Search

Abstract Glutaminase (GLS) hyperactivity was first described in 2019 in a patient with profound developmental delay and infantile cataract. Here, we describe a 4‐year‐old boy with GLS hyperactivity due to a de novo heterozygous missense variant in GLS, detected by trio whole exome sequencing. This boy also exhibits developmental delay without dysmorphic features, but does not have cataract. Additionally, he suffers from epilepsy with tonic clonic seizures. In line with the findings in the previously described patient with GLS hyperactivity, in vivo 3 T magnetic resonance spectroscopy (MRS) of the brain revealed an increased glutamate/glutamine ratio. This increased ratio was also found in urine with UPLC‐MS/MS, however, inconsistently. This case indicates that the phenotypic spectrum evoked by GLS hyperactivity may include epilepsy. Clarifying this phenotypic spectrum is of importance for the prognosis and identification of these patients. The combination of phenotyping, genetic testing, and metabolic diagnostics with brain MRS and in urine is essential to identify new patients with GLS hyperactivity and to further extend the phenotypic spectrum of this disease.

Abstract Quantifying lymphocyte vacuolization in peripheral blood smears (PBSs) serves as a measure for disease severity in CLN3 disease—a lysosomal storage disorder of childhood‐onset. However, thus far quantification methods are based on labor‐intensive manual assessment of PBSs. As machine learning techniques like convolutional neural networks (CNNs) have been deployed quite successfully in detecting pathological features in PBSs, we explored whether these techniques could be utilized to automate quantification of lymphocyte vacuolization. Here, we present and validate a deep learning pipeline that automates quantification of lymphocyte vacuolization. By using two CNNs in succession, trained for cytoplasm‐segmentation and vacuolization‐detection, respectively, we obtained an excellent correlation with manual quantification of lymphocyte vacuolization (r = 0.98, n = 40). These results show that CNNs can be utilized to automate the otherwise cumbersome task of manually quantifying lymphocyte vacuolization, thereby aiding prompt clinical decisions in relation to CLN3 disease, and potentially beyond.

Abstract Background CLN3 disease is a disorder of lysosomal homeostasis predominantly affecting the retina and the brain. The severity of the underlying mutations in CLN3 particularly determines onset and course of neurological deterioration. Given the highly conserved start codon code among eukaryotic species, we expected a variant in the start codon of CLN3 to give rise to the classical, that is, severe, phenotype. Case series We present three patients with an identical CLN3 genotype (compound heterozygosity for the common 1 kb deletion in combination with a c.1A > C start codon variant) who all displayed a more attenuated phenotype than expected. While their retinal phenotype was similar to as expected in classical CLN3 disease, their neurological phenotype was delayed. Two patients had an early onset of cognitive impairment, but a particularly slow deterioration afterwards without any obvious motor impairment. The third patient also had a late onset of cognitive impairment. Conclusions Contrasting our initial expectations, patients with a start codon variant in CLN3 may display a protracted phenotype. Future work will have to reveal the exact mechanism behind the assumed residual protein synthesis, and determine whether this may be eligible to start codon targeted therapy.