The RSH/Smith–Lemli–Opitz syndrome (RSH/SLOS) is a human autosomal recessive syndrome characterized by multiple malformations, a distinct behavioral phenotype with autistic features and mental retardation. RSH/SLOS is due to an inborn error of cholesterol biosynthesis caused by mutation of the 3β-hydroxysterol Δ7-reductase gene. To further our understanding of the developmental and neurological processes that underlie the pathophysiology of this disorder, we have developed a mouse model of RSH/SLOS by disruption of the 3β-hydroxysterol Δ7-reductase gene. Here we provide the biochemical, phenotypic and neurophysiological characterization of this genetic mouse model. As in human patients, the RSH/SLOS mouse has a marked reduction of serum and tissue cholesterol levels and a marked increase of serum and tissue 7-dehydrocholesterol levels. Phenotypic similarities between this mouse model and the human syndrome include intra-uterine growth retardation, variable craniofacial anomalies including cleft palate, poor feeding with an uncoordinated suck, hypotonia and decreased movement. Neurophysiological studies showed that although the response of frontal cortex neurons to the neurotransmitter γ-amino-n-butyric acid was normal, the response of these same neurons to glutamate was significantly impaired. This finding provides insight into potential mechanisms underlying the neurological dysfunction seen in this human mental retardation syndrome and suggests that this mouse model will allow the testing of potential therapeutic interventions.
Received 21 November 2000; Revised and Accepted 15 January 2001.
The RSH/Smith–Lemli–Opitz syndrome (RSH/SLOS; OMIM 270400) is an autosomal recessive malformation/mental retardation syndrome caused by an inborn error of cholesterol metabolism. RSH/SLOS was first described by Smith, Lemli and Opitz in 1964 (1) and has an estimated incidence of 1/10 000 to 1/60 000 in populations of northern and central European heritage (2). The RSH/SLOS phenotype is variable, but typically includes craniofacial malformations, limb anomalies, growth retardation, behavioral problems and mental retardation (3,4). The cause of RSH/SLOS syndrome remained unknown until 1993 when Irons et al. (5) found decreased cholesterol and increased 7-dehydrocholesterol (7-DHC) levels in two RSH/SLOS patients. Subsequent work showed that RSH/SLOS is an inborn error of cholesterol biosynthesis due to a deficiency of 3β-hydroxysterol Δ7Hreductase activity, which results in impaired conversion of 7-DHC to cholesterol (Fig. 1A) (6–9). In the central nervous system (CNS), the 3β‐hydroxysterol Δ7-reductase also functions to reduce 7‐dehydrodesmosterol (7-DHD) in the biosynthesis of desmosterol (Fig. 1A). The gene encoding the human 3β‐hydroxysterol Δ7-reductase (DHCR7) was identified based on homology to the sterol Δ7-reductase from Arabidopsis thaliana and was mapped to human chromosome 11q12–13 (10,11). To date, more than 70 different mutations of DHCR7 have been identified in patients with RSH/SLOS (2,11–13).
The functions of cholesterol during development and the biochemical mechanisms that underlie the dysmorphogenesis and neurological deficits seen in RSH/SLOS are only now being elucidated. Besides being a precursor for bile acids, neurosteroids and steroid hormones, cholesterol is a major lipid component of myelin and plasma membranes, it is necessary for the maturation of the hedgehog family of morphogens (14) and forms cholesterol rafts, which are involved in signal transduction (15). To further characterize the pathophysiology and neurophysiology underlying RSH/SLOS, and to provide a model system for testing therapeutic interventions, we produced a mouse model of RSH/SLOS by disruption of the mouse 3β-hydroxysterol Δ7-reductase gene (Dhcr7). Similar to RSH/SLOS infants, Dhcr7–/– pups have variable craniofacial anomalies, demonstrate growth failure, are hypotonic and fail to feed.
Disruption of the Dhcr7 gene
We disrupted the mouse 3β-hydroxysterol Δ7-reductase gene (Dhcr7) in mouse embryonic stem (ES) cells using targeted homologous recombination (Fig. 1B). The replacement targeting vector was designed to disrupt Dhcr7 by insertion of the neomycin phosphotransferase gene (PGKneo) and deletion of coding exons III, IV and part of exon V. Diphtheria toxin A subunit (DTA) was used to select against random insertional events. Cells that undergo non-homologous recombination with the targeting vector express DTA and are thus killed. Homologous recombination between the targeting vector and the Dhcr7 allele was found in 12% (11/94) of G418-resistant ES cell clones. PCR amplification between primers located outside of the targeting flanks and internal to PGKneo were used to confirm proper homologous recombination by both the 5′ and 3′ flanks of the targeting vector with the Dhcr7 allele in two independent G418-resistant clones (Fig. 1C). Southern blot analysis confirmed a single insertion of the targeting construct in both of these clones (data not shown). Both F42 and F111 clones produced germline transmitting chimeric founders. An identical phenotype was observed with both the F42 and F111 lines.
Heterozygous mice appeared phenotypically normal. Thus, we intercrossed Dhcr7+/– mice to determine if a recessive phenotype was present. Genotyping of 310 weaned progenies from heterozygous crosses identified no Dhcr7–/– animals (38% +/+ and 62% +/–). In contrast, when all pups and embryos were genotyped (n = 587) we found a close to expected Mendelian ratio of 24% +/+, 50% +/– and 26% –/– (Fig. 1D). The Dhcr7–/– pups died during their first day of life, most likely as a consequence of their failure to feed.
Biochemical characterization of Dhcr7–/– mice
Biochemical characterization of sterol levels in tissues from Dhcr7–/–newborn pups showed that these mice had decreased cholesterol (Fig. 2A) and markedly increased 7-DHC levels (Fig. 2B) compared with either Dhcr7+/+ or Dhcr7+/– pups. Serum 7-DHC levels in Dhcr7–/–pups were increased 1678-fold above serum levels measured in Dhcr7+/+pups. Dhcr7–/– tissue 7-DHC levels, compared with levels from Dhcr7+/+ pups, were increased 250-, 265-, 250-, 791- and 2003-fold in cortex, midbrain, kidney, liver and muscle, respectively. Although minor compared to the increase seen in mutant pups, 7-DHC levels in Dhcr7+/– pups were elevated by 1.4–4.2-fold in tissues and by 9.6-fold in serum compared with levels in Dhcr7+/+ pups. 8-Dehydrocholesterol (8-DHC), which is an isomer of 7-DHC (16), is also markedly elevated in mutant tissues (Fig. 2C). In the central nervous system, 3β-hydroxysterol Δ7-reductase reduces 7-DHD to form desmosterol, and desmosterol levels were markedly reduced in cortex and midbrain from Dhcr7–/–pups compared with levels found in samples from either Dhcr7+/+ or +/– pups (Fig. 2D). Consistent with a null mutation, 3β-hydroxysterol Δ7-reductase activity in mutant fibroblasts, as measured by reduction of ergosterol to brassicasterol (17), was indistinguishable from background at 0.3% of control levels (Fig. 2E). Thin layer chromatography of 14C-acetate-labeled steroids from Dhcr7–/– fibroblasts showed the same pattern of sterol intermediates as seen in human RSH/SLOS fibroblasts (data not shown). The accumulation of 7-DHC and 7-DHD during development of Dhcr7–/– embryos was characterized. In the livers of Dhcr7–/– embryos from embryonic day E12.5 until birth, 7-DHC accounts for ∼50% of total sterols compared with very low levels in control embryos (Fig. 2F). Analysis of sterol content of brain tissue from Dhcr7–/– embryos showed that 7-DHC levels, as a percentage of total sterols, approach the level of cholesterol found in Dhcr7+/+ embryos, whereas cholesterol levels decrease during development (Fig. 2G). In mutant embryos, 7-DHD levels progressively increased similar to the normal increase in desmosterol seen in Dhcr+/+ embryos during development (Fig. 2H).
Phenotypic characterization of Dhcr7–/– pups
Homozygous mutant pups were growth retarded, had craniofacial malformations, demonstrated decreased movement and failed to feed (Fig. 3A and B). Failure to suckle is evident in that the mutant pups lack a milk spot (Fig. 3A). Intra-uterine growth retardation was observed (Fig. 3C) and mean birth weight for Dhcr7–/– pups was 1.11 ± 0.10 g (n = 68) compared with 1.44 ± 0.18 g (n = 46) for Dhcr7+/+ pups (P < 0.001). Craniofacial anomalies were variably present in Dhcr7–/–pups. In 30% (20/67) of the Dhcr7–/– pups we noted the presence of a nasal plug and in 9% (6/67) of the animals we could not identify any nasal opening (Fig. 3D and E). The lack of a nasal opening was due to retention of a nasal plug in an otherwise normally formed nose (Fig. 3F–H). During development, a nasal plug is normally present from E16 until just prior to birth (18). The decreased movement and generalized weakness observed in the mutant pups may explain the lack of an expelled nasal plug. A cleft palate (Fig. 3I) was present in 9 of 111 Dhcr7–/– animals, compared with none in 115 control pups (P < 0.01, Fisher’s exact test). Histological examination of the cleft palate found in Dhcr7–/–pups showed that the secondary palate had elevated, but failed to close to the midline (Fig. 3J and K). Histopathological analysis of heart, liver, spleen, pancreas, kidney, adrenal, spinal and trigeminal ganglia, vomeronasal structure, eye, small intestine, colon and myenteric plexus showed no significant abnormalities in Dhcr7–/– pups, whereas analysis of the lungs was notable for diffuse alveolar atelectasis (data not shown). No limb or skeletal malformations were detected by gross examination or examination of alizarin and alcian blue stained skeletal preparations (data not shown). Consistent with pups that have not fed, the bladder was dilated and meconium was present in the colon.
Neurophysiological characterization of Dhcr7–/– pups
The mutant pups lacked a milk spot, rarely vocalized alarm, had a hypotonic appearance and had less vigorous movements compared with controls (Fig. 3A and B). The decreased movement and hypotonic appearance were present at the time of birth and mutant pups were less vigorous than control pups that had not yet fed. Thus, these findings did not appear to be secondary to lack of feeding. The mutant pups were not rejected by the mother and the lack of feeding was observed in all Dhcr7–/– pups irrespective of craniofacial anomalies. Except for size, histopathological analysis by hematoxylin and eosin, trichrome, and phosphotungstic acid hematoxylin staining of skeletal muscle showed no significant differences between control and mutant animals (data not shown). Although smaller, mutant brain weights were proportional to body weight (controls, 5.2 ± 0.2% n = 8; mutants, 5.3 ± 0.1% n = 12), and TUNEL (TdT-mediated dUTP nick end-labeling) staining showed no increased apoptosis in brains of Dhcr7–/– pups (data not shown). A stereotypic rhythmic sucking/swallowing reflex was observed in control animals fed by hand. However, although the mutant pups could swallow formula, their swallowing movements lacked this rhythmic pattern. Hand feeding of Dhcr7–/–pups led to aspiration. When the mutant animals were fed dye-labeled formula, aspiration resulted in staining of the trachea, bronchi and lung parenchyma (Fig. 3M). This was not observed in control animals fed in a similar manner (Fig. 3L). Gross and histological examination of the pharynx and trachea revealed no malformations (data not shown).
The combination of decreased movement, uncoordinated suck/swallow and aspiration suggested a neurological defect which we postulated might be due to impairment of neuronal function by the replacement of 7-DHC and 7-DHD for cholesterol and desmosterol, respectively, in neuronal membranes. To test this hypothesis, whole-cell recordings were made from cortical neurons in brain slices. Neurons from both control and mutant pups exhibited similar reversible, tetrodotoxin-sensitive sodium currents in response to a voltage step to –30 mV from a holding potential of –90 mV (Fig. 4A and B). The mean peak amplitudes of the control (–828 ± 87 pA) and the mutant (–835 ± 81 pA) sodium currents were not significantly different (mean ± standard error, P = 0.9, unpaired t-test, n = 13 for both groups). Neurons from both control and mutant brains were able to generate a single action potential upon injection of a depolarizing current from a potential of –55 mV (data not shown). These control experiments establish that the mutant neurons are functionally similar to control neurons. We then tested the ability of mutant cortical neurons to respond to the neurotransmitters γ-amino-n-butyric acid (GABA) and glutamate. Although the median inward current was lower in mutant animals, no significant difference was seen in the response of control (Fig. 4C) and mutant (Fig. 4D) cortical neurons to 7.3 µM GABA. In control neurons the median amplitude of the inward current induced by GABA was –262 pA (–189 pA 25% and –514 pA 75%, n = 13) and the median amplitude for mutant cortical neurons was –161 pA (–88 pA 25% and –323 pA 75%, n = 14). These means were not statistically different (P = 0.14, Mann–Whitney rank sum test). In contrast, the neuronal response to the application of exogenous glutamate (24 µM), which stimulates NMDA (N-methyl-d-asparate), AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid)and kainate glutamate receptors, was significantly reduced in mutant neurons (Fig. 4E and F). Application of glutamate induced a median inward current of –39 pA (–21 pA 25%, –67 pA 75%, n = 14) in control neurons, whereas Dhcr7–/– cortical neurons showed either a minimal response (n = 9) or no response (n = 8) to glutamate. In the 17 mutant cortical neurons from which glutamate responses were measured, the median current measured in response to glutamate was –4.9 pA (0 pA 25% and 11 pA 75%). Thus the median response to glutamate in mutant neurons was 12.5% of control level and the median mutant response was significantly different from the median control response (P < 0.001, Mann–Whitney rank sum test). The prolonged glutamate response recorded in control neurons is consistent with what is expected due to activation of NMDA glutamate receptors (19). One potential confounding factor is that a measured decrease in current induced by glutamate could be secondary to decreased cell surface area, which would result in decreased capacitance. However, if this result was due to a change in cell surface area, similar results would be expected for both the sodium current measurements and GABA-induced depolarization. The impaired glutamate response in Dhcr7–/– cortical neurons does not appear to be due to a developmental alteration of receptor expression. RT–PCR analysis established expression of NMDA, AMPA and kainate receptor subunits in mutant cortex (Fig. 5A) and western blot analysis detected no difference in cortical NMDAζ1, NMDAε1 or NMDAε2 subunit protein levels between mutants and controls (Fig. 5B).
RSH/SLOS is one of a few inborn errors of metabolism that present as a multiple malformation/mental retardation syndrome, and it is the prototypical example of a series of malformation syndromes now known to be due to inborn errors of post-squalene cholesterol biosynthesis. These other inborn errors of post-squalene cholesterol biosynthesis include desmosterolosis (20), X-linked dominant chondrodysplasia punctata of the Conradi–Hünermann type (21–23) and congenital hemidysplasia, ichthyosiform nevus and limb defects (CHILD) syndrome (24,25). Determination of the genetic and biochemical defects in this group of metabolic malformation syndromes has now provided the basis for understanding the biochemical, molecular, cellular and developmental processes that underlie the pathophysiology of these disorders. To further our mechanistic understanding of RSH/SLOS and to provide an experimental system in which therapeutic interventions could be tested, we produced a genetic mouse model of RSH/SLOS.
Our disruption of the Dhcr7 gene was designed to produce a null mutation and we could not detect 3β-hydroxysterol Δ7‐reductase activity in skin fibroblasts from Dhcr7–/– pups using an ergosterol reduction assay. Characterization of sterols present in tissue and serum samples from Dhcr7–/– animals showed increased 7-DHC levels and decreased cholesterol levels compared to serum and tissue samples from control pups similar to that found in humans with RHS/SLOS. In addition to an accumulation of 7-DHC in the brains of Dhcr7–/– animals, we also found marked elevations of 7-DHD levels.
Dysmorphic features in this mouse model of RSH/SLOS were limited to retention of the nasal plug, intra-uterine growth retardation and cleft palate. Cleft palate was found in 9% of the Dhcr7–/– pups. Cleft palate has been reported to be present in 37–47% of RSH/SLOS patients (3,4). No limb, renal, adrenal or CNS malformations were found. The relative lack of dysmorphic features, compared with what one expects to find in a human patient with two null mutations, may be due to placental transfer of maternal cholesterol during critical periods of development. Theoretically, the cholesterol ascertained in the Dhcr7–/– embryos could either be derived from maternal sources or synthesized de novo by an alternative biosynthetic pathway. In rodents, although most fetal cholesterol is derived by de novo synthesis (26), maternal cholesterol can cross the placenta and appears to be essential for early development (27). Placental transfer of maternal cholesterol to the developing fetus may explain the marked physical differences between our genetic mouse model and previously published teratogenic rodent models of RSH/SLOS. Fetuses from teratogenic rodent models using either AY9944 (28,29) or BM15.766 (30–32) to inhibit 3β-hydroxysterol Δ7-reductase activity have major developmental malformations. The finding that maternal cholesterol supplementation partially ameliorates the teratogenic phenotype caused by AY9944 supports the idea that inhibition of maternal cholesterol biosynthesis is an etiological factor in these teratogenic model systems (28,29). A second possible explanation for the marked difference between these teratogenic models and our genetic model is that these teratogenic inhibitors lack specificity for the 3β-hydroxysterol Δ7-reductase. In addition to inhibition of the 3β-hydroxysterol Δ7-reductase, AY9944 has been reported to inhibit both the 3β-hydroxysterol Δ14‐reductase (33) and the 3β-hydroxysterol Δ8-isomerase (34). In contrast to the major malformations induced by both AY9944 and BM15.766, teratogenic studies using YM9429 (35), another inhibitor of 3β-hydroxysterol Δ7-reductase (36), report a cleft palate phenotype similar to that found in our genetic mouse model.
Squalene, which is synthesized via condensation of two molecules of farnesyl diphosphate, is the first cholesterol-specific intermediate of cholesterol biosynthesis. In addition to the Dhcr7 mutant mouse, three additional post-squalene cholesterol biosynthesis mouse mutations have been described. These include disruption of squalene synthase, mutation of 3β-hydroxysteroid dehydrogenase (bare patches mouse) and mutation of sterol Δ8–Δ7 isomerase (tattered mouse). Disruption of the squalene synthase gene results in an autosomal recessive embryonic lethal phenotype. Many of the squalene synthase mutant embryos die prior to E9.5 and those that are recovered at E9.5 are developmentally retarded (37). Both the bare patches (Bpa) and tattered (Td) mutations are X-linked. Hemizygous Bpa males die shortly after implantation, whereas heterozygous females have hyperkeratotic skin lesions, asymmetric cataracts and chondrodysplasia (38). Male hemizygous Td embryos are growth retarded, edematous, have a cleft palate and craniofacial anomalies, lack intestines, have limb anomalies and die prenatally (23). Although a deficiency of any of these enzymes should inhibit endogenous cholesterol biosynthesis, the squalene synthase, Bpa, Td and Dhcr7 mutant mice all have distinct phenotypes. This may be due to variable teratogenic effects of different precursor molecules or variable ability of the sterols that are synthesized to substitute for cholesterol.
Dhcr7–/– pups demonstrate marked neurological abnormalities. The neurological findings in the mutant pups, which include decreased movement and lack of suckling, are reminiscent of problems frequently encountered in RSH/SLOS infants. RSH/SLOS infants are frequently hypotonic and weak. Poor feeding and suck, which in many patients necessitates placement of a gastrostomy tube, is a frequently encountered clinical problem. The lack of a coordinated suck is consistent with an impaired response to glutamate during development. Pharmacological blockade of NMDA receptors in normal newborn mice impairs suckling (39) and disruption of the NMDAε2 (40) subunit results in uncoordinated suckling similar to that seen in Dhcr7–/– pups. Null mutations of the NMDAζ1 subunit (41,42) disturb both respiratory function and suckling (39) and a point mutation in the glycine binding site of the NMDAζ1 subunit results in pups that do not feed (43). Our neurophysiological investigations demonstrate that there is a decreased response of neurons from the frontal cortex of Dhcr7–/– mice to glutamate. This is a specific finding given that mutant neurons exhibit a normal sodium current, can generate normal action potentials and respond to GABA. In a recently published rat model of fetal alcohol syndrome, an elevated GABA response in conjunction with a decreased NMDA glutamate response was associated with increased neuronal apoptosis (44). This potential mechanism does not appear to be an etiological factor in our murine RSH/SLOS model. In mutant pups, brain weight was proportional to body weight, and no increased apoptosis was detected in brains from Dhcr7–/– pups. Both RT–PCR and western blot analysis established that the decreased glutamate response was not due to a developmental change in the expression of glutamate receptor subunits.
There are several possible explanations for the impaired glutamate response in Dhcr7–/– pups. The substitution of 7-DHD and 7-DHC for desmosterol and cholesterol in neuronal membranes could directly inhibit receptor function, inhibit subunit interactions or impair translocation of the receptor from the microsomal compartment to the synaptosomal membrane. Alternatively, decreased cholesterol levels or increased levels of cholesterol biosynthetic intermediates could indirectly modulate NMDA receptor function by affecting neurosteroid synthesis or function. Neurosteroids are thought to be endogenous ligands for sigma 1 (σ1) receptors (45) and to modulate NMDA receptor function (46). Thus, decreased cholesterol levels could affect synthesis of neurosteroids. Additionally, the σ1 receptor and enzymes of late cholesterol biosynthesis have significant structural and pharmacological similarities (47). Thus, elevations of sterol biosynthetic intermediates may competitively inhibit neurosteroid synthesis or affect neurosteroid interactions with σ1 receptors.
We have established a genetic mouse model of RSH/SLOS. In addition to similar biochemical parameters, similarities between RSH/SLOS and this mouse model include cleft palate, growth retardation and neurological deficits. The impaired glutamate response observed in this mouse model may yield insight into the etiology of some of the neurological dysfunction seen in RSH/SLOS. Thus, this mouse model will be helpful in elucidating the biochemical and cellular basis of the neuropathophysiology underlying the behavioral and learning disabilities seen in this multiple malformation/mental retardation syndrome, and will also allow us to test potential therapeutic interventions.
MATERIALS AND METHODS
Construction of the Dhcr7 targeting vector and mouse manipulation
Mouse expressed sequence tag 875323 was identified based on homology to the A.thaliana sterol Δ7-reductase sequence. A genomic bacterial artificial chromosome (BAC) clone was isolated by PCR screening of a mouse 129 BAC library 5′-GCAGCCTTGGGAAGCATAGAGAGG-3′ and 5′-ACCAGGATGGCCCAATCATCGGAG-3′ (Genome Systems). To generate the targeting construct, a 6.5 kb BamHI/XbaI 5′ flank fragment was cloned into pBS. The neomycin phosphotransferase gene was cloned into this construct as an XbaI/EcoRI fragment. The 3′-flank was generated as a 2.1 kb PCR fragment (5′-CCGGAATTCCCTCATTAACCTGTCCTTCGC-3′ and 5′-ATAAGAATGCGGCCGCCGATGGTCTTCAGATACCAGG-3′) and was cloned as an EcoRI/NotI fragment. The negative selectable marker, DTA, was cloned into a SalI site. The targeting vector pAH-7 was linearized with NotI. The targeting construct was electroporated into J1 ES cells (Bio-Rad Electroporator, 400 V, 25 µF), which were then plated at a density of 5 × 105 cells into a 60 mm dish containing a layer of G418-resistant embryonic fibroblasts. Selection for G418 (350 mcg/ml; Gibco BRL) resistant colonies was begun 24 h after plating. Culture conditions and media were as previously described (48). Screening of G418-resistant clones was done by PCR using primers internal to the neomycin resistance gene and primers external to the targeting construct flank. To establish homologous recombination on the 5′ flank, primers A and B were used (A, 5′-CAGTGAGGCTCGGGAGTGGAAGG-3′; B, 5′-CAAATTAAGGGCCAGCTCATTCCTCC-3′). Homologous recombination of the 3′ flank was established using primers C and D (C, 5′-GAGGCCAGAGGCCACTTGTGAGCG-3′; D, 5′-GCCTAGGTACCACCCAAAGTGG-3′). Clones F42 and F111 were microinjected into C57/B6 blastocysts using standard techniques. Four PCR primers were used in a combined reaction to genotype the mice in this study. These were mutant 5′-CATCCCCTCAGAAGAACTCGT-3′, mutant 5′-CTGTGCTCGACGTTGTCACTG-3′, which produced a 600 bp fragment originating from the neomycin phosphotransferase gene, wild-type 5′-GGTGAATGGGCTGCZAAGCCTGG-3′ and wild-type 5′-GGTAGCCCTTGATCATTGCG-3′, which produced a 200 bp fragment originating from exon IV which is deleted in the mutant allele. PCR cycling conditions consisted of 5 min of denaturation at 94°C followed by 35 cycles of 30 s at 94°C, 60 s at 63°C, 60 s at 72°C and a final extension of 10 min at 72°C. For timed matings, the identification of a copulatory plug was considered to be E0.5 and embryonic age was confirmed by inspection. This animal work was done under NICHD and NIAAA approved animal study protocols.
Sterol, enzymatic, histological, RT–PCR and western blot analysis
Sterols were extracted from serum and tissues and analyzed by GC/MS as previously described (8). The rate of conversion of ergosterol to brassicasterol was measured as previously described (17). Standard histologic techniques and staining were used. For RT–PCR, total RNA was isolated from cerebral cortex of wild-type (n = 3), heterozygous (n = 1) and mutant (n = 2) mice, subjected to DNase I (Amplification grade, Life Technologies) to eliminate potential genomic DNA contamination, reverse transcribed with random hexamers and Moloney Murine Leukemia Virusreverse transcriptase (Life Technologies) and amplified by PCR for 40 cycles of 30 s at 92°C, 90 s at 55°C and 60 s at 72°C using the following primer sets: β-actin forward 5′-ACAGCTGAGAGGGAAATCGTG-3′ and reverse 5′-CTAGGAGCCAGGGCAGTAATCT-3′ (expected product 362), NMDARζ1 forward 5′-ATGCCCCTGCCACCCTCACTTTTG-3′ and reverse 5′-GCAGCTGGCCCTCCTCCCTCTCA-3′ (expected products 381 and 270 bp), NMDARε1 forward 5′-CGGGAGGGATGAAGGCTGTAA-3′ AC-3′ and reverse 5′-TGTAGATGCCCCTGCTGATGGAGA-3′ (390 bp), NMDARε2 5′-CTGCCGGACACCATCACAACAATC-3′ and reverse 5′-CACCGTGGGCTGCCTGAAGAAGTA-3′ (157 bp), GluR1 forward 5′-CAGCGGAGGAAGTGGCAGTG-3′ and reverse 5′-GAGGCGGTGTTTCATTTCTTTGTT-3′ (222 bp), GluR2 forward 5′-CAAGGCAAGGCTGTCAAT-3′ and reverse 5′-GATATCGGATGCCTCTCACC-3′ (219 bp), GluR3 forward 5′-AACTCACAAAGAACACCCAAAACT-3′ and reverse 5′-TCATGCCCGACACCAAGGAG-3′ (216 bp), GluR4 forward 5′-AGCAGGCGTCTTCTACATTC-3′ and reverse 5′-CACGGCCGTTTTCTCCCACACT-3′ (176 and 289 bp), GluR5 forward 5′-CACGACACATATTTGGAGGAC-3′ and reverse 5′-CCGCTTGTTTATTATTATTTAGAA-3′ (244 bp), GluR6 forward 5′-AAGAGGTTTTTAATGATGAGGTC-3′ and reverse 5′-CAGTTTGTGCTTGGGTGAT-3′ (215 bp) and KA2 forward 5′-GCGTGCGCCGACTGACT-3′ and reverse 5′-GCCACGGTGCGGATAGAAT-3′ (168 bp). PCR products were visualized on 2% agarose gels containing ethidium bromide. For western blot analysis, tissue extracts were made in RIPA buffer (1× PBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS). Protein was determined using the Bio-Rad DC Protein Assay. Twenty-five micrograms of protein were separated on 7% NuPage Tris–acetate gels (Invitrogen) according to the manufacturer’s protocol. Equivalent protein loading was confirmed by Coomassie Blue staining of a parallel gel. Proteins were then transferred to a nitrocellulose membrane and chemoluminescent western blotting was performed using the WesternBreeze kit by Invitrogen. Affinity purified goat polyclonal IgG primary antibodies from Santa Cruz Biotechnology [NMDAε1 (sc-1468), NMDAε2 (sc-1469) and NMDAε3 (sc-1470)] were used at a dilution of 1:100.
Whole-cell patch clamp of cerebral cortical neurons in brain slices
Horizontal brain slices (450–600 µm) were prepared within 24 h after birth. The composition of the external recording buffer (ACSF) was: 124 mM NaCl, 3 mM KCl, 2 mM CaCl2, 1.3 mM MgSO4, 1.2 mM Na2HPO4, 25 mM NaCO3 and 10 mM glucose. The slice was continuously perfused (4 ml/min) with ACSF and oxygenated with a 95%/5% mixture of O2/CO2 at room temperature (23°C) in a recording chamber. Drugs were applied by gravity flow through a micropipette placed ~150 µm from the recording site. Frontal cerebral cortex neurons were visualized under an Axioskop 2FS fixed-stage microscope (Zeiss). Patch pipettes were made by a two-stage microelectrode puller (PC-10, Narishige) and filled with internal solution containing 140 mM CsCl, 10 mM HEPES, 5.5 mM BAPTA, 2.0 mM MgCl2 and 2.0 mM Mg–ATP. Whole-cell current was measured with conventional patch clamp techniques (Axopatch 200B, Axon Instruments). Holding potential was –60 mV. The data were acquired by pClamp8 software through a Digidata 1200 interface. Data were plotted with SigmaPlot. GABA, tetrodotoxin and glutamate were obtained from Sigma.
We would like to thank Sing-Ping Huang and Dr Heiner Westphal for their assistance with production of the Dhcr7 mutant mice and Dr Diana Hanes (SAIC) and Dr Raj P. Kapur for their assistance with the histopathological analysis. R.D.S. was supported by the Oregon Child Health Research Center grant PHS 5P30-HD33703-04 from the NICHD and, as a clinical associate physician investigator from the NCRR, grant RR00334-33S3.
To whom correspondence should be addressed. Tel: +1 301 435 4432; Fax: +1 301 480 5791; Email: email@example.com