Altered muscle niche contributes to myogenic deficit in the D2-mdx model of severe DMD

Lack of dystrophin expression is the underlying genetic basis for Duchenne muscular dystrophy (DMD). However, disease severity varies between patients, based on specific genetic modifiers. D2-mdx is a model for severe DMD that exhibits exacerbated muscle degeneration and failure to regenerate even in the juvenile stage of the disease. We show that poor regeneration of juvenile D2-mdx muscles is associated with an enhanced inflammatory response to muscle damage that fails to resolve efficiently and supports the excessive accumulation of fibroadipogenic progenitors (FAPs), leading to increased fibrosis. Unexpectedly, the extent of damage and degeneration in juvenile D2-mdx muscle is significantly reduced in adults, and is associated with the restoration of the inflammatory and FAP responses to muscle injury. These improvements enhance regenerative myogenesis in the adult D2-mdx muscle, reaching levels comparable to the milder B10-mdx model of DMD. Ex vivo co-culture of healthy satellite cells (SCs) with juvenile D2-mdx FAPs reduces their fusion efficacy. Wild-type juvenile D2 mice also manifest regenerative myogenic deficit and glucocorticoid treatment improves their muscle regeneration. Our findings indicate that aberrant stromal cell responses contribute to poor regenerative myogenesis and greater muscle degeneration in juvenile D2-mdx muscles and reversal of this reduces pathology in adult D2-mdx muscle, identifying these responses as a potential therapeutic target for the treatment of DMD.


151
SCs were plated at 2000 cells/cm 2 and EdU incorporation was performed after 48 h and 152 detected using Click-iT™ EdU Cell Proliferation Kit (Thermo Fisher, C10337). For differentiation 153 assay, SCs were plated at 10,000 cells/cm 2 and myogenin (Santa Cruz, SC-12732) staining was 154 used to evaluate differentiation. For fusion assay, SC were plated at 50,000 cells/cm 2 and 155 desmin (Abcam, ab32362) staining was performed to quantify fusion index. RAW data acquired 156 from different experiments was normalized to no FAP controls.

157
Statistics. GraphPad Prism 9.2.0 was used for all statistical analyses of data. Statistical 158 analysis was performed using the non-parametric Mann-Whitney test. Data normality was 159 assessed for all statistical comparisons. All p-values less than 0.05 were considered statistically 160 significant; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Data plots reported as scatter 161 plot with mean ± SD.

165
We have described the sudden onset of histological damage, and rapid disease 166 progression in muscles of juvenile (3-4 wk) D2-mdx 24 , and reported triceps as amongst the most 167 severely affected muscles in this model 24 . Consistent with other reports in adult D2-mdx 19-22 , we 168 observed extensive networks of endomysial and perimysial fibrosis in triceps of adult (>5 169 months-old) D2-mdx (Fig. 1A,B). The interstitial fibrosis nearly doubled between juvenile and 170 adult D2-mdx muscles, while only a modest increase was observed between juvenile and adult 171 B10-mdx muscles (Fig. 1A,B). Despite the large increase in fibrosis in adult D2-mdx triceps, 172 macroscopic examination revealed unexpected improvements in pathological features, 173 prompting a detailed histological examination. H&E staining showed reduced spontaneous 174 myofiber damage and infiltrating mononuclear cells in adult D2-mdx than in juvenile D2-mdx, to 175 levels comparable to the B10-mdx muscles (Fig. 1C,D, Supplemental Fig. 1). Alizarin red 8 staining identified a notable decrease in areas of myofiber damage and calcified replacement 177 from ~15% in juvenile D2-mdx to <5% in adult D2-mdx (Fig. 1E,F). Overall, our analysis 178 revealed that while there is progressive increase in endomysial fibrosis from juvenile to adult 179 D2-mdx, surprisingly the extent of damage in the adult D2-mdx is reduced as compared to the 180 juvenile D2-mdx to levels observed in either juvenile or adult B10-mdx.

182
Juvenile D2-mdx muscle exhibits a regenerative deficit that is reversed in adult muscle

183
When assessing the adult D2-mdx histopathology, we observed a notable increase in 184 the frequency of centrally nucleated fibers (CNFs) (Fig. 1C). Quantifying myofibers with internal 185 nuclei as a percentage of total myofibers per cross-section revealed nearly 3-times more CNFs 186 in adult D2-mdx as compared to juvenile D2-mdx ( Fig. 2A,B). Consequently, while juvenile D2-187 mdx have 5-fold fewer CNFs than juvenile B10-mdx, this difference is only 2-fold between the 188 adult D2-mdx and B10-mdx ( Fig. 2A,B). As juvenile D2-mdx muscles show minimal 189 regenerative ability 21,23,24 , we examined if regenerative capacity improved in adult D2-mdx 190 muscle, leading to the observed reduction in histopathology (Fig. 1). To investigate whether the 191 earlier myogenic deficit in D2-mdx is reversed in adulthood, we used notexin (NTX) to acutely 192 injure the tibialis anterior (TA) muscle of D2 wild-type (D2-WT) to avoid confounding effects of 193 chronic muscle injury in D2-mdx. To monitor myogenesis that follows this acute in vivo injury,

198
WT, mirroring results following spontaneous injury in D2-mdx ( Fig. 2A,B). Nearly 60% of all the 199 myofibers in adult D2-WT muscles were regenerated (CNFs), which was not different from the 200 level of CNFs in B10-WT adult muscle but roughly 3-times greater than in juvenile D2-WT ( Fig.   201 2D,E). These numbers mirrored the extent of BrdU-labeled myofibers over 2 dpi, again 202 9 revealing a similar trend -greater regeneration in adult D2-WT muscles comparable with adult 203 B10-WT. Meanwhile, juvenile D2-WT showed only minor (<5%) BrdU-labeling in randomly 204 dispersed, small-caliber myofibers that constituted large areas of unresolved inflammation even 205 after 6 dpi (Fig. 2D,F). Overall, we observed that compared to juvenile D2-WT, there is a greatly 206 improved regenerative response in adult D2-WT muscles (Fig. 2).

208
Stromal alterations mark the regenerative deficit of juvenile D2-mdx muscles 209 Skeletal muscle regeneration is a multicellular response where SCs interact with the 210 ECM, macrophages, and FAPs to regulate SC proliferation, differentiation, and fusion. To 211 assess the involvement of SC, macrophage, or FAP dysregulation in the regenerative deficit 212 observed in the juvenile D2-mdx muscles, we examined the expression of genes associated 213 with these different cell types in triceps (Fig. 3). Analysis of the activated SC marker -myoblast 214 determination protein 1 (MyoD), showed that the robust myogenesis observed in B10-mdx 215 muscle, was associated with higher levels of MyoD transcript, while the increased damage and 216 regeneration in juvenile (as compared to adult) mdx mouse muscle, was associated with higher 217 myogenin (MyoG) transcript in these muscles (Fig. 3A,B). To assess whether improvements in 218 regeneration in adult D2-mdx was a consequence of increased SCs, we monitored total levels 219 of Paired Box 7 (Pax7) transcript; however, we observed no consistent strain-or age-specific 220 difference for Pax7 transcript (Fig. 3C).

221
We thus turned to examine FAP/ECM related markers and their dynamics with age and 222 disease progression. TGF-b serves as a master modulator of ECM remodeling and composition 223 during muscle repair and we previously demonstrated its heightened activity in juvenile D2-mdx 224 at disease onset 24 . We observed higher TGF-b protein activity in the D2-mdx as compared to 225 B10-mdx, however, the TGF-b activity levels did not change between juvenile and adult D2-mdx 226 (Fig. 3D). Due to the extensive effects TGF-b exerts on the regulatory and structural

247
Thus, while we observed no consistent change in SC and ECM markers between the 248 juvenile and adult D2-mdx or between the juvenile and adult B10-mdx, we observe consistent 249 dysregulation of Spp1, and inflammatory markers corresponding to both pro-inflammatory and 250 pro-regenerative macrophages in juvenile D2-mdx muscle as compared to B10-mdx and adult 251 D2-mdx ( Fig. 3F-I). This implicates changes in the muscle inflammatory niche in the poor 252 11 myogenic response specific to juvenile D2-mdx muscles. Analysis of the local muscle niche 253 requires spatial exploration of the inflammatory response to monitor the histologically defined 254 damaged regions of the muscle.

256
Regenerative deficit of juvenile D2-mdx is linked to heightened pro-inflammatory response

257
The dynamic interplay between pro-inflammatory and pro-regenerative macrophages is 258 critical for timely resolution and repair of the muscle tissue. To examine the inflammatory 259 response to spontaneous injury of mdx muscle we used the pan macrophage marker, F4/80, in 260 conjunction with pro-inflammatory (iNOS) and pro-regenerative (CD206) macrophage markers,

261
to quantify the proportions of pro-inflammatory and pro-regenerative macrophages at and away 262 from the sites of muscle damage (Fig. 4). F4/80 immunostaining shows widespread 263 macrophage infiltration in juvenile D2-mdx muscle, which is decreased by more than half in 264 muscles from adult D2-mdx (Fig. 4A,B). Focusing exclusively on the damaged areas

269
As F4/80 does not distinguish between pro-inflammatory and pro-regenerative 270 macrophages, we next evaluated the contribution of these macrophage subtypes to the total 271 macrophage response observed in juvenile D2-mdx by co-labelling tissue sections for iNOS + , 272 F4/80 + pro-inflammatory, and CD206 + , F4/80 + pro-regenerative macrophages (Fig. 4A). The 273 sum of counts of each of these macrophage types in damaged areas corresponded to our 274 finding with (F4/80 + ) macrophage labeling in juvenile D2-mdx muscle, where we observed the 275 highest macrophage density per unit damaged area, which was reduced in adult D2-mdx 276 muscles (Fig. 4D). Monitoring individual macrophage population revealed that the damaged 277 areas of the juvenile D2-mdx muscles were enriched in iNOS + and CD206 + macrophages. In the 12 adult D2-mdx muscle, these pro-inflammatory macrophages in areas of damaged muscle had 279 returned to levels comparable to B10-mdx, while the level of pro-regenerative CD206 + 280 macrophages remained elevated (Fig. 4E,F). Examination of the relative proportion of pro-281 inflammatory to pro-regenerative macrophages (iNOS + /CD206 + macrophages), showed that, 282 inflammation in the juvenile D2-mdx muscles, relative to adult D2-mdx muscle, is skewed 283 towards the pro-inflammatory status (Fig. 4G). Together, these analyses indicate that juvenile 284 D2-mdx muscles are abnormally inundated with pro-inflammatory macrophages, which 285 correlates with poor myogenic capacity of these muscles.

286
To assess whether the heightened pro-inflammatory response in juvenile D2-mdx is on 287 account of increased entry or greater retention of inflammatory macrophages, we examined the 288 kinetics of the inflammatory response. As mdx muscle suffers from spontaneous injuries, and 289 regenerative myogenic deficit is also noted in D2-WT muscle (Fig. 2E,F), to achieve a controlled 290 injury scenario we performed acute focal NTX injury to TA muscles of D2-WT and age matched 291 B10-WT and then compared the resulting inflammatory and myogenic response (Fig. 5-6). As 292 injury-triggered muscle inflammation progresses from predominantly pro-inflammatory to 293 predominantly pro-regenerative over the week following injury, we monitored total (F4/80 + ) 294 macrophages, as well as levels of pro-inflammatory (iNOS + ) and pro-regenerative (CD206 + ) 295 macrophages at both an earlier (5 dpi) and later (8 dpi) time point. F4/80 staining at 5 dpi 296 indicated ~2-fold higher level for juvenile and adult D2-WT than B10-WT counterparts (Fig. 5A-

297
C). This indicated that the muscles of D2 mice are predisposed to a stronger inflammatory 298 response irrespective of age. Subsequent assessment of the status of the F4/80 response at 8 299 dpi showed that the inflammation was largely resolved in the juvenile B10-WT and fully resolved 300 in adult B10-WT and D2-WT muscles (Fig. 5A-B,D). In contrast, the extent of inflammation in 301 juvenile D2-WT was much higher than B10-WT, remaining comparable to the levels seen at 5 302 dpi ( Fig. 5A-C). Concomitant with the resolution of inflammation in juvenile B10-WT and adult 303 D2-WT muscles at 8 dpi, we observed regenerating myofibers in the site of injury, which were 304 13 lacking in the juvenile D2-WT muscle, mirroring our earlier observations (Fig. 2D). Next, we 305 examined the nature of the macrophages in the areas of inflammation in the acutely injured 306 muscles. Our assessments were limited to 5 dpi as inflammation had resolved by 8 dpi in all 307 cohorts except juvenile D2-WT. We found that both juvenile and adult D2-WT mice mounted a 308 strong inflammatory response that was comparably represented by pro-inflammatory and pro-309 regenerative macrophages (Fig. 5E,F).

310
As an independent measure for the formation of nascent myofibers we stained acutely 311 injured D2-WT and B10-WT muscles for embryonic myosin heavy chain (eMHC) and monitored 312 these 5dpi in the damaged sites (Fig. 6). This showed widespread eMHC expression in small-

313
caliber CNFs throughout the site of injury in all cohorts except juvenile D2-WT (Fig. 6A, B). The 314 number of eMHC + fibers in the adult cohort was no different from each other, and the density of 315 eMHC + fibers in adult D2-WT was comparable to juvenile and adult B10-WT muscles 5 dpi, but 316 eMHC + fibers were lacking in juvenile D2-WT muscles (Fig. 6A, B). Further, such fibers were 317 notably smaller (< 200 µm 2 ) and did not fuse together, even when present within the same 318 basement membrane (Fig. 6A, C). Together, these results indicate that D2-WT mice mount a 319 more robust inflammatory response as compared to B10-WT, which fails to resolve in a timely  receptor-α (PDGFRα). This revealed nearly 2-fold more FAPs in juvenile D2-mdx muscle, as 14 compared to the adult D2-mdx or the juvenile/adult B10-mdx muscle (Fig. 7A, B). The 331 observation that FAP abundance in adult D2-mdx declines to levels seen in B10-mdx muscle 332 suggests that the dysregulated FAP response in the juvenile D2-mdx muscles may contribute to 333 the myogenic deficit in these muscles.

334
To investigate whether juvenile D2-mdx FAPs impair SC function, we performed co-335 culture assays and compared the effect of FAPs from juvenile D2-mdx, adult D2-mdx, and from 336 acutely injured WT mice on the proliferation, differentiation, and fusion of WT SCs (Fig. 7C).

337
SCs were plated in the presence of FAPs isolated from either juvenile D2-mdx muscles or adult 338 D2-mdx muscles exhibiting spontaneous muscle injury, or from juvenile WT muscles that were 339 acutely injured by cardiotoxin (CTX) (Fig. 7C-G). Assessment of proliferation rate of WT SCs by 340 5'-ethynyl-2'-deoxyuridine (EdU) incorporation showed co-culturing with FAPs enhanced SC 341 proliferation, but no difference in proliferation was observed in co-cultures with the different 342 FAPs -CTX-injured WT, juvenile D2-mdx, adult D2-mdx (Fig. 7D). Next, to examine SC 343 differentiation we quantified the number of myogenin-expressing SCs and found no difference in 344 SC differentiation potential after 48 h when cultured without FAPs or co-cultured with the WT or 345 juvenile or adult D2-mdx FAPs (Fig. 7E). Finally, we examined fusion capacity of the SCs 346 cultured in the absence of FAPs or in the presence of WT versus D2-mdx FAPs harvested from 347 juvenile or adult muscles. This showed a reduction in the fusion index of SCs when co-cultured 348 for 48 h with juvenile D2-mdx FAPs, as compared to the no FAP control, CTX-injured WT FAPs, 349 or adult D2-mdx FAPs (Fig. 7F, G). This final observation recapitulates the above in vivo 350 observation that 5-dpi juvenile D2-WT muscle have the smallest (< 200 µm 2 ) nascent myofibers 351 that fail to fuse with the adjacent myofibers. Together, these results identify that poor 352 myogenesis in the juvenile D2 muscles is attributable to a muscle stromal cell niche that inhibits  To address whether the altered inflammatory and FAP response specific to juvenile D2-357 mdx muscles is directly responsible for impaired regeneration, we employed an anti-358 inflammatory glucocorticoid deflazacort treatment regimen in conjunction with acute focal NTX 359 injury to the TA muscles of D2-WT mice to assess potential influence on myogenesis in a 360 controlled injury scenario. Deflazacort (1 mg/kg) treatment was initiated within 24 h of an acute 361 NTX injury and administered daily for 7 d in conjunction with our 3 d (+1 d to +4 d) BrdU-labeling 362 protocol (Fig. 8A). Assessment of pro-inflammatory macrophage markers (Nos2, 363 showed that deflazacort treatment reduced expression of these markers (Fig. 8B-D), while pro-364 regenerative macrophage marker (Cd163) was significantly increased relative to controls ( Fig.   365   8E). This reflected a change in the macrophage polarization and was associated with reduction 366 in the markers of fibrotic FAPs (Fn1, Col1a1) in the deflazacort-treated injured muscles (Fig. 8F-

368
With improvements in macrophage polarization and fibrotic response of the FAPs, we 369 next assessed if these stromal changes caused by glucocorticoid treatment improved 370 regenerative capacity of the juvenile D2 muscles. Analysis of BrdU-labeled CNFs showed that, 371 deflazacort treatment enhanced the myogenic capacity of juvenile D2 muscles, leading to ~1.6-372 fold increase in the numbers of regenerated myofibers compared to the control (Fig. 8H-I). To 373 evaluate if this improvement in regenerative capacity was the result of reduced macrophage and

374
FAPs within the sites of damage and repair, we also quantified the and PDGFRα + area within 375 the damaged site occupied by macrophages (F4/80 + ) and FAPs (PDGFRα + ) and observed 376 these were no different between deflazacort-treated and control cohorts (Fig. 8J-K). These incorporation in adult D2-mdx muscles than in juvenile D2-mdx muscles (Fig. 2). Similar to 393 previous studies 22 , we find that the adult D2-mdx mice remain less myogenic than the adult 394 B10-mdx. However, the improved myogenic ability of the adult D2-mdx helps to explain the 395 previous report of amelioration of disease pathology with age in the D2-mdx model 20 . It also 396 explains our observation that the extent of muscle damage in adult D2-mdx mice is comparable 397 to the less severe B10-mdx model (Fig. 1).

398
Disturbances of asymmetric cell division and SC depletion in older individuals have been 399 described as intrinsic impairments in SC that compromise regeneration of the dystrophic 400 muscles 45 . However, we find that the improved myogenesis of the adult D2-mdx muscle occurs 401 despite no depletion of SCs in juvenile muscles (indicated by SC-specific markers, Pax7 and 402 MyoD), as compared to adult D2-mdx muscle. Concomitantly, expression of myogenin (indicator 403 of myogenic differentiation) is comparable between the mild (B10-mdx) and the severe (D2-404 mdx) models (Fig. 3). These findings agree with prior work showing comparable SC pool and 405 myogenic activity between dystrophic and WT muscles 46,47 . Based on in vivo SC transplant and 406 in vitro analysis of stromal interaction with SCs, it is clear that the muscle niche also plays an 407 important role in SC-mediated regenerative myogenesis 16,35,48,49 . In support of the role played by 17 SC extrinsic factors (muscle niche) in the regulation of myogenesis, we observed that higher 409 expression of ECM and inflammatory regulators including Spp1, Arg1, Tnf-a, Il-10 is robustly 410 aligned with the regenerative failure observed in the juvenile D2-mdx muscle (Fig. 3).

411
Analyses of muscle ECM and inflammatory regulators have established the importance 412 of these factors in regulating SC quiescence, activation and myogenic differentiation 32,50 . ECM 413 components and stromal cell response to injury has been observed to be altered in the D2-mdx 414 mice 19,24 . In agreement with these changes, we observed a distinct inflammatory response to 415 muscle injury in the D2 (WT and mdx) models, such that juvenile D2-mdx muscles exhibit a 416 stronger inflammatory response to injury (Fig. 4). In adult D2-mdx muscle, the inflammatory 417 response is restored to levels comparable to B10-mdx, implicating the excessive inflammatory 418 response in myogenic deficit seen in the juvenile D2-mdx mice. Analysis of timed muscle injury 419 in D2-WT mice showed that the excessive inflammatory response to muscle injury in the 420 juvenile D2-WT mice is caused by delayed clearance of inflammatory macrophages that 421 intravasate into the injured tissue (Fig. 5), which caused them to adopt an anti-myogenic state 422 hindering regeneration of these inflamed lesions (Fig. 6). Such aberrant clearance of 423 inflammatory cells is a hallmark of asynchronous regeneration and was previously implicated in 424 excessive fibrosis and failed regeneration in mdx and DMD patient muscles 8,34 . Concomitant 425 with the co-occurrence of altered ECM and inflammatory responses, we previously 426 demonstrated increased FAP accumulation in the damaged areas of D2-mdx muscles, where 427 aberrant FAP responses are inhibitory to regenerative myogenesis 24,51,52 . We found that the 428 aberrant stromal (ECM and inflammatory) response alters FAP activity in the juvenile D2-mdx 429 such that even in an ex vivo co-culture assay, these FAPs significantly suppressed SC-430 mediated myogenesis. Our analysis determined that it is not the proliferation or differentiation of 431 SCs, but the stage of SC fusion that is diminished selectively by the FAPs derived from the 432 juvenile D2-mdx but not from the injured WT muscles or the adult D2-mdx muscles (Fig. 7). In 433 support of this, we observed that treatment of injured muscles in juvenile D2 mice with 434 18 deflazacort inhibits the aberrant inflammatory and fibrotic response and improves the stromal 435 cell niche that is more supportive of regenerative myogenesis (Fig. 8).

436
These studies identify the aberrant muscle niche as the driver for myogenic deficit in the

577
Osteopontin ablation ameliorates muscular dystrophy by shifting macrophages to a pro-

636
Mice were euthanized and tissues harvested 6d post-injury. D. IF images and quantification of