Dystroglycan N-terminal domain enables LARGE1 to extend matriglycan on 2 α -dystroglycan and prevents muscular dystrophy

23 Dystroglycan (DG) requires extensive post-translational processing to function as a 24 receptor for extracellular matrix proteins containing laminin-G-like (LG) domains. Matriglycan 25 is an elongated polysaccharide of alternating xylose and glucuronic acid that is uniquely 26 synthesized on α -dystroglycan ( α -DG) by like-acetylglucosaminyltransferase-1 (LARGE1) and 27 binds with high affinity to matrix proteins like laminin. Defects in the post-translational 28 processing of α -DG that result in a shorter form of matriglycan reduce the size of α -DG and 29 decrease laminin binding, leading to various forms of muscular dystrophy. However, little is 30 known regarding mechanisms that generate full-length matriglycan on α -DG (~150-250 kDa). 31 Here, we show that LARGE1 can only synthesize a short, non-elongated form of matriglycan in 32 mouse skeletal muscle that lacks the DG N-terminus ( α -DGN), resulting in a ~100-125 kDa α - 33 DG. This smaller form of α -DG binds laminin and maintains specific force but does not prevent 34 muscle pathophysiology, including reduced force induced by eccentric contractions and 35 abnormalities in neuromuscular junctions. Collectively, our study demonstrates that α -DGN is 36 required for LARGE1 to extend matriglycan to its full mature length on α -DG and thus prevent 37 muscle pathophysiology. 38 lengthening

contraction. Our studies show that DG lacking the α -DGN expresses a short form of matriglycan; this 2 5 1 suggests that α -DGN is necessary for the production of full-length matriglycan. To test this 2 5 2 hypothesis, we determined if matriglycan expression could be restored in mice lacking α -DGN.

5 3
We injected M-α-DGN KO mice with an AAV expressing α -DGN (AAV-CMV α -DGN) and 2 5 4 harvested the skeletal muscles of these mice eight to ten weeks after injection. H&E staining in analysis of these mice showed that matriglycan had a molecular weight of ~100-125 kDa and the 2 5 9 size of α -DG was shifted down, whereas β -DG remained unchanged ( Figure 6B). Laminin- binding was observed at ~100-125 kDa in M-α-DGN KO skeletal muscle infected with AAV- CMVα-DGN ( Figure 6B). Collectively, this phenotype is similar to that observed in the skeletal force or improve force deficits induced by lengthening contractions (Figure 6C, 6D) To determine if excess LARGE1 produces full-size matriglycan in M-α-DGN KO MCK-Large1 demonstrated no change in the molecular weight of matriglycan, α -DG, and β -DG Force deficit and force recovery after lengthening contractions in EDL muscles from 12-to 17- The length of matriglycan is correlated with its ability to bind ECM ligands (Goddeeris et   2  9  4 al., 2013). Therefore, we hypothesized that the susceptibility to force decline by lengthening 2 9 5 contractions would differ depending on the length of matriglycan. To test this, we performed 2 9 6 physiological muscle tests in three different mouse models to determine the difference in 2 9 7 susceptibility to lengthening contraction-induced reduction in force. Specifically, we used: 1) M-2 9 8 α -DGN KO mice, which express a short form of matriglycan, 2) Dag1 T190M mice, which harbor a 2 9 9 knock-in mutation (T190M) in DAG1 that inhibits the DG-LARGE1 interaction and leads to (41.6 ± 6.7) mice, with no difference observed between the latter groups. Immunoblot analysis of from Dag1 T190M and M-α-DGN KO mice was higher than that of myd muscle (Figure 7E). M-α-3 1 7 DGN KO and Dag1 T190M also displayed an increase in dissociation constant ( Figure 7E).

1 8
Collectively, these results suggest that α -DG positive matriglycan of at least ~150 kDa is 3 1 9 sufficient to prevent force decline from lengthening contractions and significant dystrophic 3 2 0 changes, despite a 45% reduction in laminin-binding activity.  laminin-binding level is only 9% relative to that of C57 mice and leads to a marked increase in that the production of ~120-150 kDa α -DG significantly prevents dystrophic change, suggesting Collectively our study demonstrates that α -DG with α -DGN is required for the synthesis show that matriglycan length regulates the severity of muscular dystrophy and may serve as a were bred to female Dag1 flox/flox mice. A Cre PCR genotyping protocol was used to genotype the 4 2 3 Cre allele using standard Cre primers. The primers used were Sense: Male mice expressing the Pax7-Cre transgene were bred to female mice that were heterozygous Dag1 wt/Δα-DGN were bred to female mice homozygous for the floxed Dag1 allele (Dag1 flox/flox ).

3 0
Genotyping of Pax7 Cre Dag1 flox/Δα-DGN mice was performed by Transnetyx using real-time PCR. and experience with standard deviations of the given techniques.

3 6
Forelimb grip strength test 4 3 7 Forelimb grip strength was measured at three months using previously published methods (de software. Student's t-test was used (two-sided). Differences were considered significant at a p- present study are shown as the means + / -SD unless otherwise indicated. Corporation, Parsippany, NJ), and the tester was blinded to genotype. Statistics were calculated 4 5 3

Body weight measurements
using GraphPad Prism 8 software and Student's t-test was used (two-sided). Differences were 4 5 4 considered significant at a p-value less than 0.05. Graph images were also created using To compare the contractile properties of muscles, EDL muscles were surgically removed as of childhood onset limb-girdle muscular dystrophy without mental retardation.