Mechanism of disease and therapeutic rescue of Dok7 congenital myasthenia

Congenital myasthenia (CM) is a devastating neuromuscular disease, and mutations in DOK7, an adaptor protein that is crucial for forming and maintaining neuromuscular synapses, are a major cause of CM1,2. The most common disease-causing mutation (DOK71124_1127 dup) truncates DOK7 and leads to the loss of two tyrosine residues that are phosphorylated and recruit CRK proteins, which are important for anchoring acetylcholine receptors at synapses. Here we describe a mouse model of this common form of CM (Dok7CM mice) and a mouse with point mutations in the two tyrosine residues (Dok72YF). We show that Dok7CM mice had severe deficits in neuromuscular synapse formation that caused neonatal lethality. Unexpectedly, these deficits were due to a severe deficiency in phosphorylation and activation of muscle-specific kinase (MUSK) rather than a deficiency in DOK7 tyrosine phosphorylation. We developed agonist antibodies against MUSK and show that these antibodies restored neuromuscular synapse formation and prevented neonatal lethality and late-onset disease in Dok7CM mice. These findings identify an unexpected cause for disease and a potential therapy for both DOK7 CM and other forms of CM caused by mutations in AGRIN, LRP4 or MUSK, and illustrate the potential of targeted therapy to rescue congenital lethality.

Congenital myasthenia is a group of diseases caused by mutations in genes that are important for the formation, function, and maintenance of neuromuscular synapses 1,2 . Mostly, mutations in these genes are recessive and diminish gene activity, thereby causing synaptic deficits that lead to early onset structural and functional deficits in the neuromuscular synapse, which are responsible for muscle weakness throughout life.
The formation and maintenance of neuromuscular synapses requires the assembly of highly specialized presynaptic and postsynaptic membranes, which involves the coordinated action of several key molecules [3][4][5] . AGRIN, which is released from motor nerve terminals, binds to the lipoprotein receptor-related protein 4 (LRP4) in muscle, stimulating the formation of a complex between LRP4 and muscle-specific kinase (MUSK), a receptor tyrosine kinase that acts as a master regulator of synaptic differentiation [4][5][6][7][8][9] . LRP4, once clustered in the postsynaptic membrane as a consequence of MUSK activation, also signals directly back to motor axons to stimulate presynaptic differentiation 10 . Mutations in AGRIN, LRP4 and MUSK, as well as in the genes that encode subunits of acetylcholine receptors (AChRs), also cause CM 2,11 .
Activation of MUSK also depends on the adaptor protein DOK7 12 . Mutations in Dok7 are responsible for 10-20% of all cases of CM [13][14][15] . The disease is debilitating-causing weakness in limb, neck and facial muscles-and one-quarter of patients with DOK7 CM require non-invasive ventilation at some point during their lifetime. Few treatments abate the clinical symptoms 16 . The N-terminal region of DOK7 contains pleckstrin homology (PH) and phosphotyrosine-binding (PTB) domains (Fig. 1a), which function to dimerize DOK7 and bind a phosphorylated tyrosine motif in the MUSK juxtamembrane ( JM) region 17 . A failure of DOK7 to bind MUSK leads to a failure of AGRIN to stimulate MUSK phosphorylation 12,18 , demonstrating that DOK7 is essential to stabilize phosphorylation of MUSK, probably by promoting its dimerization 19 . In addition, AGRIN-stimulated MUSK phosphorylation leads to phosphorylation of two tyrosine residues in the C-terminal region of DOK7, which triggers the recruitment of CRK and CRK-L-proteins that participate in the clustering of AChRs 15,20 .
The most common cause of Dok7 CM is a four-base-pair duplication (residues 1124-1127, TGCC), which leads to a frameshift and premature termination of DOK7 13,21 . Some individuals with Dok7 CM are homozygous for this mutant allele, whereas others carry this mutant allele in combination with a different mutant allele of Dok7. The truncated form of DOK7 retains the PH and PTB domains and binds to the tyrosine-phosphorylated JM region of MUSK 13 , but lacks the two tyrosine residues that are phosphorylated and recruit CRK proteins, suggesting that the loss of these tyrosine residues is responsible for the synaptic deficits in this common form of Dok7 CM 2,15,20 .

Dok7 CM lowers DOK7 levels and MUSK phosphorylation
To determine how loss of the DOK7 C-terminal region caused the synaptic defects, we measured expression of Dok7 mRNA and truncated DOK7 protein in Dok7 CM mice using antibodies against the DOK7 PTB domain that detected the truncated and wild-type proteins equally (Extended Data Fig. 2). Dok7 mRNA levels were normal in muscle from Dok7 CM mice, whereas the truncated DOK7 protein was expressed at threefold lower levels than the wild-type DOK7 protein (Fig. 1e, Extended Data Fig. 3). By contrast, DOK7(2YF) was expressed at normal levels ( Fig. 1f) and, as expected 15,20 , was not tyrosine phosphorylated (Extended Data Fig. 4).
Because DOK7 functions as a dimer to dimerize MUSK, thereby stabilizing MUSK tyrosine phosphorylation 19 , we determined whether MUSK tyrosine phosphorylation was diminished in Dok7 CM mice. MUSK phosphorylation was reduced sevenfold in Dok7 CM mice but was normal in Dok7 2YF mice (Fig. 1g, h).

CRK proteins are recruited directly to MUSK
We anticipated that recruitment of CRK proteins to the synapse would be absent or severely reduced in both Dok7 CM and Dok7 2YF mutant mice. Indeed, CRK recruitment to the synapse and to the MUSK complex was substantially diminished (2.8-fold) in Dok7 CM mice (Fig. 2a, b), but to our surprise was only modestly reduced (by 28%) in Dok7 2YF mice Plots show individual data and mean ± s.e.m.; *P < 0.05, ***P < 0.0005, ****P < 0.00005; two-sided Student's t-test.

Article
( Fig. 2a, b). These findings suggest that CRK is recruited to a tyrosine phosphorylated synaptic protein(s) in addition to DOK7. The three activation loop tyrosines and Y553 in MUSK become phosphorylated after stimulation by AGRIN 12,18,22,23 . We found that Y553 in the MUSK JM region is within not only a PTB-binding site that recruits DOK7, but also a potential SH2-binding motif for CRK proteins (Fig. 2c). Both CRKI and DOK7 bound the MUSK JM site in a phosphorylation-dependent manner (Fig. 2d). Mutation of amino acids that compose the SH2-binding motif, but not the PTB-binding site, impaired binding of CRKI (Fig. 2d). Thus, CRK can bind not only to the phosphorylated C-terminal region of DOK7 but also directly to the tyrosine-phosphorylated JM region of MUSK. This redundancy for recruiting CRK to the synapse is likely to explain the near-normal association of CRK with the MUSK complex in Dok7 2YF mice and the difference in the phenotypes of Dok7 CM and Dok7 2YF mice.

Agonist antibodies against MUSK
If diminished MUSK phosphorylation caused disease in Dok7 CM, we reasoned that stimulation of MUSK might rescue the synaptic defects and overcome lethality. We explored this idea by generating agonist antibodies targeting MUSK and treating Dok7 CM mice with the agonist antibodies.
We screened a phage-display library for antibodies that bound the Fz-like domain in the extracellular region of mouse and human MUSK. We targeted the Fz-like domain because this domain is not essential for MUSK function and antibodies against the Fz-like domain cause no obvious harm in mice 24,25 .
We identified high-affinity antibodies that bound the Fz-like domain in human and mouse MUSK and stimulated MUSK phosphorylation independent of AGRIN ( Fig. 3a, b, Extended Data Fig. 5), thereby overcoming the shortcomings of previously described agonist antibodies that recognized mouse but not human MUSK 25 .
We injected the MUSK agonist antibody X17, in a mouse IgG2a format with mutations that reduce Fc domain effector function 26 , interperitoneally into wild-type mice and found that 10 mg kg −1 of X17 had a half-life of 5 days in blood and that it saturated synaptic MUSK (Fig. 3c, Extended Data Fig. 6a). Chronic injection of X17 (10 mg kg −1 at postnatal day 4 (P4), P24 and P44) in wild-type mice over two months had no effect on the organization of neuromuscular synapses, weight gain or motor behaviour (Extended Data Fig. 6).

X17 rescues synapse formation and lethality
Although Dok7 CM mice on a C57BL/6 background died at birth, we found that Dok7 CM mice on a mixed genetic background survived for one to two weeks after birth (Extended Data Table 1), which facilitated experiments to study the therapeutic efficacy of X17. We injected these Dok7 CM mice with 10 mg kg −1 of X17 or an isotype-matched negative control antibody at P4, when they showed signs of disease (they were runted and had synaptic deficits) (Extended Data Fig. 7). Dok7 CM mice injected with the control antibody continued to lose weight and died within a week (Fig. 4a, b). Injection of X17 reversed the weight loss and rescued the Dok7 CM mice from this early lethality (Fig. 4a, b). Over the next few weeks, the weight gain was continuous in nine of the twelve Dok7 CM mice injected with X17; the weight gain slowed in three of the X17-injected mice, and they died at P23-P24. Another antibody, X3, rescued Dok7 CM mice from early postnatal lethality when injected at 20 mg kg −1 but not at 10 mg kg −1 (Extended Data Fig. 8), suggesting that a higher initial dose of a MUSK agonist antibody may be more effective during early postnatal development, when synapses are undergoing critical steps in maturation.
We investigated whether chronic dosing could lead to long-term survival. Repeated injections of X17 in the nine surviving Dok7 CM mice at P24 and P44 led to survival of these mice for at least two months ( Fig. 4a, b), at which point we assessed their motor performance and the structure of their neuromuscular synapses. X17 rescued synapse formation and maturation, as the neuromuscular synapses of these mice had developed the complex pretzel-like shape characteristic of fully mature mouse neuromuscular synapses (Fig. 4c). Moreover, X17 rescued the recruitment of CRK proteins to the neuromuscular synapse (Fig. 4d).
Antibody X17 rescued the motor function of Dok7 CM mice, as assessed by grip strength and rotarod assays (Fig. 4e). Moreover, Dok7 CM mice injected with X17 were fertile and produced offspring at the expected frequency. Together, these findings indicate that reduced MUSK tyrosine phosphorylation is central to disease in Dok7 CM mice. Even if the C-terminal region of DOK7 has an additional role in synapse formation, this function can be overridden by stimulating MUSK.

Therapeutic reversal in adult Dok7 CM mice
We next sought to determine whether X17 could reverse neuromuscular deficits that develop during adulthood, a question particularly relevant to developing a human therapy as DOK7 CM in humans would probably be treated during adult life. We treated Dok7 CM mice with X17 either at P4, P24 and P44 or at P4 and P18 but then discontinued antibody treatment. Both groups of Dok7 CM mice continued to maintain their weight and mobility for 2-3 months (Fig. 4f), indicating that the effects of the antibody lasted for longer than its lifetime in the blood. However, mice ultimately began to lose weight and display motor deficits ( Fig. 4f, g, Supplementary Video 1). When the mice were losing weight at a rate of about 0.4 g per day, we injected X17 once again and monitored their weight and mobility. Two days after the resumption of X17 treatment, the Dok7 CM mice began to regain weight (by about 0.4 g per day over the next week) (Fig. 4f). Within one week of reinitiating antibody treatment, the motor performance of the Dok7 CM mice had been restored (Fig. 4g, Supplementary Video 1). The mice continued to gain weight and their motor performance continued to improve for at least one additional week after antibody treatment, when the experiment was ended (Fig. 4f, g).

Discussion
Stimulation of MUSK with an agonist antibody rescued synapse formation and motor function, prevented lethality and allowed Dok7 CM mice to thrive postnatally as fertile adults. Moreover, the motor deficits that developed in adult Dok7 CM mice after withdrawal of antibody treatment were readily reversed by reinitiating antibody treatment. Thus this therapeutic strategy, which avoids the complex requirements for gene therapy 27 , might be beneficial for humans with DOK7 CM or other neuromuscular diseases.
Most previous studies of DOK7 have relied upon analysis of transfected muscle and non-muscle cells that overexpress DOK7 12,15,20 . In this context, which bypasses the normal requirement for AGRIN and LRP4 to stimulate MUSK, the in vivo consequences of Dok7 mutations might have been masked by the overexpression of DOK7.
Inbred C57BL/6 mice containing the Dok7 CM mutation showed more severe functional deficits than humans with the same mutation. The mutant phenotype was less severe in mice with a mixed genetic background, as outbred mice survived for up to three weeks postnatally, whereas inbred mutant mice died at birth. Modifiers in the hybrid strains may lessen disease severity, or C57BL/6 mice may contain genes that worsen the phenotype. In either case, the modestly prolonged lifespan of Dok7 CM mice on the mixed background offers a mouse model that presents a longer temporal window in which to assess therapeutic approaches.
These experiments demonstrate full rescue from congenital lethality by targeted therapy. Our findings point to an unforeseen therapeutic approach, as this strategy does not directly target the mutant protein but rather targets a wild-type protein that has diminished activity caused by the mutation of an upstream gene, in this case DOK7. Epistatic rescue in this way could also provide therapy for CM caused by mutations in AGRIN, LRP4 or MUSK, in addition to DOK7, as well as for other neuromuscular diseases. Moreover, this strategy has the potential for widespread use to treat genetic disorders in humans for which the disease mechanism is understood and suitable targets have been identified.

Online content
Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41586-021-03672-3.  Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
An earlier study described a similar mouse model, generated using classic embryonic stem cell gene targeting, for this common form of DOK7 CM 27 . Although the lethality of these mutant mice could be rescued by an adenoviral-associated vector expressing wild-type DOK7, establishing a gene therapy approach to treat DOK7 CM 27 , this study did not examine the cause of disease in the Dok7 1124_1127,dup mouse model.
To generate Dok7 2YF mice, we injected an sgRNA (5′ TTCGAGGTGTG TCATAG 3′, 15 ng/μl) and in vitro-transcribed Cas9 RNA (30 ng/μl), together with the DNA repair template (5′ ATGCCGGCAGCAACCTGGA CGTGTGGCGGGCCGGTGAGGAATTCGGTTCTCTGCTCAGTCTGCCTG CCCCCTGGAGCCAGCGCACCTGAGCCCAGACTGTGTGCCTGCCCACC TGGGGCGGCCGAGTA 3′, 30 ng/μl) to convert tyrosine 396 and tyrosine 406 to phenylalanine, into the cytoplasm of C57BL/6 mouse zygotes. We analysed 33 mice that were born from injected zygotes by sequencing tail DNA (primer: 5′ TGGCATTGCCACAGGCAG 3′). One mouse carried a Dok7 allele with the desired tyrosine-to-phenylalanine substitutions. We crossed the founder mouse with wild-type C57BL/6 mice to generate the Dok7 2YF line. DNA sequencing from these lines confirmed the sequence of the Dok7 mutation. Mice were housed and maintained according to Institutional Animal Use and Care Committee (IACUC) guidelines.

Growth of cultured cells
C2C12 mouse muscle cells, purchased from and authenticated by ATCC (Cat CRL-1772), were grown at 37 °C in growth medium (GM): Dulbecco's modified Eagle's medium (DMEM) containing 4.5 g/l glucose, l-glutamine and sodium pyruvate (Corning cellgro), supplemented with 10% fetal bovine serum (FBS; GemCell). Myoblast fusion and myotube differentiation were induced when myoblasts were 70% confluent by switching to differentiation medium (DM): DMEM with 4.5 g/l glucose and 1 mM l-glutamine, supplemented with 2% heat-inactivated horse serum. Immortalized myoblasts were isolated from wild-type and Dok7 2YF embryos and grown as described previously 29 .
HEK 293 cells were purchased from and authenticated by ATCC (ATCC Cat CRL-1573). Cells were grown at 37 °C in the same medium as described above for C2C12 myoblasts and transfected using Lipofectamine 3000 Transfection Reagent Kit (Thermofisher Scientific). All cell lines tested negative for mycoplasma contamination using the e-Myco Plus PCR detection kit.

Antibody treatment of C2 myotubes
Three days after C2C12 myotubes had formed, the cultures were treated for 30 min with 10 nM biotinylated Fabs in complex with 2.5 nM streptavidin, 10 nM IgGs, or 0.5 nM recombinant neural AGRIN-B8 (R&D Systems). Myotubes were homogenized at 4 °C in lysis buffer (50 mM sodium chloride, 30 mM triethanolamine, pH 7.5, 50 mM sodium fluoride, 5 mM EDTA, 5 mM EGTA, 2 mM sodium orthovanadate, 1 mM N-ethylmaleimide, 1 mM sodium tetrathionate, 10 μM pepstatin, plus complete protease inhibitor mix (Roche)). NP-40 was added to a final concentration of 1%, and the extract was incubated with rocking for 30 min at 4 °C. Insoluble proteins were removed by centrifugation at 12,000 rpm for 20 min at 4 °C. The supernatant was precleared for 1 h at 4 °C with Protein G-agarose beads (Sigma-Aldrich) before incubation overnight at 4 °C with antibodies against MUSK (MUSK 1A) 30 . Complexes were incubated for 4 h with Protein G-agarose beads. The beads were subsequently washed (three times for 9 min) in lysis buffer containing 1% NP-40. Proteins were eluted from the beads with 1% SDS in lysis buffer.

Isolation of MUSK and DOK7 from muscle
Whole leg muscles or cultured muscle cells were homogenized at 4 °C in lysis buffer (50 mM sodium chloride, 30 mM triethanolamine pH 7.5, 50 mM sodium fluoride, 5 mM EDTA, 5 mM EGTA, 2 mM sodium orthovanadate, 1 mM N-ethylmaleimide, 1 mM sodium tetrathionate, 10 μM pepstatin, plus complete protease inhibitor mix (Roche)). NP-40 was added to a final concentration of 1%, and the extract was incubated with rocking for 30 min at 4 °C. Insoluble proteins were removed by centrifugation at 12,000 rpm for 20 min at 4 °C. The supernatant was pre-cleared for 1 h at 4 °C with Protein G-agarose beads (Sigma-Aldrich) before overnight incubation at 4 °C with antibodies against MUSK (MUSK 1A) 30 or goat anti-DOK7 (R&D Systems, AF 6398), followed by incubation for 4 h with Protein G-agarose beads. The beads were subsequently washed (three times for 9 min) in lysis buffer containing 1% NP-40. Proteins were eluted from the beads with 1% SDS in lysis buffer.

Western blotting
Proteins were fractionated by SDS-PAGE and transferred to PVDF membranes. Blots were probed with antibodies against MUSK (R&D Systems, AF562), phosphotyrosine (Millipore, 05-321) or DOK7 (1916), as described previously [18][19][20]24 . Antibodies against CRK (BD Bioscience, 610035) and CRK-L (Santa Cruz Biotechnology, sc-365092) have been described previously 20 . We quantified the band intensities with a ChemiDoc imaging system (BioRad), as described previously 24 . The graphs show the mean values from at least three separate experiments. A two-sided Student's t-test was used to determine statistical significance and was conducted using GraphPad Prism 9.0 software.

Whole-mount muscle immunohistochemistry
Diaphragm muscles were dissected from E18.5 embryos and postnatal mice in oxygenated L-15 medium. The muscles were pinned onto Sylgard-coated dissection dishes, fixed for 1.5 h in 1% PFA and blocked for 1 h in PBS with 3% BSA (Sigma IgG free) and 0.5% Triton X-100 (PBT). Diaphragm muscles were stained with Alexa 488-conjugated anti-BGT (Invitrogen) to label AChRs and with antibodies against neurofilament-L (Synaptic Systems, 171002), β-TUBIII (Synaptic Systems 302302) or synapsin 1/2 (Synaptic Systems, 106002) to label motor axons and nerve terminals, respectively 31 . The antibodies were force-pipetted into the muscle, and the muscles were incubated overnight at 4 °C on an orbital shaker in a humidified chamber. Diaphragm muscles were washed 10 times over the course of 5 h with PT (PBS with Triton X-100) at room temperature and rinsed in PBS before the muscle was whole-mounted in 50% glycerol. Muscles from at least three mice of each genotype were analysed for each experiment. Images were acquired with a Zeiss LSM 800 confocal microscope using ZEN software. Adjustments to detector gain and laser intensity were made to avoid saturation. The number and size of synapses, the density of synaptic AChRs, the width of the endplate zone, the extent of denervation and the co-localization index (synapsin/AChRs) were quantified using FIJI/ImageJ software, as described previously 32 . A two-sided Student's t-test was used to determine statistical significance and was conducted using GraphPad Prism 9.0 software.

Article
Staining single muscle fibres Tibialis anterior muscles were dissected in oxygenated L-15 medium, pinned to a Sylgard-coated dish and fixed in 2% PFA (in PBS) for 2 h. After several rinses in PBS, one to three myofibres were manually teased with fine forceps. Fixed myofibres were blocked for 2 h at room temperature in PBS containing 5% BSA, 1% normal goat serum, and 0.04% saponin. Fibres were then incubated with primary antibodies overnight at 4 °C, washed three times for 5 min with PBS containing 0.04% saponin, incubated with secondary antibodies for 2 h at room temperature, washed again, and mounted in VectaShield (Vector Laboratories). Antibodies against CRK-L (Santa Cruz Biotechnology, sc-365092) were used and the postsynaptic membrane was visualized by staining with Alexa Fluor 488-anti-BGT (Invitrogen).

Cryosection immunohistochemistry
Limb muscles were embedded in optimal cutting temperature (OCT) medium and frozen on a dry-ice platform. Ten-micrometre sections, collected onto poly-l-lysine-coated glass slides, were fixed in 1-4% PFA for 10 min, washed in PBS with 3% BSA (PB) three times for 5 min, permeabilized with PB + 0.5% X-Triton (PBT) for 10 min, washed in PB and incubated overnight at 4 °C with primary antibodies against CRK-L (Santa Cruz Biotechnology, sc-365092) in PBT in a humidified chamber. Sections were washed in PB three times for 5 min before overnight incubation at 4 °C with secondary antibodies and Alexa Fluor 488anti-BGT (Invitrogen), diluted in PBS, in a humidified chamber. Sections were washed three times for 5 min in PB, then PBS, before mounting in Vectashield anti-fade mounting medium.

Behaviour
All-limb grip strength was measured using a grip-strength apparatus (Bioseb). Mice were allowed to grip the grid with both forelimbs and hindlimbs, and the mouse was pulled back steadily, until the mouse lost grip on the grid. The grip strength meter digitally displayed the maximum force applied (in grams) as the grasp was released. The mean measurement from six consecutive trials was taken as an index of all-limb grip strength. Mice were given an interval of 10-15 s between trials. Body weight was determined after all grip-strength measurements to analyse for potential co-variability. To enhance the robustness and reliability of the grip-strength assessment, all measurements were taken by the same experimenter 33 .
Motor function of male and female mice at P60 was assessed on a rotarod (AccuRotor four-channel, Omnitech Electronics). Mice were placed on the rotarod (3.0-cm rotating cylinder) rotating at 2.5 rpm, and the speed of rotation was increased linearly to 40 rpm over the course of 5 min. The time to fall from the rod was measured. Each mouse was subjected to three trials with 5-min intervals, and we recorded the longest latency to fall from the three trials. A two-sided Student's t-test was used to determine statistical significance and was conducted using GraphPad Prism 9.0 software.

Development of synthetic antibodies
The full-length extracellular region (E22 to T494 of mouse MUSK and E22 to T495 of human MUSK), including the Fz domain and the C-terminal flanking sequence (D307 to T494 of mouse MUSK and K314 to T495 of human MUSK) were expressed as a C-terminal fusion with the Avi and His 6 tags using the secretion signal sequence of mouse IgkVIII in EXPI293 cells with the ExpiFectamine 293 Transfection kit (Thermo Fisher Scientific) using standard procedures provided by the vendor. The proteins were purified from the filtered culture supernatant using a HiTrap Nickel column (GE Healthcare) and biotinylated in vitro using the BirA enzyme in the presence of 0.5 mM biotin and 10 mM ATP. The biotinylated proteins were further purified using a Superdex S75 10/300 column (GE Healthcare).
Sorting of an antibody phage-display library was performed as described previously 34 . In brief, a phage-display library was first sorted with all four antigens at 100 nM in the first round, followed by sorting with a single antigen at 100, 50 and 20 nM in the second, third and fourth rounds, respectively. To enrich for clones that bind to both human and mouse Fz domains, we used multiple sorting strategies in which alternate antigens were used in successive rounds (for example, human Fz; mouse ECD; human ECD). Individual clones were screened using phage enzyme-linked immunosorbent assay (ELISA) with the four antigens 34 , and the DNA sequences of clones bound to all of the antigens were determined.
The Fab proteins with the Avi tag at the C terminus of the heavy chain of selected clones were produced from Escherichia coli and biotinylated as described previously 34 . The mouse IgG2a-LALAPG sample of clone X17 was produced using a modified version of the pFUSE-mIgG2a-Fc vector (InvivoGen) containing the LALAPG mutations in the Fc region 26 and human CH1 domain and the pFUSE-CLIg vector (InvivoGen). This chimeric antibody consisted of a human Fab and mouse Fc sequences. In addition, we exchanged the mouse Fc sequences with those from human IgG1, containing LALA mutations, to generate hIgG1-X17, hIgG1-X2 and hIgG1-X3 antibodies.

Affinity measurements
The affinities of antibody clones in the Fab and IgG formats were measured using a bead-binding assay [35][36][37] . A biotinylated human antigen protein was immobilized on Dynabeads M280 streptavidin beads (Thermo Fisher Scientific) by rapidly mixing 100 μl of tenfold diluted beads in PBSB (PBS containing 0.5% bovine serum albumin (BSA, GeminiBio)) and 100 μl of 50 nM protein. The beads were then blocked with 2 μM biotin, washed twice with PBSB and resuspended in 1 ml PBSB. This reaction was appropriately scaled for the number of measurements when necessary. Five microlitres of the diluted beads and 20 μl of an antibody sample were mixed in a well of a 96-well polypropylene plate (Greiner Bio-One, catalogue number 650261) and incubated at room temperature for 30 min with gentle shaking. Samples were transferred to the wells of a 96-well filter plate (Millipore MultiScreen HTS HV, 0.45 mm, Thermo Fisher); the liquid was removed using a vacuum manifold and the wells were washed three times with 200 μl ice-cold PBSB using the vacuum manifold. The beads were stained with anti-human Fab antibody labelled with Alexa Fluor 647 ( Jackson Immuno Research, Alexa Fluor 647 AffiniPure Goat Anti-Human IgG, F(ab′) 2 fragment specific, 109-605-097). Following washing, the beads were suspended in 70 μl PBSB and analysed using an iQue screener (Sartorius) or an Intellicyt HTFC system. The resulting titration curves were analysed by nonlinear least-squared fitting of a 1:1 binding model using the GraphPad Prizm software.

Half-life of antibody in blood
Mice were injected intraperitoneally with antibodies. Mouse blood samples were centrifuged, and supernatants were diluted 2,000-fold in PBSB. Antibody levels were measured using the bead assay described above except that the binding reaction was performed at 4 °C. The half-life was determined by nonlinear least squares fitting of the median fluorescence intensities with a single exponential curve.
Phosphopeptide pull-down assay HEK 293 cells were transfected with plasmids encoding HA-tagged DOK7 and HA-tagged CRKI at 37 °C for 48h (Lipofectamine 3000, Thermofisher Scientific). After 48 h, the transfected cells were homogenized at 4 °C in lysis buffer; NP-40 was added to a final concentration of 1%, and the extract was incubated with rocking for 30 min at 4 °C. Insoluble proteins were removed by centrifugation at 12,000 rpm for 20 min at 4 °C. The supernatants were precleared for 1 h at 4 °C with streptavidin-agarose beads (Sigma-Aldrich).

Statistics and reproducibility
No statistical method was used to predetermine sample size. No data were excluded from the analyses. The experiments were not randomized. The investigators were not blinded to the genotype of the mice with the exception of the motor performance experiments.

Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this paper.

Data availability
Raw data generated from this study are available upon a reasonable request. Source data are provided with this paper.
Corresponding author(s): Steven J. Burden and Shohei Koide Last updated by author(s): 5/14/2021 Reporting Summary Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Research policies, see our Editorial Policies and the Editorial Policy Checklist.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.

n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one-or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A list of figures that have associated raw data -A description of any restrictions on data availability Raw data generated from this study are available upon a reasonable request.