Neuronal Ceroid Lipofuscinoses type 7 (CLN7) is a paediatric lysosomal storage disease caused by mutations of the MFSD8 gene. Affected children have normal early development, but then suffer from progressive cognitive, motor, verbal, and visual decline. Ataxia and myoclonic epilepsy are predominant features of the condition, and there are no effective therapies. Death usually occurs by approximately age 11 years. While adeno-associated virus serotype 9 (AAV9) based gene therapy holds promise for treating monogenetic neurologic disorders, the impact of this intervention is limited by the maximum safe tolerable dose and the host immune response to the capsid and gene product. This study sought to confirm the safety of high dose intrathecal AAV-based gene therapy under a comprehensive immunosuppression regimen. This was a two-year open label, dose escalation, phase 1 first-in-human study of AAV9-based intrathecal gene therapy for CLN7. 4 participants (1 low dose, 3 high dose) were followed at regular intervals with blood work, CSF analysis, EEG, MRI, and measures of neurologic and neuropsychological function. This study provided evidence of safety for high dose intrathecal AAV9 based gene therapy in CLN7 disease under a specific immunosuppression regimen. Additionally, this study provides preliminary evidence of efficacy for this gene therapy. High dose intrathecal AAV based gene therapy can be pursued with adequate immunosuppression and monitoring for immune responses to the gene product. Additional long-term monitoring of the immune system during tapering of immunosuppression is needed to identify potential reactions to the gene product. This study was funded by The Batten's Hope Foundation, Mila's Miracle Foundation, Children's Health Dallas and Philanthropic Gifts to UT Southwestern. In addition, Emily R. Nettesheim received funding from NIH training grant 5T32GM131945-03 and Hamza Dahshi was supported in part by NIH award T32 GM152319.

CLN7 disease is a neurodegenerative lysosomal storage disorder caused by defects in MFSD8. We performed a comprehensive analysis of patient mutations causing CLN7 disease, variant late-infantile and non-syndromic adult phenotypes. Our analyses of protein expression and post-translational modifications, such as proteolytic cleavage and complex type N-linked oligosaccharide processing, along with double immunofluorescence analyses, demonstrated that the nonsense mutations p.Q206X, p.W456X, p.Q474X, and p.R482X, or the in-frame deletion mutation p.V109_I113del, resulted in decreased protein levels at steady state compared with wild type CLN7 and showed mistargeting and ER retention as the primary cause for loss of CLN7 function. We also investigated several missense mutations clustered in transmembrane domain 11 that affect conserved residues, which are believed to be important for CLN7 function. Analysis of protein levels, complex type N-glycosylation, proteolytic cleavage in lysosomes, and colocalization with lysosomal marker proteins in double immunofluorescence analyses showed that patient mutations p.T458L, p.R465Q, and p.R465W did not affect protein stability or correct lysosomal targeting of CLN7, indicating functional impairment. The missense mutation p.M454T resulted in increased cysteine protease-mediated turnover of mutant CLN7 in lysosomes. Using an assay to measure the generation of an enlarged endosome phenotype in cells overexpressing CLN7 carrying missense mutations, a loss of CLN7 function could not be detected. The effects of missense mutations in transmembrane domain 11 on CLN7 function remain to be investigated. In summary, our study revealed mistargeting and ER retention of nonsense and in-frame deletion mutations in MFSD8 as a cause of CLN7 disease, variant late-infantile phenotype.

CLN7 is a lysosomal storage disease caused by pathogenic variants in the MFSD8/CLN7 gene. Typically neurodegenerative, patients present seizures and developmental delay since 2-6 years of age and a rapid psychomotor, verbal, and visual deterioration that leads to premature death. However, 'atypical' cases have also been reported. Although more than 80 DNA variants in the MFSD8/CLN7 gene have been reported, no data about a genotype/phenotype correlation is available. Here, we analyze five 'classical' and 'atypical' CLN7 patients by molecular and computational methods. Four variants have been found: c.103C > T (p.Arg35*, pathogenic), c.1394G > A (p.Arg465Gln, pathogenic), c.863 + 1G > A (likely pathogenic), and c.863 + 4A > G (of uncertain significance). Both splice variants showed altering of the splicing process on a minigene reporter assay. Furthermore, exon 8 was deleted in the MFSD8/CLN7 cDNA of blood samples from two patients carrying the splicing variants, demonstrating their effect. The c.863 + 4A > G variant also showed a residual wildtype MFSD8/CLN7 expression and, thus, explaining the milder phenotype. Finally, a clustered geographical distribution of the c.103C > T and c.863 + 4A > G variants was observed in the northeast and center of Argentina, respectively. Our data confirm the pathogenicity of the c.863 + 1G > A variant and reclassify the c.863 + 4A > G variant as pathogenic by adding experimental data, offering new information for a precise prognosis, and expanding the genetic and epidemiological spectrum of CLN7 in the South American region. Ultimately, we seek to raise awareness about the existence of this pathology in the region to reduce the so-called 'diagnostic odyssey' in pediatric patients.

Neuronal ceroid lipofuscinosis type 7 (CLN7) is a devastating paediatric neurodegenerative disorder with no cure and limited natural history data to guide therapeutic development. Here, we present the first multimodal characterization of prodromal and early-stage CLN7 disease in Japanese macaques carrying a spontaneous CLN7 -/- mutation. Using structural T 2 -weighted MRI for volumetry, [ 18 F]FDG PET for glucose metabolism, and [ 11 C]PBR28 PET for neuroinflammation, we observed region-dependent patterns in volumetric, molecular, and metabolic alterations. MRI confirmed the presence of disease-associated atrophy in many cortical and subcortical brain regions, consistent with human pathology, [ 18 F]FDG PET revealed early cortical and subcortical widespread hypometabolism, and [ 11 C]PBR28 PET imaging detected progressive neuroinflammation in the same brain areas. CSF analyses further showed age-dependent increases in neurofilament light (NfL), providing convergent evidence for neurodegeneration. Together, these results define a prodromal trajectory in CLN7 disease, establish sensitive imaging and fluid biomarkers, and validate the macaque model as a powerful platform for testing interventions.

Adult-onset recessive cerebellar ataxias comprise a heterogeneous group of disorders. To describe a founder MFSD8 variant in adult-onset cerebellar ataxia. We describe three unrelated Indian patients and one sibling pair (n = 5; median age 31 years) who exhibited progressive gait ataxia, limb dysmetria, titubation, gaze-evoked nystagmus, hypermetric saccades, and brisk reflexes without seizures, cognitive decline, visual impairment, or autonomic dysfunction. Brain MRI revealed moderate cerebellar atrophy. They were all homozygous for MFSD8 c.935T>C(p.Ile312Thr) presenting with pure cerebellar syndrome in the third decade. This ultra-rare p.Ile312Thr variant was predicted as deleterious by in silico tools and absent in homozygous state in population databases. Runs of homozygosity indicated a shared ~1.3-12Mbp haplotype with a common ancestor ~620 years ago. These findings expand MFSD8-related CLN7 disease to include adult-onset isolated ataxia and support inclusion of MFSD8 in adult ataxia gene panels, particularly in South Asian populations.