Human myogenic precursor cells have been isolated and expanded from a number of skeletal muscles, but alternative donor biopsy sites must be sought after in diseases where muscle damage is widespread. Biopsy sites must be relatively accessible, and the biopsied muscle dispensable. Here, we aimed to histologically characterize the cremaster muscle with regard number of satellite cells and regenerative fibres, and to isolate and characterize human cremaster muscle-derived stem/precursor cells in adult male donors with the objective of characterizing this muscle as a novel source of myogenic precursor cells. Cremaster muscle biopsies (or adjacent non-muscle tissue for negative controls; N = 19) were taken from male patients undergoing routine surgery for urogenital pathology. Myosphere cultures were derived and tested for their in vitro and in vivo myogenic differentiation and muscle regeneration capacities. Cremaster-derived myogenic precursor cells were maintained by myosphere culture and efficiently differentiated to myotubes in adhesion culture. Upon transplantation to an immunocompromised mouse model of cardiotoxin-induced acute muscle damage, human cremaster-derived myogenic precursor cells survived to the transplants and contributed to muscle regeneration. These precursors are a good candidate for cell therapy approaches of skeletal muscle. Due to their location and developmental origin, we propose that they might be best suited for regeneration of the rhabdosphincter in patients undergoing stress urinary incontinence after radical prostatectomy.
In striated muscle, adult myogenic stem cells are known as satellite cells, due to their superficial position on muscle fibres1. The myogenic process is a multifaceted transition between precursor states (quiescence, activation, proliferation and differentiation) that precede fusion of the myoblasts to regenerative muscle fibres2. Besides, satellite cells reside in a complex niche, which includes other precursors such as fibro-adipogenic precursor cells (FAPs) that modulate the regenerative response3, along with signals arising from nerve and capillary terminals and other interstitial cells. For cell-based therapeutic purposes, it would thus be desirable to obtain and characterize the diverse types of human muscle precursor cells from an accessible source.
Most protocols of human satellite cell isolation rely on the purification of cell subpopulations by flow cytometry or magnetic separation of muscle-derived cell suspensions through differential expression of membrane markers4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21. Despite the important recent advances in the purification and characterization of human satellite cells, they are still isolated in small numbers out of muscle biopsies of a limited size (typically of 50–100 mg; there are between 500–1,000 satellite cells per mm3 20), and the stem cells present restricted expansion capacities in vitro22,23. For these reasons, myoblasts (a heterogeneous mixture of semi-purified CD56+ cells24,25) have been mostly used in clinical trials to date, since they expand well in vitro26,27. However, the results of expanded myoblast cell infusion in a plethora of clinical trials targeting muscle regeneration have been disappointing28,29,30,31,32. Some authors argue that the large expansion rates of myoblasts have brought up excessive differentiation in culture and hence low survival, migration and fusogenic capacities when cells are transplanted in vivo33,34,35. Of note, more experiments must be done with human myoblasts to ensure that this is also the case in humans36.
A still relatively unexplored possibility is the growth of human muscle precursor cells in three-dimensional myosphere cultures37. In mice, myospheres represent a mixture of ITGA7+ Myf5+ MyoD+Pax7+ myogenic (possibly satellite cells) precursor cells and non-myogenic (possibly FAPs) precursors characterized as PDGFRα+Sca−1+38,39. Since both precursor cell populations are required for muscular regeneration, myosphere cultures would have the advantage of providing two precursor cell types instead of one, when compared to alternative satellite cell isolation strategies. Human myosphere cultures present at least a CD31−CD34−CD45−CD56−CD117−CD29+CD73+CD90+CD105+ “mesenchymal” stem cell population, which possibly is non-myogenic40 and could be equivalent to FAPs, and a CD34−CD45−PAX7+CD56+ALDH1+ myogenic precursor cell population that possibly corresponds to satellite cells41.
In this article, we propose human cremaster muscle as a convenient source of muscle-derived stem/precursor cells in male donors. This is a striated muscle not inserted through a tendon, and which also contains a variable number of smooth muscle fibres42. Predominantly, it is composed of type I (slow) fibres, although it also contains some of IIB (very fast) type. The function of this muscle in the adult is to contribute to the thermoregulation and protection of testicles, and we postulate it should be classified as an evolutionary remnant of mammalian Panniculus carnosus muscle43. Thanks to the cremasteric reflex, its electrophysiological properties are well known. The muscle is densely innervated and presents numerous motor endplates, which may be the reason underlying its abundant spontaneous discharges42. In children, no sexual dimorphism was observed in cremaster muscle except for a larger diameter of fibres in males, as it is commonly observed in most muscular groups44. In embryonic development, cremaster muscle derives from the gubernaculum, independent of the internal oblique muscle of the abdomen, and it performs a key function in testicular descent45,46,47. However, some authors propose that striated cremaster fibres transdifferentiate from smooth muscle instead48, as it may happen in other muscles of the genitourinary tract, such as the rhabdosphincter49.
Since alternative donor biopsy sites must be identified in diseases where muscle affection is widespread, we here aimed to histologically characterize the cremaster muscle with regard number of satellite cells and regenerative fibres, and to isolate and characterize human cremaster muscle-derived stem/precursor cells in adult male donors to evaluate this muscle as a novel source of myogenic precursor cells.
Histological characterization of human cremaster muscle
The cremaster muscle is surgically accessible in the context of male patients undergoing routine surgery for urogenital pathology (mainly hydrocele and varicocele). Histological characterization (haematoxylin and eosin stain) of cremaster muscle biopsies of these patients (Table 1) showed the presence of a discrete percentage (0.5–3%) of centrally nucleated, regenerative striated fibres as well as some interspersed smooth muscle fibres (Fig. 1), as expected. By immunofluorescence, striated fibre sarcomeres were clearly delineated by myosin heavy chain (MYHC all fibres) antibody staining, and muscle fibres were surrounded by LAMININ positive basal membrane (Fig. 2A,B). Predominance of type I (slow) fibres and the presence of fewer number of type II (fast) fibres was corroborated by the expression of specific MYHC I and MYHC II isoforms, respectively (Fig. 2C–F). The existence of newly formed fibres was confirmed by expression of the embryonic isoform of MYHC, MYH3 (Fig. 3A,B, arrows). To quantify the number of satellite cells in situ, the number of PAX7+ nuclei that were surrounded by LAMININ+ basal membrane was determined (Fig. 3C,D, arrow). A proportion of 1.8 ± 0.3% satellite cells were calculated.
In vitro myotube formation from human cremaster muscle-derived cells
To evaluate in vitro myogenic potential of human cremaster muscle-derived cells, a protocol previously used in mouse cultures50 was adapted to human biopsies (Fig. 4A). At day 0 (d0), suspension cultures presented abundant cellular debris and dead cells as well as unicellular suspensions and muscle tissue remnants (Fig. 4B, arrow). After 7 days of myosphere culture, cells formed spheres of variable size (Fig. 4C). These spheres were then put into differentiation culture51. The cells adhered to the substrate and started forming multinucleated myotubes by d2 (Fig. 4D) and, by d9, myotubes occupied most of the culture surface (Fig. 4E,F, arrows). These results suggested that human cremaster muscle-derived cells adopted a myogenic commitment in vitro in response to appropriate cues.
To demonstrate the presence of myogenic cells in these cultures, myogenic markers PAX7 and MYOGENIN were detected by immunofluorescence in d7 myosphere cultures, showing discrete numbers of positive cells in the myospheres (Fig. 5A,B, arrows). Satellite (PAX7+) cells were quantified at day 15 of suspension culture (Fig. 5C). A median proportion of 16.2% PAX7+ cells (Fig. 5D) indicated that the sphere culture differentially enriched satellite cells over other cell types present at day 0, as previously observed in mouse dermosphere cultures50. To determine if endothelial (CD31+) cells and mesenchymal (CD90+) cells were also present in the cremaster-derived myosphere cultures, the presence of cells positive for these markers was analysed by immunofluorescence and confocal microscopy. Both CD31+ and CD90+ cells were easily detected, although quantification was difficult due to overlapping of the signal with numerous nuclei in the confocal sections (Fig. 5E–J). Interestingly, CD31+ cells seemed to form tube-like structures, reminiscent of angiotubes (Fig. 5E–G, arrows).
After 9 days in differentiation culture (Fig. 6), single cells expressing PAX7 were still detected (Fig. 6A), with a median proportion of 6.9% PAX7+ cells (Fig. 6B). They were often observed in positions adjacent to MYHC+ multinucleated myotubes, which presented characteristic sarcomeric striations (Fig. 6C,D, arrows). The expression of myogenic genes was confirmed by qRT-PCR after 9 days in differentiation culture (Fig. 7). As negative control, patient biopsies that were taken from non-muscle tissue adjacent to the cremaster were used (Table 1). The cultures that were derived from cremaster muscle, but never those derived from non-muscle tissue, expressed variable but consistent amounts of the myogenic genes PAX7, MYF5, MYOD1, MYOGENIN, MYH3 and MYH2 (Fig. 7A–F). These results confirmed that cells of myogenic commitment were maintained by myosphere culture and were differentiated to myotubes in adhesion culture.
In vivo regenerative potential of human cremaster muscle-derived cells in a mouse model of muscular damage
To evaluate the potential use of human cremaster muscle-derived cells in future cell-based therapy trials, we analysed their regenerative potential in a proof-of-concept preclinical assay of muscle regeneration (Fig. 8A)52,53. Four weeks after cell injection, TA muscles were extracted and analysed by immunofluorescence. In the experimental TA muscle group (N = 6), a variable number of human cells was detected by their reactivity for the highly specific anti-human LAMIN A/C antibody, which was absent in the control leg (Fig. 8B,C). Of the LAMIN A/C+ cells, a relatively small percentage co-expressed the satellite cell marker PAX7 (Fig. 8D–F, arrows). The number of muscle fibres of human origin, as determined by the expression of human DYSTROPHIN by immunofluorescence (Fig. 8G,H), correlated quite well to the number of human cells detected per mice. Quantification of these human-specific markers (Fig. 9A,B) showed an average number of 1864 ± 2247 LAMIN A/C+ cells per section, of which 1.8 ± 0.6% were also PAX7+. The number of fibres of human origin (detected as hDYSTROPHIN+) was highly variable, 34.2 ± 34.0 fibres per section (Fig. 9C). These results suggested that human cremaster muscle-derived stem cells survived to the transplants and contributed to muscle regeneration in response to cardiotoxin damage.
Cell-based therapeutic approaches for muscle regeneration may be applicable in congenital myopathies54, muscular dystrophies and other neuromuscular diseases55, cardiac dysfunction56, volumetric muscle mass loss, cachexia and sarcopenia57. However, several of these pathologies affect relevant volumes of muscle tissue and this fact compromises the feasibility of cell-based approaches. For this reason, smaller muscle groups such as those affected in oculopharyngeal or in facioscapulohumeral muscular dystrophies may be approached with greater chances of success35,58. Anal and urinary sphincter deficiencies may also be excellent targets for these therapies59,60 since the muscle volume to be regenerated is small and the defect accessible via minimally invasive surgical approaches. For any of these pathologies it will be instrumental that adequate animal models are developed due to the inherent variability seen in the clinical setting61.
Despite the fact that the Panniculus carnosus muscle is vestigial in humans43, we were interested in testing if those evolutionary remnants still available in human beings would be of use to isolate muscle satellite cells of possible use in cell therapy, as we had previously done in the mouse50. The cremaster muscle was selected because it can be biopsied through a small inguinal incision which is routinely performed in several common urogenital surgeries in males. It is a predominantly slow twitch muscle and thus the number of satellite cells would be expected to be higher, although this seems to be dependent on muscle loading62. In this work, we found a satellite cell proportion in situ that seems to be below what has been established for other muscle groups63. The percentage of centrally nucleated (regenerative) fibres was on a similar range to other muscle groups64,65.
Nevertheless, we have also shown that cremasteric myogenic precursor cells can be maintained and expanded in vitro reaching sizable numbers (about 15% of the myosphere cells), and that the expanded cells retain muscle regenerative capacities in vivo. The main limitations of this initial study are: (i) the low number of samples analysed; (ii) the fact that this muscle can only be found in males, leaving aside half of the adult population; and (iii) that we characterized samples from patients and not from healthy individuals. For instance, higher grade varicoceles might be associated with denervation of cremaster muscle, causing small group atrophy66. These issues will be addressed by future investigations.
To apply cremaster-derived myogenic precursor cells in muscle pathology, a number of obstacles must be overcome: first, it must be studied if this muscle is affected by the relevant pathology at study, and second, an analogous muscle group should be found in adult women. A possible candidate could be the round ligament of the uterus, which is also composed of striated and smooth muscle fibres and that extends from uterus to deep inguinal ring, sometimes inserting itself into adipose tissue and skin of labia majora.
Importantly, the human cremaster muscle-derived myospheres were also shown to present mesenchymal (CD90+) and endothelial (CD31+) cells, and we postulate that, most likely, they will contain Schwann cells as well67. Human bone marrow-derived mesenspheres are CD90+, and contain bone marrow-derived stem cells which preserve an immature phenotype68. Tissue-resident CD31+ endothelial precursor cells have previously been isolated from mouse muscle69. Endothelial cells in co-culture spheres may self-assemble and form reticulated structures, as demonstrated here and previously seen in other systems70,71. The presence of both of these cell types might contribute to an improved vascularization of the affected area, as demonstrated in bone defects72. Finally, if Schwann cells and other peripheral nerve-derived cells were also present in these spheres, they might support reinnervation and regeneration of the degenerated tissue73. We would like to propose that, due to their origin and location, the mix of precursor cells present in the myospheres may be a good candidate to be used in cell therapy approaches of stress urinary incontinence after radical prostatectomy. Ideally, the isolation protocol should be adjusted so that precursor cell extraction and treatment may be done in the course of a single intervention74,75. Similarly, CD56+ myoblasts from pyramidal muscle have been obtained from radical prostatectomy patients76 and myoblasts derived from rectus muscles in patients undergoing open abdominal surgery77. Our approach would provide a more comprehensive pro-regenerative cellular mix.
Human cremaster is a predominantly low twitch muscle with relatively few satellite cells and some regenerative fibres in situ. Cremaster-derived myogenic precursor cells can be isolated and expanded as myospheres, and the expanded cells retain muscle regenerative capacities in vivo. We propose that these precursors are a good candidate for cell therapy approaches of skeletal muscle. Due to their location and developmental origin, they might be best suited for the treatment of stress urinary incontinence after radical prostatectomy.
Aim, design and setting of the study
This study aimed to characterize in vitro and in vivo human cremaster muscle stem cells isolated from male donor biopsies (N = 19) and propagated in vitro in the form of myospheres.
Ethics approval and consent to participate
The protocol to obtain cremaster and non-muscle tissue biopsies from patients after signature of the informed consents followed all relevant legal and ethical regulations, and was approved by the Ethics and the Scientific Committees of the HUVV (CEUMA No. 79-2015-A). Animal experiments were carried out following all relevant legal and ethical regulations, and according to protocols approved by the Biodonostia Animal Care Committee (CEEA16/008).
Human cremaster muscle biopsies
Upon informed consent signature from male patients undergoing inguinal hernia (N = 10), hydrocele (N = 4), varicocele (N = 4), or spermatic cord cyst (N = 1) surgeries at the Virgen de la Victoria University Hospital (HUVV), a 1–3 cm2 cremaster muscle biopsy (or adjacent non-muscle tissue for negative controls) was collected during regular surgery procedure without adding any morbidity to the process. Biopsy samples were managed by the HUVV-IBIMA Biobank. Men with uro-oncologic disease or other diseases that could affect the scrotal area were not included as donors. The characteristics of biopsy donors are described in Table 1.
Histology and immunohistochemistry
Deidentified human cremaster biopsies from surgery room (processed within 30 min of the procedure) were embedded in OCT medium (Q-Path, VWR), frozen by immersion in isopentane previously cooled in liquid nitrogen and stored at −80 °C until usage. Cryostat sections (7–10 μm) were stained with haematoxylin and eosin (H&E; Panreac), and mounted with Shandon Consul-Mount mounting media (Fisher Scientific) according to standard procedures. For immunohistochemistry, biopsy sections from OCT blocks were dried; and fixed in 4% Paraformaldehyde aqueous solution (Electron Microscopy Sciences) 10 min at room temperature (RT) or directly blocked with 5% bovine serum albumin (BSA) in PBS for 1 h at RT; and incubated overnight at 4 °C with the primary antibody in 1% BSA solution. Slides were rinsed in PBS containing 0.025% Triton X-100 (Amresco) and incubated for at least 1 h with the appropriate secondary antibody in the same solution used for blocking. The slides were rinsed with PBS containing 0.025% Triton X-100. Nuclei were stained with 10 μg/ml Hoechst solution (Santa Cruz, SC-394039) for 2 min and slides were mounted with Fluoro-Gel mounting media (Electron Microscopy Sciences, Cat #17985–10). Sections were imaged with a Nikon Eclipse 80i fluorescence microscope.
Isolation, proliferation, and striated muscle differentiation of muscle precursor cells
Deidentified human cremaster biopsies from surgery room (processed within 30 min of the procedure) were transported immersed on HBSS (ThermoFisher, 14170-088) with 1% Fungizone and 1% Penicillin on ice. Biopsies were mechanically dissociated and digested with a collagenase type IA solution (Sigma-Aldrich, 650 CDU/mg, 2 mg/ml) for 1–2 h at 37 °C under gentle shaking. Collagenase was inactivated with culture media and the resulted digested tissue was filtered through 40-μm Nylon Cell Strainers (Corning). Collagenase was removed by centrifugation at 1500 rpm for 5 min at RT.
The cellular pellet containing muscle-derived cells was resuspended and cultured in suspension as described50, to promote myosphere expansion in proliferation medium [Neurobasal A (ThermoFisher) with 2% 50X B27 supplement (ThermoFisher), 1% 200 mM L-glutamine [Sigma-Aldrich], and 1% penicillin/streptomycin (100×)], supplemented every two days with 2% low serum growth supplement (1×, ThermoFisher), 40 ng/mL epidermal growth factor (R&D Systems), and 80 ng/mL basic fibroblast growth factor (FGF2; R&D Systems). Differentiation of myospheres was performed in adherent culture as described51. First, extracellular matrix-coated glass coverslips were prepared by incubating a solution of Cultrex basement membrane extract (2.77 mg/mL; Trevigen), Netrin-4 (0.83 μg/mL), Netrin-G1a (0.83 μg/mL), and low molecular weight hyaluronan (2.5 mg/mL; R&D Systems) in phosphate-buffered saline (PBS; pH 7.4). For muscle induction, primary myospheres were gently disaggregated with a 0.25% trypsin-EDTA solution (Sigma-Aldrich) and resuspended in proliferation medium without added growth factors plus 10% fetal bovine serum (ATCC), before plating onto coated coverslips at a density of 79,000 cells/cm2. Every 2 days, half of the medium was replaced with a fresh medium.
Immunofluorescence of myogenic markers in myospheres
Immunofluorescence was performed in suspended sphere cultures. For suspended spheres, 1 ml of suspended culture was collected and centrifuged at 1500 rpm for 5′ at RT. The cell pellet was washed with 1X PBS and fixed in 4% Paraformaldehyde aqueous solution (Electron Microscopy Sciences) for 10 min at RT in a rotating station. Spheres were washed in PBS, permeabilized for 1 h in 0.3% Triton X-100 (Amresco) in PBS (PBS-T) and 5% normal donkey serum (Santa Cruz, SC-2044) at RT in the rotation station. Spheres were incubated with the appropriate primary antibody diluted in PBS-T overnight at 4 °C, rotating. Next, cells were 1X PBS-washed 3 times (5 min each) and incubated with the appropriate secondary antibody diluted in PBS-T for 1 h at RT in darkness and rotating. Prior to mounting in Fluoro-Gel, spheres were incubated with 10 μg/ml Hoechst (Santa Cruz, SC-394039) for 2 min rotating at RT and in darkness and washed with 1X PBS. Images were obtained by using a Nikon Eclipse 80i microscope coupled to a Nikon Digital Sight camera or a Zeiss LSM800 Confocal Laser Scanning Microscope. Confocal z-stack images were taken with the 25x objective (1.96 μm between slides). Images were processed using ZEN blue software and ImageJ software.
Gene expression in differentiated cultures
Total RNA was extracted from differentiated cultures from cremaster biopsies and non-cremasteric biopsies by miRNeasy Mini Kit (Qiagen) and converted into complementary DNA with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Each cDNA sample was amplified in triplicates for the real-time quantitative PCR (qRT-PCR) analysis which was carried out using Taqman gene expression assays in the 7900 HT Fast Real-Time PCR System (Applied Biosystems). The cycling conditions were 95 °C/10 min followed by 40 cycles at 95 °C/15 s, 60 °C/1 min in a reaction mixture that contained 1x Taqman Universal PCR Master Mix and 1x Assay Mix in a final volume of 20 μl. The expression of the genes was represented relative to the housekeep gene Tbp expression.
Intramuscular cell transplantation
Animal experiments were carried out following the experimental design of Darabi and colleagues52,53. One day before cell transplantation, 10-week old immunocompromised NSG mice (NOD.Cg-PrKdcsidIl2rgtm1Wjl/SzJ, JAX 005557) (N = 6) were anesthetized with isoflurane and Tibialis anterior (TA) muscle damage was induced in both rear limbs through injection of 7.5 μL of 100 μM cardiotoxin (CTX, Latoxan) by using a 26 s gauge Hamilton syringe (bevelled tip). 24 h later, cells were injected (235,000 cells in 15 μl of PBS pH 7.2) into right TA muscles, whereas the left TA muscles received the same volume of PBS (negative controls). Before transplantation, human cremaster-derived precursor cells had been cultured as myospheres for 7 days, dissociated with 0.25% Trypsin-EDTA, filtered through 70 μm Nylon Cell Strainers (Corning) and resuspended in 1X PBS. Animals were caged by groups with ad libitum access to food and water and they were monitored for 4 weeks until sacrifice. Animals were sacrificed with CO2 and engrafted muscles were removed and processed for histological analysis. Seven-micrometer serial transverse cryosections were cut at intervals of 100 μm throughout the entire muscles.
Primary antibodies used were anti-CD31 (CD31) (M0823; 1:30; Dako); anti-CD90 (CD90) (ab133350; 1:200; Abcam); anti-hDystrophin (DYSTROPHIN) (NCL-DYS3; 1:20; Leica); anti-hLamin A/C (LAMIN A/C) (ab108595; 1:200; Abcam); anti-Laminin (LAMININ) (L9393; 1:200; Sigma); anti-MYOD (MyoD1) (SC-377460; 1:50; Santa Cruz); anti-myosin heavy chain all fibers (MYHC) (A4-1025-c; 1:200; DSHB); anti-myosin heavy chain embryonic (MYH3) (F1.652; 1:50; DSHB); anti-myosin heavy chain fast fibers (MYHC II) (A4.74; 1:200; DSHB); anti-myosin heavy chain slow fibers (MYHC I) (A4.840; 1:200 M; DSHB); and anti-PAX7 (Pax7) (Pax7-c; 1:50; DSHB). Secondary antibodies used were donkey anti-mouse Alexa Fluor 488 (A21202; 1:500; Thermo Fisher), donkey anti-mouse Alexa Fluor 488 (A21042; 1:500; Thermo Fisher), donkey anti-mouse Alexa Fluor 555 (A31570; 1:500; Thermo Fisher), donkey anti-rabbit Alexa Fluor 488 (A21206; 1:500; Thermo Fisher), and donkey anti-rabbit Alexa Fluor 555 (A31572; 1:500; Thermo Fisher).
All data generated or analysed during this study are included in this published article.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank patients and medical personnel for their generous involvement in the study. We also acknowledge the help of Biodonostia Animal and Experimental Operations Facility. This work was supported by grants from Ministerio de Economía y Competitividad (RTC-2015-3750-1) and Instituto de Salud Carlos III (PI13/02172, PI16/01430) to A.I., co-funded by the European Union (ERDF/ESF, ‘Investing in your future’). N.N.-G. received a studentship from the Department of Education, University and Research of the Basque Government (PRE2013-1-1168). A.L.M. was funded by grants from FIS (PI17/01841 and PI14/00436), CIBERNED and the Basque Government (2015/11038, RIS3 2017222021 and BIO16/ER/022). M.F.L.-C. was supported by the Servicio Andaluz de Salud from the Consejería de Salud de la Junta de Andalucía, grant PI 0222-2014, co-funded by the European Union (ERDF/ESF). I.M.A was funded by grants from Ministerio de Economia y Competitividad (PEJ-2014-P-01215 and FJCI-2016-28121).