POMK regulates dystroglycan function via LARGE1-mediated elongation of matriglycan

Matriglycan [-GlcA-β1,3-Xyl-α1,3-]n serves as a scaffold in many tissues for extracellular matrix proteins containing laminin-G domains including laminin, agrin, and perlecan. Like-acetyl-glucosaminyltransferase 1 (LARGE1) synthesizes and extends matriglycan on α-dystroglycan (α-DG) during skeletal muscle differentiation and regeneration; however, the mechanisms which regulate matriglycan elongation are unknown. Here, we show that Protein O-Mannose Kinase (POMK), which phosphorylates mannose of core M3 (GalNAc-β1,3-GlcNAc-β1,4-Man) preceding matriglycan synthesis, is required for LARGE1-mediated generation of full-length matriglycan on α-DG (~150 kDa). In the absence of Pomk gene expression in mouse skeletal muscle, LARGE1 synthesizes a very short matriglycan resulting in a ~ 90 kDa α-DG which binds laminin but cannot prevent eccentric contraction-induced force loss or muscle pathology. Solution NMR spectroscopy studies demonstrate that LARGE1 directly interacts with core M3 and binds preferentially to the phosphorylated form. Collectively, our study demonstrates that phosphorylation of core M3 by POMK enables LARGE1 to elongate matriglycan on α-DG, thereby preventing muscular dystrophy.


Introduction
The extracellular matrix (ECM) is essential for development, regeneration and physiological function in many tissues, and abnormalities in ECM structure can lead to disease (Rowe and Weiss, 2008;Hudson et al., 2003). The heteropolysaccharide [-GlcA-b1,3-Xyl-a1,3-] n (called matriglycan) is a Figure 1. Synthesis of the a-DG Laminin-Binding Modification and Enzymes Involved. Synthesis of the laminin-binding modification begins with the addition of the core M3 trisaccharide (GalNAc-b3-GlcNAc-b4-Man) on a-DG by the sequential actions of Protein O-Mannosyltransferase 1 and 2 (POMT1/2), Protein O-linked Mannose N-Acetyl-glucosaminyltransferase 2 (POMGNT2), and b1,3-N-Acetylgalactosaminyltransferase 2 (B3GALNT2), in the ER. POMK phosphorylates the C6 hydroxyl of mannose after synthesis of core M3. The phosphorylated core M3 is further elongated in the Golgi by Fukutin (FKTN), Fukutin related protein (FKRP), Transmembrane Protein 5 (TMEM5), b1,4-Glucuronyltransferase 1 (B4GAT1), and Like-acetylglucosaminyltranserase 1 (LARGE1). Isoprenoid synthase domain-containing (ISPD) produces cytidine diphosphate (CDP)-ribitol in the cytosol, and this serves as a sugar donor for the reactions catalyzed by FKTN and FKRP. LARGE1 synthesizes matriglycan, which directly interacts with the LG domains of matrix ligands. laminin-binding capacity; however, POMK activity is required for LARGE1 to generate full-length matriglycan on a-DG (~150 kDa). In the absence of the phosphorylated core M3, the non-extended matriglycan on~90 kDa a-DG binds laminin and maintains specific force but cannot prevent eccentric contraction-induced force loss or skeletal muscle pathology. Furthermore, solution NMR studies demonstrated that LARGE1 directly interacts with core M3, binding preferentially to the phosphorylated form. Therefore, our study shows that phosphorylation of core M3 by POMK enables LARGE1 to elongate matriglycan on a-DG. Collectively, our work demonstrates a requirement for POMK in the LARGE1-mediated synthesis of full-length matriglycan and proper skeletal muscle function.

Results
To determine if matriglycan can be expressed in the absence of POMK function, and therefore better understand the role of POMK in matriglycan synthesis, we studied skeletal muscle from a patient (NH13-284) with a homozygous POMK (D204N) mutation ( Figure 2A) and congenital muscular dystrophy (CMD) accompanied by structural brain malformations. D204 serves as the catalytic base in the phosphorylation reaction catalyzed by the kinase (Figure 2A; Figure 2-figure supplement 1), and its mutation is predicted to eliminate POMK activity (Figure 2-figure supplement 1; Zhu et al., 2016). POMK activity from skin fibroblasts and skeletal muscle of patient NH13-284 (POMK D204N) was undetectable when compared to control fibroblasts and muscle, respectively ( Figure 2B). Fibroblast LARGE1 activity and skeletal muscle B4GAT1 activity of patient NH13-284 were similar to those of a control (Figure 2-figure supplement 2A and B). Immunofluorescence analyses of POMK D204N muscle demonstrated partial immunoreactivity to IIH6 (anti-matriglycan), while the transmembrane subunit of DG, b-DG, was expressed normally in POMK D204N muscle ( Figure 2C). Flow cytometry using IIH6 also demonstrated partial immunoreactivity in POMK D204N fibroblasts ( Figure 2-figure supplement 2C). To test the effect of the POMK mutation on ligand binding, we performed a laminin overlay using laminin-111. Control human skeletal muscle showed the typical broad band of a-DG laminin binding centered at~150 kDa range; in contrast, laminin binding at~90 to 100 kDa range with reduced intensity was observed in POMK D204N skeletal muscle ( Figure 2D).
To understand the biochemical basis of the~90 to 100 kDa laminin binding in the absence of POMK activity, we targeted Pomk using LoxP sites and Cre driven by the muscle creatine kinase (Mck) promoter, or both the Mck promoter and the paired box 7 (Pax7) promoter (Figure 3-figure supplements 1 and 2; Brüning et al., 1998;Cohn et al., 2002;Han et al., 2009;Keller et al., 2004) to generate muscle-specific Pomk-null mouse models. Histologic analyses of Mck Cre ; Pax7 Cre ; Pomk LoxP/LoxP (M-POMK KO) quadriceps muscles revealed hallmarks of a mild muscular dystrophy ( Figure 3A). Quadriceps muscle extracts of Mck Cre ; Pomk LoxP/LoxP mice showed reduced POMK activity compared to Pomk LoxP/LoxP muscle but had similar levels of LARGE1 activity ( Figure 3B and C). M-POMK KO mice also showed reductions in 2-limb grip strength and body weight, and elevations in post-exercise creatine kinase (CK) levels compared to littermate control Pomk LoxP/LoxP mice ( Figure 3D; Figure 3-figure supplement 3). Immunofluorescence analysis of M-POMK KO muscle showed that b-DG is expressed at the skeletal muscle sarcolemma ( Figure 3A); however, like patient NH13-284 IIH6 immunoreactivity persisted in M-POMK KO muscle, but at a reduced intensity ( Figure 3A).
We next examined ex vivo force production in extensor digitorum muscles (EDL) muscles of 18-20-week-old Control and M-POMK KO mice. EDL muscle mass and cross-sectional area (CSA) were reduced in M-POMK KO mice compared to control mice ( Figure 4A and B). Additionally, M-POMK KO EDL absolute isometric tetanic force production was significantly lower than that of controls ( Figure 4C). However, when normalized to muscle CSA, force production was comparable to control values ( Figure 4D). We also sought to determine if M-POMK KO muscle could withstand repeated eccentric contractions. EDL muscles of M-POMK KO mice demonstrated greater force deficits after five and eight lengthening contractions (LC) and recovered to a lower level after 45 min compared to Control EDL ( Figure 4E). Together, the isometric and eccentric contractile studies suggest that the M-POMK KO EDL muscles display a specific force similar to controls ( Figure 4D); however, muscle integrity is compromised following the stress of repeated eccentric contractions, as displayed by the slow, but progressive decline in force production and hampered recovery ( Figure 4E). Thus, the current results demonstrate that the short matriglycan in POMK-deficient skeletal muscle can maintain specific force but cannot prevent eccentric contraction-induced force loss or muscle pathology.
Biochemical analysis of control and M-POMK KO muscle showed a typical, lower molecular weight (MW) a-DG with anti-core DG antibody ( Figure 5A), however, on laminin overlay, we and a kinase domain (N-lobe and C-lobe). The kinase domain contains the catalytic loop (orange) and activation segment (green). (below) Alignment of protein sequences flanking the D204N mutation. The mutation alters a highly conserved aspartate that is the catalytic base of the phosphorylation reaction catalyzed by the kinase. (B) POMK activity in control and patient NH13-284 (POMK D204N) fibroblasts (left) and skeletal muscle (right). n = 3 experiments were performed in fibroblasts. Triple asterisks: statistical significance with Student's unpaired t-test (p-value<0.0001). Due to limited skeletal muscle, n = 1 experiment was performed. (C) Histology and immunofluorescence of control and POMK D204N skeletal muscle using IIH6 (antimatriglycan) and a b-DG antibody.        observed laminin binding at 90-100 kDa ( Figure 5B), similar to POMK D204N skeletal muscle ( Figure 2D). IIH6 also showed binding at 90-100 kDa ( Figure 5C). Solid-phase binding analyses of M-POMK KO and Mck Cre ; Pomk LoxP/LoxP skeletal muscle demonstrated a reduced binding capacity (relative B max ) for laminin-111 compared to control muscle (Figure 5-figure supplement 1A), but higher than that of Large myd muscle, which lacks matriglycan due to a deletion in Large. EDL muscles tested for force production. ***Statistical significance with Student's unpaired t-test with p-value<0.05 (p=0.0031). (B) Cross-sectional area (CSA) of EDL muscles. ***Statistical significance using Student's unpaired t-test with p-value<0.05 (p=0.0463). (C) Maximum Absolute Tetanic Force production by Control and M-POMK KO EDL muscles. ***Statistical significance using Student's unpaired t-test with a p-value<0.05 (p=0.0395). (D) Specific Force production in Control and M-POMK KO EDL muscles (p=0.921). (E) Force deficit and force recovery in Control (n=3) and M-POMK KO (n=4) mice after eccentric contractions. EDL muscles from 18to 20-week-old male mice were tested and are represented by open (Control) or closed (M-POMK KO) circles. ***Statistical significance using Student's unpaired t-test (p-value<0.0001) compared to Control EDL at given LC cycle. **Statistical significance using Student's unpaired t-test (p-value=0.0027) compared to Control EDL at given LC cycle. Error bars represent SD.
To determine if matriglycan is responsible for the laminin binding at 90-100 kDa in POMK-null muscle, we treated glycoproteins enriched from skeletal muscles of M-POMK KO and Mck Cre ; Pomk LoxP/LoxP mice with two exoglycosidases, a-Xylosidase and b-Glucuronidase, which in combination digest matriglycan (    overlay and solid-phase analysis showed a reduction in laminin binding from these muscles after dual exoglycosidase digestion (Figure 5D and E;. To study the role of POMK further, we used human POMK KO HAP1 cells, which have undetectable levels of POMK activity and expression ( Figure 6A; Figure 6-figure supplement 1A; Zhu et al., 2016). A mass spectrometry (MS)-based glycomic analysis of O-glycans carried by recombinantly-expressed DG mucin-like domain indicated the near complete absence of an MS peak at m/z 873.5 corresponding to phosphorylated core M3 O-glycan ( Figure 6D and E; Figure 6figure supplement 2A and B), consistent with an undetectable level of POMK activity in POMK KO HAP1 cells. Compared to WT HAP1 cells, immunoblots of POMK KO HAP1 cells showed a reduction in IIH6 immunoreactivity, a decrease in MW of core a-DG, and the presence of laminin binding at~90 kDa on laminin overlay ( Figure 6C; Figure 6-figure supplement 1B and C). Laminin binding on overlay was rescued only after adenoviral transduction with wild-type (WT) POMK (POMK WT), but not with POMK containing D204N (POMK D204N) or D204A (POMK D204A) mutations ( Figure 6C). POMK D204N also lacked POMK activity in vitro but showed normal B4GAT1, B3GALNT2, and LARGE1 activity, thus confirming the pathogenicity of the D204N mutation ( Figure 6A  , demonstrating that a-DG is the glycoprotein that binds laminin in the absence of POMK. We, therefore, refer to this glycoprotein as POMK-null a-DG (a-DG(POMK)). Since the length of matriglycan correlates with its binding capacity for ECM ligands (Goddeeris et al., 2013), we hypothesized that, given the MW of a-DG(POMK) at 90 kDa, the glycan must be shorter than full-length matriglycan, and therefore, have a lower B max for laminin. We measured the binding capacity of HAP1 a-DG using solid-phase binding assays. B max of a-DG(POMK) for laminin-111 was reduced compared to wildtype a-DG (a-DG(WT)) but was greater than that of a-DG from LARGE1 KO HAP1 cells ( Figure 7B). POMK/DAG1 KO HAP1 cells showed a reduction in B max compared to POMK KO HAP1 cells, but similar to the low levels observed in LARGE1 KO HAP1 cells ( Figure 7B). These data indicate that a short, non-extended form of matriglycan is synthesized on a-DG(POMK), and this short form has a lower binding capacity for laminin-111, thus exhibiting a reduced level of a-DG receptor function.
After POMK phosphorylates core M3, Fukutin (FKTN) modifies GalNAc with ribitol-phosphate for synthesis of full-length matriglycan ( Figure 1; Yoshida-Moriguchi and Campbell, 2015;Hohenester, 2019;Kanagawa et al., 2016). Overexpression in POMK KO HAP1 cells of ISPD, which synthesizes the substrate (CDP-ribitol) of FKTN (Figure 1), increases the amount of matriglycan (without changing its migration on SDS-PAGE) responsible for laminin binding at 90 kDa (Figure 7figure supplement 2A, B and C; Willer et al., 2012;Gerin et al., 2016;Riemersma et al., 2015). Overexpression of LARGE1 can rescue the defect in matriglycan synthesis in distinct forms of CMD as well as in LARGE1 KO HAP1 cells by generating very high molecular weight matriglycan (Figure 7-figure supplement 5A; Barresi et al., 2004). However, overexpression of LARGE1 in POMK or POMK/LARGE1 KO HAP1 cells did not produce very high molecular weight matriglycan ( Figure 7D   These findings indicate that LARGE1 requires phosphorylated core M3 to extend matriglycan on a-DG to its mature and high molecular weight forms. To understand why phosphorylated core M3 is needed for LARGE1 to elongate matriglycan, we measured the binding affinity of LARGE1, as well as POMK, for the phosphorylated core M3 using solution NMR. We previously showed that the unphosphorylated core M3 binds to POMK with high affinity (Zhu et al., 2016). The mannose anomeric proton (Man H1) is well resolved and its intensity decreases only slightly with increasing POMK protein concentration ( These results indicate that, compared to the unphosphorylated core M3 of GGM-MU, the phosphorylated core M3 of GGMp-MU binds to POMK with a much weaker affinity. Then, we measured the binding affinities of LARGE1 for GGMp-MU and GGM-MU in a similar manner. Our results showed that LARGE1 binds with greater affinity to GGMp-MU compared to GGM-MU (K d = 11.5 ± 1.2 mM for GGMp-MU compared to K d >90 mM for GGM-MU) ( Figure 8A, B and D). This indicates that the core M3 phosphate increases the binding affinity of LARGE1 for core M3 and could explain the ability of LARGE1 to elongate matriglycan in the presence of POMK.

Discussion
POMK is a novel muscular dystrophy gene that phosphorylates mannose of the core M3 trisaccharide (GalNAc-b1,3-GlcNAc-b1,4-Man) on a-DG during synthesis of the O-mannose-linked polysaccharide ending in matriglycan. LARGE1 is responsible for the synthesis of matriglycan, and addition of matriglycan enables a-DG to serve as a predominant ECM receptor in many tissues, in particular, skeletal muscle and brain. Over eighteen genes are implicated in matriglycan synthesis, and complete loss-of-function mutations in these genes abrogate synthesis of the O-mannose linked modification and preclude the addition of matriglycan, thereby leading to dystroglycanopathies, congenital and limb-girdle muscular dystrophies with or without structural brain and eye abnormalities. Here, we have used a multidisciplinary approach to show that the absence of POMK activity does not preclude addition of matriglycan. Instead, in the absence of core M3 phosphorylation by POMK, LARGE1 synthesizes a short, non-extended form of matriglycan on a-DG (~90 kDa). However, in order to generate full-length mature matriglycan on a-DG (~150 kDa), LARGE1 requires phosphorylation of core M3 by POMK ( Our study shows that the short form of matriglycan is able to bind to laminin with high affinity and thus enables a-DG(POMK) to function as an ECM receptor. Given the very small increase in apparent MW in a-DG(POMK) compared to a-DG from cells and muscle lacking LARGE1 ( , the short, non-extended form of matriglycan likely contains few Xyl-GlcA repeats. However, it can still bind laminin since only a single Xyl-GlcA repeat is needed for laminin binding (Briggs et al., 2016), but it cannot function as an ECM scaffold. This short matriglycan likely attenuates muscular dystrophy in patient NH13-284 with a complete loss-of-function mutation in POMK, preventing the severe WWS phenotype that is observed in the complete absence of the other known dystroglycanopathy genes.
Muscle-specific POMK KO mice express the short, non-extended form of matriglycan on~90 kDa a-DG and develop a mild muscular dystrophy phenotype. Muscle physiology studies demonstrate that the short matriglycan expressed in the absence of POMK can maintain specific force but cannot         The peak Man H1 is derived from the mannose anomeric H1 proton. Stars indicate impurity peaks derived from buffer. (C, D) Fitting of the NMR binding data of POMK (C) and LARGE1 (D) to core M3 glycans of GGM-MU and GGMp-MU, respectively. The bound fraction was obtained from the NMR titration data by measuring the difference in the peak intensity of the anomeric proton Man H1 in the absence (free form) and presence (bound form) of POMK or LARGE1, then divided by the peak intensity of the free form. Figure 8 continued on next page prevent eccentric contraction-induced force loss or skeletal muscle pathology. Interestingly, missense mutations in FKRP that cause LGMD2I also show reduced expression of matriglycan (Yoshida-Moriguchi and Campbell, 2015) and exhibit a milder muscular dystrophy. Thus, M-POMK KO mice are an excellent model of milder forms of dystroglycanopathy in which short matriglycan is expressed and will be useful for future studies of these forms of dystroglycanopathy.
a-DG is composed of three domains: the DGN, which undergoes cleavage at arginine-312 by a furin-like convertase during a-DG post-translational processing, a central mucin-like domain, and a C-terminus Singh et al., 2004). The natural C-terminal domain boundary of DGN, arginine-312 in humans, is proximal to three sites of matriglycan synthesis (threonines-317, 319, 379) within the mucin-like domain of a-DG. Biochemical studies using various POMK KO HAP1 cell lines demonstrated that the synthesis of the short, non-extended form of matriglycan occurs on threonine-317 of the mucin-like domain and, like full-length matriglycan, requires LARGE1, DGN, and CDP-ribitol. Cell biological experiments demonstrated that the DGN is necessary for synthesis of the short form of matriglycan. As the binding of LARGE1 to the DGN is essential for the synthesis of full-length matriglycan on a-DG Hara et al., 2011a), it is required for synthesis of the short form of matriglycan as well. Solution NMR studies revealed that LARGE1 binds to core M3, and the binding affinity increases in the presence of the mannose phosphate. The phosphorylated core M3, could, therefore serve to recruit DGN-bound LARGE1 to the proper residue during the initiation of full-length matriglycan synthesis. In the absence of the mannose phosphate, the DGN-bound LARGE1 may instead act only upon the matriglycan acceptor added to threonine-317, the threonine nearest to the DGN. Synthesis of full-length matriglycan may, therefore, proceed through a complex of DGN, LARGE1, and phosphorylated core M3. The phosphorylated core M3 may also serve to anchor LARGE1 to a-DG during matriglycan elongation. In the absence of POMK, the binding of LARGE1 to the DGN and the unphosphorylated core M3 may only be sufficient for synthesis of a short form of matriglycan. Further structural and biochemical studies will be required to understand the precise interactions between DGN, LARGE1, and the phosphorylated core M3. Taken together, our results indicate that LARGE1 requires DGN to synthesize the short, nonextended form of matriglycan but needs both the DGN and the phosphorylated core M3 to generate full-length matriglycan on a-DG.
Our study demonstrates that POMK is required for the synthesis of full-length and high-molecular weight forms of matriglycan (Figure 8-figure supplement 2A). In the absence of POMK, LARGE1 generates a short, non-extended form of matriglycan (Figure 8-figure supplement 2B). Collectively, our work provides the first insights into the pathogenic mechanism behind POMK-deficient muscular dystrophy and better elucidates how full-length matriglycan is synthesized so it can act as a scaffold for ECM proteins, thereby enabling proper skeletal muscle function and preventing muscular dystrophy.

Patient information
Patient NH13-284 received a diagnosis of congenital muscular dystrophy (CMD) with brain malformations.

Generation of Pomk LoxP/LoxP mice
The Pomk gene consists of five exons, exons 1, 2, and 3, which are non-coding and exons 4 and 5, which are coding (Zhu et al., 2016;Di Costanzo et al., 2014). We used Clustered Regularly Interspersed Short Palindromic Repeats (CRISPR)-Cas9 to insert LoxP sites around exon 5.  Animals B6SJLF1/J mice were purchased from Jackson Labs (100012; Bar Harbor, ME). Male mice older than 8 weeks were used to breed with 3-5 week-old super-ovulated females to produce zygotes for electroporation. Female ICR (Envigo, Indianapolis, IN; Hsc:ICR(CD-1)) mice were used as recipients for embryo transfer.
Sanger sequencing of tail DNA was performed by the University of Iowa Genome Editing Core Facility to confirm incorporation of 5' and 3' LoxP sites. PCR probes were developed at Transnetyx to genotype mice expressing both Pax7-Cre and Mck-Cre. Genotyping of Mck Cre ; Pax7 Cre ; Pomk LoxP/LoxP mice was performed by Transnetyx using real-time PCR.
All mice were socially housed in a barrier-free, specific pathogen-free conditions as approved by the University of Iowa Animal Care and Use Committee (IACUC). All animals were maintained in a climate-controlled environment at 25˚C and a 12/12 light/dark cycle. Animal care, ethical usage, and procedures were approved and performed in accordance with the standards set forth by the

Preparation of Cas9 RNPs and the microinjection mix
Chemically modified CRISPR-Cas9 crRNAs and CRISPR-Cas9 tracrRNAs were purchased from Integrated DNA Technologies (IDT) (Alt-R CRISPR-Cas9 crRNA; Alt-R CRISPR-Cas9 tracrRNA (Cat# 1072532)). The crRNAs and tracrRNA were suspended in T10E0.1 and combined to 1 mg/mL (~29.5 mM) final concentration in a 1:2 (mg: mg) ratio. The RNAs were heated at 98˚C for 2 min and allowed to cool slowly to 20˚C in a thermal cycler. The annealed cr:tracrRNAs were aliquoted to single-use tubes and stored at À80˚C.
Cas9 nuclease was also purchased from IDT (Alt-R S.p. HiFi Cas9 Nuclease). Cr:tracr:Cas9 ribonucleoprotein complexes were made by combining Cas9 protein and each cr:tracrRNA; final concentrations: 60 ng/mL (~0.4 mM) Cas9 protein and 60 ng/mL (~1.7 mM cr:tracrRNA). The Cas9 protein and annealed RNAs were incubated at 37˚C for 10 min. The two RNP mixes were combined and incubated at 37˚C for an additional 5 min. The single stranded oligonucleotide donors (ssODN) were purchased from IDT as Ultramers. The ssODNs were added to the RNPs and the volume adjusted to the final concentrations in the injection mix were 10 ng/mL each ssODN; 20 ng/mL each guide RNA and 40 ng/mL Cas9 protein.

Collection of embryos and microinjection
Pronuclear-stage embryos were collected using previously described methods (Pinkert, 2002). Embryos were collected in KSOM media (Millipore, Burlington, MA; MR101D) and washed three times to remove cumulous cells. Cas9 RNPs and ssODNs were injected into the pronuclei of the collected zygotes and incubated in KSOM with amino acids at 37˚C under 5% CO 2 until all zygotes were injected. Fifteen to 25 embryos were immediately implanted into the oviducts of pseudo-pregnant ICR females.
Insertion of loxP1 (5') and loxP2 (3') sites was confirmed by cloning and sequencing of genomic PCR products (Figure 3-figure supplement 2) from tail DNA of filial 0 (F0) Pomk LoxP/+ mice using primers flanking the 5' loxP site, ACTCCAGTTGGTTTCAGGAAG and GAGGGAAGAGAAGTCAG-GAAAG. For the 3' loxP site, primers of sequence ACCGAGTGTGAGATTCAAGTG and GGTTGC TGGTAGGGTTAAGAG were used. The 5' loxP site contains a KpnI cleavage site, and the 3' loxP site contains a BamHI site. The screen of the 5' loxP site gives a product of 803 base pairs for the LoxP allele when uncut. Kpnl digestion of the 5' loxP site gives three products of 381, 355, and 67 base pairs. A screen of the 3' loxP site gives a product of 396 base pairs for the uncut allele with loxP site, while BamHI digestion of the 3' loxP site gives products of 273 and 123 base pairs.
Genotyping was carried out using primers flanking the exon five loxP1 site or the (TTCTTTCTG TGATGTGTGCTTATTC) or loxP2 (CAGACACTCACCCTTTACCTTAG) site. The wild-type allele is 197 bp while the floxed allele is 235 bp. Pomk LoxP/+ mice were backcrossed five generations onto a C57BL/6J background and backcrossed mice used whenever possible.

Forelimb grip strength test
Forelimb grip strength was measured at 1 month and 4 months of age using previously published methods (de Greef et al., 2016). A mouse grip strength meter (Columbus Instruments, Columbus, OH) was mounted horizontally, with a nonflexible grid connected to the force transducer. The mouse was allowed to grasp the grid with its two front paws and then pulled away from the grid by its tail until the grip was broken. This was done three times over five trials, with a one-minute break between each trial. The gram force was recorded per pull, and any pull where only one front limb or any hind limbs were used were discarded. If the mouse turned, the pull was also discarded. After 15 pulls (5 sets of 3 pulls), the mean of the three highest pulls of the 15 was calculated and reported. Statistics were calculated using GraphPad Prism eight software. Student's t-test was used (twosided). Differences were considered significant at a p-value less than 0.05. Graph images were also created using GraphPad Prism and the data in the present study are shown as the means + / -SD unless otherwise indicated.

Creatine kinase assay
Creatine kinase levels were measured in 8-week-old mice 2 hr after mild downhill run (three meters per minute for 5 min followed by 15 m per minute for 10 min) at a 15-degree downhill incline as previously described (de Greef et al., 2016;Goddeeris et al., 2013). Blood was collected by tail vein bleeds from non-anesthetized, restrained mice using a Microvette CB300 (Sarstedt AG and Co, Newton, NC). Samples were centrifuged at 12,000 rpm for 10 min and prepared using an enzymecoupled CK kit (Stanbio Laboratory, Boerne, TX) using the manufacturer's instructions. Absorbance was measured using a plate reader at 340 nm every 30 s for 2 min at 37˚C. Statistics were calculated using GraphPad Prism software and Student's t-test was used (two-sided). Differences were considered significant at a p-value less than 0.05. Graph images were also created using GraphPad Prism eight and the data in the present study are shown as the means + / -SD unless otherwise indicated.

Body weight measurements
Mice were weighed as previously described (de Greef et al., 2016). Weights were measured after testing grip strength using a Scout SPX222 scale (OHAUS Corporation, Parsippany, NJ), and the tester was blinded to genotype. Statistics were calculated using GraphPad Prism eight software and Student's t-test was used (two-sided). Differences were considered significant at a p-value less than 0.05. Graph images were also created using GraphPad Prism and the data in the present study are shown as the means + / -SD unless otherwise indicated.

Measurement of in vitro muscle function
To compare the contractile properties of muscles, extensor digitorum longus (EDL) muscles were surgically removed as described previously (Rader et al., 2016;de Greef et al., 2016). The muscle was immediately placed in a bath containing a buffered physiological salt solution (composition in mM: NaCl, 137; KCl, 5; CaCl 2 , 2; MgSO 4 , 1; NaH 2 PO 4 , 1; NaHCO 3 , 24; glucose, 11). The bath was maintained at 25˚C, and the solution was bubbled with 95% O 2 and 5% CO 2 to stabilize pH at 7.4. The proximal tendon was clamped to a post and the distal tendon tied to a dual mode servomotor (Model 305C; Aurora Scientific, Aurora, ON, Canada). Optimal current and whole muscle length (L 0 ) were determined by monitoring isometric twitch force. Optimal frequency and maximal isometric tetanic force (F 0 ) were also determined. The muscle was then subjected to an eccentric contraction (ECC) protocol consisting of eight eccentric contractions (ECCs) at 3 min intervals. A fiber length L fto-L 0 ratio of 0.45 was used to calculate L f . Each ECC consisted of an initial 100 millisecond isometric contraction at optimal frequency immediately followed by a stretch of L o to 30% of L f beyond L o at a velocity of 1 L f /s at optimal frequency. The muscle was then passively returned to L o at the same velocity. At 3, 15, 30, 45, and 60 min after the ECC protocol, isometric tetanic force was measured. After the analysis of the contractile properties, the muscle was weighed. The CSA of muscle was determined by dividing the muscle mass by the product of L f and the density of mammalian skeletal muscle (1.06 g/cm 3 ). The specific force was determined by dividing F o by the CSA (kN/mm 2 ). 18-20week-old male mice were used, and right and left EDL muscles from each mouse were employed whenever possible, with n = 5 to 8 muscles used for each analysis. Each data point represents an individual EDL. Statistics were calculated using GraphPad Prism eight software and Student's unpaired t-test was used (two-sided). Differences were considered significant at a p-value less than 0.05.

H&E and immunofluorescence analysis of skeletal muscle
Histology and immunofluorescence of mouse skeletal muscle were performed as described previously (Goddeeris et al., 2013). Mice were euthanized by cervical dislocation and directly after sacrifice, quadriceps muscles were isolated, embedded in OCT compound and then snap frozen in liquid nitrogen-cooled 2-methylbutane. 10 mM sections were cut with a cryostat (Leica CM3050S Research Cryostat; Amsterdam, the Netherlands) and H&E stained using conventional methods. Whole digital images of H&E-stained sections were taken by a VS120-S5-FL Olympus slide scanner microscope (Olympus Corporation, Tokyo, Japan). For immunofluorescence analyses, a mouse monoclonal antibody to glycoepitopes on the sugar chain of a-DG (IIH6, 1:100 dilution, Developmental Studies Hybridoma Bank, University of Iowa; RRID:AB_2617216) was added to sections overnight at 4˚C followed by Alexa Fluor-conjugated goat IgG against mouse IgM (Invitrogen, Carlsbad, CA, 1:500 dilution), for 40 min. The sections were also stained with rabbit polyclonal antibody to b-DG (AP83; 1:50 dilution) followed by Alexa Fluor-conjugated 488 Goat anti-rabbit IgG (1:500). Whole sections were imaged with a VS120-S5-FL Olympus slide scanner microscope. Antibody IIH6 is a monoclonal to the glycoepitope of a-DG (Ervasti and Campbell, 1991), and AP83 is a polyclonal antibody to the C-terminus of b-DG (Ervasti and Campbell, 1991), both of which have been described previously.
For histologic analysis of human skeletal muscle, H&E staining on 10 mm frozen section was performed using the Leica ST5020 Multistainer workstation (Leica Biosystems, Buffalo Grove, IL) according manufacturer's instructions. For immunofluorescence analysis, unfixed frozen serial sections (7 mm) were incubated with primary antibodies for 1 hr, and then with the appropriate biotinylated secondary antibodies for 30 min followed by streptavidin conjugated to Alexa Fluor 594 (ThermoFisher Scientific, UK) for 15 min. Primary antibodies used were mouse monoclonal: a-DG IIH6 (clone IIH6C4) (Ervasti and Campbell, 1991), b-DG (Leica, Milton Keynes, UK; clone 43DAG1/8D5). All washes were made in PBS and incubations were performed at room temperature. Sections were evaluated with a Leica DMR microscope interfaced to MetaMorph (Molecular Devices, Sunnyvale, CA).

Tissue biochemical analysis
30 slices of 30 mM thickness were taken with a cryostat (Leica CM3050S Research Cryostat) from skeletal muscle or heart that had been frozen in liquid nitrogen-cooled 2-methylbutane. For biochemical analysis of murine skeletal muscle, quadriceps muscle were used.
Samples were solubilized in 500 mL of 1% Triton X-100 in 50 mM Tris pH 7.6 and 150 mM NaCl with protease inhibitors (per 10 mL buffer: 67 mL each of 0.2 M phenylmethylsulfonylfluoride (PMSF), 0.1 M benzamidine and 5 mL of each of leupeptin (Sigma/Millipore) 5 mg/mL, pepstatin A (Millipore) 1 mg/mL in methanol, aprotinin (Sigma-Aldrich) 5 mg/mL, calpeptin (Fisher/EMD Millipore) 1.92 mg/ mL in Dimethyl Sulfoxide (DMSO), Calpain Inhibitor 1 (Sigma-Aldrich) 1.92 mg/mL in DMSO). Samples were vortexed for 4 min and solubilized for 2.5 hr at 4˚C with rotation. Samples were then spun down at 12,000 rpm for 30 min at 4˚C on a Beckman Tabletop Centrifuge. The supernatant was incubated with 100 mL WGA-Agarose slurry (Vector Biolabs, Malvern, PA, AL-1023) overnight at 4˚C with rotation. The next day samples were washed three times in 50 mM Tris pH 7.6 and 150 mM NaCl with 0.1% TX-100 and protease inhibitors. 100 mL of 5X Laemmli Sample Buffer (LSB) was added, samples boiled for 10 min, and 125 mL of this was loaded in each lane of gels for western blotting.
For flow cytometry analyses, fibroblasts cultured from skin biopsies were grown in Dulbecco's modified Eagles medium (Invitrogen) with 20% fetal bovine serum (FBS, Life Technologies), 1% glutamax (Thermo Fisher Scientific) and 1% penicillin/streptomycin (Sigma-Aldrich). Upon approximately 90% confluence, cells were washed with PBS without Ca and Mg, detached with non-enzymatic dissociation solution (Sigma-Aldrich cat. C5914) and fixed in 2% paraformaldehyde for 10 min. Cells were subsequently incubated on ice with the following antibodies diluted in PBS/0.1% FBS: anti-a-DG IIH6 (Millipore) for 30 min, anti-mouse biotinylated IgM (Vector Labs, Burlingame, CA) for 20 min, Streptavidin-Phycoerythrin (BD Pharmingen) for 15 min. Negative controls for each fibroblast population were incubated with 0.1% FBS/PBS without the primary antibodies. Cells were washed twice and centrifuged at 1850 g for 4 min, after each incubation step. After the last wash, cell pellets were resuspended in 500 mL of PBS. A total of 10,000 event were acquired using the Cyan ADP analyser (Beckman Coulter, Brea, CA) and analysed using FlowJo software version 7.6.5 (Tree Star, USA).

Generation and characterization of HAP1 mutant cell lines
HAP1 cells (RRID:CVCL_Y019) are a haploid human cell line with an adherent, fibroblast-like morphology, originally derived from parent cell line KBM-7 (RRID:CVCL_A426). WT C631 cells (a diploid cell line containing duplicated chromosomes of HAP1) were purchased from Horizon Discovery, and gene-specific knockout (KO) HAP1 cells were generated by Horizon Discovery. Absence of the gene was confirmed via PCR amplification and Sanger sequencing. The identity of the cells has been authenticated by the company using the STR profiling method. Mycoplasma testing of the cells was performed on a routine basis to ensure the cells are not contaminated. HAP1 KO cell lines have complete loss of gene function and are validated in the lab by performing western blot analysis before and after gene transfer with the appropriate gene. For each HAP1 KO cell line, a matched WT control parental cell line (WT C631) was provided, ensuring that phenotypes can be attributed directly to the genetic modification. These cells are cultured in Iscove's Modified Dulbecco's medium (IMDM) supplemented with 10% fetal calf serum and 1% Pen-Strep antibiotics.

POMK knockout (KO) HAP1
HAP1 cells bearing a 10 bp deletion of exon 4 of POMK, generated using the CRISPR/Cas9 system, were purchased from Horizon Discovery (HZGHC001338c004, clone 1338-4) and were previously described (Zhu et al., 2016). POMK knockout (KO) HAP1 cells lack the single copy of the wild-type POMK allele and are therefore null at the POMK locus. The sequence of the guide RNA used is TGA-GACAGCTGAAGCGTGTT. Absence of the wild-type POMK allele was confirmed by Horizon Discovery via PCR amplification and Sanger sequencing. PCR primers used for DNA sequencing are POMK Forward 5'-ACTTCTTCATCGCTCCTCGACAA-3', and POMK Backward 5'-GGATGCCACACTGC TTCCCTAA-3'. The identity of the cells has been authenticated by the company using the STR profiling method. Mycoplasma testing of the cells was performed on a routine basis to ensure the cells are not contaminated.

POMK/LARGE1 KO HAP1
HAP1 cells lacking both POMK and LARGE1 expression (POMK/LARGE1 KO HAP1 cells) (HZGHC007364c011) were generated using CRISPR/Cas9 by Horizon Discovery. A 43 bp deletion of exon 3 of LARGE1 was introduced into the POMK KO HAP1 line (HZGHC001338c004). The guide RNA sequence was CTCGGCGATGGGATGGGGCT and the primer sequence was PCR forward GAGGCATGGTTCATCCAGATTAAAG and PCR reverse CTTTACCTCGCATTTCTCCACGA.

POMK/ISPD KO HAP1
HAP1 cells containing a 1 bp insertion of exon 4 of the POMK gene, generated using the CRISPR/ Cas9 system, were purchased from Horizon Discovery (HZGHC001338c001, clone 1338-1). The mutation in POMK is predicted to lead to a frameshift. These cells also lacked expression of ISPD. The guide RNA sequence was TGAGACAGCTGAAGCGTGTT. The sequences of PCR primers were PCR forward ACTTCTTCATCGCTCCTCGACAA and PCR reverse GGATGCCACACTGCTTCCCTAA.

LARGE1 KO HAP1
HAP1 cells (clone 122-6, HZGHC000122c006) were purchased from Horizon Discovery. Cells were generated using a CRISPR/Cas9-mediated 1 bp deletion of exon 3 of LARGE. The guide RNA sequence was GCTCTCGCGCTCCCGCTGGC and the primer sequence for 122-7 was PCR forward ATGGAGTAGGTCTTGGAGTGGTT and PCR reverse GAGGCATGGTTCATCCAGAGTTAAAG.

FKTN KO HAP1
HAP1 cells (clone 721-10, catalog number 32597-10) were purchased from Horizon Discovery. CRISPR/Cas9 was used to introduce a 16 bp deletion of exon 3 of FKTN. The sequence of the guide RNA was CAGAACTTGTCAGCGTTAAA and the sequences of PCR forward CAGATCAAAGAA TGCCTGTGGAAAT and PCR reverse TGCAAAGAGAAGTGTGATCAGAAAA.
Adenovirus production DGE (Delta H30-A316) was generated and described previously (Hara et al., 2011a;Kanagawa et al., 2004;Kunz et al., 2001). DG T317A, DG T319A, and DG T317A/T319A were first subcloned into an Fc-tagged DG construct (DGFc3) (Hara et al., 2011b;Kanagawa et al., 2004;Kunz et al., 2001). The KpnI-XhoI fragments from the DGFc3 mutants corresponding to the mutant constructs (DG T317A, DG T319A, or DG T317A/T319A) were then subcloned into pAd5RSVK-NpA (obtained from the University of Iowa Viral Vector Core) as was the XhoI-XbaI fragment from an adenovirus encoding dystroglycan WT. E1-deficient recombinant adenoviruses (Ad5 RSV DG WT, DG T317/T319, DG T317A, DG T319A, DGE, Ad-POMK WT) were generated by the University of Iowa Viral Vector Core (VVC) using the RAPAd system (Anderson et al., 2000). Assays for replication competence of adenoviruses were performed to check for contamination. Ad-POMK WT and Ad-POMK D204A were generated by ViraQuest Inc (North Liberty, IA) using the RAPAd system and was described previously (Zhu et al., 2016). Ad-POMK D204N was also generated by ViraQuest Inc Absence of the viral E1 DNA sequence was confirmed by ViraQuest Inc after PCR amplification of the viral DNA and staining on DNA agarose gel electrophoresis. Replication competence of adenoviruses was negative as assessed by plaque forming assays in cells performed from 10 9 viral particles up to 14 days. Adenoviral Fukutin (FKTN) and ISPD have been described previously (Willer et al., 2012). Adenoviral LARGE1 has been described previously . DGFc340TEV was cloned into the pUC57-mini vector by GenScript (Hara et al., 2011a;Kanagawa et al., 2004;Kunz et al., 2001). The insert includes TEV protein cleavage site between amino acids (AAs) 1-340 of rabbit DG and human IgG1 Fc. The insert was subcloned in pcDNA3 expression vector with EcoRI. Subsequently, FseI-x-340 AAs DG-TEV-6xHis-NotI fragment was obtained using pcDNA3DGFc340-TEV as a PCR template. FseI-x-340 AAs DG-TEV-6xHis-NotI was ligated into pcDNA3DGFc340TEV digested with FseI and NotI to construct DG340TEVHis, which includes 1-340 AAs of rabbit DG, TEV site, and 6x Histidine. The construct was also inserted in pacAd shuttle plasmid from the VVC to generate the adenoviral vector. Next, FseI-x-390 AAs-TEV-6xHis-NotI was obtained using pcDNA3rbtDG as a PCR template and ligated into the pcDNA3DG340TEVHis digested with FseI and NotI to construct DG390TEVHis, which includes 1-390 AAs of rabbit DG, TEV site, and 6x Histidine. The construct was also inserted in pacAd shuttle plasmid from the VVC to generate the Ad virus vector. E1-deficient recombinant adenoviruses were generated by the University of Iowa Viral Vector Core using the RAPAd system (Kunz et al., 2001).

HAP1 cell culture and adenovirus infection
HAP1 cells were maintained at 37˚C and 5% CO 2 in Iscove's Modified Dulbecco's Medium (IMDM, Gibco) supplemented with 10% Fetal Bovine Serum (FBS) and 1% penicillin/streptomycin (Invitrogen). Cells were split every 3 days at 1:10 using Trypsin-EDTA (ThermoFisher Scientific). On day one for adenovirus transfection experiments, media was changed to 2% IMDM, and an average of 5.9 Â 10 6 POMK KO HAP1 cells were infected at the indicated multiplicity of infection (MOI) with the indicated adenovirus. On day 2, infection medium was replaced with 10% IMDM, and on day three the cells were processed for biochemical analyses.

Glycoprotein isolation and biochemical analyses from cultured cells
For western blots and laminin overlay, HAP1 cells and fibroblasts were washed twice in ice-cold Dulbecco's phosphate-buffered saline (DPBS, Gibco). The second PBS wash contained the protease inhibitors (0.23 mM PMSF and 0.64 mM benzamidine). Plates were scraped, spun down for 5 min at 14, 000 rpm at 4˚C, and pellets were solubilized in 1% Triton X-100 in Tris-buffered saline (TBS, 50 mM Tris-HCl pH 7.6, 150 mM NaCl) with protease inhibitors (0.23 mM PMSF and 0.64 mM benzamidine) for 1 hr at 4˚C. Samples were then spun down at 14,000 rpm for 5 min, and supernatants incubated in 200 mL wheat-germ agglutinin (WGA)-agarose (Vector Laboratories, AL-1023) as previously described Goddeeris et al., 2013). The following day, WGA beads were washed three times with 0.1% Triton X-100-TBS plus protease inhibitors and heated to 99˚C for 10 min with 250 mL of 5X Laemmli sample buffer. Samples were run on SDS-PAGE and transferred to PVDF-FL membranes (Millipore) as previously published Goddeeris et al., 2013).

EDTA treatment of ligand overlays
EDTA treatment of laminin overlay assays was performed as described above for laminin overlays; however, calcium was excluded from all buffers made with LBB (i.e. 5% milk-LBB, 3% BSA-LBB) and 10 mM EDTA was added to all LBB-based buffers, including LBB wash buffer, 5% milk-LBB, and 3% BSA-LBB buffers.
POMK assay HAP1 cells were washed twice in ice-cold PBS, scraped, and spun down at 14,000 rpm for 5 min at 4˚C. After removing supernatant, the cell pellet was resuspended in 0.1 M MES buffer pH 6.5 with 1% Triton X-100 with Protease Inhibitors (0.23 mM PMSF and 0.64 mM Benzamidine) for 1 hr at 4˚C rotating. Samples were spun down again, and the supernatant was incubated with 200 mL of WGAagarose beads (Vector Biolabs, AL-1023) overnight at 4˚C with rotation. Samples were washed the next day three times in 0.1 M MES pH 6.5 with 0.1% Triton X-100 and protease inhibitors, and 100 mL of the beads were resuspended in 100 mL of the wash buffer.
For fibroblast POMK activity measurements, cells were processed as above and solubilized in 1% TX-100 in 50 mM Tris and 150 mM NaCl pH 7.6 with protease inhibitors as described above and incubated with WGA-agarose beads. The next day, WGA beads were washed three times and resuspended in 0.1% TX-100 in 0.1 M MES pH 6.5 buffer with protease inhibitors.
For measurement of mouse and human skeletal muscle POMK activity, 30 slices of 30 mM thickness were taken using a Leica 3050 s cryostat from quadriceps muscle frozen in liquid nitrogencooled 2-methylbutane. Samples were solubilized in 250 mL of 1% Triton X-100 in 0.1 M MES pH 6.5 with protease inhibitors (per 10 mL buffer: 67 mL each of 0.2 M PMSF, 0.1 M benzamidine and 5 mL/ 10 mL of buffer of leupeptin (Sigma/Millipore) 5 mg/mL, pepstatin A (Millipore) 1 mg/mL in methanol, aprotinin (Sigma-Aldrich) 5 mg/mL, calpeptin (Fisher/EMD Millipore) 1.92 mg/mL in dimethyl sulfoxide (DMSO), calpain inhibitor 1 (Sigma-Aldrich) 1.92 mg/mL in DMSO). Samples were solubilized for 2.5 hr at 4˚C on a rotator. Samples were then spun down at 14,000 rpm for 30 min at 4˚C on a Beckman Tabletop Centrifuge. The supernatant (total lysate) was separated from the pellet, and 10 mL of this was used for POMK assays.
For POMK reaction in HAP1 cells and fibroblasts, 20 mL slurry (consisting of 10 mL beads and 10 mL MES buffer) was incubated with reaction buffer for a final reaction volume of 40 mL. For POMK assay from skeletal muscle, 10 mL of total lysate was incubated with 20 mL of reaction buffer for a reaction volume of 30 mL. The final reaction concentration was 10 mM ATP, 10 mM MnCl2, 10 mM MgCl2, 10 mM GGM-MU, 0.1% TX-100 in 0.1 M MES Buffer pH 6.5. Reactions were run at 37˚C for 24 hr for HAP1 cells, 48 hr for fibroblasts, or 16 hr for skeletal muscle. Experiments were done in triplicate, with each replicate representing a separate plate of cells or animal. After POMK reaction, 6 mL 0.5 M EDTA was added to 30 mL of reaction supernatant, and the mixture boiled for 5 min. 25 mL of this mixture and added to 30 mL ddH20 in HPLC vial and run on an LC18 column of a reversephase HPLC (Shimadzu Scientific, Columbia, Maryland) with a 16% B med sensitivity gradient. The reaction was analyzed using a 4.6 Â 250 mm Supelcosil LC-18 column (Supelco). Solvent A was 50 mM ammonium formate (pH 4.0), and solvent B was 80% acetonitrile in solvent A. Elution of the MU derivative was monitored by fluorescence detection (325 nm for excitation, and 380 nm for emission) and peak area used as a measure of activity. The enzymatic activity was calculated as the peak area of the product.

B4GAT1 assay
For the assessment of endogenous B4GAT1 activity in skeletal muscle, Triton X-100-solubilized lysates (10 ml for human skeletal muscle or 40 mL for mouse skeletal muscle) were incubated in a volume of 50 mL (human skeletal muscle) for 12 hr at 37˚C, with 0.4 mM Xylose-b-MU (Xyl-b-MU) and 10 mM Uridine diphosphate glucuronic acid (UDP-GlcA) in 0.1 M MES buffer, pH 6.0, at 5 mM MnCl 2 , 5 mM MgCl 2 , and 0.05% Triton X-100 (Willer et al., 2014). The reaction was terminated by adding 25 mL of 0.1 M EDTA and boiling for 5 min, and the supernatant was analyzed using an LC-18 column. Both the substrate Xyl-b-MU and the product GlcA-Xyl-b-MU were separated on a 16% acetonitrile isocratic gradient. Elution of the MU derivative product was monitored by fluorescence detection (325 nm for excitation, and 380 nm for emission). The percent conversion of substrate to product was used as the activity of the B4GAT1 in the 10 mL sample. The B4GAT1 activity then was normalized against the amount of protein measured in the 10 mL of sample using the DC protein assay (Bio-Rad, Hercules, CA) with BSA as the standard.
For assessment of B4GAT1 activity in HAP1 cells, the HAP1 WGA beads were incubated in a volume of 80 mL for 26 hr at 37˚C, with 0.4 mM Xyl-b-MU and 10 mM UDP-GlcA in 0.1 M MES buffer, pH 6.0, at 5 mM MnCl 2 , 5 mM MgCl 2 , and 0.05% Triton X-100. The reaction was terminated by adding 25 mL of 0.1 M EDTA and boiling for 5 min, and the supernatant was analyzed using an LC-18 column. Elution of the MU derivative was monitored by fluorescence detection (325 nm for excitation, and 380 nm for emission) and peak area used as a measure of activity. The percent product was determined by taking the product peak area and dividing by the total peak areas of substrate plus product peak. Then this number was taken and multiplied by 100 for percent conversion to product.

LARGE1 assay
For the assessment of endogenous LARGE1 GlcA-T activity in skeletal muscle, Triton X-100-solubilized lysates were incubated in a volume of 25 mL for 3 hr at 37˚C, with 0.4 mM Xyl-a1,3-GlcA-b-MU and 10 mM UDP-GlcA in 0.1 M MES buffer, pH 6.0, at 5 mM MnCl 2 , 5 mM MgCl 2 , and 0.5% Triton X-100. The reaction was terminated by adding 25 mL of 0.1 M EDTA and boiling for 5 min, and the supernatant was analyzed using an LC-18 column. Elution of the MU derivative was monitored by fluorescence detection (325 nm for excitation, and 380 nm for emission) and peak area used as a measure of activity. The GlcA-T activity was assessed by subtracting the background observed in the negative control sample without donor sugar and normalized against the amount of protein measured using the DC protein assay (Bio-Rad).
For assessment of LARGE1 enzymatic activity in HAP1 cells, the Triton X-100-solubilized HAP1 cells were loaded onto WGA beads and processed as described for POMK assay above. The next day after wash, beads were incubated in a volume of 90 mL with 0.4 mM Xyl-a1,3-GlcA-b-MU and 10 mM UDP-GlcA in 0.1 M MOPS buffer, pH 6.0, at 5 mM MnCl 2 , 5 mM MgCl 2 , and 0.05% Triton X-100. The samples were run for 46 hr at 37˚C. The reaction was terminated by adding 25 mL of 0.25 M EDTA and boiling for 5 min, and the supernatant was analyzed using an LC-18 column.
For the assessment of endogenous LARGE1 activity in fibroblasts, supernatants from Triton X-100-solubilized fibroblasts were (20 mL) directly used. Supernatants were incubated in a volume of 100 mL for 24 hr at 37˚C, with 0.4 mM Xyl-a1,3-GlcA-b-MU and 10 mM UDP-GlcA in 0.1 M MES buffer, pH 6.0, at 5 mM MnCl2, 5 mM MgCl2, and 0.5% Triton X-100. The reaction was terminated by adding 25 mL of 0.1 M EDTA and boiling for 5 min, and the supernatant was analyzed using an LC-18 column.
Elution of the MU derivative was monitored by fluorescence detection (325 nm for excitation, and 380 nm for emission) and peak area used as a measure of activity. The percent product was determined by taking the product peak area and dividing by the total peak areas of substrate plus product peak. Then this number was taken and multiplied by 100 for percent conversion to product.

B3GALNT2 assay
To assess B3GALNT2 activity in HAP1 cells, 20 mL of the WGA beads from HAP1 cells were incubated with a 20 mL volume of the reaction mix. The final volume of reaction buffer was 40 mL (30 mL reaction mixture and 10 mL WGA beads). The final concentrations were 10 mM MgCl 2 , 10 mM MnCl 2 , 0.1 M MES pH 6.5, 10 mM GM-MU, and 10 mM UDP-GalNAc. Reactions were run at 37˚C for 72 hr. Experiments were done in triplicate, with each replicate representing a separate plate of cells. After B3GALNT2 reaction, 6 mL 0.5 M EDTA was added to 30 mL of reaction supernatant, and the mixture boiled for 5 min. 25 mL of this mixture and added to 30 mL ddH20 in HPLC vial and run on an LC18 column of a reverse-phase HPLC (Shimadzu Scientific) with a 16% B med sensitivity gradient. The reaction was analyzed using a 4.6 Â 250 mm Supelcosil LC-18 column (Supelco, Bellefonte, PA). Solvent A was 50 mM ammonium formate (pH 4.0), and solvent B was 80% acetonitrile in solvent A. Elution of the MU derivative was monitored by fluorescence detection (325 nm for excitation, and 380 nm for emission) and peak area used as a measure of activity. The enzymatic activity was calculated as the peak area of the product.

Digestion of a-DG with exoglycosidases
Exoglycosidase treatment was carried out as described previously (Briggs et al., 2016;Salleh et al., 2006;Moracci et al., 2000). T. maritima b-glucuronidase (Salleh et al., 2006;Moracci et al., 2000) (Bgus) and S. solfataricus a-xylosidase (Xylsa), both bearing a His-tag were overexpressed in E. coli, and purified using Talon metal affinity resin as described and activity determined as described (Salleh et al., 2006;Moracci et al., 2000) with some modifications. Briefly, the cell pellet was resuspended in 20 mM HEPES buffer (pH 7.3), 150 mM NaCl, 0.1% NP-40 and sonicated. After centrifugation (30 min at 40,000 x g), the crude extract was incubated with Benzonase (Novagen) for 1 hr at room temperature and then heat-fractionated for 10 min at 75˚C. The supernatant was purified by using Talon metal affinity resin.
Samples to be digested by Bgus and Xylsa were exchanged into 150 mM sodium acetate (pH 5.5) solution and mixed with Bgus (0.45 U) and/or Xylsa (0.09 U), or no enzymes, and incubated overnight at 65˚C. Samples were then run on SDS-PAGE, transferred to PVDF-FL (Millipore), and probed with anti-a-DG core antibody (AF6868) and anti-a-DG glycan antibody (IIH6). Enriched rabbit a-DG (100 mL of the 150 mM sodium acetate (pH 5.5) solution) was mixed with Bgus (0.45 U) and/or Xylsa (0.09 U), or no enzymes, and incubated overnight at 65˚C. Samples were then run on SDS-PAGE, transferred to PVDF-FL (Millipore), and subjected to immunoblotting.

Solid-phase assay
Solid-phase assays were performed as described previously Goddeeris et al., 2013). Briefly, WGA eluates were diluted 1:50 in TBS and coated on polystyrene ELISA microplates (Costar 3590) overnight at 4˚C. Plates were washed in LBB and blocked for 2 hr in 3% BSA/LBB at RT. The wells were washed with 1% BSA/LBB and incubated for 1 hr with L9393 (1:5000 dilution) in 3% BSA/LBB followed by incubation with Horseradish Peroxidase (HRP)-conjugated anti-rabbit IgG (Invitrogen, 1:5000 dilution) in 3% BSA/LBB for 30 min. Plates were developed with o-phenylenediamine dihydrochloride and H 2 O 2 , and reactions were stopped with 2 N H 2 SO 4 . Absorbance per well was read at 490 nm by a microplate reader.

Statistics
The included Shimadzu post-run software was used to analyze POMK, LARGE1, and B4GAT1 activity in fibroblasts and mouse skeletal muscle, and the percent conversion to product was recorded. The means of three experimental replicates (biological replicates, where each replicate represents a different pair of tissue culture plates or animals, i.e. control and knockout) were calculated using Microsoft Excel, and the mean percent conversion to product for the WT or control sample (Control human fibroblasts or Pomk LoxP/LoxP skeletal muscle, respectively) reaction was set to 1. Percent conversion of each experimental reaction was subsequently normalized to that of the control, and statistics on normalized values were performed using GraphPad Prism 8. For analysis of POMK and LARGE1 activity in fibroblasts and mouse skeletal muscle, Student's t-test was used (two-sided). Differences were considered significant at a p-value less than 0.05. Graph images were also created using GraphPad Prism and the data in the present study are shown as the means + / -SD unless otherwise indicated. The number of sampled units, n, upon which we report statistics for in vivo data, is the single mouse (one mouse is n = 1).
For measure of POMK activity in HAP1 cells, the percent conversion from GGM-MU to GGM(P)-MU was first calculated using the included Shimadzu analysis software. The means plus standard deviations of the percent conversion to GGM(P)-MU for three experimental replicates was calculated using GraphPad Prism 8. One-way ANOVA with the Dunnett's Method for Multiple Comparisons was performed, and the data for the POMK KO HAP1 sample set as the control. Differences were considered significant at a p-value less than 0.05. Graph images were created in GraphPad and show mean + / -SD.
To measure POMK activity in control and NH13-284 skeletal muscle, we only performed one experimental replicate due to the limited amount of sample available. To measure B4GAT1 activity, two technical replicates were performed from skeletal muscle. Protein concentration from control and NH13-284 skeletal muscle was also measured using two technical replicates. The percent conversion to product for the B4GAT1 reaction was divided by the protein concentration, and the values for these two technical replicates graphed using GraphPad Prism 8. The graph reported is shown as the mean + / -SD.
For flow cytometry analyses, six experimental replicates were performed, and the mean fluorescence intensity (MFI) reported. Statistics were performed using the Student's unpaired t-test, twosided in GraphPad Prism eight and the values reported as mean + / -SD.

NMR spectroscopy
1D 1 H NMR spectra of the core M3 trisaccharides GGM-MU and GGMp-MU in the absence and presence of POMK or LARGE1 were acquired at 25˚C on a Bruker Avance II 800 MHz NMR spectrometer equipped with a sensitive cryoprobe by using a 50 ms T 2 filter consisting of a train of spinlock pulses to eliminate the broad resonances from the protein (Mayer and Meyer, 2001). Danio rerio POMK titrations were performed in 25 mM Tris (pH 8.0), 180 mM NaCl, and 10 mM MgCl 2 in 98% D 2 O. LARGE1 titrations were performed in 20 mM HEPES, 150 mM NaCl, pH 7.3 in 90% H 2 O/ 10% D 2 O. The 13 C and 1 H resonances of the trisaccharides were reported previously (Yoshida-Moriguchi et al., 2010). The 1 H chemical shifts are referenced to 2,2-dimethyl-2-silapentane-5-sulfonate. The NMR spectra were processed using NMRPipe (Delaglio et al., 1995) and analyzed using NMRView (Johnson and Blevins, 1994). The glycan binding affinity to POMK and LARGE1 was determined using glycan-observed NMR experiments as described previously (Briggs et al., 2016). For the resolved anomeric trisaccharide peak, the bound fraction was calculated by measuring the difference in the peak intensity in the absence (free form) and presence (bound form) of POMK or LARGE1, and then dividing by the peak intensity of the free form. To obtain dissociation constant, the data were fitted to the standard quadratic equation using GraphPad Prism (GraphPad Software). The standard deviation from data fitting is reported.

Mass spectrometry
In order to generate DG fusion proteins for MS analyses, HAP1 cells were grown in IMDM with 10% FBS and 1% penicillin/streptomycin on p150 plates. When plates were 80% confluent, cells were washed twice with DPBS, media changed to serum-free IMDM with 1% penicillin/streptomycin (Invitrogen), and cells infected at high MOI (250-1000) of adenovirus expressing DG390TEVHis. Three days later, the medium was harvested and stored at 4˚C until samples were ready for MS analysis.
Reductive elimination. Glycans were reductively eliminated from DG390 proteins and purified on a 50WS8 Dowex column, and the purified glycans were subjected to permethylation and purified according to published methods (Jang-Lee et al., 2006;Zhang et al., 2014). Briefly, the freezedried DG390 sample was dissolved in 55 mg/mL potassium borohydride in 1 mL of a 0.1 M potassium hydroxide solution. The mixture was incubated for 18 hr at 45˚C and quenched by adding five to six drops of acetic acid. The sample was loaded on the Dowex column and subsequently eluted with 5% acetic acid. The collected solution was concentrated and lyophilized, and excessive borates were removed with 10% methanolic acetic acid.
Permethylation. For the permethylation reaction, three to five pellets per sample of sodium hydroxide were crushed in 3 mL dry dimethyl sulfoxide. Methyl Iodine (500 mL) as well as the resulting slurry (0.75 mL) were added to the sample. The mixture was agitated for 15 min and quenched by adding 2 mL ultrapure water with shaking. The glycans were extracted with chloroform (2 mL) and washed twice with ultrapure water. Chloroform was removed under a stream of nitrogen. The permethylated glycans were loaded on a C18 Sep-pak column, washed with 5 mL ultrapure water and successively eluted with 3 mL each of 15, 35, 50, and 75% aq. acetonitrile. The solutions were collected and lyophilized. The lyophilized 35 and 50% fractions were dissolved in 50% aqueous solution of methanol and combined for MALDI analysis.
Mass spectrometry. A Bruker Autoflex III MALDI-TOF/TOF was used for acquisition of all MALDI MS data. An in-house made BSA digest was used to calibrate the MS mode. 3,4-diaminobenzophenone was used as the matrix. Permethylated samples were dissolved in 10 mL of methanol, and 1 mL of this solution was premixed with 1 mL matrix. 1 mL of this mixture was spotted on the plate. patients' serum and biopsy samples. We confirm that informed consent was provided to the patient and family regarding the nature of the genetic studies to be performed upon collection of samples and is available for our patient. Animal experimentation: Animal experimentation: This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) protocols of the University of Iowa (#0081122). The mass spectra in the article were zoomed into the range of m/z 750-950 to better present MS signals corresponding to core M3 glycan structures.

Decision letter and Author response
. Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript.