygen, nutrients delivery to the muscle fiber and muscle development\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Like development of fibrosis, vasculature changes in dystrophic muscle are typically assessed using 2D analysis of tissue cross-sections that unfortunately exhibits some limitations\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Here, we introduced a novel 3D approach to analyze the structure of the vasculature and collagen network, aiming to gain more detailed insights into their spatial organization within the muscle bundles. For that, we decided to employ a multimodal imaging technique combining TPEF and SHG to investigate fluorescent lectin-stained vasculature and label-free SHG collagen on cleared rat dystrophic muscle, respectively. Unlike previous studies that relied on conventional histological methods to assess vascular and fibrotic remodeling\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e, our dual-modality imaging approach enabled simultaneous visualization of TPEF and SHG signals from both structures and analysis of their interactions within muscle fibers. Combined with iDISCO\u0026thinsp;+\u0026thinsp;and CUBIC clearing methods, it improved analysis of microvessel organization and collagen network connectivity over at least 1mm in depth. In a previous study, we have demonstrated the usefulness of tissue clearing and SHG imaging to study fibrosis in DMD rat heart using CUBIC protocol\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Here, we combine CUBIC with iDISCO\u0026thinsp;+\u0026thinsp;to firstly enhance the penetration of the fluorescent probes for the vessels thanks to the organic solvent pretreatment in iDISCO\u0026thinsp;+\u0026thinsp;protocol and secondly to improve the tissue clearing and to avoid the tissue shrinkage thanks to hyperhydration pretreatment in CUBIC protocol. The combination of iDISCO\u0026thinsp;+\u0026thinsp;and CUBIC tissue clearing methods has already been successfully described for \u003cem\u003ein toto\u003c/em\u003e imaging of whole ovarian follicle with its extrinsic vascular and neuronal networks\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn addition to this original 3D imaging approach, we developed an image analysis protocol specifically designed to handle the complexity of the objects being measured. Deep learning have emerged as powerful tools alongside traditional hand-crafted analysis workflows, offering the potential for more accurate and efficient results\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Recent studies have demonstrated the effectiveness of deep learning U-net architecture to segment blood vessels in heart tissue sections, yielding accurate vascular morphometrics such as vessel length and density\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. In our study, 3D features of the vasculature and collagen network were extracted using a U-Net model. We were able to measure vessel density, branching and sharing as well as to characterize the collagen network in terms of density packing and thickness/elongation, corroborating the conclusions of Lapierre-Landry and co-author's work on the applicability of these tools\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eRecently, several studies have demonstrated major structural and functional vascular network defects in the main model of DMD, the \u003cem\u003emdx\u003c/em\u003e mice\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Alterations that increase in severity with age correspond to abnormal flow capacity, vessel diameter change or vessel density modification\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Based on their findings, these studies emphasize the crucial role of the vascular network damage in the pathophysiology of DMD. With this study, we present significant disruptions of the vascular network in another model of DMD, namely the DMD\u003csup\u003emdx\u003c/sup\u003e rat\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Two-dimensional analysis revealed an increased sharing factor compared to that observed in WT animal, which are consistent with the previous findings in \u003cem\u003emdx\u003c/em\u003e mice\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Importantly, 3D investigations of the vasculature provided additional information, confirming the existence of highly significant morphological alterations in dystrophic rat muscle. This includes a high degree of microvessel ramifications, which could correspond to a compensatory response to the altered metabolic demands in dystrophic context\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. In parallel, we also found a decrease in vessel-muscle fiber contact, which may indicate impaired functional interactions between the vascular network and muscle fibers. Similar vascular alterations were reported by Latroche et al. in a 1-year-old \u003cem\u003emdx\u003c/em\u003e mouse model, where structural abnormalities were characterized by a significant increase in branching and functional impairments\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eRegarding fibrosis, 3D SHG imaging allowed us to show significant changes in the morphology and distribution of the collagen network, which were not detectable through 2D image analysis. Specifically, we demonstrated that dystrophic rat muscle was defined by a thicker and more densely packed connective tissue. It is important to note that the architecture of collagen fibers has been reported to be associated with increased passive stiffness in fibrotic skeletal muscles of \u003cem\u003emdx\u003c/em\u003e mice\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This is particularly interesting as the amount of fibrotic tissue in \u003cem\u003emdx\u003c/em\u003e mouse muscles correlates with contractile function. Given the much more pronounced tissue phenotype in the dystrophic rat model compared to the \u003cem\u003emdx\u003c/em\u003e mouse\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e as well as the variability observed depending on the type of muscle considered, it would be valuable to determine whether the same observation holds in the model used here. This could help to better understand potential impacts on contractile muscle function.\u003c/p\u003e \u003cp\u003eAdditionally, 3D SHG imaging revealed that the collagen network frequently encased vascular structures in dystrophic muscle, a phenomenon reported in DMD patients\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. This encasing likely reflects a pathological interaction between fibrosis and vasculature. Excessive collagen deposition may restrict blood flow, contributing to the increased distance between capillaries and muscle fibers. These findings suggest a pathological feedback loop, where fibrosis exacerbates vascular pathology, which in turn drives further muscle degeneration.\u003c/p\u003e \u003cp\u003eThe combined analysis of fibrosis and vascular alterations using SHG and TPEF imaging underscores the complex interplay between these two processes in DMD. The findings suggest that therapeutic strategies targeting both extending the vascular network and fibrosis reduction could be effective in managing DMD. Enhancing vascular integrity could improve tissue perfusion and mitigate muscle damage, potentially alleviating some symptoms of the disease. Concurrently, reducing fibrosis could restore the structural and functional integrity of the muscle. Future studies should investigate the temporal dynamics of vascular and fibrotic remodeling in DMD to identify critical windows for intervention. Early therapeutic strategies targeting these processes may help slow or halt disease progression. Additionally, advanced imaging approaches, such as those used in this study, could be used to evaluate the efficacy of potential therapies in preclinical and clinical settings.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eCombination of multimodal 3D SHG/TPEF imaging with iDISCO\u0026thinsp;+\u0026thinsp;and CUBIC tissue clearing and deep learning-based image analysis provided new tools to get valuable insights into the structural and spatial features of microvessels and collagen network in DMD. This approach is particularly well suited for producing an in depth characterization of the extent of deterioration observed in the vasculature and collagen network. Through this original approach, we demonstrated that dystrophic muscle features a spatially well-developed vascular network made up of small microvessels whose interactions with muscle fibers are limited by an encasing connective tissue. The new quantitative 3D imaging technique described here may greatly enhance the understanding of the natural progression of this disease. Additionally, it could serve as a relevant tool for assessing the impact of biotherapies on vascularization and fibrosis in preclinical studies on DMD.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eAnimals and tissue sample\u003c/strong\u003e. All experiments were carried out in accordance with the ARRIVE guidelines (\u003ca href=\"https://arriveguidelines.org\"\u003ehttps://arriveguidelines.org\u003c/a\u003e). Wild-type (WT) and dystrophic (DMD\u003csup\u003emdx\u003c/sup\u003e) Sprague-Dawley rats were obtained from the transgenic rat immunophenomic platform (TRIP; Nantes, France). Three animals of each, all aged 1 year, were included in this study. Animals were maintained in a controlled environment (temperature 21 \u0026plusmn; 1\u0026deg;C, 12-h light-dark cycle) at the Boisbonne Center for gene and cell therapy (Oniris, Nantes, France; agreement number: J44273). All efforts were made to minimize suffering. Rats were provided with environmental enrichment: provision of rolls is reported to potentially modify the behavior of housed animals and reduce chronic pain. Anesthesia was induced with a mixture of ketamine (100 mg/kg, Imalgene, Merial, Lyon, France) and xylazine (10 mg/kg, Rompun, Bayer, Leverkusen, Germany), after which the rats were euthanized by intravenous administration of sodium pentobarbital (300 mg Dolethal; Vetoquinol SA, Magny Vernois). The \u003cem\u003eBiceps femoris\u003c/em\u003e muscles were collected and separated into two portions: one was transferred to a tube containing 4% paraformaldehyde for fixation overnight at 4\u0026deg;C and storage. The second was embedded in paraffin for classical histological analysis. The study was approved by the Ethics Committee for Animal Experiments of the Pays de la Loire Region, France. All experiments were carried out in accordance with the French National Research Council guidelines for the care and use of laboratory animals (Permit number: APAFIS #39967-2022122112121987 v10).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eiDISCO+ clearing pretreatment.\u0026nbsp;\u003c/strong\u003eFixed muscles were cut into thick muscle sections (6 x 3 x 1 mm\u0026sup3;). After washing in PBS, sections were dehydrated stepwise in 20%, 40%, 60%, and 80% methanol (MeOH, 20846.292, VWR) / distilled water (dH₂O), with gentle shaking at room temperature (RT) for 1 hour each. Tissue sections were then washed twice in 100% methanol (1 hour, RT), followed by incubation in a 66% dichloromethane (DCM, 5895811000, Sigma-Aldrich) / 33% methanol solution overnight (ON) at RT. Tissue sections were rehydrated once again stepwise in 80%, 60%, 40%, and 20% MeOH/dH₂O with gentle shaking (1 hour, RT) each. They were then washed twice with 0.2% Triton X-100 (X100, Sigma-Aldrich) in 1X PBS (1 hour). Afterward, they were soaked (ON, 37\u0026deg;C) in a permeabilization solution containing 0.2% Triton X-100, 20% dimethyl sulfoxide (DMSO, D4540, Sigma-Aldrich), 0.3 M glycine (G7126, Sigma-Aldrich), and 0.02% sodium azide (S2002, Sigma-Aldrich) in 1X PBS, with gentle shaking.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eVascular staining.\u0026nbsp;\u003c/strong\u003eFollowing iDISCO+ clearing pretreatment, sections were incubated with Lycopersicon esculentum lectin conjugated with Dylight 594 (LEL DyLight 594, Invitrogen, L32471), diluted 1:50 in labeling solution (0.2% Tween 20, 10% DMSO, 0.02% sodium azide in 1X PBS) with gentle shaking (7 days, 37\u0026deg;C). Then, sections were washed with 0.1% Tween 20 (P1379, Sigma-Aldrich) in 1X PBS (1 day) before proceeding with the clearing procedures.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eiDISCO+ clearing protocol.\u003c/strong\u003e Stained sections were gradually dehydrated using the same procedure previously described above. They were then incubated (ON, RT) in a solution of 66% dichloromethane (DCM) / 33% methanol (MeOH), washed twice in 100% DCM (20 min) and subsequently stored in dibenzyl ether (DBE, 108014, Sigma-Aldrich) at RT.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCUBIC clearing protocol.\u0026nbsp;\u003c/strong\u003eiDISCO+ cleared sections were rehydrated stepwise in 80%, 60%, 40%, and 20% MeOH / dH₂O, with each step involving gentle shaking (RT, 1 hour). Tissue sections were then washed three times with 1X PBS for 1 hour each. Afterward, they were transferred to a 50% CUBIC-L (T3740, TCI) / 50% dH₂O solution (ON, 37\u0026deg;C), followed by incubation with 100% CUBIC-L with gentle shaking (2 days, 37\u0026deg;C). Following three washes with 1X PBS, sections were transferred to 50% CUBIC-R (T3741, TCI) (overnight, RT) and clarified a second time with CUBIC-R+ (T3741, TCI) (2 days, RT). Finally, sections were mounted with fibers oriented transversally on a cavity slide (Sigma-Aldrich, BR475565) using CUBIC-R+ (refractive index 1.52), and a 0.17 \u0026mu;m thick coverslip was placed on top before imaging.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMultiphoton microscopy imaging.\u0026nbsp;\u003c/strong\u003eImages were acquired using a laser scanning multiphoton microscope (A1R-MP+, Nikon Europe B.V., Amstelveen), coupled with a tunable laser (Insight DeepSee, Spectra Physics, France) operating in the wavelength range of 680\u0026ndash;1300 nm, with a pulse duration of 120 femtoseconds at a repetition frequency of 80 MHz. A motorized half-wave plate (MKS-Newport, USA) was used to control the laser polarization angle, which was adjustable between 45\u0026deg; and 90\u0026deg;. The objective lens employed for imaging was a Plan-Apo Lambda S 25X silicon objective (refractive index 1.406, MRD73250, Nikon Europe B.V.), with a numerical aperture (NA) of 1.05 and a working distance (WD) of 0.55 mm. The microscope was equipped with eight non-descanned detectors (NDDs), four for backward detection and four for forward detection. Further details on the setup are given in Figure 7. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3D image acquisition.\u003c/strong\u003e Excitation at 820 nm coupled with resonant scanning mode was used to acquire green autofluorescence from muscle fibers, red fluorescence from LEL DyLight 594-labeled vessels, and second harmonic generation (SHG) signals. Green autofluorescence and red fluorescence signals were collected in reflection using band-pass emission filters, 525/50BP and 629/56BP for muscle fibers and vessels, respectively. SHG signals were collected in both reflection mode (bSHG) and transmission mode (fSHG) using short band-pass filters, 415/10BP. The field of view (FoV) was 297 \u0026micro;m x 297 \u0026micro;m, scanned at 1024 x 1024 pixels, resulting in a pixel size of 0.29 \u0026micro;m. Z-series were collected with a 1 \u0026micro;m step size within a 350 \u0026micro;m depth of the specimen. Tiles and depth scanning were performed to analyze at least 100 fibers, with the step size between tiles fixed at 1 \u0026micro;m. Images were generated as 12-bit ND2 files, and stitching and 3D reconstruction were performed using NIS-Elements software (version 5.20, Nikon Europe B.V).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeep learning model training.\u0026nbsp;\u003c/strong\u003eMuscle fiber green autofluorescence was segmented using the U-Net-based deep learning model \u0026ldquo;\u003cem\u003eSegmentObject.ai\u003c/em\u003e\u0026rdquo; integrated into NIS-Elements software. This model was particularly suitable for segmenting muscle fibers due to its capability to separate densely packed objects. The training dataset comprised seven cropped images (400 \u0026times; 400 \u0026times; 10 \u0026micro;m), each containing annotations of 20 manually identified muscle fibers. The patch size was automatically set to 512 \u0026times; 512. The training process consisted of a minimum of 2,500 epochs across three batches and was conducted on an NVIDIA RTX5000 graphics unit (Driver version 516.40, NVIDIA). The trained model was applied on another dataset comprised eight manually annotated images for validation. The 4 following metrics were measured (details are provided on Supplemental data S1 and S2):\u003c/p\u003e\n\u003cp\u003e- Precision: \u0026nbsp;It evaluates the fraction of correctly predicted positive regions out of all predicted positive regions\u003c/p\u003e\n\u003cp\u003e- Recall: Also known as sensitivity or True Positive Rate, it measures how many of the true positive regions were correctly identified.\u003c/p\u003e\n\u003cp\u003e- F1 score: Harmonic mean of Precision and Recall\u003c/p\u003e\n\u003cp\u003e- Interception Over Union (IoU): Also called the Jaccard Index, it measures the overlap between the predicted segmentation and the ground truth.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable style=\"width:453.1pt;border-collapse:collapse;border:none;\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 113.25pt;border-width: 1pt 1pt 1.5pt;border-style: solid;border-color: rgb(153, 153, 153) rgb(153, 153, 153) rgb(102, 102, 102);border-image: initial;padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003ePrecision\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.25pt;border-top: 1pt solid rgb(153, 153, 153);border-left: none;border-bottom: 1.5pt solid rgb(102, 102, 102);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003eRecall\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: 1pt solid rgb(153, 153, 153);border-left: none;border-bottom: 1.5pt solid rgb(102, 102, 102);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003eF1\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: 1pt solid rgb(153, 153, 153);border-left: none;border-bottom: 1.5pt solid rgb(102, 102, 102);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003eIoU\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 113.25pt;border-right: 1pt solid rgb(153, 153, 153);border-bottom: 1pt solid rgb(153, 153, 153);border-left: 1pt solid rgb(153, 153, 153);border-image: initial;border-top: none;padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003e0,90\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.25pt;border-top: none;border-left: none;border-bottom: 1pt solid rgb(153, 153, 153);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003e0,82\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: none;border-left: none;border-bottom: 1pt solid rgb(153, 153, 153);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003e0,84\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: none;border-left: none;border-bottom: 1pt solid rgb(153, 153, 153);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003e0,74\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp style='margin-top:0in;margin-right:0in;margin-bottom:8.0pt;margin-left:0in;text-align:justify;line-height:115%;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cspan style='font-family:\"Arial\",sans-serif;'\u003eTable 1 Score metrics for \u003cem\u003eSegmentObject.ai\u003c/em\u003e model trained\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eThe LEL-DyLight594-labeled vasculature was segmented using the \u0026ldquo;\u003cem\u003eSegment.ai\u003c/em\u003e\u0026rdquo; in NIS-Elements software. The training dataset included three images (200 \u0026times; 200 \u0026times; 300 \u0026micro;m) with a patch size automatically set to 256 \u0026times; 256 pixels. Initial identification of vasculature structures was achieved through thresholding, followed by manual refinement to create accurate binary masks, which served as the ground truth for model training. The model underwent two separate training runs, each consisting of 500 epochs. To validate the vascular segmentation, the same approach previously employed for validating muscle fiber segmentation was applied.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable style=\"width:453.1pt;border-collapse:collapse;border:none;\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 113.25pt;border-width: 1pt 1pt 1.5pt;border-style: solid;border-color: rgb(153, 153, 153) rgb(153, 153, 153) rgb(102, 102, 102);border-image: initial;padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003ePrecision\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.25pt;border-top: 1pt solid rgb(153, 153, 153);border-left: none;border-bottom: 1.5pt solid rgb(102, 102, 102);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003eRecall\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: 1pt solid rgb(153, 153, 153);border-left: none;border-bottom: 1.5pt solid rgb(102, 102, 102);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003eF1\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: 1pt solid rgb(153, 153, 153);border-left: none;border-bottom: 1.5pt solid rgb(102, 102, 102);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Aptos Narrow\",sans-serif;color:black;'\u003eIoU\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 113.25pt;border-right: 1pt solid rgb(153, 153, 153);border-bottom: 1pt solid rgb(153, 153, 153);border-left: 1pt solid rgb(153, 153, 153);border-image: initial;border-top: none;padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e0,92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.25pt;border-top: none;border-left: none;border-bottom: 1pt solid rgb(153, 153, 153);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e0,77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: none;border-left: none;border-bottom: 1pt solid rgb(153, 153, 153);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e0,82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113.3pt;border-top: none;border-left: none;border-bottom: 1pt solid rgb(153, 153, 153);border-right: 1pt solid rgb(153, 153, 153);padding: 0in 5.4pt;height: 0.2in;vertical-align: top;\"\u003e\n \u003cp style='margin-top:0in;margin-right:0in;margin-bottom:0in;margin-left:0in;text-align:center;line-height:normal;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e0,71\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp style='margin-top:0in;margin-right:0in;margin-bottom:8.0pt;margin-left:0in;text-align:justify;line-height:115%;font-size:16px;font-family:\"Aptos\",sans-serif;'\u003e\u003cspan style='font-family:\"Arial\",sans-serif;'\u003eTable 2 \u0026nbsp;Score metrics for \u003cem\u003eSegment.ai\u003c/em\u003e model trained\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eBoth segmentation models demonstrated high accuracy (SegmentObject.ai in table 1, Segment.ai in table 2), with minimal false positives (precision were 0.90 and 0.92 for the muscle fiber model and the microvessel model, respectively). However, some false negatives were observed, as indicated by recall values of 0.82 and 0.77 for the respective models. The Intersection over Union (IoU) score was 0.74 and 0.71 for the muscle fiber model and the microvessel model respectively, while the F1-score was 0.84 and 0.82, further supporting the robustness of the segmentation performance.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImage analysis of the vasculature.\u003c/strong\u003e Two-dimensional images: vascularization was evaluated on five regions per animal (n = 6 rats), each measuring 1052 \u0026times; 1052 \u0026micro;m, using the General Analysis 3 (GA3) module in NIS-Elements software (version 6.10.1, Nikon Europe BV, Amstelveen). Vessels were detected using the trained \u0026quot;\u003cem\u003eSegment.ai\u003c/em\u003e\u0026quot; model, which was imported into the General Analysis 3 (GA3) module of NIS-Elements software. The GA3 pipeline (details are provided in Supplemental data S3) was then applied to 30 regions (n = 15 from 3 WT rats and 15 from 3 DMD\u003csup\u003emdx\u0026nbsp;\u003c/sup\u003erats) to analyze the vascular structures. For feature extraction, only endomysial vessels were considered using binary mask of muscle fibers. Vessel-to-fiber ratio was determined by dividing the total number of vessels by the total number of fibers. Vessel density was calculated by dividing the number of vessels by the muscle fiber area. Vessel count per fiber and sharing factor were determined by doing binary mask between muscle fibers and microvessels. For that, a proximity area was created by expanding the binary mask of vessels by 10 \u0026micro;m in order to identify fibers in close contact with vessels, defining the fiber-to-vessel association.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThree-dimensional images: vascularization was evaluated on four regions per animal (n = 6, rats), each measuring 526 \u0026times; 526 \u0026times; 350 \u0026micro;m, using the GA3 module. Vascular network was detected using the \u0026ldquo;\u003cem\u003eSegment.ai\u003c/em\u003e\u0026rdquo; model imported into the GA3 pipeline (details are provided in Supplemental data S4) and applied to 24 regions (n = 12 from 3 WT rats and 3 DMD\u003csup\u003emdx\u003c/sup\u003e rats). This enabled the extraction of key features: vascular network density, vascular branching and surface area of vessels embedded within muscle fibers. The extracted data were exported for statistical analysis.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThree-dimensional image analysis of the\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ecollagen network.\u0026nbsp;\u003c/strong\u003eConnective tissue remodeling was evaluated through analysis of SHG endomysial collagen network on four regions of interest (ROIs) per animal (n = 6, rats), each measuring 526 \u0026times; 526 \u0026times; 350 \u0026micro;m. 3D rendering and image analysis were performed using the General Analysis 3 (GA3) module in NIS-Elements software (version 6.10.1, Nikon Europe B.V.). A pre-trained deep learning denoising algorithm was applied to the SHG collagen fiber signal to remove shot noise. Subsequently, a spatial \u0026ldquo;Laplacian of Gaussian\u0026rdquo; (LoG) filter (Gaussian \u0026sigma; = 2.7, kernel size = 5 \u0026times; 5) was applied to identify SHG collagen network boundaries. A manual threshold based on edge detection was used for segmenting the SHG\u003csup\u003e+\u003c/sup\u003e objects. Muscle fibers were segmented using the previously trained deep learning model specifically designed for muscle fiber segmentation. To restrict the analysis to SHG collagen network in the endomysial region, the segmented binary mask of autofluorescent muscle fibers was expanded by a 3 \u0026micro;m kernel diameter. A logical \u0026ldquo;AND\u0026rdquo; operator was then applied between the binary masks of muscle fibers and SHG collagen network. The resulting SHG collagen network binary masks in the endomysial region were used to extract features of interest: density, volume, elongation and inter-fiber distance of SHG\u003csup\u003e+\u003c/sup\u003e objects, overlapping surface between collagen network and vessels. The GA3 pipeline is described in detail in Supplementary Data S5.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e. Statistical analyses were performed using R software (version 4.4.1). Data are presented as the mean \u0026plusmn; standard error of the mean (SEM). Comparisons between groups were conducted using a linear mixed-effects model, with random effects on muscle, implemented in the lme4 package. All models presented in this study were assessed for independence and normality of residuals, as recommended. For datasets that did not meet these assumptions, a Wilcoxon test was used instead. For SHG dataset, loglinear models were used to compare the distribution of measures previously discretized into classes. A significant level of 0.05 was assessed in all statistical tests.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank the staff of the Boisbonne Center (Oniris, Nantes, France) for animal care and the APEX platform (INRAE/Oniris, Center of Excellence Nikon Nantes [CENN], Nantes, France) from UMR 0703 PAnTher (INRAE/Oniris, Nantes, France) for their valuable technological support. We also extend our gratitude to the FAIR CHARM consortium (H2020 program, in which UMR 703 PAnTher is a partner) for fostering a collaborative scientific environment through discussions and the sharing of biological materials. We gratefully acknowledge financial support from R\u0026eacute;gion Pays de la Loire and NeurATRIS: A Translational Research Infrastructure for Biotherapies in Neurosciences\u0026nbsp;,\u0026nbsp;\u0026ldquo;Investissement d\u0026rsquo;Avenir-ANR-11-INBS-0011\u0026rdquo;. The authors also thank Biogenouest (the network of technology core facilities in Western France in life sciences and the environment, supported by the Conseil R\u0026eacute;gional des Pays de la Loire) for supporting APEX.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIbrahim Hassani received a PhD grant from the Association Nationale de la Recherche et de la Technologie (ANRT, grant n\u0026deg; 2022/0868) in collaboration with Nikon France Healthcare, succursale Nikon Europe BV.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eI.H. conceived, designed, and performed the experiments, analyzed the data, and wrote the manuscript. M.L. assisted with sample clearing. C.T. performed statistical analysis, reviewed, and edited the manuscript. T.F. supervised the study, assisted with data acquisition, and reviewed and edited the manuscript. M.A.C. supervised the study, reviewed, and edited the manuscript. K.R. supervised the study, contributed to data interpretation, and participated in manuscript writing. L.D. supervised the study, conceived and designed the experiments, contributed to data interpretation, and participated in manuscript writing. All authors read and approved the final manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBushby, K. M. D. Genetic and clinical correlations of Xp21 muscular dystrophy. \u003cem\u003eJ. Inher Metab. Disea\u003c/em\u003e. \u003cb\u003e15\u003c/b\u003e, 551\u0026ndash;564 (1992).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoffman, E. P., Brown, R. H. \u0026amp; Kunkel, L. M. Dystrophin: The Protein Product of the Duchenne Muscular Dystrophy Locus.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonilla, E. et al. Duchenne muscular dystrophy: Deficiency of dystrophin at the muscle cell surface. \u003cem\u003eCell\u003c/em\u003e \u003cb\u003e54\u003c/b\u003e, 447\u0026ndash;452 (1988).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesguerre, I. et al. Endomysial Fibrosis in Duchenne Muscular Dystrophy: A Marker of Poor Outcome Associated With Macrophage Alternative Activation. \u003cem\u003eJ. Neuropathol. Exp. Neurol.\u003c/em\u003e \u003cb\u003e68\u003c/b\u003e, 762\u0026ndash;773 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKohler, M. et al. Disability and survival in Duchenne muscular dystrophy. \u003cem\u003eJ. Neurol. Neurosurg. Psychiatry\u003c/em\u003e. \u003cb\u003e80\u003c/b\u003e, 320\u0026ndash;325 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMazur, W. et al. Patterns of left ventricular remodeling in patients with Duchenne Muscular Dystrophy: a cardiac MRI study of ventricular geometry, global function, and strain. \u003cem\u003eInt. J. Cardiovasc. Imaging\u003c/em\u003e. \u003cb\u003e28\u003c/b\u003e, 99\u0026ndash;107 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGillies, A. R. \u0026amp; Lieber, R. L. Structure and function of the skeletal muscle extracellular matrix. \u003cem\u003eMuscle Nerve\u003c/em\u003e. \u003cb\u003e44\u003c/b\u003e, 318\u0026ndash;331 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrashear, S. E. et al. Collagen cross-links scale with passive stiffness in dystrophic mouse muscles, but are not altered with administration of a lysyl oxidase inhibitor. \u003cem\u003ePLoS ONE\u003c/em\u003e. \u003cb\u003e17\u003c/b\u003e, e0271776 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoufrani, L. et al. Flow (Shear Stress)\u0026ndash;Induced Endothelium-Dependent Dilation Is Altered in Mice Lacking the Gene Encoding for Dystrophin. \u003cem\u003eCirculation\u003c/em\u003e \u003cb\u003e103\u003c/b\u003e, 864\u0026ndash;870 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLatroche, C. et al. Structural and Functional Alterations of Skeletal Muscle Microvasculature in Dystrophin-Deficient mdx Mice. \u003cem\u003eAm. J. Pathol.\u003c/em\u003e \u003cb\u003e185\u003c/b\u003e, 2482\u0026ndash;2494 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKodippili, K., Thorne, P. K., Laughlin, M. H. \u0026amp; Duan, D. Dystrophin deficiency impairs vascular structure and function in the canine model of Duchenne muscular dystrophy. \u003cem\u003eJ. Pathol.\u003c/em\u003e \u003cb\u003e254\u003c/b\u003e, 589\u0026ndash;605 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLatroche, C. et al. Structural and Functional Alterations of Skeletal Muscle Microvasculature in Dystrophin-Deficient mdx Mice. \u003cem\u003eAm. J. Pathol.\u003c/em\u003e \u003cb\u003e185\u003c/b\u003e, 2482\u0026ndash;2494 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen, F., Guigand, L., Goubault-Leroux, I., Wyers, M. \u0026amp; Cherel, Y. Microvessel density in muscles of dogs with golden retriever muscular dystrophy. \u003cem\u003eNeuromuscul. Disord.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 154\u0026ndash;163 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoufrani, L. et al. Absence of Dystrophin in Mice Reduces NO-Dependent Vascular Function and Vascular Density: Total Recovery After a Treatment with the Aminoglycoside Gentamicin. \u003cem\u003eATVB\u003c/em\u003e 24, 671\u0026ndash;676 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrupler, M. et al. Second harmonic imaging and scoring of collagen in fibrotic tissues. \u003cem\u003eOpt. Express\u003c/em\u003e. \u003cb\u003e15\u003c/b\u003e, 4054 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFuentes-Corona, C. G. et al. Second harmonic generation signal from type I collagen fibers grown in vitro. \u003cem\u003eBiomed. Opt. Express\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e, 6449 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAghigh, A. et al. Second harmonic generation microscopy: a powerful tool for bio-imaging. \u003cem\u003eBiophys. Rev.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 43\u0026ndash;70 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePichon, J. et al. Label-free 3D characterization of cardiac fibrosis in muscular dystrophy using SHG imaging of cleared tissue. \u003cem\u003eBiol. Cell\u003c/em\u003e. \u003cb\u003e114\u003c/b\u003e, 91\u0026ndash;103 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCampagnola, P. J. et al. Three-Dimensional High-Resolution Second-Harmonic Generation Imaging of Endogenous Structural Proteins in Biological Tissues. \u003cem\u003eBiophys. J.\u003c/em\u003e \u003cb\u003e82\u003c/b\u003e, 493\u0026ndash;508 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAltendorf, H. et al. Imaging and 3D morphological analysis of collagen fibrils: Imaging and morphological analysis of collagen fibrils. \u003cem\u003eJ. Microsc.\u003c/em\u003e \u003cb\u003e247\u003c/b\u003e, 161\u0026ndash;175 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBianchini, P. \u0026amp; Diaspro, A. Three-dimensional (3D) backward and forward second harmonic generation (SHG) microscopy of biological tissues. \u003cem\u003eJ. Biophotonics\u003c/em\u003e. \u003cb\u003e1\u003c/b\u003e, 443\u0026ndash;450 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTilbury, K. et al. Second harmonic generation microscopy analysis of extracellular matrix changes in human idiopathic pulmonary fibrosis. \u003cem\u003eJ. Biomed. Opt.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 086014 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDenk, W., Strickler, J. H. \u0026amp; Webb, W. W. Two-Photon Laser Scanning Fluorescence Microscopy. \u003cem\u003eScience\u003c/em\u003e \u003cb\u003e248\u003c/b\u003e, 73\u0026ndash;76 (1990).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLang, R. T. \u0026amp; Spring, B. Q. Two-photon peak molecular brightness spectra reveal long-wavelength enhancements of multiplexed imaging depth and photostability. \u003cem\u003eBiomed. Opt. Express\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, 5909 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLagerweij, T. et al. Optical clearing and fluorescence deep-tissue imaging for 3D quantitative analysis of the brain tumor microenvironment. \u003cem\u003eAngiogenesis\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 533\u0026ndash;546 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichardson, D. S. \u0026amp; Lichtman, J. W. \u003cem\u003eClarifying Tissue Clearing Cell.\u003c/em\u003e \u003cb\u003e162\u003c/b\u003e, 246\u0026ndash;257 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilvestri, L., Costantini, I., Sacconi, L. \u0026amp; Pavone, F. S. Clearing of fixed tissue: a review from a microscopist\u0026rsquo;s perspective. \u003cem\u003eJ. Biomed. Opt.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 081205 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCostantini, I., Cicchi, R., Silvestri, L., Vanzi, F. \u0026amp; Pavone, F. S. In-vivo and ex-vivo optical clearing methods for biological tissues: review. \u003cem\u003eBiomed. Opt. Express\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e, 5251 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTainaka, K., Kuno, A., Kubota, S. I., Murakami, T. \u0026amp; Ueda, H. R. Chemical Principles in Tissue Clearing and Staining Protocols for Whole-Body Cell Profiling. \u003cem\u003eAnnu. Rev. Cell. Dev. Biol.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 713\u0026ndash;741 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRenier, N. et al. A Simple, Rapid Method to Immunolabel Large Tissue Samples for Volume Imaging. \u003cem\u003eCell\u003c/em\u003e \u003cb\u003e159\u003c/b\u003e, 896\u0026ndash;910 (2014). iDISCO.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSusaki, E. A. et al. Advanced CUBIC protocols for whole-brain and whole-body clearing and imaging. \u003cem\u003eNat. Protoc.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 1709\u0026ndash;1727 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGranger, D. N. \u0026amp; Senchenkova, E. \u003cem\u003eInflammation and the Microcirculation\u003c/em\u003e. vol. 2 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLatroche, C. et al. Skeletal Muscle Microvasculature: A Highly Dynamic Lifeline. \u003cem\u003ePhysiology\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 417\u0026ndash;427 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHudlick\u0026aacute;, O. Growth of capillaries in skeletal and cardiac muscle. \u003cem\u003eCirc. Res.\u003c/em\u003e \u003cb\u003e50\u003c/b\u003e, 451\u0026ndash;461 (1982).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeveci, D., Marshall, J. M. \u0026amp; Egginton, S. Relationship between capillary angiogenesis, fiber type, and fiber size in chronic systemic hypoxia. \u003cem\u003eAm. J. Physiol. Heart Circ. Physiol.\u003c/em\u003e \u003cb\u003e281\u003c/b\u003e, H241\u0026ndash;H252 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNederveen, J. P., Betz, M. W., Snijders, T. \u0026amp; Parise, G. The Importance of Muscle Capillarization for Optimizing Satellite Cell Plasticity. \u003cem\u003eExerc. Sport Sci. Rev.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, 284\u0026ndash;290 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcKey, J., Cameron, L. A., Lewis, D., Batchvarov, I. S. \u0026amp; Capel, B. Combined iDISCO and CUBIC tissue clearing and lightsheet microscopy for in toto analysis of the adult mouse ovary\u0026dagger;. \u003cem\u003eBiol. Reprod.\u003c/em\u003e \u003cb\u003e102\u003c/b\u003e, 1080\u0026ndash;1089 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlancy, B. \u003cem\u003eVivo\u003c/em\u003e Microscopy Reveals Extensive Embedding of Capillaries within the Sarcolemma of Skeletal Muscle Fibers. \u003cem\u003eMicrocirculation\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 131\u0026ndash;147 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKharraz, Y., Guerra, J., Pessina, P., Serrano, A. L. \u0026amp; Mu\u0026ntilde;oz-C\u0026aacute;noves, P. Understanding the Process of Fibrosis in Duchenne Muscular Dystrophy. \u003cem\u003eBiomed. Res. Int.\u003c/em\u003e \u003cb\u003e2014\u003c/b\u003e, 1\u0026ndash;11 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiovarelli, M. et al. Characterisation of Progressive Skeletal Muscle Fibrosis in the Mdx Mouse Model of Duchenne Muscular Dystrophy: An In Vivo and In Vitro Study. \u003cem\u003eIJMS\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 8735 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePodkalicka, P. et al. Age-Dependent Dysregulation of Muscle Vasculature and Blood Flow Recovery after Hindlimb Ischemia in the mdx Model of Duchenne Muscular Dystrophy. \u003cem\u003eBiomedicines\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 481 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePodkalicka, P., Mucha, O., Dulak, J. \u0026amp; Loboda, A. Targeting angiogenesis in Duchenne muscular dystrophy. \u003cem\u003eCell. Mol. Life Sci.\u003c/em\u003e \u003cb\u003e76\u003c/b\u003e, 1507\u0026ndash;1528 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLattouf, R. et al. Picrosirius Red Staining: A Useful Tool to Appraise Collagen Networks in Normal and Pathological Tissues. \u003cem\u003eJ. Histochem. Cytochem.\u003c/em\u003e \u003cb\u003e62\u003c/b\u003e, 751\u0026ndash;758 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLapierre-Landry, M., Liu, Y., Bayat, M., Wilson, D. L. \u0026amp; Jenkins, M. W. Digital labeling for 3D histology: segmenting blood vessels without a vascular contrast agent using deep learning. \u003cem\u003eBiomed. Opt. Express\u003c/em\u003e. \u003cb\u003e14\u003c/b\u003e, 2416 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarcher, T. et al. Characterization of Dystrophin Deficient Rats: A New Model for Duchenne Muscular Dystrophy. \u003cem\u003ePLoS ONE\u003c/em\u003e. \u003cb\u003e9\u003c/b\u003e, e110371 (2014).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"muscular dystrophy, vasculature, fibrosis, clearing, multiphoton microscopy, image analysis","lastPublishedDoi":"10.21203/rs.3.rs-6156479/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6156479/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDuchenne muscular dystrophy (DMD) is characterized by repeated cycles of muscle fiber necrosis/regeneration and their progressive replacement by fibrous and adipose tissues. Vascular abnormalities are also reported as pathophysiological hallmark. Traditional investigations of these morphological changes rely on two-dimensional (2D) assessments of thin tissue sections. However, they fail to capture the global spatial distribution and structural disorganization within the muscle.\u003c/p\u003e \u003cp\u003eIn this study, we developed a 3D approach combining multimodal microscopy and tissue-clearing methods to investigate the properties of microvascular and connective tissue networks in a dystrophic context. By using segmentation techniques based on deep learning models, we established a dedicated 3D image analysis workflow. We analyzed samples from healthy and dystrophic rats, quantifying key parameters for vascular and connective tissue compartments.\u003c/p\u003e \u003cp\u003eWe showed a profound spatial reorganization of the vascular network in dystrophic muscle, characterized by its embedding within connective tissue and a consequent reduction in physical interactions with muscle fibers.\u003c/p\u003e \u003cp\u003eOur findings demonstrate that this novel imaging approach provides detailed insights into the extent of remodeling in the vascular system and connective tissue of dystrophic muscle. It holds significant potential as a powerful tool for monitoring disease progression and evaluating the impact of therapeutic interventions.\u003c/p\u003e","manuscriptTitle":"Multiphoton microscopy and tissue clearing for 3D characterization of the vasculature and fibrosis remodeling in rat dystrophic skeletal muscle","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-11 19:55:32","doi":"10.21203/rs.3.rs-6156479/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-09T06:43:12+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-02T18:56:37+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"29314983901458200829774816929953383595","date":"2025-06-19T16:42:42+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-08T15:52:07+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"214344231415136878330051646123271521417","date":"2025-03-31T11:50:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"312971873764749033738332393528750960910","date":"2025-03-21T13:25:57+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-17T14:36:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-03-17T14:31:28+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-03-13T03:56:42+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-12T05:27:05+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-03-04T18:06:43+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ff1240dd-b0aa-4493-8f32-14e1767c7a96","owner":[],"postedDate":"June 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":46866303,"name":"Physical sciences/Optics and photonics/Optical techniques/Microscopy/Multiphoton microscopy"},{"id":46866304,"name":"Biological sciences/Computational biology and bioinformatics/Image processing"},{"id":46866305,"name":"Health sciences/Anatomy/Musculoskeletal system/Muscle/Skeletal muscle"}],"tags":[],"updatedAt":"2025-10-27T16:24:17+00:00","versionOfRecord":{"articleIdentity":"rs-6156479","link":"https://doi.org/10.1038/s41598-025-20335-9","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-10-21 16:16:44","publishedOnDateReadable":"October 21st, 2025"},"versionCreatedAt":"2025-06-11 19:55:32","video":"","vorDoi":"10.1038/s41598-025-20335-9","vorDoiUrl":"https://doi.org/10.1038/s41598-025-20335-9","workflowStages":[]},"version":"v1","identity":"rs-6156479","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6156479","identity":"rs-6156479","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}