2019 年 42 巻 6 号 p. 1054-1057
Mental retardation is the most common feature among inborn errors of amino acid metabolism. Patients with homocystinuria/homocysteinemia caused by cystathionine β-synthase (CBS) deficiency suffer from thromboembolism and mental retardation from early ages; therefore, detection by newborn screening is performed. Furthermore, elevated levels of serum homocysteine during pregnancy are associated with the occurrence of neural tube defects (NTDs) in newborns. However, the causes of such central nervous system (CNS) defects are unknown. We found previously impaired learning abilities in Cbs-deficient (Cbs−/−) mice (but not NTD births). Here, we investigated the amino acid profiles of serum and cerebrospinal fluid (CSF) from Cbs−/− mice. Mice deficient in cystathionine γ-lyase (Cth), a downstream enzyme of CBS in transsulfuration, as well as wild-type mice, were analyzed as controls. Cbs−/− and Cth−/− mice were smaller than wild-type mice, and CSF yields in Cbs−/− mice were lower than the others. CSF amino acid levels were generally lower than those in serum, and compared with the dramatic amino acid level alterations in Cbs−/− mouse serum, alterations in CSF were less apparent. However, marked upregulation (versus wild-type) of aspartic acid/asparagine (Asp/Asn), glutamine (Gln), serine (Ser), threonine (Thr), phenylalanine (Phe), tyrosine (Tyr), methionine (Met), total homocysteine, and citrulline, and downregulation of lysine (Lys) were found in Cbs−/− mouse CSF. Because similar regulation of total homocysteine/citrulline/Lys was observed in the CSF of Cth−/− mice, which are free of CNS dysfunction, the reduced CSF volumes and the level changes of other amino acids could be relevant to Cbs−/−-specific CNS defects.
Genetic deletion of cystathionine β-synthase (CBS), a transsulfuration enzyme involved in methionine metabolism, is known to cause homocystinuria (OMIN 236200), an autosomal recessive inborn error with high levels of blood homocysteine and urinary homocystine.1,2) Untreated CBS-deficient (classical) homocystinuria patients display various severe clinical symptoms including ectopia lentis, atherothrombosis, osteoporosis, and mental retardation, and could die from myocardial infarction or stroke in their thirties.2) For this reason, this disease has been listed among newborn screening tests in most developed countries, and the patients receive early treatment to prevent progression of their central nervous system (CNS) defects.3)
The early onset of atherothrombosis in homocystinuria patients is attributed to the elevated levels of blood homocysteine.2) Consistently, its elevation is considered as an independent risk factor for cardiovascular disease (CVD). Although endothelial dysfunction mediated by reduced nitric oxide bioavailability appears to be implicated in CVD onset,4) neurodegenerative actions of homocysteine seem to be responsible for the CNS disorders in these patients.5) The prevalence of various psychological/psychiatric symptoms (including behavioral disorders, depression, obsessive-compulsive disorder, and personality disorder) has been reported in either treated or untreated homocystinuria patients, and the elevations of both homocysteine and methionine in the blood may be involved in such CNS phenotypes.6)
To gain an understanding of the CNS dysfunctions (impaired learning abilities and cerebellar malformation) that were observed in our Cbs-deficient (Cbs−/−) mice,7) we examined their amino acid profiles of both blood and cerebrospinal fluid (CSF). Although our Cbs−/− mice are free of neural tube defects (NTDs) such as spina bifida and hydrocephalus,7) some dysregulation in the blood–CSF barrier could exist; the CSF amino acid profiles could be reflections of the CNS amino acid metabolism and the blood–CSF barrier function. Mice deficient for cystathionine γ-lyase (Cth), a downstream enzyme of Cbs in transsulfuration, were also analyzed, because while Cbs−/− mice display homocysteinemia/methioninemia, Cth−/− mice show homocysteinemia/cystathioninemia and are apparently normal and free of any CNS disorders.8) We herein observed low CSF volume yields and high levels of some CSF amino acids in Cbs−/− mice.
Heterozygous Cbs+/− mice9) were obtained from the Jackson Laboratory (Bar Harbor, ME, U.S.A.) and Cth+/− mice were generated by our group.8) Heterozygous mice on a C57BL/6J background were bred to obtain Cbs−/− or Cth−/− mice. Blood and CSF samples were collected from 2-week-old pups before Cbs−/− mice start to die.7,9) Mice were anesthetized using isoflurane; serum and CSF samples were collected by cardiocentesis (ca. 15 µL per pup) and ventricle puncture with glass capillaries (2–4 µL per pup), respectively. Serum/CSF amino acid levels were measured using HPLC, amino acid labeling with 4-fluoro-7-nitrobenzofurazan (NBD-F [Dojindo, Japan]), and thiol labeling with 4-fluoro-7-sulfobenzofurazan (SBD-F [Dojindo]) as described previously.7,8,10) Because 10 µL is the minimum volume required for our assays, equal volumes of CSF from 4–6 mice were pooled together. Aspartic acid/asparagine (Asp/Asn) peaks were indistinguishable and tryptophan (Trp) was not derivatized with NBD-F. For the detection of thiol-containing amino acids, serum/CSF samples were reduced to cleave disulfide bonds before SBD-F labeling, and the total levels of homocysteine, cysteine, and glutathione (tHcy, tCys, and tGSH, respectively) were measured. All animal procedures conformed to the Guide for the Care and Use of Laboratory Animals, 8th Edition published by the U.S. National Research Council, and were approved by the Animal Care Committees of Showa Pharmaceutical University (No. P-2016-10 and P-2018-07).
Cbs−/− and Cth−/− mice were significantly smaller than wild-type mice at 2 weeks of age (Fig. 1A). CSF volume yields and CSF volume yield/body weight ratios were the lowest in Cbs−/− mice (Figs. 1B and C, respectively). When mice were divided into 3 groups by body weight, the low CSF yields in the smaller groups of Cbs−/− mice were remarkable (Fig. 1D). Serum levels of 17 out of 24 amino acid species measured were significantly upregulated (×1.45–25.1) in Cbs−/− mice, whereas only the levels of arginine (Arg), tHcy, and citrulline were elevated in Cth−/− mice (Table 1). The upregulation of CSF amino acids was more limited in their amounts and variations; Asp/Asn, glutamic acid (Gln), serine (Ser), threonine (Thr), phenylalanine (Phe), tyrosine (Tyr), methionine (Met), tHcy, and citrulline levels were upregulated in Cbs−/− mice and only tHcy and citrulline levels in Cth−/− mice (Table 1). Serum levels of glycine (Gly) were higher in Cbs−/− mice and lower in Cth−/− mice, and those of taurine were lower in Cth−/− mice than those in wild-type mice; however, there were no significant differences in CSF Gly/taurine levels between the three genotypes (Table 1). Exceptionally, CSF levels of lysine (Lys) were significantly lower in both Cbs−/− and Cth−/− mice than those in wild-type mice although their serum levels of Lys were equivalent (Table 1). Amino acid differences in both quantity and quality between serum and CSF ruled out the possibility of blood contamination of CSF during sample preparation.
Two-week-old wild-type (WT), Cbs−/−, and Cth−/− mice were anesthetized, weighed, and subjected to craniotomy for CSF collection. BW (A), CSF volume yields (B), BW/CSF yield ratio (C), and CSF volume yields divided to 3 pools by BW (≤5.0 g, 5.1–6.0 g, and ≥6.1 g in D). Data are mean ± S.D. from independent mouse samples (sample numbers in parentheses) with significant differences between the genotypes (* p < 0.05, ** p < 0.01, and *** p < 0.001) and the BW groups (#p < 0.05 and ###p < 0.001) using Student’s t-test.
Mean ± S.D. from independent (serum) and pooled (CSF) samples (n: sample/pool numbers), and fold changes versus wild-type samples are presented in parentheses. * p < 0.05, ** p < 0.01, and *** p < 0.001 versus wild-type samples; ## p < 0.01 and ### p < 0.001 versus serum samples using Student’s t-test.
Cbs−/− mice were generated as homocystinuria models9) and have been analyzed by research groups worldwide, including us. They exhibit some features similar to homocystinuria patients including homocysteinemia,7,9) methioninemia,7) homocystinuria,11) and hepatic steatosis7,9); however, most of them die by 2 weeks of age due to unknown mechanisms.7,9) Although serum levels of alanine aminotransferase, aspartate aminotransferase, and triglyceride were all higher in 2-week-old Cbs−/− pups than wild-type pups (369 versus 13.2 [IU/L], 432 versus 61.8 [IU/L], and 226 versus 91.9 [mg/dL], respectively),8) and some differential proteins exist in serum, including α-fetoprotein,12) such changes should not be the primary cause of the death. Meanwhile, Cth−/− mice were generated as cystathioninuria (OMIN 219500) models, another inborn error with high levels of blood/urinary cystathionine.8,13) In contrast to homocystinuria, cystathioninuria has been considered free of any striking manifestations, and likewise, our Cth−/− mice appeared normal without intervention.8,14−17) The study aimed to reveal Cbs−/−-specific CSF amino acid profiles that possibly affect the CNS and become the cause of the early deaths in this model. CSF is a clear liquid that is formed/secreted continuously from the choroid plexus ependymal cells, circulated around the brain/spinal cord, and absorbed into the blood. Routine CSF analyses in the clinic investigate its color, clarity, pressure, protein/glucose levels, and exuded (red and white) blood cell counts, but not amino acid levels.
We found that CSF volume yields and amino acid profiles of Cbs−/− mice were significantly different from wild-type and Cth−/− mice. Although we could neither find any NTD-like phenotypes nor measure CSF pressures in tiny Cbs−/− pups, slight and gradual CSF leaks to the surrounding protective spinal dural sac could exist. Clinical features of spontaneous CSF leaks involve orthostatic headache, nausea, vomiting, dizziness and fatigue, and such conditions could be the primary cause of deaths in Cbs−/− pups. Moreover, amino acid levels in CSF were generally much lower than those in serum but some amino acids (Asp/Asn, Gln, Ser, Thr, Phe, Tyr, and Met) were accumulated in Cbs−/− mice while not in CNS dysfunction-free Cth−/− mice. The level changes of those CSF amino acids (and the reduced CSF volumes) could be relevant to Cbs−/−-specific CNS defects. Such CSF alterations could reflect some degree of dysfunction in the blood–CSF barrier or amino acid metabolism within the brain; Cbs is substantially expressed within the brain.11) These findings may contribute to the understanding of CNS disorders in Cbs−/− mice and homocystinuria patients.
This work was partly supported by Grants-in-Aid for Scientific Research (17K08287 [NA] and 16H05107 [II]) from the Japan Society for the Promotion of Science, and a Grant-in-Aid for Young Scientists of Showa Pharmaceutical University [SK].
The authors declare no conflict of interest.