The Journal of Nutrition Vol. 128 No. 10 October 1998, pp. 1819-1828

Reduced Glucose Consumption in the Curly Tail Mouse Does Not Initiate the Pathogenesis Leading to Spinal Neural Tube Defects^1,2

Marian C. E. Peeters, Jan L.M.C. Geelen^*, Johan W. M. Hekking, Niels Chavannes, Joep P. M. Geraedts^*, and Henny W. M. van Straaten³

Department of Anatomy and Embryology and ^* Department of Genetics, Faculty of Medicine, University of Maastricht, NL-6200 MD Maastricht, Netherlands

	ABSTRACT

Abstract Introduction Methods Results Discussion References

At embryonic stages of neural tube closure, the mouse embryo exhibits a high rate of glycolysis with glucose as the main energy source. In the curly tail mouse, often used as model system for study of human neural tube defects, a delay in closure of the posterior neuropore (PNP) is proposed to be indirectly caused by a proliferation defect in the caudal region. Because glucose is important for proliferation, we tested glucose uptake in curly tail and control embryos, and in a BALB/c-curly tail recombinant strain. The structure and expression of Glut-1, a glucose transporter molecule that is abundantly present during those embryonic stages and that has been mapped in the region of the major curly tail gene, were also studied; however, no strain differences could be demonstrated. Glucose uptake was determined by measuring glucose depletion from the medium in long-term embryo cultures that encompassed the stages of PNP closure and by measuring accumulation of ³H-deoxyglucose in short-term cultures at the stages of early and final PNP closure. Both approaches indicated a reduced glucose uptake by curly tail and recombinant embryos. Surprisingly, the uptake per cell appeared normal, accompanied by a significantly lower DNA content of the mutant embryos. Therefore, it is unlikely that reduced cell proliferation is caused by a reduction in glucose supply during the pathogenesis of the defects in curly tail embryos. The reduced DNA content as well as the reduced glucose uptake per embryo are likely downstream effects of the aberrant proliferation pattern.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Multivitamin supplementation containing folic acid prevents the occurrence of neural tube defects (NTD),⁴ such as spina bifida in humans, in 70% of the cases, leaving a substantial number of NTD that do not respond (Czeizel and Dudas 1992, Wald et al. 1991). The mouse mutant curly tail exhibits NTD that closely resembles human NTD (Seller and Adinolfi 1981). However, this NTD cannot be prevented by supplementation of folic acid or its metabolites. Thus, the curly tail is a genetic model of folate-resistant NTD and might therefore be used to unravel the pathogenesis of this subcategory of human NTD (Greene and Copp 1997, Seller 1994, Van Straaten et al. 1995).

The curly tail mutant arose spontaneously in 1950 and was first reported as an autosomal recessive trait with incomplete penetrance (Grüneberg 1954). More recently, the defect has been considered polygenic, involving an unidentified major gene on chromosome 4 and probably two or three modifier genes (Letts et al. 1995, Neumann et al. 1994). The curly tail phenotypes comprise curled tail (52%) and spina bifida (8%), both of which result from a delay in closure of the posterior neuropore (PNP) and can be recognized in embryos at the developmental stages of 27-29 somites (Copp 1985, Van Straaten et al. 1992). It has been proposed that a temporary enhancement of axial curvature in the caudal region of curly tail embryos induces the closure delay (Brook et al. 1991, Peeters et al. 1996 and 1997, Van Straaten et al. 1993). The aberrant curvature is likely the result of dorsoventral proliferation differences in the caudal region; ventral tissues such as notochord and gut endoderm exhibit a decreased proliferation in embryos showing a delayed PNP closure compared with embryos with a normal closure, whereas the proliferation of the dorsal tissues such as neuroepithelium did not differ (Copp et al. 1988).

During the process of neurulation, the rodent embryo is highly dependent on glucose as its main energy source. Before establishment of a functional chorioallantoic circulation (up to the 30-somite stage), >90% of the consumed glucose is used for glycolysis, whereas the pentose phosphate shunt (PPS) is the major oxidative pathway for the remainder. A gradual shift to oxidative glucose metabolism based on the citric acid cycle and electron transport pathway, and a simultaneously relative reduction in glycolysis and PPS, occur during the final stage of PNP closure and thereafter (Akazawa et al. 1994, Clough and Whittingham 1983, Hunter and Sadler 1988, Shepard et al. 1970). The importance of glucose for neurulation-stage embryos is also demonstrated by the ease with which NTD can be induced by aberrant glucose concentrations, i.e., both hyperglycemic and hypoglycemic circumstances disturb neurulation (Akazawa et al. 1989, Maeda et al. 1993, Reece et al. 1996, Sadler 1993). Because the aberrant proliferation pattern in curly tail embryos is related to the PNP closure delay at a stage during which the energy metabolism is dependent on glycolysis and PPS, glucose uptake might be reduced in those embryos, causing the reduced proliferation.

Glucose can be utilized by a cell only after it is transported across the cell membrane. The carrier Glut-1 is abundantly present in the neurulation-stage embryo and, moreover, is expressed in neuroepithelial cells at an increasing level during neural tube closure (Maeda et al. 1993, Trocino et al. 1994). Because of the close proximity of curly tail and Glut-1 on chromosome 4 (Hogan et al. 1991, Neumann et al. 1994), Glut-1 might be involved in a reduced glucose uptake in curly tail.

In this study, we tested glucose uptake in curly tail and control (BALB/c and CBA/J) embryos and in a BALB/c-curly tail recombinant strain, in which curly tail has been placed on a BALB/c genetic background. The last-mentioned strain was constructed to provide a background-matched comparison between BALB/c-curly tail and BALB/c, because the parental curly tail strain has a different background from all existing inbred strains. It appeared that the glucose consumption, which was determined in vitro from the stages of early to final PNP closure, was lower in curly tail and recombinant embryos than in control embryos. To determine more precisely the developmental stage and the embryonic region of altered consumption, we measured separately the accumulation of 2-deoxy-D-[1-³H] glucose (DO-glucose) at the stages of early and final PNP closure for the caudal and cranial embryonic regions and for the extra-embryonic membranes.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Mouse strains.  The mouse mutant curly tail arose spontaneously in a female of the GFF inbred stock; this female was subsequently mated with a CBA/Gr male (Grüneberg 1954). The curly tail stock was derived from this mating and is kept as a closed, random-bred colony. The nonmutant inbred BALB/c and CBA/J mice were obtained commercially (Broekman, Someren, Netherlands). The BALB/c-curly tail recombinant strain was constructed via cross-intercross cycles as follows: curly tail mice were mated with BALB/c mice, followed by an intercross of the heterozygous offspring; the descendants of the intercross were subdivided into those with a curly tail phenotype (curled or kinked tail, with or without spina bifida) and those with a normal phenotype (straight tail without spina bifida). Mice with a curly tail phenotype were mated with BALB/c, again followed by an intercross. Offspring of this intercross, with a curly tail phenotype (genetic background of 75% BALB/c and 25% curly tail) were kept as a separate colony (recombinant-2). Recombinant-3 mice (87.5% BALB/c and 12.5% curly tail) were obtained after a third cross-intercross cycle. Mice consumed water and commercial diet pellets (SRM-A diet, Hope Farms, Woerden, Netherlands) ad libitum and were maintained on a light:dark cycle with the dark period from 2000 to 600 h. Animal care and handling were in compliance with applicable guidelines from the Dutch Ministry of Agriculture.

Glucose consumption during PNP closure.  Males and females of the mouse strains curly tail, recombinant-2, BALB/c and CBA/J were paired for mating overnight; females were checked for copulation plugs the following morning (d 0). On d 9, female mice were killed by cervical dislocation and the uterine horns were explanted into prewarmed (37°C) Hanks' buffer containing 1% penicillin (Gibco, Life Technologies, Breda, Netherlands), 1% streptomycin (Gibco) and 10% newborn calf serum (Gibco). Myometrium, endometrium and Reichert's membrane were removed, leaving intact extra-embryonic membranes (amnion and yolk sac with ectoplacental cone) around the embryo. Head and crown-rump lengths were measured as described previously (Van Straaten et al. 1992). These lengths were subsequently used to select embryonic developmental stages varying from 18 to 22 somites (the stage of early PNP closure) (Peeters et al. 1996).

Embryos were cultured according to the method of New et al. (1973) as follows. Each embryo was transferred with the use of a glass pipet to a separate culture bottle containing 1.0 mL culture medium that contained a minimum amount of Hanks' buffer. The culture medium consisted of immediately centrifugated, heat-inactivated serum from adult Wistar rats, supplemented with 1% penicillin (Gibco) and 1% streptomycin (Gibco). Culture bottles were gassed and placed on a rotator at 38°C. After 16.5 h of culture, embryos were checked for viability (presence of heart beat and yolk sac circulation) and transferred to prewarmed Hanks' buffer. The culture medium was immediately frozen and stored at 20°C for determination of the glucose level. Extra-embryonic membranes were dissected from the cultured embryos, and the number of somites was counted. Only embryos of the developmental stages of 27-33 somites were processed further because these are the stages of final PNP closure. Head length, crown-rump length and PNP length were measured, and the PNP was classified according to the degree of delay of PNP closure (Copp 1985). Because glucose is the main energy source at the stages studied, we expected the glucose consumption to be related to the amount of increase in embryonic size. Therefore, embryonic growth had to be determined. Growth was defined as gain in head length during the culture period (head length at the end minus head length at the start of the culture).

Control samples of culture medium were taken before and after culture to determine the initial glucose concentration and the effect of culture conditions on the glucose concentration. Glucose concentration in the culture media was determined by using COBAS BIO (Roche diagnostica, Basel, Switzerland), according to a standard spectrophotometric hexokinase method (Boehringer Mannheim, Almere, Netherlands). Glucose consumption per embryo was defined as the percentage depletion of glucose from the culture medium during the culture period, calculated as follows:

Intra-embryonic accumulation of DO-Glucose at two developmental stages.  On d 9 of pregnancy, embryos of the curly tail, recombinant-3 and BALB/c mice were collected as described above (stage of 16-23 somites, early PNP closure). Recombinant-3 instead of recombinant-2 embryos were used in this part of the study because they had become available. On d 10, embryos of curly tail and BALB/c were collected (stage of 26-33 somites, final PNP closure). In each culture bottle, two embryos were placed in 1.0 mL culture medium and kept for an adaptation period of at least 30 min at 38°C on a rotator. The culture medium was then supplemented with 20 µL 2-deoxy-D-[1-³H] glucose (Amersham, 's-Hertogenbosch, Netherlands) (14 nmol/L final concentration, 7.4 MBq/L). Embryos were further cultured for 1.5 h. After being checked for viability (see above) and rinsed in Hanks' buffer, the extra-embryonic membranes were removed and collected in a scintillation tube containing 200 µL Hanks' buffer. The embryo was measured and staged as described above, and then bisected transversely at the level of the penultimate somite; the length of the caudal part was measured. The cranial and caudal fractions were collected separately in scintillation tubes containing 200 µL Hanks' buffer. The cranial, caudal and membrane fractions were sonicated (5-10 s, 20 kHz, 2 µm amplitude; MSE sonicator type L667); samples of 20, 100 and 40 µL, respectively, were taken for DNA assay. To the remainder of each fraction, 4 mL formula 989 (Packard, Groningen, Netherlands) was added and radioactivity was measured by using a Liquid Scintillation Counter 1414 (Win Spectral, Wallac EG&G, Turku, Finland).

The accumulation of DO-glucose was evaluated as an accumulation per fraction and as an accumulation related to the cellular part of each fraction. For the latter evaluation, DNA contents of the fractions had to be determined. DNA samples were supplemented with an autoclaved saline buffer (10 mmol/L NaH₂PO₄, 40 mmol/L Na₂HPO₄, 2 mol/L NaCl, 2 mmol/L EDTA, pH 7.4) to a total volume of 1000 µL. After the addition of 50 µL Hoechst 33258, the DNA content was determined spectrofluorimetrically (type ZFM4, Zeiss, Weesp, Netherlands). A dilution series of calf thymus DNA was used for calibration.

DO-glucose accumulation was evaluated separately for the caudal, cranial and membrane fractions and was expressed as mean DO-glucose (pmol) per fraction. Moreover, these accumulations per fraction were corrected for head length (pmol/mm), crown-rump length (pmol/mm) and somite number. The accumulation in the caudal fraction was also corrected for the length of the caudal part (pmol/µm). Additionally, the accumulations per fraction were corrected for DNA content (pmol/µg).

cDNA sequencing and probe construction.  Total cellular RNA was extracted from spleens of adult female curly tail, BALB/c and CBA/J mice using the Qiashredder and RNeasy system according to the manufacturer's protocol (Qiagen GmbH, Hilden, Germany). First strand cDNA synthesis was performed using the total RNA sample and SUPERSCRIPT II RT (Qiagen).

The sequence of the primers for amplification and sequencing were derived from the published Glut-1 cDNA sequence (Kaestner et al. 1989), as shown in Table 1 and Figure 1. PCR amplification was performed in a 100-µL reaction with Taq polymerase. Samples were initially denatured for 2 min at 95°C, followed by 35 cycles of 1 min at 94°C, 1 min at 65°C and 2 min at 72°C, and an extension after the last cycle of 7 min at 72°C. Smaller fragments for sequencing were constructed under the same PCR conditions for 25 cycles. Before sequencing, the PCR products were purified on an agarose gel by using the QIAquick gel extraction kit (Qiagen). Products were sequenced automatically using an ABI 377 Sequencer (Perkin-Elmer, Norwalk, CT).

View larger version (5K): [in this window] [in a new window]   Fig 1. Position of the primers used for Glut-1 cDNA amplification and sequencing.

RFLP analysis.  DNA was isolated from spleens of adult female curly tail, BALB/c and CBA/J mice by a high salt extraction procedure (Müllerbach et al. 1989). The DNA was tested for restriction fragment length polymorphisms (RFLP) at the Glut-1 locus, using the enzymes HindIII, XbaI, BglII, BamI, EcoRI, StuI and PvuII (Boehringer Mannheim).

The cDNA probe (total fragment CBA/J mouse; Table 1, Fig. 1) was labeled with [-³²P] dCTP by the RadPrime DNA labeling system (Gibco). The probe was purified on a Sephadex column and used for hybridization to the Southern blots in Rapid-hyb buffer (Amersham).

Immunohistochemistry.  On d 9 and 10 of pregnancy, embryos of curly tail, BALB/c and recombinant-3 were collected as described above and, additionally, extra-embryonic membranes were removed. Embryos were fixed for 1 h by using 4% paraformaldehyde in PBS, rinsed in PBS, immersed in 20% sucrose (overnight), embedded in O.C.T. compound (Tissue-Tek, Miles, Elkhart, IN) and quick-frozen in liquid nitrogen. Sections of the PNP region were cut at 10 µm on a cryostat (type 600, Anglia, Cambridge, UK), collected on polylysine-coated glass slides and air-dried. After pretreatment with 0.09 mol/L H₂O₂ (10 min) and goat serum in 10 g/L bovine serum albumin (BSA) (1:5), sections were incubated overnight at 4°C with anti-Glut-1 (rabbit polyclonal antiserum AB1340, Chemicon, Temecula, CA) at a dilution of 1:5000. Control slides were incubated with 1% BSA only. Subsequently, the slides were incubated with the secondary antibody (peroxidase-conjugated goat anti-rabbit; P0448, DAKO, Glostrup, Denmark) at a dilution of 1:100 for 90 min at room temperature. Immunolabels were stained by incubation with 3, 3'-diaminobenzidine with 9 mmol/L H₂O₂ for 5 min.

Staining intensities of neuroepithelium, gut endoderm, dorsal and ventral parts of mesoderm were measured in at least three sections per embryo of three to five embryos per strain at both developmental stages using a Quantimet 500 image-analyzer (Leica, Rijswijk, Netherlands).

Statistical analysis.  Glucose consumption by the embryos of the four different mouse strains from the stage of early to the stage of final PNP closure was compared using an ANOVA model with the mouse strains as factor, embryonic growth (gain in head length) as covariate and Bonferroni correction. The covariate was included in the model because embryos had differential growth during the culture period and, moreover, glucose consumption appeared closely related to growth.

To test for a difference in glucose consumption between embryos with normal versus delayed PNP classifications, the data from curly tail and recombinant-2 strains were pooled and PNP classifications were compared with a t test for independent samples.

The intra-embryonic accumulation of DO-glucose by embryos of the different mouse strains at the stage of early PNP closure was compared using a one-way ANOVA with Bonferroni correction. The accumulation of DO-glucose of curly tail and BALB/c embryos at final PNP closure was evaluated using the t test for independent samples.

Staining intensities after Glut-1 immunolabeling were compared using the Kruskal-Wallis test. Statistical analysis were performed using SPSS for Windows (Chicago, IL), release 6.0.

	RESULTS

Abstract Introduction Methods Results Discussion References

Glucose consumption during PNP closure.  In a long-term culture of 16.5 h, the glucose consumption of embryos during the critical phase of PNP closure was established for the two mutant and two wild-type mouse strains. Figure 2 shows a positive relationship (r = 0.68; P < 0.001) between embryonic growth and glucose consumption, as expected. The increase in consumption was similar in the four mouse strains used. However, strain differences were observed in the amount of glucose consumption. Consumption was ~10% lower for curly tail and recombinant-2 embryos than for BALB/c and CBA/J embryos (P < 0.001).

View larger version (16K): [in this window] [in a new window]   Fig 2. Glucose consumption in mutant and control embryos during closure of the posterior neuropore. The relationship between the consumption of glucose and embryonic growth (as defined by head length gain obtained during the 16.5-h culture period) is shown. Data were evaluated statistically using an ANOVA model with growth as covariate and strain as factor. Glucose consumption appeared to be positively related with embryonic growth (r = 0.68; P < 0.001). The amount of increase in glucose consumption was similar for the mutant and control embryos, as appears from the similar slopes of the regression lines. However, the glucose consumption was lower (~10%) for the curly tail and recombinant-2 embryos, compared with the nonmutant BALB/c and CBA/J embryos (P < 0.001). For clarity of graphical presentation, a regression line is drawn for each mouse strain.

Both curly tail and recombinant-2 embryos had normal as well as delayed PNP, whereas the control embryos were all of the normal phenotype (Table 2). When the data for the mutant embryos were pooled and analyzed separately for each phenotype, we observed no differences in glucose consumption. Similarly, no correlation could be demonstrated between the length of the PNP and the glucose consumption (data not shown).

The mean initial glucose concentration in the culture medium was 9.84 mmol/L; after the culture period, the mean concentration in the control samples, in which no embryo had been present, was 10.15 mmol/L. This implies that culture conditions caused a 3% increase (P < 0.001) in glucose concentration, probably caused by evaporation at the moment at which the gas mixture was added to the culture bottle.

Intra-embryonic accumulation of DO-glucose at two developmental stages.  The accumulation of DO-glucose was evaluated in several ways. At the stage of early PNP closure (16-23 somites), curly tail embryos accumulated less DO-glucose, expressed as mean accumulation per embryonic fraction, than BALB/c embryos in both their cranial and caudal fractions (Table 3). The accumulation in the membrane fraction was not different between those mouse strains. The cranial fraction of recombinant-3 embryos showed that DO-glucose accumulation was comparable to that of curly tail, whereas the accumulation in the caudal fraction was significantly higher than in the curly tail (Table 3). When DO-glucose accumulation in the caudal fraction at the stage of early PNP closure was corrected for head length or length of the isolated caudal part, curly tail embryos showed reduced accumulation compared with BALB/c embryos (data not shown). The same tendencies were observed after correction for crown-rump length and somite number. However, the cranial and membrane fractions did not show a consistently different accumulation for the mouse strains. Evaluating those corrected accumulations in the fractions of the recombinant-3 embryos demonstrated similar accumulations as for the curly tail embryos (data not shown).

However, in contrast to our findings for DO-glucose accumulation per fraction, accumulation per µg DNA showed no significant differences in the caudal, cranial and membrane fractions for the curly tail, recombinant-3 and BALB/c embryos, except for the caudal region, which had a higher accumulation for recombinant-3 than for BALB/c embryos (Table 4).

In embryos at the stage of final PNP closure (26-33 somites), accumulation of DO-glucose in curly tail and BALB/c embryos was not different, either when expressed per fraction (Table 3), or after correction for embryo size or somite number (data not shown), or after correction for DNA (Table 4).

Structure and expression of Glut-1.  Before constructing smaller fragments of Glut-1 cDNA for sequencing, the total cDNA fragment constructed by PCR with Glut-1 primers (Table 1, Fig. 1) was separated on an agarose gel and appeared to have the expected 1.4-1.5 kb. Sequence analysis of Glut-1 cDNA resulted in identical sequences for the curly tail, BALB/c and CBA/J mice.

Figure 3 shows the results of the test for Glut-1 polymorphisms using the restriction enzymes Eco RI, Stu I and Pvu II. Identical patterns were found for the mouse strains curly tail, CBA/J and BALB/c. With the enzymes Hind III, Xba I, Bgl II and Bam I, identical patterns were found for those mouse strains as well (data not shown). This means that for these enzymes, no restriction fragment length polymorphisms exist at the Glut-1 locus.

View larger version (67K): [in this window] [in a new window]   Fig 3. Analysis of polymorphisms at Glut-1 using restriction enzymes. Size of the DNA fragments on an agarose gel after incubation of total DNA with restriction enzymes, electrophoresis and labeling with the Glut-1 probe is shown. The restriction enzymes Eco RI, Stu I and Pvu II each resulted in identical fragment lengths for the curly tail, CBA/J and BALB/c mice. This means that the sites of recognition of the enzymes were identical in the three mouse strains, indicating that no mutation was present at those sites.

Use of Glut-1 antiserum to stain cross sections of the PNP region at the early and final PNP closure stages gave strongest staining in gut endoderm and neuroepithelium, weaker staining in dorsal mesoderm and weakest staining in ventral mesoderm (Fig. 4). The expression at the stage of final PNP closure was significantly reduced compared with the younger stage. Within the PNP region, a craniocaudal gradient of expression was observed that was most pronounced at the stage of final PNP closure, i.e., a reduction in the expression in gut endoderm and neuroepithelium and an increase in the dorsal and ventral mesoderm. No strain differences in Glut-1 expression could be detected.

View larger version (151K): [in this window] [in a new window]   Fig 4. Glut-1 immunostaining in the posterior neuropore (PNP) region of a curly tail mouse embryo at the stage of early PNP closure. High levels of Glut-1 staining are observed in the gut endoderm (g) and neuroepithelium (n), lower levels in the dorsal mesoderm (d), whereas the ventral mesoderm (v) is most weakly stained. Bar = 50 µm.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Consumption of glucose during PNP closure was determined in a long-term embryo culture by measuring the depletion of glucose from the culture medium. The depletion was less in curly tail cultures than in BALB/c and CBA/J cultures. This means that the glucose consumption of curly tail embryos was less during the developmental stages at which the closure of the caudal neural tube occurs. However, it was not clear whether glucose consumption was reduced during the whole period of PNP closure. Therefore, a short-term embryo culture using DO-glucose was set up to test the glucose uptake of embryos at the stages of early and final PNP closure, which corresponded to the developmental stages at the start and end of the long-term culture. DO-glucose competes with glucose at the cellular level for binding to the glucose transporter and for phosphorylation. When the concentration of DO-glucose is kept below the level at which it has inhibitory effects, as performed in this study, DO-glucose accumulation can be used as a measure of glucose uptake (Hunter and Tugman 1995). To determine whether the reduction in glucose uptake varied in different parts of the conceptus, the DO-glucose accumulation was measured separately in the caudal, cranial and membrane fractions. Evaluating the glucose accumulation as mean per fraction and after correcting for embryonic length displayed a reduction for the caudal and cranial fractions but not for the membrane fraction of curly tail embryos at the stage of early PNP closure compared with BALB/c embryos. At the stage of final PNP closure, the accumulation of glucose by curly tail and BALB/c embryos was not significantly different.

Thus, both culture methods indicate a reduced glucose uptake by curly tail embryos compared with nonmutant embryos during the period of PNP closure when evaluated per embryo or per fraction. The reduction appears especially marked at the stage of early PNP closure and most pronounced in the caudal region, with a normalization of glucose uptake at the stage of final PNP closure.

Reduced glucose uptake is not due to the extra-embryonic membranes.  In both experimental designs used in this study, glucose in the culture medium has to be absorbed by the yolk sac in order to reach the embryo. This is facilitated by several transporter molecules of the Glut family, which transport glucose across the cell membrane according to a concentration gradient (Smith and Gridley 1992). After uptake by the yolk sac, glucose is transported to the embryo via the yolk sac vessels and via diffusion through the yolk sac cavity, amnion, amniotic cavity and embryonic surface. Inside the embryo, transport of glucose across the cell membrane is also mediated by several Glut isoforms, of which Glut-1 is abundantly present at this stage of development (Maeda et al. 1993, Matsumoto et al. 1995, Trocino et al. 1994). Subsequently, the enzyme hexokinase converts glucose to glucose-6-phosphate, which is then used in several metabolic pathways.

For the membrane fraction, it appeared that the DO-glucose accumulation was almost identical between the mouse strains. It is therefore unlikely that the reduced glucose uptake by the curly tail and recombinant embryos, as found in this study, is due to a reduced uptake by the yolk sac. A suboptimal function in any other part of the transport route could therefore be responsible for the reduced glucose uptake.

Normal structure and expression suggest a normal function of Glut-1.  The BALB/c-curly tail recombinant strain is the result of selection for the curly tail phenotype while backcrossing the curly tail gene onto the BALB/c-genetic background. The recombinant-2 embryos have 75% of the genetic background of BALB/c, but their glucose consumption is almost identical to that of the curly tail embryos and lower than the consumption of the BALB/c embryos. Moreover, 87.5% of the genetic background of the recombinant-3 embryos is identical to that of BALB/c, and yet their glucose accumulation resembles more that of the curly tail embryos. This suggests that the curly tail genetic defect itself is responsible, either directly or indirectly, for the reduced glucose uptake. The major gene responsible for the curly tail defect has been mapped on chromosome 4. Glut-1 also maps on chromosome 4, close to the curly tail gene and, moreover, is important for glucose metabolism in mouse embryos during neurulation. In humans, polymorphisms for Glut-1 have been described (Kaku et al. 1990, Shows et al. 1987, Xiang et al. 1987); these are probably related to functional properties of Glut-1 (Li et al. 1988, Tao et al. 1995). It was therefore postulated that the function of Glut-1 has a regulatory role in the curly tail defect.

In this study, restriction enzyme analysis and DNA sequencing were used to test whether the structure of Glut-1 could be involved in the curly tail defect. Both approaches failed to show differences in Glut-1 between the curly tail and nonmutant BALB/c and CBA/J mouse strains. However, five differences were found between the sequence obtained and the sequence reported by Kaestner et al. (1989) and one difference was found compared with the data of Reed et al. (1990) (Table 5). All differences result in amino acid substitutions. The substitution at codon 250, at the cytoplasmic loop between transmembrane domains 6 and 7, differs from both previously reported sequences. This substitution is within the group of positively charged polar amino acids and probably does not markedly alter the function of Glut-1. The substitution at codon 403 has also been published by Hogan et al. (1991). The amino acid substitutions found in this study do not correspond with amino acids of Glut-1 to which functional properties have been ascribed (Garcia et al. 1992, Mueckler et al. 1994, Wellner et al. 1995).

Strain differences were not detected in the expression pattern of the Glut-1 protein, which was studied immunohistochemically by using a polyclonal antiserum. Expression was strongest in the neuroepithelium and gut endoderm and weaker in the dorsal and ventral mesoderm. Moreover, there was reduced expression at the stage of final PNP closure compared with the younger stage. The expression pattern of Glut-1 found in this study closely matches that described in other studies (Maeda et al. 1993, Matsumoto et al. 1995, Trocino et al. 1994).

Because structure and expression of Glut-1 are identical for curly tail and BALB/c embryos, it is unlikely that an aberrant function of Glut-1 is causally involved in the processes underlying NTD in curly tail. However, both recombinant-2 and -3 embryos demonstrated glucose uptake that more closely resembles curly tail than BALB/c. This indicates that glucose may be linked to the pathogenetic process leading to a delay in the closure of the PNP, although Glut-1 is not the primary defect.

Relationship between glucose and neurulation.  Neurulation appears to become disturbed easily by abnormalities of the surrounding glucose level (Reece et al. 1996, Sadler 1993). Maternal hyperglycemia during pregnancy is associated with an intra-embryonic depletion of inositol, which is thought to be responsible in part for the defects of neurulation (Cockroft 1988, Sussman and Matschinsky 1988). Closure of the hindbrain in the curly tail strain is extremely sensitive to inositol deficiency compared with other mouse strains (Cockroft et al. 1992), although it has been shown recently that inositol uptake by curly tail embryos is normal. Nevertheless, administration of additional inositol can prevent spinal NTD in curly tail mice (Greene and Copp 1997). In this study, the initial glucose concentration in both long-term and short-term cultures was ~10 mmol/L, which is below the level that appears to be teratogenic for mouse embryos (Sadler et al. 1993). Thus, the curly tail embryos were not exposed to a hyperglycemic milieu in this study. In fact, the curly tail embryos demonstrated a reduced glucose uptake, thereby possibly creating an intra-embryonic hypoglycemic situation. It is unlikely therefore that the delay in PNP closure in curly tail embryos is due to a hyperglycemia-induced inositol depletion.

Glucose uptake per cell is normal in curly tail embryos.  In the short-term culture, the DO-glucose accumulation was evaluated not only per fraction but also per microgram DNA. Surprisingly, this latter analysis yielded equal results for the curly tail and nonmutant embryos. This means that the glucose uptake by curly tail embryos at the stage of early PNP closure is reduced per embryo, but normal per embryonic cell. Because of this discrepancy, the DNA values were evaluated more precisely. From Table 6, it appears that the DNA content of the curly tail embryo is reduced, indicating a reduced number of cells. At the stage of final PNP closure, the glucose uptake by curly tail and BALB/c embryos was not different, irrespective of the evaluation per embryo or per cell. The DNA content at these stages is slightly reduced in curly tail, but not significantly different (Table 6). The indication that the glucose uptake per cell is normal fits with the results of normal structure and expression of Glut-1 in this study, which imply a normal function.

View larger version (14K): [in this window] [in a new window]   Fig 5. Embryonic sizes of curly tail and BALB/c embryos. The relationships between the size (head length and crown-rump length) and the developmental stage (number of somites) in the embryo are almost identical for curly tail and BALB/c. Individual data points are derived from embryos after the short-term culture and from some embryos collected additionally. Regression lines (r > 0.90) are drawn.

A reduced number of cells in the whole embryo is the likely result of reduced cell proliferation. Previously, a reduced proliferation was described for the caudal part of the curly tail embryo at the stage of final PNP closure, especially for the gut endoderm and the notochord (Copp et al. 1988). In this study, however, the DNA content was reduced not only in the caudal part, but in the whole embryo as well. Moreover, the DNA content at the stage of final PNP closure was slightly reduced, but the effect was more pronounced at the stage of early PNP closure. Therefore, it would be worthwhile to study cell proliferation at both stages in the caudal and cranial parts of the embryo.

The reduced proliferation in the curly tail embryo has been proposed to cause the delay in closure of the PNP by its effect on the curvature of the caudal body axis (Brook et al. 1991, Peeters et al. 1997 and 1998). To resolve the basis of the proliferation defect, several growth-related molecules have been studied, and some are thought to be involved in the pathogenetic process of the curly tail. Hyaluronan, which is located in the extracellular matrix, is reduced in the PNP region (Copp and Bernfield 1988); the retinoic acid receptors RAR- and RAR- are down-regulated in hindgut and tail bud, respectively (Chen et al. 1995); transferrin is bound in reduced quantities on the luminal surface of the hindgut endoderm (Hoyle et al. 1996). At present, it is unknown how these and other molecules are related to each other and how they might determine the proliferation pattern in the curly tail embryo.

The reduced DNA content of the curly tail embryo suggests that these embryos would be smaller than the nonmutant embryos. However, embryos from curly tail and BALB/c mice had identical numbers of somites, head lengths and crown-rump lengths on embryonic d 9 and 10. Moreover, when the relationship between the embryonic size (head length and crown-rump length) and its developmental stage (number of somites) was considered, it appeared to be equal for the curly tail and BALB/c embryos (Fig. 5). Thus, overall growth retardation of the mutant embryos does not seem to occur, although the length parameters may be insufficient to detect the expected relatively small changes in volume. Moreover, the volume of the embryo is determined not only by the number of cells, but also by cell size and volume of the extracellular matrix, which might have been increased in the curly tail embryo. Whether these parameters are involved is yet to be determined.

From this study, it can be concluded that the glucose uptake by the curly tail embryo is reduced during PNP closure, especially at its early stage. Unexpectedly, the DNA content is reduced, accompanied by a normal glucose uptake per cell. In addition, the structure and expression of the glucose carrier Glut-1, which is abundantly present at this stage, appear unaffected. It is therefore unlikely that the reduced proliferation is caused by a reduced glucose supply in the pathogenetic process of the curly tail leading to NTD. On the contrary, the reduced glucose uptake per embryo during early PNP closure is probably due to a temporarily reduced general proliferation pattern in the curly tail embryo, indicating that a disturbed glucose intake by the embryo might be a downstream effect in the subcategory of folate-resistant NTD.

	FOOTNOTES

Manuscript received 7 November 1997. Initial reviews completed 20 March 1998. Revision accepted 18 May 1998.

	ACKNOWLEDGMENTS

The authors acknowledge R. Jongbloed for her assistance in sequencing, P. van Dijk for his histotechnical and photographical assistance, and E. Raadschilders for his technical assistance. We thank A. Copp for critically reading the manuscript.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Akazawa M., Akazawa S., Hashimoto M., Akashi M., Yamazaki, Tahara D., Yamamoto H., Yamaguchi Y., Nakanishi T., Nagataki S. Effects of brief exposure to insulin-induced hypoglycemic serum during organogenesis in rat embryo culture. Diabetes 1989; 38:1573-1578[Abstract]
• Akazawa S., Unterman T., Metzger B. E. Glucose metabolism in separated embryos and investing membranes during organogenesis in the rat embryo. Metabolism 1994; 43:830-835[Medline]
• Brook F. A., Shum E.S.W., Van Straaten H.W.M., Copp A. J. Curvature of the caudal region is responsible for failure of neural tube closure in the curly tail (ct) mouse embryo. Development 1991; 113:671-678[Abstract]
• Chen W. H., Morriss-Kay G. M., Copp A. J. Genesis and prevention of spinal neural tube defects in the curly tail mutant mouse: involvement of retinoic acid and its nuclear receptors RAR- and RAR-. Development 1995; 121:681-691[Abstract]
• Clough J. R., Whittingham D. G. Metabolism of [¹⁴C]glucose by postimplantation mouse embryos in vitro. J. Embryol. Exp. Morphol. 1983; 74:133-142[Medline]
• Cockroft D. L. Changes with gestational age in the nutritional requirements of postimplantation rat embryos in culture. Teratology 1988; 38:281-290[Medline]
• Cockroft D. L., Brook F. A., Copp A. J. Inositol deficiency increases the susceptibility to neural tube defects of genetically predisposed (curly tail) mouse embryos in vitro. Teratology 1992; 45:223-232[Medline]
• Copp A. J. Relationship between timing of posterior neuropore closure and development of spinal neural tube defects in mutant (curly tail) and normal mouse embryos in culture. J. Embryol. Exp. Morphol. 1985; 88:39-54[Medline]
• Copp A. J., Bernfield M. Accumulation of basement membrane-associated hyaluronate is reduced in the posterior neuropore region of mutant (curly tail) mouse embryos developing spinal neural tube defects. Dev. Biol. 1988; 130:583-590[Medline]
• Copp A. J., Brook F. A., Roberts H. J. A cell-type-specific abnormality of cell proliferation in mutant (curly tail) mouse embryos developing spinal neural tube defects. Development 1988; 104:285-295[Abstract]
• Czeizel A. E., Dudas I. Prevention of the first occurrence of neural tube defects by periconceptional vitamin supplementation. N. Engl. J. Med. 1992; 327:1832-1835[Abstract]
• Garcia J. C., Strube M., Leingang K., Keller K., Mueckler M. M. Amino acid substitutions at tryptophan 388 and tryprophan 412 of the HepG2 (GLUT1) glucose transporter inhibit transport activity and targeting to the plasma membrane in Xenopus oocytes. J. Biol. Chem. 1992; 267:7770-7776[Abstract/Free Full Text]
• Greene N.D.E., Copp A. J. Inositol prevents folate-resistant neural tube defects in the mouse. Nature Med. 1997; 3:60-66[Medline]
• Grüneberg H. Genetical studies on the skeleton of the mouse. VIII curly tail. J. Genet. 1954; 52:52-67
• Hogan A., Heyner S., Charron M. J., Copeland N. G., Gilbert D. J., Jenkins N. A., Thorens B., Schultz G. A. Glucose transporter gene expression in early mouse embryos. Development 1991; 113:363-372[Abstract]
• Hoyle C., Henderson D. J., Matthews D. J., Copp A. J. Transferrin and its receptor in the development of genetically determined neural tube defects in the mouse embryo. Dev. Dyn. 1996; 207:35-46[Medline]
• Hunter E. S., Sadler T. W. Embryonic metabolism of foetal fuels in whole-embryo culture. Toxicol. In Vitro 1988; 2:163-167
• Hunter E. S. III, Tugman J. A. Inhibitors of glycolytic metabolism affect neurulation-staged mouse conceptuses in vitro. Teratology 1995; 52:317-323[Medline]
• Kaestner K. H., Christy R. J., McLenithan J. C., Braiterman L. T., Cornelius P., Pekala P. H., Lane M. D. Sequence, tissue distribution, and differential expression of mRNA for a putative insulin-responsive glucose transporter in mouse 3T3-L1 adipocytes. Proc. Natl. Acad. Sci. U.S.A. 1989; 86:3150-3154[Abstract/Free Full Text]
• Kaku K., Matsutani A., Mueckler M., Permutt M. A. Polymorphisms of HepG2/Erythrocyte glucose-transporter gene. Diabetes 1990; 39:49-56[Abstract]
• Letts V. A., Schork N. J., Copp A. J., Bernfield M., Frankel W. N. A curly-tail modifier locus, mct1, on mouse chromosome 17. Genomics 1995; 29:719-724[Medline]
• Li S. R., Baroni M. G., Oelbaum R. S., Stock J., Galton D. J. Association of genetic variant of the glucose transporter with non-insulin-dependent diabetes mellitus. The Lancet 1988; 2:368-370[Medline]
• Maeda Y., Akazawa S., Akazawa M., Takao Y., Trocino R. A., Takino H., Kawasaki E., Yokota A., Okuno S., Nagataki S. Glucose transporter gene expression in rat conceptus during early organogenesis and exposure to insulin-induced hypoglycemic serum. Acta Diabetol. 1993; 30:73-78[Medline]
• Matsumoto K., Akazawa S., Ishibashi M., Trocino R. A., Matsuo H., Yamasaki H., Yamaguchi Y., Nagamatsu S., Nagataki S. Abundant expression of Glut-1 and Glut-3 in rat embryo during early organogenesis period. Biochem. Biophys. Res. Commun. 1995; 209:95-102[Medline]
• Mueckler M., Weng W., Kruse M. Glutamine 161 of Glut1 glucose transporter is critical for transport activity and exofacial ligand binding. J. Biol. Chem. 1994; 269:20533-20538[Abstract/Free Full Text]
• Müllerbach R., Lagoda P.J.L., Weller C. An efficient salt-chloroform extraction of DNA from blood and tissues. Trends Genet. 1989; 5:391[Medline]
• Neumann P. E., Frankel W. N., Letts V. A., Coffin J. M., Copp A. J., Bernfield M. Multifactorial inheritance of neural tube defects: localization of the major gene and recognition of modifiers in ct mutant mice. Nature Genet. 1994; 6:357-362[Medline]
• New D.A.T., Coppola P. T., Terry S. Culture of explanted rat embryos in rotating tubes. J. Reprod. Fertil. 1973; 35:135-138[Abstract/Free Full Text]
• Peeters M.C.E., Hekking J.W.M., Vainas T., Drukker J., Van Straaten H.W.M. Spatio-temporal curvature pattern of the caudal body axis for non-mutant and curly tail mouse embryos during the period of caudal neural tube closure. Anat. Embryol. 1997; 195:259-266[Medline]
• Peeters M.C.E., Shutte B., Lenders M.H.Y.N., Hekking J.W.M., Drukker J., Van Straaten H.W.M. Role of differential cell proliferation in the tail bud in aberrant mouse neurulation. Dev. Dyn. 1998; 211:382-389[Medline]
• Peeters M.C.E., Shum A.S.W., Hekking J.W.M., Copp A. J., Van Straaten H.W.M. Relationship between altered axial curvature and neural tube closure in normal and mutant (curly tail) mouse embryos. Anat. Embryol. 1996; 193:123-130[Medline]
• Reece E. A., Wiznitzer A., Homko C. J., Hagay Z., Wu Y. K. Synchronization of the factors critical for diabetic teratogenesis: an in vitro model. Am. J. Obstet. Gynecol. 1996; 174:1284-1288[Medline]
• Reed B. C., Shade D., Alperovich F., Vang M. 3T3-L1 adipocyte glucose transporter (HepG2 class): sequence and regulation of protein and mRNA expression by insulin, differentiation, and glucose starvation. Arch. Biochem. Biophys. 1990; 279:261-274[Medline]
• Sadler T. W., Denno K. M., Hunter E. S. Effects of altered maternal metabolism during gastrulation and neurulation stages of embryogenesis. Ann. N.Y. Acad. Sci. 1993; 678:48-61[Medline]
• Seller, M. J. (1994) Vitamins, folic acid and the cause and prevention of neural tube defects. In: Neural Tube Defects, Ciba Foundation Symposium 181, pp.161-179. Wiley, Chichester, UK.
• Seller M. J., Adinolfi M. The curly-tail mouse: an experimental model for human neural tube defects. Life Sci. 1981; 29:1607-1615[Medline]
• Shepard T. H., Tanimura T., Robkin M. A. Energy metabolism in early mammalian embryos. Dev. Biol. (suppl.) 1970; 4:42-58
• Shows T. B., Eddy R. L., Byers M. G., Fukushima Y., Dehaven C. R., Murray J. C., Bell G. I. Polymorphic human glucose transporter gene (GLUT) is on chromosome 1p31.3-p35. Diabetes 1987; 36:546-549[Abstract]
• Smith D. E., Gridley T. Differential screening of a PCR-generated mouse embryo cDNA library: glucose transporters are differentially expressed in early postimplantation mouse embryos. Development 1992; 116:555-561[Abstract]
• Sussman I., Matschinsky F. M. Diabetes affects sorbitol and myo-inositol levels of neuroectodermal tissue during embryogenesis in rat. Diabetes 1988; 37:74-81[Abstract]
• Tao T., Tanizawa Y., Matsutani A., Kaneko T., Kaku K. HepG2/erythrocyte glucose transporter (GLUT1) gene in NIDDM: a population association study and molecular scanning in Japanese subjects. Diabetologia 1995; 38:942-947[Medline]
• Trocino R. A., Akazawa S., Takino H., Takao Y., Matsumoto K., Maeda Y., Okuno S. I., Nagataki S. Cellular-tissue localization and regulation of the GLUT-1 protein in both the embryo and the visceral yolk sac from normal and experimental diabetic rats during the early postimplantation period. Endocrinology 1994; 134:869-878[Abstract/Free Full Text]
• Van Straaten H.W.M., Blom H. J., Peeters M.C.E., Rousseau A.M.J., Cole K. J., Seller M. J. Dietary methionine does not reduce penetrance in curly tail mice but causes a phenotype-specific decrease in embryonic growth. J. Nutr. 1995; 125:2733-2740
• Van Straaten H.W.M., Hekking J.W.M., Consten C., Copp A. J. Intrinsic and extrinsic factors in the mechanism of neurulation: effect of curvature of the body axis on closure of the posterior neuropore. Development 1993; 117:1163-1172[Abstract]
• Van Straaten H.W.M., Hekking J.W.M., Copp A. J., Bernfield M. Deceleration and acceleration in the rate of posterior neuropore closure during neurulation in the curly tail (ct) mouse embryo. Anat. Embryol. 1992; 185:169-174[Medline]
• Wald N., Sneddon J., Densem J., Frost C., Stone R., Group M.V.R.S. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. The Lancet 1991; 338:131-137[Medline]
• Wellner M., Monden I., Mueckler M. M., Keller K. Functional consequences of proline mutations in the putative transmembrane segments 6 and 10 of the glucose transporter GLUT1. Eur. J. Biochem. 1995; 227:454-458[Medline]
• Xiang K., Cox N. J., Karam J. H., Bell G. I. Bgl II RFLP at the human erythrocyte/HepG2-type glucose transporter (GLUT) locus on chromosome 1. Nucleic Acid Res. 1987; 15:9101[Free Full Text]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences