Hyperprolinemia type I caused by homozygous p.T466M mutation in PRODH

Hyperprolinemia type I (HPI) is an autosomal recessive metabolic disorder caused by defects in proline oxidase. We herein describe a case of a patient with HPI and harboring the NM_016335.4 (PRODH_v001):c.1397 C > T (p.T466M) mutation and polymorphisms in the PRODH gene, as detected by plasma amino acid analysis and Sanger sequencing. The patient presented with short stature, carbohydrate-rich dietary preferences, and mild intellectual disability that was suggestive of a neurodevelopmental or learning disorder.

suspected. Nevertheless, he presented fasting normoglycemia (91 mg/dL) at admission, and his adrenal function test was normal. Acylcarnitine profile analysis revealed no abnormalities. However, plasma amino acid analysis detected high proline levels (530 μmol/L; reference: 78 − 273 μmol/L) (Table 1A). An intelligence assessment using the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) at the age of 6 years and 3 months revealed the following: a score of 83 in Full Scale IQ (FSIQ), 80 in Verbal Comprehension Index (VCI), 115 in Perceptual Reasoning Index (PRI), 68 in Working Memory Index (WMI), and 78 in Processing Speed Index (PSI). These characteristics suggest a tendency toward neurodevelopmental disorders, including autism, attention-deficit hyperactivity disorder, and learning disorders.
Upon referral to our team, we further investigated the patient's developmental history. His mother reported that he could not keep up with regular class studies in primary school or perform detailed work. He had difficulty in reading and writing Kanji. WISC-IV intelligence assessment at 9 years and 9 months was 73, 86, 91, 63 and 67 for FSIQ, VCI, PRI, WMI, and PSI, respectively, indicating mild intellectual disability. Neurological examination revealed slightly poor coordination, but there was no history of impulsive behavior. Brain magnetic resonance imaging (MRI) (Fig. 1A) and magnetic resonance spectroscopy revealed nonspecific findings, and his electroencephalogram was unremarkable. However, Sanger sequencing detected the homozygous mutation c.1397 C > T (p.T466M) in the PRODH gene as well as some homozygous variants (Fig. 1B and  Table 1B), confirming the diagnosis of HPI. No pathogenic variants were detected in ALDH4A1, which encodes P5C dehydrogenase (EC 1.2.1.88) (data not shown), defects of which cause hyperprolinemia type II. At the age of 10 years and 3 months, the patient was doing well in a supported education class and did not require medication. Written informed consent was obtained from the patient's parents for this report. This study was approved by the Institutional Ethics Committee of the Faculty of Life Science, Kumamoto University.
In Table 1C, we summarize nine cases, including ours, of the POX p.T466M variant 1,6,7 . The PRODH gene, which is located in the 22q11 chromosomal region, is hemideleted in 22q11 deletion syndrome, also known as velo-cardio-facial syndrome (VCFS) 8 . VCFS shares some clinical features with HPI.
A combination of mutations and polymorphisms in the PRODH gene cause dysregulation of POX enzyme activity and may lead to a variety of phenotypes in HPI, including neuronal function disorders. The proline metabolic pathway links cellular proline levels with glutamate and glutamine levels in neurons, and POX has been proposed to play a regulatory role in glutamatergic neurotransmission by affecting the cellular concentration of glutamate 9 . Moreover, proline is thought to induce oxidative stress in the rat brain 10,11 . Hyperprolinemia induces significant oxidative damage to proteins, lipids, and DNA 12 , decreases the activities of antioxidant enzymes, and induces lipid peroxidation in the blood of rats 13 .
Mouse models of POX deficiency, which is also present in individuals with schizophrenia, exhibit increases in neurotransmitter release at glutamatergic synapses as well as deficits in associative learning and response to psychomimetic drugs 14 . Furthermore, hyperprolinemia has been suggested to be a risk factor for schizophrenia 14,15 , and polymorphisms in PRODH, such as rs372055, which this case harbored, are thought to correlate with schizophrenia 16 .
Some pharmacological, biochemical, and behavioral studies have suggested the involvement of the glutamatergic system in ASD pathology 17 , and ASD has been considered a common clinical manifestation of HPI 1,2,6 . Poor social, adaptive, and academic skills are often evident in patients with HPI 3 . Our patient displayed a complex learning disorder, including deficits in reading, writing, and math, and required specific educational training. Although this case and other reports suggest a correlation between intelligence and plasma proline levels 6,8 , further studies are required for confirmation.
Bender et al. proposed the following classification of PRODH mutations based on POX activity reduction: mild (<30%), moderate (30-70%), and severe (>70%) 18 . POX uses flavin adenine dinucleotide (FAD) as a cofactor. The homozygous c.1397 C > T (p. T466M) mutation results in reduced enzyme activity to less than 20% of the control 18 . T466 in POX interacts with the adenine moiety of FAD to stabilize noncovalent binding of the cofactor to the POX apoenzyme, and the T466M mutation alters the affinity of the POX apoenzyme for FAD.
Genotype-phenotype correlations in HPI have been suggested 2,6 . For example, Rosa et al. 3 reported two patients with the same PRODH genotype and the same range of plasma proline levels (376 and 493 μmol/L); these patients presented with a similar neurobehavioral profile, including aggressiveness and sexual disinhibition. Our patient first presented with short stature and mild hyperprolinemia but without obvious intellectual disability, though developmental delay became more noticeable with age. Alexandra et al. 1 described a patient with the p.T466M variant who presented with clinical manifestations and blood proline levels similar to those in our case. Although p.T466M is predicted to be damaging (0.943) in Polyphen 2 (http://genetics.bwh.harvard.edu/pph2/), ClinVar (http:// www.ncbi.nlm.nih.gov/clinvar) provides conflicting interpretations of its pathogenicity (Table 1B). Therefore, the clinical impact of p. T466M remains unclear. Although short stature has not been reported as a clinical manifestation of HPI, our case showed short stature persisting from infancy (Fig. 1C). Harries et al. 19 reported persistent short stature until the age of 27 months in a patient with HPI receiving a low-proline diet. Moreover, van de Ven et al. 20 described a patient with HPI presenting with a variable eating behavior pattern that is consistent with that of our patient. Indeed, our patient preferred a carbohydrate-rich diet and disliked proteinrich foods. The manifestations of short stature and food preference may be derived from HPI, but more case reports are needed to clarify this association.
In conclusion, we encountered a case of HPI that was first detected through plasma amino acid analysis performed during the detailed evaluation of short stature in a child. Mild intellectual disability, mild learning disorders, autism tendencies, and attention-deficit hyperactivity disorder (ADHD) tendencies are considered phenotypes related to HPI. The patient's unique dietary habit is also thought to be one of the phenotypes of HPI, and blood proline levels vary depending on the dietary content. As very few cases of HPI have been reported to date, other patients may display as-yet-unidentified phenotypes. The accumulation of more cases is essential to further our understanding of the clinical characteristics of HPI.