Proteomics by mass spectrometry has rapidly matured from a niche method into a standard tool. The recent 10-year trajectory of single-cell proteomics has opened a new biological dimension for studying disease. Mitochondrial diseases, with their pronounced cell-to-cell heterogeneity, are particularly, well-suited to these methods. Here, we discuss how this approach can serve as an orthogonal functional layer for rare disease diagnostics. We trace the evolution of rare disease diagnostics from biochemical enzyme assays through genomics, transcriptomics, proteomics and metabolomics, highlighting incremental gains in diagnostic yield from individual omics layers and their integration. We discuss the limitations of bulk approaches in capturing the functional consequences of genetic perturbations, the new opportunities opened up by single-cell measurements and how spatial single-cell proteomics can further enrich the biological signal of affected cells in diagnostic tissues. We observe that the persisting diagnostic gap reflects not only technological limitations but, increasingly, challenges in data sharing and infrastructure as well as interpretive frameworks for functional molecular evidence. In this context, we consider opportunities for artificial intelligence and the ethical dimensions of single-cell proteomics in rare disease diagnostics. Finally, we propose a single-cell deep visual proteomics (scDVP) framework for clinical diagnostics of rare diseases with cell-to-cell variability, arguing that mitochondrial diseases are an ideal proof-of-concept.

Inherited metabolic disorders (IMDs) arise from defects in metabolic pathways essential for normal biochemical function. Although the pathogenic genetic variants are present at birth, clinical manifestations may not emerge until adolescence or adulthood. Symptoms may be more subtle compared to those of childhood onset IMDs and may have overlap with signs and symptoms of acquired disorders. This makes diagnosing IMDs in adulthood challenging. To identify how routine diagnostic care for adults suspected of having an IMD could be improved, we conducted a reanalysis study in a cohort of 31 adult participants. The participants, who remained undiagnosed after standard diagnostic care, were seen by a clinical geneticist, adult metabolic specialist and neurologist for deep phenotyping and data collection. Subsequently, exome sequencing data were reanalyzed and possible disease-causing variants were prioritized for further evaluation by a multidisciplinary team including both clinicians and geneticists (biochemical and molecular). This approach led to a (partial) diagnosis for nine participants from six families, giving a 25% diagnostic yield. We could distinguish four possible reasons why the identified disease-associated variants were not detected during routine diagnostic care. Beyond the expected impact of technical advancements and emerging knowledge of disease genes, we found that the most impactful factor was of organizational nature. Our findings underscore the value of reviewing all strong genetic variants within a multidisciplinary team setting, facilitating a bidirectional exchange of expertise. This approach enhances diagnostic accuracy, particularly when the underlying cause of the disease is not immediately apparent.

Male patients with the classical phenotype of Fabry disease (FD) are at risk of developing inhibiting antidrug antibodies (iADAs) against recombinant α-galactosidase-A, administered in the form of enzyme replacement therapy (ERT). The presence of ADAs is linked to infusion-associated reactions (IARs) and reduced ERT effectiveness. This study evaluates the incidence of iADAs and IARs in ERT treated classical male FD patients, their interrelatedness and the effect of an adjusted ERT initiation protocol, aimed at inducing immune tolerance. In this monocenter cohort study, a retrospective analysis was performed to evaluate the occurrence of iADAs and IARs. Prospectively, a stepwise dose escalation ERT initiation protocol was evaluated in classical male FD patients with a calculated high risk of iADA formation and compared to the standard initiation protocol. Retrospectively (n = 52), the proportion of patients experiencing IARs was higher in patients that developed iADAs (20/27 seropositive patients, 0/19 seronegative, p < 0.001). Higher iADA titers were associated with persisting iADA titers (p = 0.004). Prospectively (n = 17), none of the patients in the intervention group (stepwise dose escalation) experienced IARs (0/8), compared to half of the patients in the control group (5/9, p = 0.029). The proportion of patients developing iADAs was not different between the two groups (respectively, 4/8 and 6/9, p = 0.52). To conclude, iADA development is associated with the occurrence of IARs. The stepwise dose escalation at ERT initiation used in the current study in a single cohort seems to prevent IARs in male patients with classical FD, but does not seem to prevent iADA development.

Coenzyme A (CoA) biosynthesis is a conserved, dynamically regulated pathway essential for mitochondrial energy production, fatty acid oxidation, lipid biosynthesis and protein acylation. Biallelic variants in PANK2, PPCS, PPCDC, and COASY cause rare inborn errors of CoA biosynthesis, associated with markedly different clinical phenotypes: PANK2 and COASY defects predominantly cause neurological disorders within or adjacent to the neurodegeneration with brain iron accumulation (NBIA) spectrum, whereas PPCS and PPCDC deficiencies present mainly as severe early-onset dilated cardiomyopathy. However, COASY variants can also cause pontocerebellar hypoplasia and riboflavin-responsive lipid storage myopathy. This review examines these four disorders from a metabolic perspective, integrating clinical features, experimental models, biochemical data and emerging therapeutic approaches. Current evidence indicates that disease pathogenesis cannot be explained only by global CoA depletion. Total CoA levels may be reduced in PPCS and PPCDC deficiency, but are often preserved under basal conditions in PKAN and COASY-related models. Instead, impaired compartment-specific CoA handling and failure to sustain CoA-dependent flux under increased metabolic demand are emerging as central pathogenic concepts. Perturbation of fatty acid handling, acyl-CoA/acylcarnitine balance, mitochondrial function, iron homeostasis, protein acylation and 4'-phosphopantetheinylation may contribute to tissue-selective vulnerability. Therapeutic strategies are therefore likely to require disease-specific approaches, including precursor bypass or PANK activation where pathway flux can be restored, early pantethine supplementation in cardiomyopathic forms, and downstream or gene-directed strategies for COASY-related disorders. Understanding CoA as a regulator of metabolic adaptability provides a unifying framework for interpreting both shared mechanisms and disease divergence.

Phosphomannomutase deficiency (PMM2-CDG), the most common congenital disorder of glycosylation (CDG), is characterized by multisystem involvement and a lack of disease-modifying therapies. While previous transcriptomic studies have uncovered disrupted cellular pathways, the functional consequences of these alterations remain poorly understood. To further investigate PMM2-CDG pathophysiology, we integrated proteomic and metabolomic profiling of patient-derived fibroblasts with previously published transcriptomic data. Proteomic analysis was performed using Tandem Mass Tag-based mass spectrometry, while metabolomics was conducted via Nuclear Magnetic Resonance spectroscopy. Multi-omics integration was performed using principal component analysis-based dimensionality reduction, incorporating clinical metadata as supplementary variables. Proteomic analysis identified 43 significantly altered proteins, with enrichment in the retinoic acid synthesis pathway, wound healing, and cytoskeletal organization. Metabolomic profiling revealed altered amino acid levels and elevated concentrations of UDP-GlcNAc, consistent with perturbation of the hexosamine biosynthesis pathway, and increased levels of myo-inositol. Notably, myo-inositol levels showed a strong association with disease severity in the integrated analysis. RT-qPCR confirmed the upregulation of GFPT2. This integrative multi-omics study identifies consistent alterations in the retinoic acid synthesis pathway and hexosamine biosynthesis in PMM2-CDG patient-derived fibroblasts and reveals an association between intracellular myo-inositol levels and disease severity. These findings provide new insights into PMM2-CDG-associated molecular alterations and illustrate the value of multi-omics integration for hypothesis generation in rare diseases.

Glycogen storage disease type IIIa (GSDIIIa) causes progressive cardiomyopathy, and current high-fat dietary strategies lack consensus regarding long-term cardiovascular safety. We evaluated the efficacy and safety of high-protein versus high-fat diets in a novel cardiac-specific AGL knockout (CKO; AGL[flox/flox]/MHC-Cre) mouse model to specifically assess isolated cardiac responses. CKO mice were randomized at weaning to High-Protein (HPD), High-Protein High-Fat (HPHFD), or Low-Protein (LPD) diets, with approximate protein/fat/carbohydrate distributions of 40%/4%/50%, 40%/50%/10%, and 10%/4%/80%, respectively; CKO mice on normal diet (ND) and AGL[flox/flox] mice served as controls. Cardiac phenotypes and hepatic gluconeogenic enzymes were evaluated longitudinally up to 24 weeks. CKO-ND mice developed progressive cardiomyopathy with elevated myocardial glycogen at 24 weeks (32.01 ± 3.22 vs. 5.90 ± 2.21 mg/g in controls, p < 0.001), reduced left ventricular ejection fraction, and elevated serum creatine kinase. Both HPD and HPHFD significantly reduced myocardial glycogen burden (12.82 ± 3.58 and 15.07 ± 4.40 mg/g, p < 0.001), restored systolic function, and normalized hypertrophy. However, HPHFD induced distinct hyperlipidemia, whereas HPD maintained a stable lipid profile. Furthermore, therapeutic benefits in high-protein groups were associated with upregulated hepatic rate-limiting gluconeogenic enzymes (FBP2, PCK1). In this cardiac-specific AGL knockout model, a high-protein, non-high-fat diet attenuated cardiac glycogen accumulation and systolic dysfunction without the hyperlipidemia observed with the high-fat regimen. These preclinical findings support further evaluation of high-protein dietary strategies for GSDIIIa cardiomyopathy and suggest a possible liver-heart metabolic axis involving hepatic gluconeogenesis.

Phenylketonuria (PKU) is a rare metabolic disorder resulting from a mutation in the gene encoding the enzyme phenylalanine hydroxylase (PAH), resulting in very high phenylalanine (Phe) levels in blood and brain. A PKU mutant mouse model was developed via N-ethyl-N-nitrosourea (ENU) mutagenesis, mimicking the high brain Phe content seen in classic PKU patients. A large variety of behavioral tasks have been employed to explore the behavioral phenotype of the Pah[enu2] mouse. The purpose of this scoping review is to provide an overview of the behavioral performance of PKU mutant mice. We addressed three aims: (1) provide an overview of the behavioral tasks that have been performed with these mice, (2) summarize the reported findings across behavioral domains, (3) identify gaps and future perspectives for behavioral testing in the field of preclinical PKU. Peer-reviewed studies published in English up to May 2026 were identified through structured searches in PubMed and Web of Science, supplemented by free-text searches. Studies were eligible if they used adult untreated Pah[enu2] mice and included a comparison with wild-type controls. Twenty-two studies were included. Nest-building behavior and balance beam performance always reveal deficits in PKU mutant mice, whereas most studies found no significant anxiety-like phenotypes. A striking variation in cognitive test performance was found across studies, despite high Phe levels. In conclusion, balance beam and nest-building tests reliably demonstrate deficits in Pah[enu2] mice, while other cognitive findings remain variable. The Pah[enu2] model retains value for future behavioral research and benchmarking of novel PKU models.

Bile acid Coenzyme A: amino acid N-acyltransferase (BAAT) catalyzes the conjugation of bile acids with taurine or glycine, a process essential for bile acid solubility and intestinal lipid absorption. Mutations in BAAT cause an inborn error of bile acid metabolism, typically characterized by reduced conjugated bile acids and fat-soluble vitamin deficiency. However, several clinical features of BAAT deficiency cannot be fully explained by impaired conjugation alone, suggesting the presence of broader metabolic disturbances. In this study, we analyzed serum bile acid and intermediate profiles in a Japanese patient with genetically confirmed BAAT deficiency and identified a marked elevation of 7α-hydroxy-3-oxo-4-cholestenoic acid (7-HOCA), a key intermediate in bile acid synthesis. To investigate the metabolic consequences of BAAT loss, we generated a hepatic Baat knockdown mouse model using adeno-associated virus serotype 8-mediated delivery of Nme2Cas9 and Baat-targeting sgRNA. This model faithfully recapitulated the accumulation of 7-HOCA observed in the patient, together with a significant reduction in amino acid-conjugated bile acids. Integrated analyses combining bile acid profiling, quantitative PCR, and stable-isotope tracing of cholesterol demonstrated enhanced cholesterol flux toward bile acid biosynthetic pathways and upregulation of bile acid synthetic enzymes in Baat-deficient livers. These findings indicate that BAAT deficiency leads to dysregulated bile acid synthesis in addition to defective conjugation. Together, our results reveal a previously unrecognized metabolic phenotype of BAAT deficiency and provide mechanistic insight into its pathophysiology, establishing a novel in vivo model for studying bile acid metabolism beyond simple conjugation defects.

The cornerstone of treatment in creatine synthesis defects is the supplementation of creatine (CT). The treatment leads to partial replenishment of creatine; it also leads to the reduction of guanidinoacetate (GAA). Considering the neurotoxic accumulation of GAA in patients with guanidinoacetate methyltransferase (GAMT) deficiency, understanding the process by which CT reduces GAA is crucial. Ten-week-old GAMT mice were fed with different diets, either CT-free or containing 2% and 4% CT for 10 weeks. Subsequent investigations included the analysis of CT and GAA in urine, blood, and mouse organs, and the analysis of AGAT expression in mouse organs by qPCR, western blotting, and enzymatic activity. CT supplementation in GAMT mutant mice led to a significant reduction of GAA in body fluids and organs, with the exception of the liver. The GAA-lowering effect of CT was mediated through downregulation of AGAT expression. The CT supplementation was sufficient to correct CT deficiency in all organs. Even in wild type mice, CT supplementation demonstrated a consistent reduction of GAA and increased the CT concentration in kidney and liver. CT supplementation in wild type mice did not change the CT concentration in brain, heart, and skeletal muscle. The liver exhibited a peculiar situation characterized by resistance to the GAA-lowering effects of CT. Our mouse study provides new insights in the effects of CT supplementation in wild type and GAMT deficient mice and the mechanism by which CT reduces the GAA accumulation in GAMT deficiency via downregulation of AGAT. This mechanism should be further explored for the treatment of GAMT deficiency.

Hepatic glycogen storage disease type IX (GSD IX) is due to a deficiency of phosphorylase kinase and is one of the most common types of GSD. We conducted a retrospective, observational cohort study on individuals with GSD IX from across the United Kingdom. We describe the natural history and long-term outcomes for 89 individuals with GSD IX with a median age of 16.4 years (range 4 months to 73 years). This included 60 patients with IXα2, 12 with IXβ and 13 with IXγ2. We report 49 novel alleles in PHKA2, PHKB and PHKG2. The median age at initial presentation was 2.5 years, with 82% (n = 68) presenting with hepatomegaly and 46% (n = 38) presenting with ketotic hypoglycaemia. Steatosis was reported in 71% (n = 12) who had a liver biopsy. The mean Z-score for height at initial presentation was -1.21, and this significantly improved over childhood. A spectrum of severity was seen in all subtypes, with some requiring more intensive dietary management after initial presentation and others requiring no formal dietary treatment at all. No patients developed adenomas or hepatocellular carcinomas, and 75% (n = 33) of adults were on no dietary therapy. Individuals with IXγ2 had a more severe disease course and had significantly worse biochemistry at initial presentation compared to other subtypes. They were diagnosed at a younger age and required more intensive dietary management across the lifespan. We believe that GSD IXγ2 is an ideal target for novel therapies. We recommend regular monitoring and multidisciplinary input in all patients with GSD IX and encourage more formal assessment of protein intake with each review.

In methylmalonic (MMA) and propionic acidemias (PA), liver or liver-kidney transplantation (Tx) is indicated for metabolic decompensations, kidney failure (MMA), and to improve quality of life. Liver cancer was reported in five patients with MMA. We characterized the pathology of 23 explanted livers and one cancer to investigate for pre-cancerous changes. We included seven patients with PA, 16 with MMA, and a patient with cancer after kidney Tx for MMA. Liver function tests, alpha-foetoprotein, and liver ultrasound were collected. Routine and special stains were performed. Abnormalities were observed in liver tests or ultrasound in half of the patients. Two had cirrhosis (one MMA, one PA). The maximum alpha-foetoprotein was 28 ng/mL. The key lesion was clusters and nodules of clear cells in 83%: distended hepatocytes with central nuclei, sharply demarcated from the parenchyma, in the periportal area. These cells contained less glycogen than the surrounding liver; macro-vacuolar steatosis was observed in 20%. Fibrosis was present in all but two, mostly stage 1 (67%), and mild lymphocytic inflammation in the portal tracts. Large-cell dysplasia was observed in the three oldest patients (one PA, two MMA). The phenotype of the clusters and nodules highlighted mitochondrial and LFABP loss. Abnormal labelling of glutamine synthetase was seen at distance from the nodules. The liver cancer was a hepatocellular carcinoma. Liver abnormalities were observed in all patients. The clusters and nodules of clear cells likely originate from propionyl-CoA accumulation and mitochondrial dysfunction. This abnormal pathology pleads for early liver Tx. Regular liver monitoring is recommended with alpha-foetoprotein and ultrasound.

Inborn errors of metabolism affecting the urea cycle are rare severe conditions caused by impaired nitrogen detoxification, leading to hyperammonemia and neurological morbidity across a broad clinical spectrum. Current biochemical diagnostics largely rely on quantification of metabolites. While indispensable for diagnosis and acute management, these snapshot readouts are highly dependent on external factors and often fail to reflect urea cycle function in metabolically compensated states, limiting their usefulness for disease stratification, therapeutic monitoring, and preclinical evaluation of treatments. Here, we reflect on the rationale, development, and application of a functional ureagenesis assay based on administration of [[15]N]-labeled ammonium chloride, enabling quantitative assessment of total in vivo urea cycle function. This assay development was prompted by discordant biochemical and phenotypic findings in a gene therapy-treated mouse model of ornithine transcarbamylase deficiency where supraphysiological enzyme activity did not translate into clinical improvement. We demonstrate that the [[15]N]-isotope based ureagenesis assay is applicable to in vivo models and enables detection of functional correction of the urea cycle. Importantly, we show its applicability in metabolically stable individuals affected by urea cycle disorders, where classical biochemical markers are often normal despite significant enzyme or transporter impairment. Together, these data establish the [[15]N]ammonium chloride-based functional ureagenesis assay as a robust and sensitive tool for quantifying urea cycle function. This approach offers clear advantages for preclinical research and translational studies, particularly in the evaluation of novel therapeutic strategies where traditional biomarkers lack discriminatory power.