Mechanism of hyperthermia effects on CNS development Rostral gene expression domains remain despite severe head truncation and the hindbrainotocyst relationship is alteredкод для вставкиСкачать
TERATOLOGY 59:139–147 (1999) Mechanism of Hyperthermia Effects on CNS Development: Rostral Gene Expression Domains Remain, Despite Severe Head Truncation; and the Hindbrain/Otocyst Relationship Is Altered DANIELA BUCKIOVÁ1* AND NIGEL A. BROWN2 of Experimental Medicine, Academy of Sciences of the Czech Republic, 142 20 Prague, Czech Republic 2Department of Anatomy and Developmental Biology, St. George’s Hospital Medical School, London SW17 ORE, United Kingdom 1Institute ABSTRACT To study the mechanism of hyperthermia on the development of the rostral neural tube, we used a model in which closely-staged presomite 9.5-day rat embryos were exposed in culture to 437C for 13 min, and then cultured further for 12–48 hr. This treatment had little effect on the development of the rest of the embryo, but resulted in a spectrum of brain defects, the most severe being a lack of all forebrain and midbrain structures. Whole-mount in situ hybridisation was used to monitor the expression domains of Otx2, Emx2, Krox20, and hoxb1. These showed that there were no ectopic expression patterns, for any gene at any stage examined. Even in those embryos which apparently lacked all forebrain and midbrain structures, there were expression domains of Otx2 and Emx2 in the most rostral neural tissue, and these retained their nested dorso-ventral boundaries, showing that cells fated to form rostral brain were not wholly eliminated. Thus, heat-induced rostral neural tube truncation is of a quite different mechanism from the respecification proposed for retinoic acid, despite their very similar phenotypes. In the hindbrain region of treated embryos, we observed decreased intensity of Krox20, staining and an abnormal relationship developed between the position of hoxb1 expression and the otocyst and pharyngeal arches. In the most extreme cases, this domain was shifted to be more caudal than the rostral edge of the otocyst, while the otocyst retained its normal position relative to the pharyngeal arches. We interpret this as a growth imbalance between neuroepithelium and overlying tissues, perhaps due to a disruption of signals from the midbrain/ hindbrain boundary. Teratology 59:139–147, 1999. r 1999 Wiley-Liss, Inc. Hyperthermia can be teratogenic in all experimental species tested, and is probably also teratogenic in humans (reviewed by Edwards et al., ’95). The development of the brain is particularly sensitive to hyperthermia. In rats, the most sensitive period for the induction r 1999 WILEY-LISS, INC. of head defects is during late gastrulation, between 9–10 days of embryonic age (Webster et al., ’85). Lack of development of eyes from the forebrain, and open neural tube in the midbrain region were the most commonly observed defects at term. This sensitivity of the rostral neural tube is also seen when gastrulating rat embryos are exposed to hyperthermia in culture (Walsh et al., ’87, ’97). When observed at about the 25-somite stage, the most severely affected embryos appeared to lack the whole of the fore- and midbrain (Walsh et al., ’87, ’97). The cellular response to heat is increasingly welldocumented (reviewed by Nover and Scharf, ’97), but the role of heat shock proteins and other aspects of the response in the mechanism of developmental disruption remain obscure, despite much study (see Thayer and Mirkes, ’97). Recent research on the cellular response in heat-exposed embryos has concentrated on the pathway leading to apoptosis (e.g., Mirkes et al., ’97). It is well-known that hyperthermia induces cell death in embryos (see Walsh et al., ’97), but again, the mechanism of induction is unknown, and the role in teratogenesis is unclear. The induction, by heat, of embryos that lack rostral neural structures is reminiscent both of retinoic acid (RA) treatment, and of the targeted mutation of genes required for rostral development. When gastrulating vertebrate embryos are exposed to high levels of exogenous RA, the development of the rostral neural tube can be severely affected or truncated (Durston et al., ’89; Conlon and Rossant, ’92; Simeone et al., ’95; Avantaggiato et al., ’96). This is generally thought to be because RA causes a rostral expansion of the expression Grant sponsor: Grant Agency of the Czech Republic; Grant numbers: 304/1716/1994, 304/0909/1996; Grant sponsor: Biotech; Grant number: BIO2 CT-930107; Grant sponsor: UK Medical Research Council. *Correspondence to: Daniela Buckiová, Institute of Experimental Medicine, Academy of Sciences of the Czech Republic, Vı́deòská 1083, 142 20 Prague 4, Czech Republic. E-mail: bucki@ biomed.cas.cz Received 15 July 1997; Accepted 24 September 1998 140 D. BUCKIOVÁ AND N.A. BROWN domains of certain specifying genes, which results in the transformation of rostral cells to a more caudal phenotype (Simeone et al., ’95; Avantaggiato et al., ’96). Similarly, the rostral neural tube does not develop in mice with targeted mutations of Lim1 or Otx2, transcription factors required in head specification (Shawlot and Behringer, ’95; Acampora et al., ’95; Matsuo et al., ’95; Ang et al., ’96). Hyperthermia is known to affect gene and protein expression, both as general repression in response to heat shock, and in more specific ways: e.g., the protein product of the fushi tarazu gene in Drosophila is stabilized by heat shock, resulting in an overexpression phenotype (Welte et al., ’95). To investigate the possibility that hyperthermia causes abnormalities of the rostral neural tube by altering the expression of developmental control genes, we monitored the temporal and spatial expression of key genes in heat-treated rat embryos. We utilized whole-embryo culture to allow precise staging of embryos at the time of hyperthermia, and used whole-mount in situ hybridization to detect the expression of: Emx2, which is expressed in dorsal forebrain (Simeone et al., ’92); Oxt2, which is expressed throughout the fore- and midbrain (Simeone et al., ’93); Krox20, which is expressed in rhombomeres 3 and 5 of the hindbrain (Wilkinson et al., ’89); and hoxb1, which is restricted to rhombomere 4 in cranial expression (Murphy et al., ’89). Each of these genes encodes a transcription factor, of the homeobox or zinc finger (Krox20) type. Expression of receptor tyrosine kinase gene EphA4 (Sek-1) is restricted to rhombomeres 3 and 5 after the segmentation of hindbrain (Irwing et al., ’96). bottles were briefly cooled, returned to the 38°C incubator, and cultured for 12, 24, or 48 hr. MATERIALS AND METHODS Twenty-four hours following heat treatment Whole-embryo culture Forty control and 98 treated embryos were morphologically examined. After 24 hr of culture, untreated embryos had developed 10–12 pairs of somites (Fig. 2A–D). The head folds were closed, but the caudal neuropore was wide open. The otic pit and optic cup were marked, and the first pharyngeal arch was visible. Embryos had almost completed turning, to be dorsally convex. Heat-treated embryos fell into two distinct groups. About half the embryos were not markedly affected (Fig. 2E–H). In these, the head folds had closed normally. Some were slightly retarded, in that the otic pit was less indented and turning was incomplete (Fig. 2E,F). Also, the optic cup may have been of a slightly different shape compared to the controls in some embryos (e.g., Fig. 2F), but objective measures were not taken. There were irregularly shaped somites in a proportion of embryos (Fig. 2H). In the remaining heat-treated embryos, the head folds were not fused (Fig. 2I–K). In some embryos the folds were completely open (Fig. 2I,J), while in others the folds were fused in some segments of the brain (Fig. 2K, for an example of open hindbrain). In those embryos with wholly unfused head folds (Fig. 2L) , the fore- and midbrain region was Embryos were cultured as described previously (Brown et al., ’91). Briefly, Wistar rats were mated overnight and conceptuses were explanted in the afternoon of day 10 of gestation (plug ⫽ day 1). Embryos were staged using criteria previously described (Fujinaga et al., ’92), and presomite embryos of late allantoic bud, early head fold, and late head fold stages (nomenclature of Downs and Davies, ’93) were selected. These were cultured (4 conceptuses per 4 ml medium) in 5O-ml bottles, on rollers, at 38°C. The culture medium was 75% immediately centrifuged, heat-inactivated, 0.2-µm filtered rat serum, and 25% Eagle’s minimal essential medium with Earle’s salts. Embryos were gassed initially with 5% 02, changed to 20% at 16 hr and 40% at 26 hr, with 5% CO2 and the balance N2. Exposure to heat Bottles containing embryos were placed in a 38°C incubator and rotated for 30 min–1 hr to equilibrate. Bottles were then placed upright in a water bath for 13 min at 43°C, or 38°C for controls. After treatment, Whole-mount in situ hybridization Embryos harvested at 12, 24, or 48 hr following treatment were rinsed in phosphate-buffered saline and then fixed in 4% (w/v) paraformaldehyde overnight. Whole-mount in situ hybridization was performed essentially as described by Wilkinson and Green (’90). Digoxigenin-labeled antisense RNA probes were synthesized from cDNA fragments of the following lengths, cloned into the EcoR1 site of pKS: Otx2, 260 bp; Emx2, 600 bp; Krox20, 700 bp; hoxb1, 800 bp; and EphA4, 1,500 bp. RESULTS Twelve hours following heat treatment The sample sizes, i.e., numbers of heat-exposed and control embryos morphologically examined, were 44 controls and 82 exposed. After 12 hr of culture, untreated embryos had developed, on average, six pairs of somites (Fig. 1A–D). The head folds were elevated, but not opposed, and embryos had not begun body turning. Compared to controls, heat-treated embryos were not of markedly different morphology (Fig. 1E,F). Some embryos were retarded, and in some cases this appeared more pronounced in the most rostral portion of the head (e.g., Fig. 1F), but this was not dramatic in any embryo. The neural groove in the caudal portion of the embryo was irregular in a small proportion of embryos (e.g., Fig. 1H). Although there was some variability in morphology of the rostral neural tubes, embryos did not fall into obviously different groups. HYPERTHERMIA AND ROSTRAL CNS GENE EXPRESSION DOMAINS 141 Fig. 1. Embryos of 4–6 somites, 12 hr following heat treatment. Whole-mount in situ hybridization of Otx2 (A, E), Emx2 (B, F), Krox20 (C, G), and hoxb1 (D, H) in control (A–D) and treated (E–H) embryos. A, B, E, F: Right lateral views. C, D, G, H: Dorsal views. The rostral (arrow) and caudal (arrowhead) margins of Otx2 and Emx2 expression are marked. Note the diminished expression of Krox20 in rhombomere 3 of the treated embryo (double arrowhead, G), and the irregular neural groove (H). small, the optic cup was usually not visible, and the otic placode was not indented (Fig. 2I,J). Appearance of growth retardation was not dependent on embryonic stage at time of treatment, as seen in comparison of 16 controls and 38 treated embryos. The ratio of retarded and fully developed embryos (evaluation based on level of open anterior neural plate) was 1.0. about 30% of embryos the head region was grossly normal (Fig. 3D). In contrast, there was a spectrum of head defects in the remaining 70% of treated embryos. In the most severe cases, the whole of the fore- and midbrain region appeared to be missing (Fig. 3J,K). Obviously, there was no optic primordium in these embryos, but the otocyst had developed. The optic primordium was also not visible in slightly less affected embryos, which had a neural tube rostral to the hindbrain, but without any of the morphological features of the mid- or forebrain (Fig. 3G,L). Still less affected embryos had an optic primordium, but reduced foreand midbrain regions, which were unfused in some cases (Fig. 3F, for an example of unfused midbrain). Treated embryos were significantly retarded in any Forty-eight hours following heat treatment After 48 hr of culture, untreated embryos (n ⫽ 40) had developed 22–24 pairs of somites (Fig. 3A–C). The optic stalk was closing, the otocyst was closed, and three pharyngeal arches were visible. In all heattreated embryos (n ⫽ 136), the heart and structures more caudal were overtly normal (not shown), and in 142 D. BUCKIOVÁ AND N.A. BROWN Fig. 2. Embryos of 10–12 somites, 24 hr following heat treatment. Whole-mount in situ hybridization of Otx2 (A, E, I), Emx2 (B, F, J), Krox20 (C, G, K), and hoxb1 (D, H, L) in control embryos (A–D), treated embryos with mild phenotype with closed head folds (E–H), and treated embryos with severe phenotype with open head folds (I–L). Right lateral views. The rostral (arrows) and caudal (arrowheads) margins of Otx2 and Emx2 expression are marked. Note the expression of Krox20 in migrating crest cells in the control embryo (arrow in C) and the diminished expression of Krox20 in rhombomere 3 of the treated embryos (arrow in K). tested parameters by morphological scoring system (yolk sac diameter, crown-rump, head length, or somites) (Brown, ’90). The proportions of embryos with differing phenotypes were compared with the exact embryonic stage at the time of explantation and heat treatment (Table 1). Embryos were categorized into three groups, based on rostral brain morphology: grossly normal; reduced; or truncated/missing (no optic primordium). All control embryos were normal. Each phenotype was found following treatment at late allantoic bud, early head fold, or late head fold stages. The proportions were similar following exposure at late allantoic bud, and early head fold stages, with 70–75% of embryos affected, most of HYPERTHERMIA AND ROSTRAL CNS GENE EXPRESSION DOMAINS Fig. 3. Cranial portions of embryos of 22–24 somites, 48 hr following heat treatment. Whole-mount in situ hybridization of Otx2 (A, D, G, J), Emx2 (B, E, K), and hoxb1 (C, F, H, L) in control embryos (A–C) and treated embryos (D–L) of various phenotypes. Right lateral views, 143 except for ventral views in J and K. The rostral (black arrowheads) and caudal (black arrows) margins of expression are marked. For the hoxb1 embryos, the rostral edge of the otocyst (white arrowheads) and the middle of the first pharyngeal arch are also marked (white arrow). 144 D. BUCKIOVÁ AND N.A. BROWN TABLE 1. Stage variation in the proportions of phenotypes 48 hr after exposure of presomite rat embryos to hyperthermia in culture* Fore/midbrain phenotype Total embryos Normal Reduced Truncated (%) (%) (%) exposed Stage at exposure Late allantoic bud Early head fold Late head fold 35 40 41 27 32 42 13 10 34 60 58 24 *The truncated phenotype was defined as absence of the optic primordium, while the reduced phenotype had no telencephalic evaginations. which had missing or severely truncated fore- and midbrain regions. Following treatment at the late head fold stage, fewer embryos were affected (about 60%), less than half of which had missing or truncated rostral brain structures. Because the phenotypes did not differ markedly between the different treatment stages, these are pooled in the following descriptions of gene expression patterns. Finally, it is worth noting that there were no overt signs of necrosis, observable as opaque areas following some chemical treatments, in any heattreated embryo. Gene expression patterns Twelve hours following heat treatment. In 4–6somite embryos 12 hr following heat treatment, there was no obvious difference in the expression domains of Otx2, Emx2, or hoxb1 compared to control embryos. Otx2 was expressed from the most rostral tip of the neural tube, including the ventral portion of the forebrain, to a caudal margin just rostral to the preotic sulcus (Fig. 1A,E). The domain of Emx2 was smaller, did not include the ventral lip of the forebrain, and had a distinctly more rostral caudal margin than Otx2 (Fig. 1B,F). Hoxb1 was expressed in a band just caudal to the preotic sulcus, representing rhombomere 4, throughout the latero-medial extent of the neural tube (Fig. 1D,H). In control embryos, Krox20 was expressed in the characteristic pattern of two bands, representing rhombomeres 3 and 5, again throughout the latero-medial extent of the neural tube (Fig. 1C). In some embryos, a population of expressing cells was observed emanating from the neural crest at the caudal edge of the rhombomere 5 tube (Fig. 1C). In many treated embryos, the Krox20 expression domain in r3 was very faint (double arrow in Fig. 1G) or absent, and that in rhombomere 5 was reduced in the latero-medial plane (Fig. 1G). However, the rostro-caudal dimension of the two domains, and the gap between them, as well as their location relative to the preotic sulcus, appeared normal. Twenty-four hours following heat treatment. In 10–12-somite control embryos, Otx2 was expressed throughout the dorsal and ventral forebrain and midbrain, with a caudal border in the region of the midbrain/ hindbrain junction (arrowhead in Fig. 2A). Emx2 expres- sion was restricted to the dorsal forebrain, and was also seen in the lateral ridge, adjacent to the most caudal somites (Fig. 2B). In the brain, the caudal margin of expression was just caudal to the forebrain/midbrain sulcus (arrowhead in Fig. 2B). As before, Krox20 was observed in rhombomeres 3 and 5, which overlay the rostral two thirds of the otic pit, and in migrating neural crest cells (arrow in Fig. 2C). Hoxb1 was expressed in rhombomere 4, rostral to the otic pit, and in the primitive streak (Fig. 2D). In those heat-treated embryos in which the rostral neural tube was closed, the expression patterns of Otx2, Emx2, and hoxb1 were not noticeably different from controls (Fig. 2E,F,H). In those embryos with smaller fore- and midbrains, compared to controls, domains were correspondingly smaller, but not ectopic (see Fig. 2F for Emx2). However, in these embryos, as at 12 hr, Krox20 was markedly fainter, particularly in rhombomere 3 (Fig. 2G). In those heat-treated embryos in which the rostral neural tube was open and truncated, the domains of Otx2 and Emx2 were abnormally shaped, but this appeared to be in register with the abnormal morphology (Fig. 2I,J). In some cases, Otx2 expression appeared to be stronger in the ventral neural tube (although still with a caudal margin at the presumed midbrain/ hindbrain junction, Fig. 2I) but this seemed to be characteristic of the elevated neural fold stage, and was seen in control embryos in this phase of neurulation (not shown). Even in embryos with a severely reduced rostral head, Emx2 remained restricted to the most rostral region, and was excluded from the ventral portion (Fig. 2J). Krox20 was reduced in the more severely affected embryos, as before (Fig. 2K), and again hoxb1 was not affected (Fig. 2L). Forty-eight hours following heat treatment. In 22–24-somite embryos, the patterns of Otx2 and Emx2 in control and heat-treated embryos were of the same profile as observed in the 10–12-somite embryos (Fig. 3). The one exception was some heat-treated embryos with fused head folds, in which Otx2 expression was less intense in the fore- and midbrain (arrow in Fig. 3D), but which had staining in the roof of the hindbrain (Fig. 3D). The hindbrain staining was considered to be unspecific, although this was not observed with other probes. The reduced staining more rostrally is of unknown significance, since other embryos in this group had overtly normal levels of expression (not shown). The most striking finding was that Otx2 and Emx2 domains were seen even in those embryos which apparently lacked fore- and midbrain structures (Fig. 3J,K). In these cases, Otx2 was expressed in the rostralmost portion of the truncated head, in a strip from the dorsal side of the neural tube (arrow in Fig. 3J), to the most ventral part, adjacent to the pharyngeal cavity (arrowhead in Fig. 3J). Emx2 was also expressed in the most rostral portion, and from the dorsal edge of the truncated head (arrow in Fig. 3K), but in contrast to Otx2, expression stopped about halfway to the pharyngeal HYPERTHERMIA AND ROSTRAL CNS GENE EXPRESSION DOMAINS cavity (arrowhead in Fig. 3K), leaving a ventral region without staining. Krox20 is no longer expressed in the hindbrain at this stage (not shown). A hoxb1 domain of normal rostrocaudal width was observed in all heat-treated embryos, but in the more severely affected embryos, this was reduced in the latero-medial plane (Fig. 3H,L). Strikingly, the position of the hoxb1 domain relative to the otocyst was shifted in heat-treated embryos. In control embryos, hoxb1 expression, representing rhombomere 4, lies immediately dorsal to the first pharyngeal arch (white arrow in Fig. 3C), with a caudal boundary that is more rostral than the rostral edge of the otocyst (white arrowhead in Fig. 3C). There was a relative caudal shift of hoxb1 expression in heat-treated embryos, which was graded in proportion to the severity of head truncation. In mildly affected embryos, the expression domain just overlapped the rostral edge of the otocyst, and was dorsal to the first pharyngeal arch (Fig. 3F). In more severely affected embryos, the domain overlapped the otocyst, and was dorsal to the second pharyngeal arch (Fig. 3H,L). DISCUSSION As we reported previously (Williams et al., ’97), and is to be expected, the expression patterns we observed in cultured rat embryos are indistinguishable from those described for mouse embryos in vivo. Also, the morphological effects of heat exposure that we observed in cultured rat embryos are very similar to those previously described (Walsh et al., ’87, ’97). The brief heatshock at immediately presomite stages induced a spectrum of rostral truncation of the neural tube, ranging from apparent absence of the whole of the fore- and midbrain, to overtly normal heads. The reason for this variability is unclear, but is not simply stage-dependence. Similarly, the cause of the particular sensitivity of the rostral neural tube, more caudal structures being essentially unaffected, remains obscure. The major objective of this study was to investigate the possibility that respecification of tissue, due to ectopic expression of control genes, might be responsible for rostral truncation. We found no evidence to support this theory. For all of the genes examined, and at all stages, there was no evidence for anterior expansion of expression domains, nor any other ectopic expression. This contrasts with retinoic acid-induced head truncation, in which marked shifts in both Otx2, Emx2, and other genes have been reported (Durston et al., ’89; Conlon and Rossant, ’92; Simeone et al., ’95; Avantaggiato et al., ’96). In some of these studies, retinoic acid also caused a relative expansion of hindbrain tissues, which we did not observe. It is notable that we found regions of Otx2 and Emx2 expression even in those embryos that appeared to lack all foreand midbrain tissue. In addition, the spatial relationships of these genes were unaltered, despite the grossly abnormal structures in which they were located. Thus, 145 Otx2 was expressed throughout the dorso-ventral extent of the most rostral tissue, while Emx2 was excluded from the most ventral portion. This suggests that a very small fraction of the cells normally fated to become fore- and midbrain remain in these embryos. It was not possible to reliably compare the relative caudal boundaries of Otx2 and Emx2, so it is possible that the midbrain was eliminated completely, but this was not the case in slightly less affected embryos (e.g., Fig. 3G). The very severe reduction in the numbers of fore- and midbrain cells could be due to induced cell death, or impaired proliferation, or a combination of both. We have no direct evidence in either regard, but the relative lack of phenotypic effects at 4–6 somites, 12 hr following exposure, suggests that the action of heat was not an immediate large-scale induction of cell death. That heads could appear relatively normal at this stage, yet be severely truncated later, is probably explained by the differential growth of the brain regions over this phase. In the period following 4–6 somites there is a very rapid expansion in the forebrain (MorrissKay and Tuckett, ’87). The very severe reduction in the numbers of fore- and midbrain cells could be due to induced cell death, impaired proliferation, or a combination of both. By the 20-somite stage, the most affected embryos showed a domain of hoxb1 expression that was significantly more caudal than normal, relative to both the pharyngeal arches and the otocyst. In contrast, there was no indication of a change in the position of the otocyst, relative to the pharyngeal arches. This is another effect of heat that contrasts with effects of exogenous retinoic acid, for which a rostral shift in the position of the otocyst relative to the pharyngeal arches is well-known (Webster et al., ’86). The distance of the heat-induced caudal shift in the relative position of the hoxb1 domain was, by 20 somites, equivalent to 1 or 2 rhombomeres, from the normal position where the caudal margin of expression is rostral to the otocyst, to one in which the rostral margin is registered with the rostral edge of the otocyst. Caudal shift of hoxb1 expression is the only finding of ectopic expression in this study. We assume that the expressed domain still represents rhomobomere 4, because hoxb1 (Gavalas et al., ’98; Studer et al., ’98) is responsible for the establishment and maintenance of this segment. Expression of other spatially restrictedspecific genes, e.g., Krox20 and EphA4, have been downregulated at 20 somites embryos. There was no apparent shift in the positions of either hoxb1 or Krox20 at 10–12 somites, 24 hr after exposure. Also, EphA4 (Sek-1) expression that was segmentally restricted in rhombomere 3 and rhombomere 5 in the stages (Irwing et al., ’96); confirmed the unchanged identity of the area (not shown). However, reductions in the intensity of expression of Krox20 in rhombomere 3, and in the latero-medial extent of expression in rhombomere 5, were consis- 146 D. BUCKIOVÁ AND N.A. BROWN tently observed at both 4–6- and 10–12-somite stages. The exact significance of this is not clear, but may reflect an effect on proliferation in these regions. Whatever the significance, there was not an equivalent latero-medial reduction in hoxb1 expression in the interposed rhombomere 4 at the 4–6- and 10–12-somite stages, but this was observed at 22–24 somites. It is possible that the reduction in the midbrain population, as shown by the much reduced Otx2 domain, may have affected signals from the midbrain/ hindbrain border region. Although the caudal limit of Otx2 expression is at the midbrain/hindbrain boundary (Millet et al., ’96), null mutants of Otx2 also lack rhombomeres 1 and 2 (Acampora et al., ’95; Matsuo et al., ’95; Ang et al., ’96). This has been suggested to be an effect on the early precaudal plate, but an alternative explanation is an action of the isthmus-organizing region located at the midbrain/hindbrain boundary, which provides instructive signals that pattern the adjacent areas (Ang, ’96). Further insight may be gained by study of Fgf8, a major signal from this region. Endogenous retinoic acid also plays a role in patterning the hindbrain, but current evidence does not support a role for alterations in retinoid signaling in heat-induced defects. Exogenous retinoic acid can cause ectopic expression, predominantly anteriorization, and transformation in rhombomeres 1–4 (Morris-Kay et al., ’91; Conlan and Rossant, ’92; Marshall et al., ’92), while retinoid deficiency, at least in the quail, causes loss of rhombomeres 4–8 (Maden et al., ’96). Neither of these responses is similar to our observations of heat-induced effects. ACKNOWLEDGMENTS We are grateful to E. Boncinelli, R. Krumlauf, and D. Wilkinson for the gift of probes. We thank Fiona Mann and Ian Andrew for technical help and valuable advice. LITERATURE CITED Acampora D, Mazan S, Lallemand Y, Avantaggiato V, Maury M, Simeone A, Brulet P. 1995. Forebrain and midbrain regions are deleted in Otx2-/- mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121:3279– 3290. Ang SL. 1996. The brain organization. Nature 380:25–27. Ang SL, Jin O, Rhinn M, Daigle N, Stevenson L, Rossant J. 1996. A targeted mouse Otx2 mutation leads to severe defects in gastrulation and formation of axial mesoderm and to deletion of rostral brain. Development 122:243–252. Avantaggiato V, Acampora D, Tuorto F, Simeone A. 1996. Retinoic acid induces stage-specific repatterning of the rostral central nervous system. Dev Biol 175:347–357. Brown NA. 1990. Routine assessment of morphology and growth: Scoring systems and measurements of size. In: Copp AJ, Cocroft DL, editors. Postimplantation mammalian embryos: A practical approach. Oxford: IRL Press. p 93–108. Brown NA, Clarke DO, McCarthy A. 1991. Adaptation of postimplantation embryos to culture: Membrane lipid synthesis and response to valproate. Reprod Toxicol 5:245–253. Conlon RA, Rossant J. 1992. Exogenous retinoic acid rapidly induces anterior ectopic expression of murine Hox-2 genes in vivo. Development 116:357–368. Downs KM, Davies T. 1993. Staging of gastrulating mouse embryos by morphological landmarks in the dissecting microscope. Development 118:1255–1266. Durston AJ, Timmermans JP, Hage WJ, Hendriks HF, de Vries NJ, Heideveld M, Nieuwkoop PD. 1989. Retinoic acid causes an anteroposterior transformation in the developing central nervous system. Nature 340:140–144. Edwards MJ, Shiota K, Smith MSR, Walsh DA. 1995. Hyperthermia and birth defects. Reprod Toxicol 9:411–425. Fujinaga M, Brown NA, Baden JM. 1992. Comparison of staging systems for the gastrulation and early neurulation period in rodents: A proposed new system. Teratology 46:183–190. Gavalas A, Studer M, Lumsden A, Rijli FM, Krumlauf R, Chambon P. 1998. Hoxa and Hoxb1 synergize in patterning the hindbrain, cranial nerves and second pharyngeal arch. Development 125:1123– 1136. Irwing C, Nieto A, Das Gupta R, Charnay P, Wilkinson DG. 1996. Progressive spatial restriction of Sek-1 and Krox-20 gene expression during hindbrain segmentation. Dev Biol 173:26–38. Maden M, Gale E, Kostetskii I, Zile M. 1996. Vitamin A-deficient quail embryos have half a hindbrain and other neural defects. Curr Biol 6:417–426. Marshall H, Nonchev S, Sham MH, Muchamore I, Lumsden A, Krumlauf R. 1992. Retinoic acid alters hindbrain Hox code and induces transformation of rhombomeres 2/3 into a 4/5 identity [see comments]. Nature 360:737–741. Matsuo I, Kuratani S, Kimura C, Takeda N, Aizawa S. 1995. Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev 9:2646–2658. Millet S, Bloch-Gallego E, Simeone A, Alvarado-Mallart RM. 1996. The caudal limit of Otx2 gene expression as a marker of the midbrain/hindbrain boundary: A study using in situ hybridisation and chick/quail homotopic grafts. Development 122:3785–3797. Mirkes PE, Huynh LT, Little SA. 1997. Activation of CPP 32 and internucleosomal DNA fragmentation and inactivation of poly (ADPribose) polymerase (PARP) during hyperthermia and sodium arsenite-induced apoptosis in early post-implantation mouse embryos. Teratology 55:44. Morriss-Kay G, Tuckett F. 1987. Fluidity of the neural epithelium during forebrain formation in rat embryos. J Cell Sci [Suppl] 8:433–449. Morriss-Kay GM, Murphy P, Hill RE, Davidson DR. 1991. Effects of retinoic acid excess on expression of Hox-2.9 and Krox-20 and on morphological segmentation in the hindbrain of mouse embryos. EMBO J 10:2985–2995. Murphy P, Davidson DR, Hill RE. 1989. Segment-specific expression of a homoeobox-containing gene in the mouse hindbrain. Nature 341:156–159. Nover L, Scharf KD. 1997. Heat stress proteins and transcription factors. Cell Mol Life Sci 53:80–103. Shawlot W, Behringer RR. 1995. Requirement for Lim1 in headorganizer function [see comments]. Nature 374:425–430. Simeone A, Gulisano M, Acampora D, Stornaiuolo A, Rambaldi M, Boncinelli E. 1992. Two vertebrate homeobox genes related to the Drosophila empty spiracles gene are expressed in the embryonic cerebral cortex. EMBO J 11:2541–2550. Simeone A, Acampora D, Mallamaci A, Stornaiuolo A, D’Apice MR, Nigro V, Boncinelli E. 1993. A vertebrate gene related to orthodenticle contains a homeodomain of the bicoid class and demarcates anterior neuroectoderm in the gastrulating mouse embryo. EMBO J 12:2735–2747. Simeone A, Avantaggiato V, Moroni MC, Mavilio F, Arra C, Cotelli F, Nigro V, Acampora D. 1995. Retinoic acid induces stage-specific antero-posterior transformation of rostral central nervous system. Mech Dev 51:83–98. Studer M, Gavalas A, Marshall H, Ariza-McNaughton L, Rijli FM, Chambon P, Krumlauf R. 1998. Genetic interactions between Hoxa1 and Hoxb1 reveal new roles in regulation of early hindbrain patterning. Development 125:1025–1036. HYPERTHERMIA AND ROSTRAL CNS GENE EXPRESSION DOMAINS Thayer JM, Mirkes PE. 1997. Induction of Hsp72 and transient nuclear localization of Hsp73 and Hsp72 correlate with the acquisition and loss of thermotolerance in postimplantation rat embryos. Dev Dyn 208:227–243. Walsh DA, Klein NW, Hightower LE, Edwards MJ. 1987. Heat shock and thermotolerance during early rat embryo development. Teratology 36:181–191. Walsh DA, Li Z, Wu Y, Nagata K. 1997. Heat shock and the role of the HSPs during neural plate induction in early mammalian CNS and brain development. Cell Mol Life Sci 53:198–211. Webster WS, Germain MA, Edwards MJ. 1985. The induction of microphthalmia, encephalocele, and other head defects following hyperthermia during the gastrulation process in the rat. Teratology 31:73–82. Webster WS, Johnston MC, Lammer EJ, Sulik KK. 1986. Isotretinoin 147 embryopathy and the cranial neural crest: An in vivo and in vitro study. J Craniofac Genet Dev Biol 6:211–222. Welte MA, Duncan I, Lindquist S. 1995. The basis for a heat-induced developmental defect: Defining crucial lesions. Genes Dev 9:2240– 2250. Wilkinson DG, Green J. 1990. In situ hybridization and the threedimensional reconstruction of serial sections. In: Copp AJ, Cocroft DL, editors. Postimplantation mammalian embryos: A practical approach. Oxford: IRL Press. p 155–172. Wilkinson DG, Bhatt S, Chavrier P, Bravo R, Charnay P. 1989. Segment-specific expression of a zinc-finger gene in the developing nervous system of the mouse. Nature 337:461–464. Williams JA, Mann FM, Brown NA. 1997. Gene expression domains as markers in developmental toxicity studies using mammalian embryo culture. Int J Dev Biol 41:359–364.