Preclinical efficacy of adipose-derived cell therapies for the treatment of myositis

Abstract

Importance

Idiopathic inflammatory myopathies, commonly referred as myositis, are autoimmune diseases that cause muscle damage, progressive weakness, and disability. Current treatments, including corticosteroids and immunosuppressants, have significant limitations, highlighting the need for new therapies.

Objective

This preclinical study explored the therapeutic potential of adipose tissue–derived cell therapies, specifically stromal vascular fraction (SVF) and adipose-derived stem cells (ADSC), using an Icos^−/− NOD mouse model of spontaneous myositis.

Design

SVF and ADSC were extracted from CD1 female mice adipose tissue and cultured. Various doses were injected intramuscularly into the right hind limb of 20- to 22-week-old female Icos^−/− NOD mice with a control group. The therapeutic effects were assessed through clinical scoring, grip strength test, and motor function analysis using Catwalk system. Muscle atrophy was evidenced by histology, and systemic inflammation was analyzed by flow cytometry.

Results

Mice treated with either SVF or ADSC showed a dose-dependent slowdown in disease progression and improvements in motor functions, such as gait, movement, speed, and weight distribution between the legs. Histological analysis showed a reduction in muscular atrophy, particularly in the injected limb. Flow cytometry analysis on lymph nodes showed shifts in leukocyte populations, with reduced expression of inflammatory and activation markers.

Conclusions and relevance

Overall, this study demonstrated the therapeutic potential intramuscular injection of SVF or ADSC in the Icos^−/− NOD mouse model of myositis, providing a proof-of-concept for the use of adipose tissue–derived cell therapies in the treatment of idiopathic inflammatory myopathies.

Graphical abstract

Graphical Abstract.

Significance Statement:

Myositis is a group of autoimmune muscle diseases that need new therapeutic options. This preclinical study aims to demonstrate the therapeutic potential of stromal vascular fraction and adipose-derived stem cells. We studied intramuscular injection of these therapies in a mouse model of spontaneous myositis and showed for each of them a dose-dependent slowing of disease progression and improvements in motor function. Histological analysis showed a reduction in muscle atrophy, and flow cytometry showed a decrease in the expression of inflammatory and activation markers on leukocyte populations. These results form the basis of a clinical trial in inclusion body myositis.

Introduction

Idiopathic inflammatory myopathies, also known as myositis, are characterized by autoimmune muscle damage, which results in progressive muscle weakness and disability. These conditions include polymyositis, dermatomyositis, immune-mediated necrotizing myositis, inclusion body myositis (IBM), and overlap myositis.^1–3 The primary treatment for myositis involves the use of corticosteroids either alone or in combination with immunosuppressive agents.⁴ For patients who do not respond to these treatments, alternative therapies are being investigated, including immunosuppressive biologics frequently used to treat autoimmune and chronic inflammatory diseases.⁵ These medications are designed to reduce leukocyte infiltration and humoral autoreactivity.^6–8 One of the promising approaches under investigation is the application of mesenchymal stem cells (MSCs)-based therapies, which are used for their ability to regenerate the damaged tissues and to regulate the immune system.^9–11 Previously limited by the difficulties in extracting MSCs from bone marrow or cord blood, the discovery of their presence in adipose tissue has expanded their therapeutic potential revived interest in these cell therapies.^12,13

Two types of cell pools can be extracted from adipose tissue for cell therapy: the stromal vascular fraction (SVF), which is a mixture of various cell populations playing a role in maintaining adipose tissue homeostasis, and adipose tissue-derived stem cells (ASDCs), mainly MSCs obtained from SVF culture.¹⁴ Criteria for identifying ADSCs have been proposed to ensure consistency of preparation and use,^15,16 and these guidelines are continually evolving in line with advances in the field.¹⁷ Previous research has demonstrated that adipose-derived cell therapies (ADCT) possess immunomodulatory, regenerative, pro-angiogenic, and anti-fibrotic properties, which have been shown to be effective in the treatment of autoimmune diseases.¹⁸ These properties may also make them suitable options for the treatment of myositis.¹³

Although indispensable, preclinical models used for the development of myositis treatments have limitations. For example, most murine models are induced by immunization against an antigen, which can distort evaluation of immunomodulatory therapies.^19,20 In addition, these models often display acute rather than chronic disease features, with moderate clinical or biological manifestations. The Icos^−/− NOD mouse model offers significant advantages, as it mimics key features of myositis. The onset of myopathy symptoms in these mice generally occurs between 20 and 25 weeks of age, with full disease at around 35 weeks of age. It is characterized by severe myopathy accompanied by significant muscle atrophy and loss of motor function, ultimately leading to significant walking disability.^21,22 Here, we used Icos^−/− NOD mice to provide proof-of-concept for the therapeutic potential of SVF and ADSCs in the treatment of myositis.

Methods

Mice

CD1 mice were purchased from Charles River Laboratories. Icos^−/− NOD mice were bred and housed in our facility under specific pathogen-free conditions in secure rodent facilities, maintained on a 12-hour light/dark cycle with constant access to food and water. They were generated as previously described.²³ The animal protocol was designed to minimize pain and discomfort for animals. All experimental procedures complied with the European Community guiding principles on the care and use of animals, the French Decree No. 97/748 of October 19, 1987 and the recommendations of the Cenomexa Ethics Committee. The animal experiments were approved on October 2019 by the Cenomexa Ethics Committee (#2019-083016072100). At the endpoint of protocols and for adipose tissue harvest, mice were euthanized by cervical dislocation. Icos^−/− NOD mice were included in the study and injected before disease onset. No inclusion or exclusion criteria were applied. Mice included in therapeutic protocols were randomized to have the same age (median) between different groups in the same protocol. The investigators were unblinded to the tests and analyses. The clinical score of injected Icos^−/− NOD mice was assessed on a weekly basis by evaluating muscle disability, as previously described.²¹ Briefly, this score is based on the number of legs showing impaired extension/flexion when suspended by the tail (from 0 to 4), combined with the number of legs showing a defect when walking on a grid (0 to 4). An extra point was added if the mouse had breathing difficulties. Mice with a clinical score >8 were euthanized, and muscle and lymph nodes were harvested for analysis when possible. In fact, some mice were found dead during daily control due to the disease and were excluded from the endpoint analysis. Muscle strength on the front limbs was assessed using a grip test (Bioseb). Locomotor activity was evaluated 14 weeks after injection using the CatWalk XT gait analysis system (Noldus).

Stromal vascular fraction extraction and injection

Inguinal and perianal white adipose tissue was collected after lymph node extraction from CD1 female mice between of 8 and 12 weeks of age. CD1 mice were chosen because they share a close genetic background with NOD mice and therefore Icos^−/− NOD mice. Adipose tissues were washed three to four times with phosphate-buffered saline (PBS; (Gibco) and finely minced. Adipose tissues were then placed in an incubator shaker set to 4 G at 37 °C for 10 minutes before addition of volume to volume of a digestion solution containing 0.25 U/mL of collagenase NB5 type I and II (Nordmark) in PBS. Incubation was performed with shaking for 1 hour. The digestion reaction was neutralized with PBS containing 10% fetal bovine serum (FBS; Gibco). The digested tissue was centrifuged and the supernatant containing adipocytes was removed. The solution was filtered through a 100 µm strainer and centrifuged to obtain pellets of SVF cells. After counting the living cells, the cellular suspension was diluted in 50 µL of PBS to different concentrations for injection of 10⁶ or 5 × 10⁵ cells distributed between the two heads of the right gastrocnemius muscle of 20- to 22-week-old isoflurane anesthetized mice. Each group of doses was composed of eight mice, and a control group was injected with PBS. Stromal vascular fraction (SVF) of Icos^−/− NOD female mice aged 8-12 or 22 weeks was also prepared.

Adipose-derived stem cell isolation and injection

For the isolation of ADSC, SVF was placed in a 75 cm² flask containing complete culture medium (CCM) consisting of Dulbecco’s modified Eagle’s medium combined with the nutrient mixture F12 (DMEM-F12; Lonza) with 20% of FBS at 37 °C with 5% humidified CO₂. After 24 hours, the flasks were washed with PBS and fresh CCM was added to each flask. Half the volume of the culture medium was replaced every 2-3 days with fresh CCM. At 70% confluence, the cells were detached by trypsinization with 0.25% trypsin/1 mM EDTA (Gibco), neutralized with CCM, centrifuged, and reseeded onto culture flasks at a density of 100–200 cells per cm². This process was repeated for subsequent passages, after which cells were injected. After counting, the cellular suspension was diluted in 50 µL of PBS to different concentrations (2 × 10⁵ or 10⁵ cells) distributed between the two heads of the right gastrocnemius of 20- to 22-week-old isoflurane anesthetized mice. Each group of doses was composed of six mice, and a control group was injected with PBS.

Muscle histology

At endpoint, the gastrocnemius and quadriceps muscles were harvested, weighed, and frozen in cooled isopentane. Immunofluorescence was performed on 7 µm muscle sections after blocking with PBS containing 1% bovine serum albumin and 10% neonatal goat serum. Sections were labelled with primary antibodies directed against laminin (Dako) and then incubated with appropriate secondary antibodies (Invitrogen). Slides were mounted with DAPI Fluoromount-G (Southern Biotech). Images were taken with a Thunder Imaging Tissue 3D wide field microscope (Leica Microsystems), and the immunoreactive area of two sections per muscle was quantified using Fiji software (ImageJ, NIH).

Flow cytometry

Cells were washed and resuspended in FACS buffer consisting of 1× PBS with 0.5% BSA (Sigma-Aldrich) and 5 mM EDTA. They were then incubated for 15 minutes with a rat anti-mouse Fc block (BD Bioscience), followed by 30 minutes with the antibody mix according to the panel (Table 1) combined with a viability marker (Fixable blue, Thermo). After incubation, the cells were fixed using Fix/lyse solution (BD). All analyses were performed on a BD Fortessa^TM cytometer (Becton Dickinson) and post-analyses using FlowJo software (FlowJo).

Table 1.

Flow cytometry antibody panel.

Panel	Antibody	Laboratory	References	Clone
SVF/ADSC	CD45 BV510	SONY	1115685	30-F11
	CD34 APC	SONY	1243060	HM34
	CD31 PE	SONY	1112040	390
	CD146 A488	SONY	1273540	ME-9F1
	CD73 BV605	SONY	1236075	TY/11.8
	CD90.1 BV711	SONY	1612695	OX-7
	CD105 PERCPC5.5	SONY	1202080	MJ7/18
Leukocytes	CD45 BV650	SONY	1115755	30-F11
	CD4 APC EFLUOR 780	EBIOSCIENCE	47-0042-82	RM4-5
	CD8 PRC5.5	SONY	1103670	53-6.7
	CD19 BV605	SONY	1177700	6D5
	CD62L FITC	SONY	1122030	MEL-14
	CD44 PC7	SONY	1115150	IM7
	CD11B A700	SONY	1106110	M1/70
	CD11C APC	SONY	1186550	N418
	LY6C BV421	SONY	1240160	HK1.4

Abbreviations: ADSC = adipose-derived stem cells; SVF = stromal vascular fraction.

Statistical analysis

All experiments were repeated at least twice, and the results are expressed as the mean ± standard deviation. Mann–Whitney test or Wilcoxon tests were used for intergroup comparisons depending on whether data are paired or not, Kruskal–Wallis test with Dunn post hoc test was used for multigroup column comparisons, two-way ANOVA with Holm-Sidak post hoc test was used for multigroup curve comparisons, and Spearman test was used for correlation, with differences considered statistically significant at P < 0.05 (*P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001). The statistical analyses and generation of images were performed with GraphPad Prism 8.0 (GraphPad).

The work has been reported in line with the ARRIVE guidelines 2.0. No human cells, tissues, samples, or cell lines were used in this research.

Results

Stromal vascular fraction phenotype from different mouse strains

The SVF cell populations extracted from the inguinal and perianal fat pads of the mice were classified using the gating strategy shown in Figure 1A. The cell composition revealed a significantly lower proportion of leukocytes (P = 0.0136) in CD1 mice than in Icos^−/− NOD mice of the same age (8-10 weeks old) (Figure 1B). This difference primarily affected the CD45⁺CD34⁺ cells (P = 0.0573). However, among the stromal cell populations, CD1 mice showed a significantly higher proportion of MSC (P = 0.0123) and nearly significative for EPC (P = 0.0683), but not pericytes. Some age- or disease-related differences were also observed in Icos^−/− NOD mice. For instance, 22-week-old mice had a significantly lower proportion of CD45⁺CD34⁺ cells than 8-week-old mice (P = 0.0077), despite the absence of a significant difference in leukocyte count (Figure 1B). No significant variation was observed in the other populations.

Figure 1.

Stromal vascular fraction phenotype and comparison between strains: (A) Gating strategy and (B) graphical representations of comparison of main cell population proportion for SVF harvested and extracted from CD1 (8 weeks old, N = 5) or Icos^−/− NOD mice (8 or 22 weeks old, N = 5) and analyzed by flow cytometry. (EPC = endothelial progenitor cell; MSC = mesenchymal stem cell. (*p<0,05; **p<0,01).

SVF collected from 20 CD1 mice was diluted for intramuscular injection. A portion of SVF was cultured as previously described. After 6 weeks, the cultured ADSC displayed a very low signal for negative markers such as CD45, CD34, and CD31 and a high signal for positive markers such as CD73, CD90, and CD105, which are characteristic of the ADSC phenotype (Figure S1, see online supplementary material for a color version of this figure).

Adipose-derived cell therapy injections stabilize myositis progression in Icos  ^−/− NOD mice

The administration of SVF and ADSC from CD1 mice in a preventive manner to Icos^−/− NOD mice demonstrated the ability to control the progression of myositis over time through both the clinical score and muscle strength parameters. A significant difference in the clinical score was observed from 15 weeks post-injection of a high dose of SVF (10⁶ cells) compared to untreated mice (P = 0.0094) (Figure 2A). Mice treated with ADSC exhibited a dose-dependent effect as early as 7 weeks post-injection for both doses when compared with untreated mice (P < 0.0345 and 0.0137 for low and high doses, respectively) (Figure 2B). However, the low dose slowed the progression of the disease, whereas the high dose appeared to block it. This stabilization of the disease was further supported by muscle strength measurements, which revealed a significant difference at 17 and 6 weeks after injection in mice receiving the highest dose of SVF and ADSC, respectively (P < 0.0425 and 0.0399, respectively) (Figure 2C and D). It is noteworthy that the onset of therapeutic effects for both clinical scores and muscle strength required several weeks for both treatments. However, the results indicate that ADSC exhibited a more rapid response than SVF.

Figure 2.

Clinical follow-up of Icos^−/− NOD mice after treatment with adipose-derived cell therapies: Graphical representation of (A, B) clinical score, followed weekly, (C, D) grip strength, followed at 3-week intervals, from mice treated with stromal vascular fraction (SVF) (on the left, N = 8 in each group) or adipose-derived stem cells (ADSC) (on the right, N = 6 in each group), respectively. (*p<0,05, **p<0,01, ***p<0,001, ****p<0,00001)

Fourteen weeks following ADCT administration, catwalk analysis revealed an enhancement in locomotor function. When compared with untreated mice, treated mice exhibited a significant or nearly higher average speed with both the highest dose of SVF (P = 0.0678) (Figure 3A) and ADSC (P = 0.0405) (Figure 3C). Similarly, the step cadence of the mice that were treated with the highest dose showed a nearly significant increase for SVF (P = 0.0678) (Figure 3C) and a significant increase with ADSC (P = 0.0060) (Figure 3D) when compared to the control group. In addition to the difference in speed, a variation in the step sequence pattern was observed, with a lower proportion of rotate and cruciate sequences compared with alternate sequences in mice administered high and low doses of each therapy. Additionally, there was a statistically significant decrease in the number of rotate sequences in mice treated with ADSC (P = 0.0091 for the high dose) (Figure 3E-G). The regularity index of these step sequences was also significantly higher in mice treated with the highest dose of ADSC (P = 0.0381), but not in those treated with SVF (Figure 3F-H). The change in gait was attributable to improved balance between the legs, as evidenced by the distribution of support during walking. Both cell therapies resulted in a reduction in the support of the three paws in favor of two diagonal paws in treated mice. The differences were significant in mice treated with the highest dose of ADSC (P = 0.0108 for three paw support and P = 0.0132 for diagonal paw support) and nearly significant for three paw support in mice treated with the highest dose of SVF (P = 0.0658) (Figure 3I and J). A trend toward increased support on a single paw was also observed, although the difference was not statistically significant.

Figure 3.

Catwalk analysis of body motor function of Icos^−/− NOD mice after treatment with adipose-derived cell therapies: Graphical representation of (A, C) average speed, (B, D) cadence, (E, G) repartition of step sequence (with C, L, and H for control, low, and high dose groups, respectively), (F, H) step sequence regularity index, and (I, J) repartition of support evaluated 14 weeks after injection for mice treated with stromal vascular fraction (SVF) (on the left, N = 5 in each group) or adipose-derived stem cells (ADSC) (on the right, N = 4, 5, and 6 for control, low, and high doses, respectively); (*p<0,05, **p<0,01).

To gain a deeper understanding of the disparities in the distribution of step sequences and types of support, we endeavored to investigate the progression of motor capabilities on a step-by-step basis. Our observations revealed a reduction in the area of hind paw prints in the treated mice compared to the control group, which was noteworthy for both sides in the case of ADSC-treated mice (P= 0.0703 and 0.0117 for the right hind paw of high- and low-dose treated mice, respectively, and P = 0.0283 and 0.0469 for the left hind paw of high- and low-dose treated mice, respectively) (Figure 4B). Only the injected side (right) exhibited a significant reduction in the hind paw print area with the highest dose of SVF (P = 0.0431) (Figure 4A), indicating a more pronounced systemic effect with the use of ADSC. Additionally, both cell therapies appeared to increase the print area of the front legs, although the trend was not statistically significant. These results indicate a more equitable distribution of body weight, moving from the back to the front of the body. This shift is thought to be responsible for the observed changes in gait and the increased speed of movement.

Figure 4.

Catwalk analysis of paw motor function of Icos^−/− NOD mice after treatment with adipose-derived cell therapies: Graphical representation of (A, B) print area, (C, D) step cycle, and (E, F) stride length of the different paws evaluated 14 weeks after injection for mice treated with SVF (on the left, N = 5 in each group) or ADSC (on the right, N = 4, 5, and 6 for control, low, and high doses, respectively). RF = right front; LF = left front; RH = right hind; LH = left hind, (*p<0,05).

According to the increase in cadence, the duration of the step cycle for the four paws of mice treated with high-dose ADSC was significantly reduced compared to control mice (P = 0.0553 for the right front paw, P = 0.0256 for the left front paw, P = 0.0098 for the right hind paw, and P = 0.0122 for the left hind paw) (Figure 4D). Additionally, the right (injected side) hind paw of mice treated with the lowest dose of ADSC and with the highest dose of SVF showed a significant or nearly decrease in step cycle duration (P = 0.0746 and 0.0252, respectively), although the other legs showed a trend (Figure 4C). Furthermore, stride length significantly increased or nearly so for the four paws in mice treated with high-dose ADSC compared to the control (P = 0.0445 for the right front paw, P = 0.0746 for the left front paw, P = 0.0221 for the right hind paw, and P = 0.0182 for the left hind paw) (Figure 4F). Similar to the previous results, only the right (injected side) hind paw of mice treated with the highest dose of SVF displayed a significant increase in stride length compared to the control (P = 0.0431) (Figure 4E). Collectively, these results suggest that the acceleration of running speed is attributable to a synchronized lengthening of stride and increase in cadence. These two factors are a result of an increase in muscular strength, which facilitates the preservation of the body’s equilibrium and the appropriate distribution of weight between the legs.

Adipose-derived cell therapy injections limit muscular atrophy

To evaluate the preservation of muscle strength in treated mice, we weighed the gastrocnemius and quadriceps muscles before freezing them for histological analysis. The results showed that the gastrocnemius muscles of mice treated with SVF were significantly heavier than those of untreated mice (P = 0.048 and P = 0.0328 for the right and left gastrocnemius, respectively), as were the gastrocnemius muscles of mice treated with ADSC (P = 0.0317 and P = 0.0476 for the right and left gastrocnemius, respectively) (Figure 5A and B). Additionally, there was a trend toward increase in the weight of the right quadriceps of mice treated with ADSC but not significant (P = 0.0556). Interestingly, the right gastrocnemius muscle was significantly heavier than the left gastrocnemius muscle in mice treated with SVF (P = 0.0156), indicating a strong local effect of SVF. This result was not observed in mice treated with ADSC. To confirm these findings, we compared the average surface area of various muscle sections between the treated and untreated mice. Our analysis revealed a reduction in muscle atrophy, with a statistically significant or nearly significant increase in the average surface area of the gastrocnemius muscle in SVF-treated mice (P = 0.0357 and P = 0.0571 for the right and left gastrocnemius, respectively) and ADSC-treated mice (P = 0.0238 and P = 0.0635 for the right and left gastrocnemius, respectively) compared to untreated mice (Figure 5C and D).

Figure 5.

Atrophy comparison between adipose-derived cell therapies treated and untreated mice: Graphical representation of (A, B) muscle weight measured after their extraction and (C, D) muscle area of mice treated by SVF (on the left, N = 7 in each group) or ADSC (on the right, N = 5 in each group), respectively, compared to control. Correlation between muscle area and (E) clinical score, (F) grip strength, (G) average speed, or (H) cadence tests, (*p<0,05).

To assess the effect of atrophy reduction on the various clinical and functional signs observed in this model, we determined the correlation between muscle weight/muscle area and clinical score, grip strength, average speed, and cadence observed with the catwalk. Our results indicate a significant negative correlation between muscle weight/muscle area and the clinical score (r = −0.7156, P = 0.0002 [data not shown], and r = −0.7414, P = 0.0004 [Figure 5E], for muscle weight/muscle area, respectively), as well as a strong positive correlation between muscle weight/muscle area and grip strength (r = 0.7597, P < 0.0001 [data not shown], and r = 0.7730, P = 0.0002 [Figure 5F], for muscle weight/muscle area, respectively), average speed (r = 0.6688, P = 0.0009 [data not shown], and r = 0.7296, P = 0.0006 [Figure 5G], for muscle weight/muscle area, respectively), and cadence observed with the catwalk (r = 0.6429, P = 0.0017 [data not shown], and r = 0.7482, P = 0.0004 [Figure 5H], for muscle weight/muscle area, respectively). These results emphasize the importance of reducing muscle atrophy from a clinical standpoint. This aligns with the regenerative properties of ADCT, although it does not establish their potential immunomodulatory effects.

Adipose-derived cell therapies reduce systemic inflammation

To investigate the immunomodulatory effects of ADCT, we assessed their influence on systemic inflammation. Analysis of the phenotype of the primary leukocyte populations in the lymph nodes (Figure 6A) revealed substantial differences between treated and untreated mice. Initially, we identified differences in the proportions of the cell populations. Specifically, there was a statistically significant decrease in CD11b⁺ monocytes in mice treated with SVF and with ADSC (P = 0.0605 and P = 0.0247 for high and low doses, respectively, in SVF-treated mice, and P = 0.0392 for high and low doses in ADSC-treated mice) (Figure 6B and C). In contrast, the T cell populations were more prominent, with a significantly higher proportion of CD4⁺ T cells (P = 0.0258 and P = 0.0652 for high and low doses in SVF-treated mice and P = 0.0622 and P = 0.0196 for high and low doses in ADSC-treated mice, respectively) compared to CD8⁺ T cells (P = 0.0196 for high-dose SVF-treated mice and P = 0.0587 and P = 0.0454 for high and low doses in ADSC-treated mice, respectively). Variations in other cell populations were observed, particularly in B cells and neutrophils (data not shown), which showed a tendency to decrease, although the significance remained unclear.

Figure 6.

Analysis of the effect of adipose-derived cell therapies on immune cell populations in vivo: (A) gating strategy and (B, C) graphical representation of immunologic effects of cell therapies on leukocytes harvested from lymph node extracted from mice treated with SVF (upper panels, N = 4, 5, and 6 for control, low, and high doses, respectively) or ADSC (lower panels, N = 5 in each group), respectively, compared to control.(*p<0,05, **p<0,01)

In addition, our observations revealed a notable decrease in the proportions of proinflammatory macrophages and/or effector T cells. Specifically, we observed a significant reduction in Ly6C^Hi monocytes in mice treated with high dose of SVF (P = 0.0093) and ADSC (P = 0.0710 and P = 0.0078 at high and low doses, respectively). The distribution of naive T cells (T_N = CD62L^hi CD44^Lo), central memory T cells (T_CM = CD62L^Hi CD44^Hi), and memory effector T cells (T_EM = CD62L^Lo CD44^Hi) was also altered. Our results show a decrease in the activation level of these cellular populations in treated mice, with a significant reduction in T_EM (P = 0.0018 for high dose for CD4⁺ and CD8⁺ in the SVF treatment protocol, P = 0.0116 and P = 0.0678 for CD4⁺ and P = 0.0392 and P = 0.0218 for CD8⁺ for high and low doses, respectively, in ADSC-treated mice) and an increase in T_CM (P = 0.0025 for high dose compared to control for CD4⁺ in SVF-treated mice, P = 0.0678 and P = 0.0116 for CD4⁺ for high and low doses, respectively, in ADSC-treated mice and P = 0.0618 and P = 0.0429 for CD8⁺ in mice treated with high and low doses of ADSC, respectively). T_N also appeared to increase, but not significantly except for CD8⁺ in SVF-treated mice (P = 0.0452 and P = 0.0178 for high and low doses, respectively). This reduction in the level of activation of both T cells and monocytes suggests a systemic reduction in inflammation, indicating an immunomodulatory effect of ADCT.

Discussion

Previous preclinical myositis studies have frequently been subject to limitations owing to the design of the models used. Often, these models are based on immunization against muscle proteins, which can introduce bias into the results.¹⁹ In some cases, immunization is directed against an antigen that is specific to or associated with myositis in humans, allowing for the creation of a humanized model. However, these models do not always accurately replicate all clinicopathological features of human myositis.^24–26 The use of Icos^−/− NOD mice allowed us to avoid the bias due to the induction of the immune response and to observe the effects of ADCT in a histologically and clinically discernible disease. This condition is defined by a progressive and irreversible decline in muscle strength, attributed to chronic inflammation with leukocytic infiltrates, particularly Th1-infiltrating cells, and accompanied by evidence of muscle necrosis and regeneration.^23,27 Recent studies have demonstrated a similarity in pathophysiology between this model and human disease, involving pathogenic autoantibodies,²² IFNγ, and oxidative stress.²⁸ However, this model lacks features specific to certain subtypes of myositis, such as human-specific autoantibodies, amyloid deposits (seen in IBM), or extramuscular manifestations (as seen in dermatomyositis). In this study, mice were treated at 20 weeks of age, which corresponds to age of first disease onset in this model. The choice of this age allowed the observation of the earliest clinical signs that could be observed in patients while taking into account the time required to establish clinical effects.

In this preliminary investigation, we did not use cells from diseased mice to avoid any potential confounding factors that might influence adipose tissue and treatment efficacy. Although various studies have been conducted on this topic,^29–31 the effects of age, sex, and overall health on the effectiveness of ADCT remain largely unexplored.³² We have evidenced differences in SVF composition, particularly for leukocytes populations, between 8-week-old and 22-week-old Icos^−/− NOD mice. However, age and disease evolution are strictly associated in these mice, and therefore, we cannot draw definitive conclusions about the specific influence of one versus the other on the cell phenotype. Due to the low availability of mice, we could not use healthy Icos^−/− NOD mice as donors for ADCT therapy. To remain close to the autologous use of these therapies, specifically for SVF, we chose CD1 mice as a source of adipose tissue for cellular therapies. CD1 mice exhibit genetic characteristics similar to those of NOD mice, from which Icos^−/− NOD mice were derived. Both strains have traced their ancestry back to Swiss mice.³³ Nevertheless, to further support the findings presented here and to assess the influence of the disease on the efficacy of ADCT, it would be interesting to compare the effects of cells of both young and healthy Icos^−/− NOD mice and aged mice exhibiting myositis symptoms. This would be of interest given the potential use of autologous cells from patients in this type of therapy.

The results of this study are promising. Preclinical evaluation, including muscle strength evaluation, is often a limitation when studying new treatments for myositis. It mainly consists of establishing a clinical score specific to the model and conducting physical tests to evaluate specific criteria (grip strength, balance, and fatigue during walking).³⁴ These tests have numerous limitations, such as the animal’s willingness to perform them, and the results may vary significantly between animals and operators.³⁵ While these methods offer the advantage of regular monitoring of muscle capacity in mice, they require supplementation with a more intricate approach. In this study, we chose clinical score and grip test for longitudinal follow-up, allowing us to highlight the exacerbation of neuromuscular disorders by evaluating grip loss and the catwalk for final evaluation, providing a more comprehensive and fine assessment of mice motor capabilities. Initially used in behavioral studies,³⁶ this device has since been employed to examine a range of neurological or musculoskeletal disorders.³⁷ In a recent study, it allowed to evidence the preventive effect of an antioxidant treatment in the same myositis model.²⁸ Specifically, it has provided evidence for alterations in gait over time in Icos^−/− NOD mice, including a change in weight distribution from the rear to the front of the body, which is indicated by differences in paw print surface and body balance, with reduced support from the three legs. These changes lead to an increase in stride length while concurrently reducing the duration of a step cycle, resulting in an increase in the average speed. It is noteworthy that while there is an overall impact on speed and gait, the effect on the limb that received ADCT injections appears to be greater, suggesting both a local and distant effect of these cell therapies.

Early studies on Icos^−/− NOD model demonstrated, in addition to muscle weakness, a global weight loss in mice.²² The parallelism of muscle weakness and weight loss suggests that it may be muscle mass loss, which has long been observed in patients with myositis.³⁸ In this study, we have shown that both cell therapies help to maintain muscle mass and surface. While muscle weakness is not always indicative of disease progression, some studies have found a relationship between muscle area measured by magnetic resonance imaging and DAS score.³⁹ Additionally, muscle strength, clinical scores (MMT8, CMAS),^38,39 and quality of life scores (CHAQ)³⁹ are strongly correlated with muscle weight or surface area. These correlations were also observed in the present study. Maintenance of these parameters and muscle mass in treated mice is indicative of therapeutic efficacy at the tissue level. The decline in skeletal muscle mass, which is linked to the pathophysiological processes of myositis leading to autoimmune muscle atrophy, is also a detrimental outcome of long-term corticotherapy.⁴⁰ The use of cellular therapies derived from adipose tissue in conjunction with corticosteroids may help to minimize the adverse effects of the latter and reduce their use owing to the therapeutic benefits of these cellular preparations.

Systemic effects of ADCT can also be observed at lymph nodes level. Muscle atrophy reduction can be due to both regenerative and immunomodulatory properties of cell therapies. Here, the phenotypic changes observed in the leukocyte populations in the nodes confirm their immunomodulatory effects, since we observed significant phenotypic differences in two leukocyte populations between treated and untreated mice. First, CD11b⁺ monocytes showed a diminished expression of Ly6C. This glycoprotein, found at the surface of monocytes, has been identified as a marker of inflammation and has been shown to aggravate certain autoimmune diseases, such as rheumatoid arthritis or systemic lupus erythematosus.⁴¹ Although its exact role in human physiology remains unclear, its involvement in the inflammatory phase in various myositis virus-induced models has been demonstrated in several studies.^42,43 Second, following ADCT administration, we observed an increase in both CD4⁺ and CD8⁺ lymphocyte populations, with a reduction in T_EM and an increase in T_CM for both. This change is characterized by the maintenance of CD62L expression, a selectin primarily involved in the homing of lymphocytes to the lymph node and the restriction of cytotoxic T cell properties.⁴⁴ CD62L expression is known to vary in response to TCR stimulation. A short contact leads to the maintenance of CD62L expression and the generation of T_CM cells, as seen in acute infections, while prolonged exposure induces a loss of CD62L and the generation of T_EM cells, as seen in chronic infections or autoimmune diseases.^45–47 The observed alterations after ADCT administration reflected a reduction in inflammation and antigenic exposure of autoreactive T lymphocytes, resulting in the maintenance of CD62L expression and homing to the lymph node.

In this study, both SVF and ADSC demonstrated promising therapeutic benefits, with ADSC exhibiting a more significant effect. The muscular or immunological origin of myositis has not yet been demonstrated.⁴⁸ In this study, we were able to observe a response without disease progression but also without improvement in muscle strength, associated with a systemic immunomodulatory effect. This lack of regeneration associated with immunomodulation suggests a role for this therapy on immunity. Nevertheless, further investigations are needed to deepen our understanding of this issue. While not prevalent, the comparison between SVF and ADSC in preclinical or clinical research settings is not entirely absent from the scientific literature.⁴⁹ The primary aim of our study was not to conduct a comparative analysis of these therapeutic approaches, but rather to lay the groundwork and demonstrate the feasibility of their utilization in myositis treatment. In the literature, the efficacy of SVF or ADSC therapy depended on the therapeutic indication being addressed, with one offering greater benefit without a comprehensive understanding of the underlying mechanism.^50–53 In contrast to ADSC, SVF comprises not only mesenchymal stem cells but also endothelial progenitors and leukocytes. While the myoregenerative effect is predominantly attributed to mesenchymal stem cells, the immunomodulatory effect of SVF probably stems from all these cell populations.¹³ The encouraging outcomes observed in this study confirm the feasibility of intramuscular administration to use these therapies in myositis treatment. This approach not only demonstrates effectiveness in addressing localized muscle lesions associated with myositis but also presents already known benefits, including enhanced systemic effects and dwell time of the cells of interest.⁵⁴ The findings of this study lead us to initiate a clinical trial in IBM to assess the therapeutic potential of the SVF (NCT05032131).

Conclusion

The therapies evaluated in this study have exhibited promising outcomes in the treatment of a preclinical model of myositis. These therapies effectively reduce the severity of the disease and preserve muscle strength and motor function, notably by limiting muscle atrophy. In this study, therapies were administered as pooled cell samples from multiple mice, mitigating the cell heterogeneity commonly reported in the literature for SVF. Nevertheless, studying this heterogeneity through functional and clinical trials could facilitate the identification of valuable efficacy and quality markers for clinical application. However, these two treatments have the potential to become novel therapies for myositis, leading us and other team to start clinical trials in IBM.

Author contributions

Baptiste Pileyre (Conceptualization, Formal analysis, Investigation, Methodology, Resources, Visualization, Writing—original draft, Writing—review & editing), Silvia Gandolfi (Data curation, Formal analysis, Investigation, Methodology, Resources, Writing—original draft, Writing—review & editing), Catalina Abad (Formal analysis, Investigation, Methodology, Resources, Writing—review & editing), Thara Jaworski (Data curation, Formal analysis, Investigation, Writing—review & editing), Laurent Drouot (Conceptualization, Investigation, Methodology, Writing—review & editing), Laetitia Jean (Investigation, Resources, Writing—review & editing), Olivier Boyer (Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration , Writing—review & editing), Isabelle Dubus (Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Validation, Writing—original draft, Writing—review & editing), and Jeremie Martinet (Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—­original draft , Writing—review & editing)

Supplementary material

Supplementary material is available at Stem Cells Translational Medicine online.

Funding

This work was supported by the “Fondation pour la Recherche Médicale” (grant #FDM201806006523).

Conflict of interest

O.B. has received financial support for research from Argenx, Bristol Myers Squibb, UCB, and CSL Behring. The other authors did not disclose any competitive interest relative to this work.

Data availability

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available because of privacy or ethical restrictions.

Ethical considerations

For this study titled: « Etude préclinique de la greffe intramusculaire de cellules issues de la fraction vasculaire stromale dans le traitement des myosites; caractérisation des mécanismes immunomodulateurs et myorégénératifs impliqués », animal experiments were approved on October 2019 by a French institutional ethics committee (#2019-083016072100).

The authors declare that they have not use AI-generated work in this manuscript.