Otx

Neuron, Vol. 31, 943–956, September 27, 2001, Copyright 2001 by Cell Press

Specification of Thermosensory Neuron Fate in C. elegans Requires ttx-1, a Homolog of otd/Otx John S. Satterlee,1 Hiroyuki Sasakura,2 Atsushi Kuhara,2 Maura Berkeley,1 Ikue Mori,2 and Piali Sengupta1,3 1 Department of Biology and Volen Center for Complex Systems Brandeis University 415 South Street Waltham, Massachusetts 02454 2 Laboratory of Molecular Neurobiology Division of Biological Science Graduate School of Science Nagoya University Nagoya 464-8602 Japan

Summary Temperature is a critical modulator of animal metabolism and behavior, yet the mechanisms underlying the development and function of thermosensory neurons are poorly understood. C. elegans senses temperature using the AFD thermosensory neurons. Mutations in the gene ttx-1 affect AFD neuron function. Here, we show that ttx-1 regulates all differentiated characteristics of the AFD neurons. ttx-1 mutants are defective in a thermotactic behavior and exhibit deregulated thermosensory inputs into a neuroendocrine signaling pathway. ttx-1 encodes a member of the conserved OTD/ OTX homeodomain protein family and is expressed in the AFD neurons. Misexpression of ttx-1 converts other sensory neurons to an AFD-like fate. Our results extend a previously noted conservation of developmental mechanisms between the thermosensory circuit in C. elegans and the vertebrate photosensory circuit, suggesting an evolutionary link between thermosensation and phototransduction. Introduction Environmental temperature modulates many aspects of physiology, development, and behavior of animals. In mammals, temperature is sensed by different classes of sensory neurons, each of which responds within a distinct thermal range (Hensel, 1973; Spray, 1986; Iggo, 1990; Reichling and Levine, 2000). Nociceptive thermal stimuli are transduced via the VR1 and VRL-1 vanilloid receptor genes encoding nonselective cation channels (Caterina et al., 1997, 1999). However, the molecules required for thermosensation in other sensory neurons are largely unknown. Thermal sensory inputs are integrated in the preoptic region of the hypothalamus to alter behavior and development by poorly understood mechanisms. Behavior and development of the nematode C. elegans are also modulated by temperature. Worms grown in the presence of food at a particular temperature will 3

Correspondence: [email protected]

actively seek this temperature when presented with a thermal gradient but will avoid this temperature if it is associated with an absence of food (Hedgecock and Russell, 1975; Mori, 1999). Integration of food and thermal cues may be mediated by calcium signaling (Gomez et al., 2001). In addition, entry into and exit from the alternative dauer developmental stage is modulated by temperature (Golden and Riddle, 1984a, 1984b). Recent work suggests that, similar to vertebrates, C. elegans may also possess multiple thermosensory pathways, each of which is activated within a specific thermal range (Wittenburg and Baumeister, 1999; Ailion and Thomas, 2000). In C. elegans, the thermosensory circuit for responses to the 15⬚C–25⬚C temperature range has been well characterized (Mori and Ohshima, 1995) (see Figure 4C). The AFD sensory neurons and their major postsynaptic partners, the AIY interneurons, specify the thermophilic components of the circuit, driving movement to higher temperatures in a thermal gradient (Mori and Ohshima, 1995; Mori, 1999). This movement is counterbalanced by the AIZ interneurons and an unidentified sensory neuron(s), which drive cryophilic movement (Mori and Ohshima, 1995). Crosstalk between these two opposing circuits results in movement to the cultivation temperature. Genetic and behavioral screens to date have resulted in the identification of a small number of molecules required for the thermosensory functions of the AFD neuron type. Thermosensory signal transduction requires the function of a cyclic nucleotide-gated channel encoded by the tax-2 and tax-4 genes (Coburn and Bargmann, 1996; Komatsu et al., 1996). In addition, the ceh14 LIM homeobox gene has been shown to be required for the correct function of the AFD neurons (Cassata et al., 2000). Additional genes required for the development and/or function of the AFD neurons remain to be identified. The OTD/OTX subclass of homeodomain proteins is found across species and has been implicated in patterning anterior neural regions and development of sensory structures (Hirth and Reichert, 1999; Klein and Li, 1999; Acampora et al., 2000). In Drosophila, orthodenticle (otd) gene function is required for development of anterior structures of the embryonic brain, cell fate determination in the ventral midline, and development of visual structures (Finkelstein and Perrimon, 1990; Hirth et al., 1995; Royet and Finkelstein, 1995; Vandendries et al., 1996; Younossi-Hartenstein et al., 1997). In vertebrates, the Otx1 and Otx2 genes are expressed sequentially in overlapping domains and play critical roles in the patterning of anterior brain structures and sensory organs, such as the eye and ear (Simeone et al., 1992, 1993; Matsuo et al., 1995; Acampora et al., 1995, 1996, 2000; Morsli et al., 1999; Weimann et al., 1999). Another member of the Otx family in vertebrates, Crx, has been shown to be required specifically for the development of photoreceptors (Furukawa et al., 1997, 1999). Thus, most, if not all, members of this family

Neuron 944

Figure 1. Mutations in ttx-1 Affect Gene Expression and Morphology of the AFD Thermosensory Neurons (A and B) Expression of a stably integrated gcy-8::GFP fusion gene in wild-type (A) and in ttx-1(p767) mutant adult animals (B). Arrow points to the AFD cell body. Images were acquired using identical exposure times. (C and D) Confocal images of the sensory endings of the AFD neurons in wild-type (C) and in ttx-1(p767) mutants (D). Microvillar “fingers” (C) and elongated cilium (D) are marked with arrows. Lateral view and anterior at left in all images. Scale bars: 10 ␮m.

have been implicated in the development of sensory structures. Here we describe the cloning and characterization of the gene ttx-1. ttx-1 encodes the first characterized homolog of OTD/OTX in C. elegans. Similar to the functions of otd/Otx genes in other organisms, ttx-1 is necessary and partly sufficient for the development of a sensory cell type, the AFD thermosensory neurons. We show that ttx-1 mutations also deregulate thermosensory inputs into the dauer-regulatory neuroendocrine pathway. In addition, as initially proposed by Svendsen and McGhee (1995), we note an intriguing conservation of developmental regulatory mechanisms between the thermosensory circuit of C. elegans and the visual circuit in vertebrates, suggesting an evolutionary link between thermosensation and phototransduction. Results TTX-1 Regulates Gene Expression in the AFD Thermosensory Neurons To identify genes required for the development and differentiation of the AFD neurons, we carried out screens for mutants with altered expression of AFD-specific genes. The promoter of the receptor guanylyl cyclase gene gcy-8 drives expression of the reporter GFP exclusively in the bilateral AFD neurons (Yu et al., 1997) (Figure 1A). We isolated two mutants, oy26 and oy29, in which expression of a stably integrated gcy-8::GFP transgene was severely reduced in adult animals (Table 1). Subsequent mapping and complementation analyses showed

Table 1. Mutations in ttx-1 Affect Expression of gcy-8::GFP in the AFD Thermosensory Neurons Strain b

oyIs17 or oyIs18 ttx-1(p767);oyIs18 ttx-1(oy26) oyIs17 ttx-1(oy29) oyIs17

% Mutanta

n

0 99 61 96

250 562 244 519

a Mutant is defined as reduced expression in adult animals, as compared with that of oyIs17 or oyIs18 adults at 25⬚C. b oyIs17 and oyIs18 contain stably integrated gcy-8::GFP fusion genes on LG V and LG X, respectively (see Experimental Procedures). Expression levels of gcy-8::GFP in these strains are equivalent.

that oy26 and oy29 are allelic to the previously described ttx-1(p767) mutation (Hedgecock and Russell, 1975). As expected, we found that expression of gcy-8 is decreased in ttx-1(p767) mutants (Figure 1B and Table 1). On the basis of both the severity of the gene expression defect and behavioral defects (see below), p767 appears to be the strongest allele. In all three ttx-1 mutants, we found that gcy-8::GFP expression was largely unaffected in young (L1) larvae (data not shown), suggesting that ttx-1 is required for maintenance but not for initiation of AFD-specific gene expression. To determine the extent to which mutations in ttx-1 affect gene expression in the AFD neurons, we examined the expression of additional genes, including the nhr-38 nuclear hormone receptor gene, which is expressed exclusively in the AFD neurons (Miyabayashi et al., 1999), and the ceh-14 LIM homeobox and tax-2 cyclic nucleotide-gated channel genes, which are expressed in AFD as well as in several additional neuron types (Coburn and Bargmann, 1996; Cassata et al., 2000). In ttx-1 mutants, expression of all three genes was decreased only in the AFD neurons (Table 2). Thus, TTX-1 regulates expression of most, if not all, genes required for the differentiated functions of the AFD neurons.

The Morphology of the AFD Sensory Endings Is Defective in ttx-1 Mutants Most sensory neurons in C. elegans can be distinguished on the basis of their distinctive specialized ciliary endings (Ward et al., 1975; Perkins et al., 1986). The sensory endings of the AFD neurons are especially elaborate, consisting of a single shortened cilium and numerous microvillar “fingers” (Figure 1C). Previous electron micrograph reconstructions of the sensory endings of ttx-1 mutants showed that the sensory endings are defective (Perkins et al., 1986). We confirmed this by examining the sensory endings of ttx-1 mutants using confocal microscopy. We found that, consistent with previous reports, in most animals examined, the sensory endings of the AFD neurons were grossly abnormal, with complete loss of the finger-like projections and extension of the cilia (Figure 1D). This defect was evident even in young larvae, indicating that TTX-1 function may be required throughout development for correct morphogenesis of the AFD sensory endings.

ttx-1 Specifies Thermosensory Neuron Fate 945

Table 2. Expression of AFD-Specific Markers Is Affected in ttx-1(p767) Mutants Strain nhr-38::GFP nhr-38::GFP; ttx-1(p767) ceh-14::GFP; rol-9(sc148) ceh-14::GFP; ttx-1(p767) rol-9(sc148) tax-2::GFPb tax-2::GFP; ttx-1(p767) ttx-3::GFP ttx-3::GFP; ttx-1(p767) rol-9(sc148); str-1::GFP ttx-1(p767) rol-9(sc148); str-1::GFP Ex[ttx-1Exon-1::GFP]; ttx-1 rol-9/⫹ or ⫹/⫹c Ex[ttx-1Exon-1::GFP]; ttx-1 rol-9

Neuron Type % Mutanta n AFD AFD AFD AIY AWB AFD

0 100 2 96 14 61 0 0 0 0 4

252 221 180 390 119 117 180 191 125 150 66

90

41

a

Mutant is defined as reduced or absent expression levels as compared with wildtype. Expression was examined at 25⬚C in adult animals in all cases. b This tax-2::GFP fusion gene drives GFP expression in the AFD, AWC, ASE, ASI, BAG, and AQR neurons in the head. The AFD neurons could not be reliably identified on the basis of cell body position in adult animals. Thus, mutant is defined as animals expressing GFP in fewer than five neurons (excluding AQR) on at least one side in the head. c An extrachromosomal array containing ttx-1Exon-1::GFP, which expresses specifically in the AFD neurons, was crossed into ttx-1(p767) rol-9(sc148) animals. Segregating animals containing the array were scored for rol-9 and for expression of ttx-1Exon-1::GFP. Except as indicated, the expression of stably integrated marker::GFP fusion genes was examined in all cases.

ttx-1 Encodes a Homolog of the OTD/OTX Family of Homeodomain Proteins We mapped ttx-1 to a small interval on LG V using threefactor mapping crosses, mapping with respect to deficiencies, and single nucleotide polymorphisms (see Experimental Procedures) (Figure 2A). A 14 kb genomic fragment from this interval fully rescued the ttx-1 gene expression defect and partially rescued the thermotaxis behavioral phenotype (data not shown and Figure 4B; see below). This fragment is predicted to contain a single open reading frame (Y113G7A.6) encoding an OTD/OTXlike homeodomain protein of 391 residues. The ttx-1 locus is predicted to contain nine exons (Figure 2B). Analyses of ESTs (gift from Yuji Kohara) as well as RT-PCR experiments showed the presence of at least two alternatively spliced messages, which differ from each other in the presence or absence of the fourth exon (see Experimental Procedures). To determine which, if any, of these alternatively spliced messages functions in AFD, we generated a “minigene” by fusing ⵑ3 kb of upstream promoter sequences of ttx-1 to a cDNA that lacks the fourth exon. This minigene rescued the gcy8::GFP gene expression defect of ttx-1 mutants (data not shown), indicating that this cDNA contains most, if not all, sequences required for TTX-1 function in the AFD neurons. Rescue was abolished upon disruption of this coding sequence by a frameshift mutation. We determined the nature of the lesions in the ttx-1 alleles. The oy26 mutation results in a missense mutation that changes a highly conserved Ala residue in helix 2 of the predicted homeodomain to Val (Figures 2B and

2C). Sequencing of all the exons, as well as partial sequencing of the introns and promoter sequences, failed to define the lesion in the oy29 allele. The p767 allele causes a G-to-A transition at the 5⬘ splice donor site of the second intron (Figure 2B). This mutation is predicted to affect all splice variants. To determine the effect of the p767 mutation, we amplified the ttx-1 message from ttx-1(p767) mutants using RT-PCR, subcloned the amplified fragment(s), and sequenced four independent subclones (see Experimental Procedures). We detected altered splicing events in two subclones, which are predicted to result in frameshifts in the translation products, resulting in premature termination prior to the homeodomain. However, exons 2 and 3 were correctly spliced in two additional subclones sequenced. These results suggest that the p767 mutation causes altered splicing in some but not all transcribed messages. Thus, ttx-1(p767) is likely not a molecular null. In Drosophila and vertebrates, different critical thresholds of OTD/OTX proteins are required for correct brain morphogenesis and development of specific head subdomains (Royet and Finkelstein, 1995; Acampora et al., 1997; Suda et al., 1997). Because our results indicate that ttx-1(p767) severely affects AFD differentiation, it is possible that levels of wild-type TTX-1 protein present in ttx-1(p767) mutants are insufficient to direct correct AFD development. Further reduction in TTX-1 levels by placing the ttx-1(p767) mutation in trans to a deficiency (ozDf2) did not result in an increased penetrance in the loss of expression phenotype of the tax-2::GFP marker in the AFD neurons compared with ttx-1(p767) homozygotes (data not shown). The homeodomains of the OTX subclass of pairedtype homeodomain proteins are highly conserved across phyla. The homeodomain of Drosophila OTD differs from vertebrate OTX1 by only three residues, and these two proteins can in part functionally substitute for each other (Acampora et al., 1998; Leuzinger et al., 1998; Nagao et al., 1998). The homeodomain of TTX-1 is highly homologous to that of OTX1, with 80% identity to mouse and human OTX1 and Drosophila OTD and 76% identity to mouse and human CRX (Figure 2C). TTX-1 also contains a lysine at position 9 of the recognition helix of the homeodomain, a distinctive feature of this subclass (Galliot et al., 1999) (Figure 2C). TTX-1 is also predicted to contain a variation of the OTX tail sequence, a 12 amino acid peptide present at the extreme C terminus of CRX, OTX1, and OTX2 but not Drosophila OTD (Freund et al., 1997; Furukawa et al., 1997). Sequences other than those encoding the homeodomain of TTX-1 show no significant similarities to other proteins. The C. elegans genome is predicted to contain two additional homeodomain proteins (CEH-36 and CEH-37) closely related to the OTX subfamily (Ruvkun and Hobert, 1998; Galliot et al., 1999) (Figure 2C). However, on the basis of phylogenetic analysis (see http://www.sciencemag.org/ feature/data/985556s2.gif; Y102G3) (Ruvkun and Hobert, 1998) and sequence conservation in the homeodomain, ttx-1 appears to encode the C. elegans Otx ortholog. ttx-1 Is Expressed in the AFD Thermosensory Neurons and Autoregulates Expression To determine the expression pattern of ttx-1, we generated a full-length, GFP-tagged ttx-1 fusion gene con-

Neuron 946

Figure 2. ttx-1 Encodes a Homolog of OTX/ OTD (A) Genetic map position of ttx-1. ttx-1 corresponds to the ORF Y113G7A.6 (GenBank accession number AL132858). oyP46-oyP51 denote identified SNPs between the Bristol N2 strain and the CB4856 strain. The position of the SNPs relative to the ttx-1, rol-9, and unc-51 markers are shown; these have not been mapped precisely with respect to the genetic map. (B) Genomic structure of ttx-1. The exon/intron structure of ttx-1 was deduced from sequencing of cDNAs (see Experimental Procedures). Exons encoding the predicted homeodomain are hatched. The positions of the p767 splice site mutation and the oy26 missense mutation are marked. (C) Alignment of the homeodomains of TTX1 with the homeodomains of mouse OTX1 and OTX2 (GenBank accession numbers S35345 and S35346), Drosophila OTD (GenBank accession number A35912), C. elegans CEH-36 and CEH-37 (GenBank accession numbers Q93352 and Q93356), and mouse CRX (GenBank accession number O54751). Dots indicate residues identical to TTX-1. The positions of the helices are overlined. The Ala residue mutated to Val in ttx-1(oy26) is in bold and boxed. The invariant Lys at position 9 of helix 3 is shaded.

taining ⵑ3 kb of promoter sequences and all predicted exons and introns of ttx-1 (Figure 3A). This fusion gene partially rescued the gcy-8::GFP gene expression defect and the thermotaxis behavioral defect of ttx-1 mutants (data not shown and Figure 4B; see below). Transgenic animals carrying this fusion gene expressed GFP in the bilateral AFD neurons as well as in the nine pharyngeal marginal cells (Figure 3A). The TTX-1-GFP-tagged protein is localized to the nuclei of all cells. Marginal cells are structural cells of the pharynx with unknown functions (White, 1988). ttx-1(p767) mutants do not exhibit gross abnormalities in the development or function of the pharynx. Because ttx-1(p767) is likely not a molecular null mutation, it is possible that defects in marginal cell fate or function are not evident in this allele. To investigate this possibility, we performed RNA interference experiments (Fire et al., 1998). Wild-type animals were injected with double-stranded ttx-1 RNA, and progeny were examined for abnormalities. However, no ttx-1(RNAi) animals showed any morphological or behavioral defects consistent with a defect in the function of the pharynx (n ⫽ ⵑ1000 progeny; 10 independent injected animals; see Experimental Procedures). Thus, the role of ttx-1, if any, in the development or function of the marginal cells is unclear.

To delineate the sequences required for expression in the AFD and marginal cells, we generated two additional fusion genes. In the first, the same promoter sequences as present in the full-length fusion gene and part of exon 1 were fused in frame to GFP (ttx-1Exon1::GFP; Figure 3B). This construct drives GFP expression exclusively in the AFD neurons (Figure 3B). Expression of ttx-1 was first observed well after the time when the AFD neurons were born (Sulston et al., 1983) and was maintained throughout development, consistent with the proposed role of TTX-1 in maintenance of AFD fate. In the second, GFP was fused in frame to exon 4 (ttx-1Exon4::GFP). Transgenic animals carrying this fusion gene showed GFP expression in both the AFD as well as in the marginal cells, although expression was no longer restricted to the nucleus (data not shown). This suggests that ttx-1 promoter sequences drive expression solely in the AFD neurons, whereas sequences in the introns promote additional expression in the marginal cells. Positive autoregulation of expression provides a simple mechanism by which gene expression is maintained. To investigate whether TTX-1 regulates its own expression, we examined expression of the ttx-1Exon1::GFP fusion gene in ttx-1(p767) mutants. As shown in Table 2, GFP expression in the AFD neurons was dramatically

ttx-1 Specifies Thermosensory Neuron Fate 947

cue of gcy-8 expression was examined. Approximately 4%–8% of transgenic animals expressing Otx1 showed rescue of the gcy-8 gene expression defect at 25⬚C (three independent lines; n ⬎ 95 for each line), whereas ⵑ1% of nontransgenic animals showed rescue (n ⬎ 200). No rescue was observed upon expression of Crx. Thus, vertebrate Otx1 can only partly and weakly substitute for ttx-1 function in C. elegans. In ttx-1 Mutants, the AFD Neurons Express Partial AWC-like Characteristics We and others have previously shown that the AWA, AWB, and ASG chemosensory neurons in C. elegans share a common AWC olfactory neuron-like developmental default fate (Sagasti et al., 1999; Sarafi-Reinach and Sengupta, 2000; Sarafi-Reinach et al., 2001). Expression of specific transcription factors represses this default fate and promotes the expression of cell typespecific characteristics. We investigated whether the AFD neurons adopt an AWC-like fate in ttx-1 mutants. As shown in Table 3, we found that the AWC-specific marker str-2::GFP (Troemel et al., 1999) is expressed ectopically in AFD in ⵑ62% of ttx-1 mutant animals. This phenotype was not further enhanced in a ttx-1(p767); ceh-14(ch3) double mutant (data not shown). We also examined the expression of additional AWC markers, including tax-2 (Coburn and Bargmann, 1996) and the G␣ subunit gene odr-3 (Roayaie et al., 1998). In ttx-1 mutants, we detected odr-3 expression in the AFD neurons in ⵑ9% of animals examined (Table 3). tax-2 is expressed both in the AFD and AWC neurons in wildtype animals (Coburn and Bargmann, 1996). Interestingly, even though the AFD neurons adopt partial AWC fate in ttx-1 mutants, expression of tax-2 in the AFD neurons is not maintained (Table 2). Taken together with the finding that the sensory endings of the AFD neurons in ttx-1 mutants do not resemble those of the AWC neurons, these results indicate that mutations in ttx-1 result in partial adoption of the AWC fate by the AFD neurons. Figure 3. ttx-1 Is Expressed in the AFD Thermosensory Neurons (A) Expression of a GFP-tagged full-length ttx-1 fusion gene. Shown is a schematic of the ttx-1::GFP fusion gene, which includes all predicted exons and introns of ttx-1. GFP is fused in frame to the last exon. Closed exons represent exons encoding the homeodomain. Arrows point to the nuclei of the two AFD neurons (top down view). This fusion gene is also expressed in pharyngeal marginal cells. (B) Expression of a ttx-1 promoter::GFP fusion gene. This fusion gene contains GFP fused in frame to the first exon of ttx-1. Arrow points to the cell body of the AFD neuron (lateral view). Note lack of expression in the pharynx. Anterior is at left in all images. Scale bar: 10 ␮m.

reduced, indicating that TTX-1 autoregulates to maintain expression in the AFD neurons. Otx1, but Not Crx, Partly Rescues the Gene Expression Defects of ttx-1 We investigated whether rat Otx1 or mouse Crx could functionally substitute for ttx-1. Otx1 and Crx cDNAs driven by the ttx-1 promoter were injected into ttx-1(p767) mutants carrying integrated copies of gcy-8::GFP, and res-

ttx-1 Mutants Show Defects in Thermotaxis Behavior Animals in which both AFD neurons have been killed using a laser beam show cryophilic or athermotactic thermotaxis behavior when placed on a thermal gradient (Mori and Ohshima, 1995). ttx-1(p767) animals have been previously shown to exhibit thermotaxis defects (Hedgecock and Russell, 1975; Mori and Ohshima, 1995). Because we have shown that ttx-1 mutations result in severe defects in the differentiation of the AFD neurons, we further investigated the thermotaxis behaviors of ttx-1 mutants. Single, well-fed animals grown at 20⬚C were placed on a thermal gradient ranging from 17⬚C–25⬚C. The behaviors of animals on this gradient were scored as wild type (moving to 20⬚C), cryophilic (moving to temperatures ⬍20⬚C), thermophilic (moving to temperatures ⬎20⬚C), or athermotactic (moving randomly on the gradient) (Mori and Ohshima, 1995) (Figure 4A). Animals were also scored for isothermal tracking, where they move in concentric circles at a given temperature (Mori and Ohshima, 1995). As shown in Figure 4B, ttx-1 mutants were strongly cryophilic. Approximately 83% of ttx-1(p767)

Neuron 948

Figure 4. Thermosensory Behaviors of ttx-1 Single and Double Mutants (A) Movement exhibited by wild-type and mutant animals on a radial thermal gradient ranging from 17⬚C (center) to 25⬚C (outer perimeter). Animals were raised in the presence of food at 20⬚C. Movement to 25⬚C is scored as thermophilic (T); movement to 17⬚C is scored as cryophilic (C); movement across the thermal gradient (17⬚C/25⬚C) is scored as athermotactic (A). Movement at 20⬚C was scored for isothermal tracking behavior (IT) or nonisothermal tracking (NIT). Wild-type animals show isothermal tracking behavior at 20⬚C. (B) Thermosensory behaviors of ttx-1 mutants. All animals were raised in the presence of food at 20⬚C. The percentage of adult animals of a given genotype showing different thermosensory behaviors is shown. For strains carrying extrachromosomal arrays, the behaviors of individual transgenic animals from a total of three lines each were examined. Individual lines showed similar overall behaviors. Ex(ttx-1 genomic) extrachromosomal arrays contain a 14 kb genomic fragment that rescues the ttx-1(p767) gene expression defects. Ex(ttx-1::GFPtag) extrachromosomal arrays also contain the ttx-1-rescuing genomic fragment, with GFP coding sequences inserted prior to the STOP codon. Numbers shown are from three to eight independent assays. (C) The thermosensory circuit (Mori and Ohshima, 1995). Temperature inputs are sensed by the AFD and unidentified sensory neuron(s). Thermal and chemosensory information is integrated by the AIY and AIZ interneurons. The AFD-AIY arm of the circuit mediates movement to higher temperatures (T), whereas the AIZ arm mediates movement to lower temperatures (C). Mutations in ttx-1 and ceh-14 affect the thermosensory functions of the AFD neurons, while mutations in ttx-3 and lin-11 affect the thermoregulatory functions of the AIY and AIZ neurons, respectively. See text for additional details. (D) Thermosensory behaviors of double-mutant animals. All animals were raised at 20⬚C in the presence of food. Shown is the percentage of animals of indicated genotypes exhibiting different thermosensory behaviors. Numbers shown are from two to eight independent assays.

and ⵑ63% of ttx-1(oy26) mutants showed a cryophilic phenotype, whereas fewer ttx-1(oy29) mutants exhibited this behavior. Moreover, ttx-1(p767) mutants rarely exhibited isothermal tracking behavior at any temperature. Fewer ttx-1(p767) mutants exhibited athermotactic behavior, a phenotype seen in some AFD-killed animals (Mori and Ohshima, 1995). It is possible that in ttx-1 mutants, the adoption of partial AWC fate by the AFD neurons results in altered regulation of the thermosensory circuit. We also examined the thermotaxis behavior of ttx-1(p767); ceh-14(ch3) double mutants. Although mutations in ceh-14 do not affect AFD-specific gene expression, ceh-14(ch3) mutants exhibit weak athermotactic behavior (Cassata et al., 2000) (Figure 4D). The

thermotaxis behavior of ttx-1; ceh-14 double-mutant animals was indistinguishable from that of ttx-1(p767) single mutants, further confirming that TTX-1 regulates the expression of ceh-14 as well as additional genes required for AFD neuron function. ttx-1 mutants were wild type for all other sensory behaviors examined (data not shown). TTX-1 and TTX-3 Specify the Thermoregulatory Functions of Neuronal Components Driving Thermophilic Behavior The LIM-homeobox genes ttx-3 and lin-11 have been shown to be required for the thermoregulatory functions of the AIY and AIZ interneuronal components of the

ttx-1 Specifies Thermosensory Neuron Fate 949

thermosensory circuit respectively (Hobert et al., 1997, 1998) (Figure 4C). ttx-3 mutants are cryophilic, whereas lin-11 mutants have been shown to be thermophilic (Hobert et al., 1997, 1998; Cassata et al., 2000). To genetically define the thermosensory circuit, we examined the thermosensory behaviors of double mutants among ttx-1, ttx-3, and lin-11. In all cases, we used putative null alleles (ttx-3 and lin-11) or the strongest available allele (ttx-1). As shown in Figure 4D, ttx-1; ttx-3 double mutants, similar to ttx-3 single mutants, are strongly cryophilic, confirming that AFD signals primarily via the AIY interneurons to regulate upward movement in a thermal gradient. However, in our hands, lin-11 null mutants were only weakly thermophilic, with the majority of animals exhibiting wild-type behavior (Figure 4D), suggesting that unlike ttx-1 and ttx-3, lin-11 does not specify all thermoregulatory functions of the AIZ interneurons. Consistent with this, lin-11; ttx-1 double-mutant animals exhibit only weak athermotactic behavior, with most animals remaining strongly cryophilic (Figure 4D), indicating a severe disruption of only the upward, but not the downward, arm of the thermosensory circuit. tax-4 and tax-2 encode the ␣ and ␤ subunits of a cyclic nucleotide-gated channel required for thermosensation (Coburn and Bargmann, 1996; Komatsu et al., 1996). Unlike ttx-1 mutants, tax-2 and tax-4 mutants are athermotactic, suggesting that TAX-2/4 functions in additional unidentified sensory neuron(s) that mediate cryophilic sensory responses (Mori and Ohshima, 1995). As shown in Figure 4D, tax-2; ttx-1 and tax-4; ttx-1 double mutants are also athermotactic. Thus, our results confirm the identification of the components of the thermosensory circuit that mediate thermophilic behavior, as defined by laser-killing experiments.

1992; Thomas et al., 1993; Gottlieb and Ruvkun, 1994; Riddle and Albert, 1997). Temperature is a major modulator of both dauer formation and recovery. Within the temperature range of 15⬚C–25⬚C, dauer pheromone is the primary trigger for dauer formation, with increasing temperatures increasing dauer formation in the presence of a constant amount of pheromone (Golden and Riddle, 1984a, 1984b). ttx-1 mutants have previously been shown to be hypersensitive to dauer pheromone (Golden and Riddle, 1984b; Ailion and Thomas, 2000), suggesting that the AFD neurons regulate temperature input into the dauer pathway. ttx-3 mutations have been shown to decouple temperature input and dauer recovery, and it has been proposed that ttx-3 mediates temperature input into the dauer process via temperature-mediated regulation of insulin signaling (Hobert et al., 1997). However, additional thermosensory pathways are also likely to regulate dauer formation (Ailion and Thomas, 2000). To investigate the role of ttx-1 in dauer formation in greater detail, we examined the effect of ttx-1(p767) on dauer formation and recovery of animals mutant in the insulin receptor and TGF-␤ ligand genes daf-2 and daf-7, respectively. daf-2 and daf-7 mutants exhibit a dauerconstitutive or daf-c phenotype, where they form dauers inappropriately under nonadverse conditions. As shown in Figure 5A, we found that at 25⬚C, ttx-1; daf-7 double mutants formed dauers similarly to daf-7 single mutants. However, we observed strong effects of ttx-1 on dauer recovery. Only ⵑ7% of daf-7; ttx-1 double mutants, as compared with 65% of daf-7 single mutants, remained arrested in the dauer stage after 72 hr at 25⬚C (Figure 5B), whereas the remaining animals recovered and resumed normal growth. ttx-1 mutations had no effect on dauer formation or recovery at lower temperatures of 15⬚C (data not shown). These results differ from those observed with ttx-3 mutants because mutations in ttx-3 decouple temperature inputs into the dauer pathway at both 25⬚C and 15⬚C (Hobert et al., 1997). Interestingly, ttx-1 mutations appeared to delay development and dauer formation in daf-2 mutants. After 48 hr at 25⬚C, whereas 100% of daf-2 mutants form dauers, only ⵑ23% of daf-2; ttx-1 double mutants form dauers, with the remainder of the animals in an L2-like developmental stage (Figure 5A). However, after 72 hr, nearly 73% of daf-2; ttx-1 double mutants form dauers that do not recover (Figure 5B). Thus, mutations in the AFD-mediated thermosensory circuit deregulate temperature input into the dauer neuroendocrine process, likely by decoupling temperature input and regulation of insulin signaling.

Mutations in ttx-1 Affect a TemperatureRegulated Neuroendocrine Process Under adverse conditions such as overcrowding, scarcity of food, or high temperatures, C. elegans enters into the alternative dauer developmental state (Golden and Riddle, 1984a; Riddle and Albert, 1997). Exit from this state and resumption of normal reproductive growth is also regulated by sensory cues that regulate the production of neuroendocrine signals such as TGF-␤ and insulin. The TGF-␤ and insulin pathways function in parallel to regulate the dauer process (Vowels and Thomas,

Misexpression of ttx-1 Is Sufficient to Confer AFD-like Fate onto a Subset of Ciliated Neurons To determine whether ttx-1 is sufficient for AFD fate specification, we misexpressed ttx-1 using the osm-6, str-1, and odr-10 promoters in animals carrying stably integrated gcy-8::GFP fusion genes. The osm-6 promoter drives expression in most ciliated neurons in C. elegans (Collet et al., 1998), whereas the str-1 and odr-10 promoters drive expression specifically in the AWB and AWA olfactory neurons, respectively (Sengupta et al., 1996; Troemel et al., 1997). We found that gcy-8::GFP was expressed ectopically in animals misexpressing

Table 3. The AFD Neurons Partially Adopt AWC Fate in ttx-1 (p767) Mutants

Strain

% with Ectopic Expression in at Least one AFD Neuronb

n

str-2::GFP str-2::GFP; ttx-1(p767) odr-3::GFP odr-3::GFP; ttx-1(p767)

0 62 0 9c

122 169 112 114

a

a

All strains contain rol-9(sc148). L2 animals were examined at 25⬚C. AFD neurons were identified on the basis of relative cell body position. c Expression of GFP was weaker as compared with odr-3::GFP expression in the AWC neurons in wild-type animals. The expression of stably integrated marker::GFP fusion genes was examined in all cases. b

Neuron 950

Figure 5. Mutations in ttx-1 Deregulate Temperature Input into the Dauer Pathway (A) Mutations in ttx-1 weakly affect dauer formation. Shown is the percentage of dauer animals of the indicated genotypes formed after 48 hr at 25.5⬚C. More than 100 animals were counted for each genotype. Numbers shown are from a single experiment with all strains assayed in parallel. Experiments were repeated a minimum of five times on independent days and showed similar relative differences. #—Although only ⵑ23% of daf-2; ttx-1; oyIs18 animals formed dauers after 48 hr, this does not reflect a dauer formation defect. These animals were developmentally delayed such that the remainder of the animals were at an L2-like developmental stage, unlike oyIs18 or ttx-1; oyIs18 animals, which were L4s or young adults. Alleles used were ttx-1(p767), daf-7(e1372), and daf2(e1370). (B) Mutations in ttx-1 strongly affect dauer recovery of daf-7 mutants. Shown is the percentage of dauer animals after 72 hr at 25.5⬚C. More than 100 animals of each genotype were counted. Numbers shown are from a single experiment with all strains assayed in parallel. Experiments were repeated a minimum of seven times on independent days and showed similar relative differences. The recovery of daf-7; oyIs18 animals was variable, although consistently significantly less than that of daf-7; ttx-1; oyIs18 animals. #—After 72 hr, a larger number of daf-2; ttx-1; oyIs18 animals form dauers, with the remaining animals at younger stages. These dauers failed to recover even after 5 days.

ttx-1 under the osm-6, but not the str-1 or odr-10, promoters. In animals transgenic for the osm-6::ttx-1 fusion gene, we observed gcy-8::GFP expression in several ciliated neurons in the anterior, lateral, and lumbar ganglia (Figures 6A and 6B). In the lateral ganglia, ectopic gcy-8::GFP expression was observed most frequently in the ASI chemosensory neurons, with occasional expression in additional cells including the ASJ, ADF, and

AWA neurons. In the lumbar ganglia, ectopic expression was observed frequently in the PHB sensory neurons. A stably integrated nhr-38::GFP fusion gene was also ectopically expressed in animals misexpressing ttx-1, approximately to the same extent as gcy-8::GFP (data not shown). Thus, misexpression of ttx-1 is sufficient to confer AFD-specific gene expression onto several, although not all, ciliated neuron types. We next examined the morphology of the sensory endings of the misexpressing neurons. Interestingly, we found that most misexpressing cells had sensory endings similar to those of the AFD neurons, with elaborate finger-like structures (Figure 6B). This transformation was especially evident in the tail. As shown in Figure 6C, the PHB sensory neurons normally have a single ciliary sensory ending. However, upon misexpression of ttx-1, the endings adopted the complex morphology of the AFD neurons (Figure 6D). To examine the behavioral consequences of misexpression of thermosensory neuron characteristics, we examined the thermotaxis behaviors of animals expressing osm-6::ttx-1. As shown in Figure 7A, we found that an average of 58% of wild-type animals misexpressing ttx-1 showed abnormal cryophilic, thermophilic, or athermotactic thermosensory behaviors (two independent lines), as compared with 23% of wild-type animals. The AIY and AIZ interneurons are also the postsynaptic partners of additional ciliated sensory neurons (White et al., 1986). It is possible that expression of thermosensory properties in these ciliated neurons results in inappropriate thermosensory inputs into the AIY-AIZ-regulated thermotactic motor circuit. To determine whether adoption of AFD-specific characteristics corresponds with a loss of neuron-specific functions, we examined additional sensory behaviors of osm-6::ttx-1 expressing transgenic animals. Although overall chemosensory behaviors were not significantly altered in these animals (data not shown), we found that osm-6::ttx-1 exhibited a strong daf-c phenotype (Figure 7B). This could result from the expression of AFD-specific characteristics in the ASI and ADF neurons (see above) and the consequent loss of neuron-specific functions. Animals lacking ASI and ADF neuron function have been shown to be daf-c (Bargmann and Horvitz, 1991; Schackwitz et al., 1996). Alternatively, the daf-c phenotype could result from overexpression of ttx-1 in the AFD neurons and altered temperature regulation of neuroendocrine signaling. However, overexpression of a GFP-tagged, full-length ttx-1 fusion gene did not result in altered dauer formation (Figure 7B), suggesting that the daf-c phenotype of osm-6::ttx-1 misexpressing animals could result in part from defects in ASI and/or ADF neuron function. Six pairs of chemosensory neurons in the head including the ASI neurons (but not the AFD neurons) fill with the lipophilic dye DiI (Perkins et al., 1986; Starich et al., 1995). If misexpression of ttx-1 results in conversion of the ASI neurons to AFD fate, then the ASI neurons should fail to fill with DiI in osm-6::ttx1-expressing transgenic animals. We found that, although 94% of wild-type animals showed DiI filling in at least one ASI neuron (n ⫽ 71), only 10% and 20% of transgenic animals, respectively, from the two osm6::ttx-1 misexpressing lines showed filling in at least one ASI neuron (n ⬎ 50 for each line). In contrast, ASI neurons

ttx-1 Specifies Thermosensory Neuron Fate 951

Figure 6. Misexpression of ttx-1 Is Sufficient to Confer AFD-Specific Characteristics on Other Ciliated Neurons (A and B) Expression of a stably integrated gcy-8::GFP fusion gene in a wild-type animal (A) and in an animal misexpressing ttx-1 driven by the osm-6 promoter from an extrachromosomal array (B, lateral view). Note GFP expression in multiple cells in (B) and conversion of sensory endings to those resembling AFD. (C) Expression of an osm-10::GFP fusion gene in the PHB neuron of the phasmid in a wild-type animal (Hart et al., 1999). Note the single cilium of the PHB neuron. (D) Expression of gcy-8::GFP in the PHB neuron of an animal misexpressing ttx-1 driven by the osm-6 promoter. Note the presence of microvilli, similar to those of the AFD neurons. Anterior is at left in all images. Scale bar: 10 ␮m.

in nontransgenic animals from the same line showed consistent DiI filling. Taken together, our results indicate that misexpression of ttx-1 confers thermosensory neuron-specific properties onto a subset of ciliated neurons. Discussion The Functions of Members of the otd/Otx Gene Family Are Conserved The otd/Otx family of homeobox genes shows remarkable conservation of both sequence and function across

metazoan phyla. Otx orthologs have been identified in organisms ranging from the cnidarian Hydra vulgaris to chordates (Klein and Li, 1999). Otx genes are expressed in the anterior regions of most animals with an anteriorposterior body axis and are required for their patterning (Simeone, 1998; Hirth and Reichert, 1999; Klein and Li, 1999; Acampora et al., 2000). Specifically, Otx genes have been implicated in the patterning and development of anterior brain structures (Simeone, 1998; Hirth and Reichert, 1999; Acampora et al., 2000). We have now shown that ttx-1, a homolog of otd/Otx in C. elegans, is

Figure 7. Misexpression of ttx-1 Results in Altered Thermosensory Behaviors and Dauer Formation (A) Thermosensory behaviors of wild-type and transgenic animals carrying osm-6::ttx-1 on an extrachromosomal array. The thermosensory behaviors of transgenic animals from two independently generated lines are shown. The behavior of animals from a third line was indistinguishable from that of wild-type animals (not shown). All animals were raised at 20⬚C in the presence of food. Numbers shown are from a total of three independent assays. (B) osm-6::ttx-1 misexpressing animals show altered dauer formation. Shown is the percentage of dauers among wild-type or transgenic animals from the two lines examined in (A) above, after 72 hr at 25.5⬚C. No dauer animals were observed after 48 hr at this temperature, suggesting a delayed growth of these animals. No dauers were observed among nontransgenic animals or in a transgenic line carrying the coinjection marker and an unrelated construct. More than 80 animals were counted for each genotype. Numbers shown are from a single experiment with all animals examined in parallel; experiments performed a minimum of five independent times showed similar relative differences.

Neuron 952

also expressed in the anterior region and is required for the development of a sensory cell type. Interestingly, similar to ttx-1, Otx genes have been implicated in the development of sensory structures in several metazoan species. Most intriguing is the conservation of specialized function between the Otx-related gene Crx in vertebrates and that of ttx-1. CRX has been implicated in the differentiation of the vertebrate photoreceptors (Freund et al., 1997; Furukawa et al., 1997, 1999; Swain et al., 1997; Sohocki et al., 1998) and shares 76% identity in its homeodomain with TTX-1. Both ttx-1 and Crx are expressed relatively specifically in sensory cells (AFD thermosensory neurons, and photoreceptors and pineal glands, respectively) during development and are required for their functions (this work; Hedgecock and Russell, 1975; Mori and Ohshima, 1995; Furukawa et al., 1997, 1999). In addition to Crx, mouse Otx1 has been shown to be required for the development of the ciliary structures of the eye and the semicircular canals of the ear (Acampora et al., 1996, 2000). In Drosophila, hypomorphic alleles of otd affect the development of the ocelli, rudimentary light-sensing structures located on the head, and photoreceptors (Finkelstein and Perrimon, 1990; Royet and Finkelstein, 1995; Vandendries et al., 1996). Thus, several members of the Otx gene family have been implicated in the development of sensory structures and specifically of photosensory cell types.

Conservation of Developmental Regulatory Mechanisms between the Thermosensory Circuit of C. elegans and the Vertebrate Visual Circuit Interestingly, while the related genes ttx-1 and Crx have been shown to be required for the differentiation of the AFD thermosensory and the vertebrate photoreceptor cells in the retina respectively, another orthologous gene pair has been shown to be required for the development and differentiation of their primary postsynaptic partners. In the vertebrate retina, the paired-type homeobox gene Chx10/Vsx-1 has been implicated in the development of the bipolar cells, a major class of interneurons in the inner nuclear layer (Levine et al., 1994; Liu et al., 1994; Burmeister et al., 1996; Barabino et al., 1997; Chen and Cepko, 2000). In C. elegans, ceh-10, the ortholog of Chx10, is required for the differentiation of the AIY interneurons, the major postsynaptic partner of the AFD neurons (Svendsen and McGhee, 1995; Forrester et al., 1998; Altun-Gultekin et al., 2001). Photoresponses have been described for C. elegans (Burr, 1985), although the cellular bases for these responses are unknown, and these responses have not been genetically characterized. It has been proposed previously that the conservation of expression patterns and function between ceh-10 and Chx10 may suggest a common evolutionary history between the thermosensory circuit and the photosensory circuit (Svendsen and McGhee, 1995). Results in this paper further extend this hypothesis. Alternatively, this conservation of regulatory mechanisms between the thermosensory and photosensory circuits may represent an example of a developmental genetic pathway recruited independently to perform multiple tasks in different developmental contexts.

The Thermosensory Circuit The food-associated thermotaxis behavior of C. elegans is regulated by the thermosensory circuit. This circuit consists of two counterbalancing drives that promote movement either upward or downward in a thermal gradient. Crosstalk between these two opposing pathways results in the correct choice of the preferred temperature (Mori and Ohshima, 1995). Similarly, in the hypothalamus, the anterior and the posterior areas regulate decreases or increases in body temperature, respectively (Kupfermann et al., 2000). Removal of both branches of the thermosensory circuit by ablation of the AFD and AIZ neurons results in athermotactic behavior (Mori and Ohshima, 1995). However, a lin-11; ceh-14 ttx-3 triple mutant that has been suggested to also abolish the functions of the AFDAIY-AIZ circuit shows cryophilic behavior (Cassata et al., 2000). This apparent contradictory result has led to the proposal of an additional underlying cryophilic circuit, whose function is revealed only in the absence of the AFD-AIY-AIZ circuit (Cassata et al., 2000). A basic assumption of this model is that mutations in lin-11, ceh-14, and ttx-3 completely abolish all thermoregulatory functions of AIZ, AFD, and AIY neurons, respectively. Mutations in ttx-1 as well as in ttx-3 result in strong cryophilic behavior, similar to that of animals in which either the AFD or AIY neurons have been killed (this work; Mori and Ohshima, 1995; Hobert et al., 1997). Thus, it is likely that most, if not all, thermoregulatory functions of the AFD neurons in ttx-1 mutants and AIY neurons in ttx-3 mutants are abolished. However, mutations in lin-11 only partly affect AIZ thermoregulatory function (this work; Hobert et al., 1998). Consistent with this, lin-11; ttx-1 double mutants continue to exhibit strong cryophilic behavior. Similarly, mutations in ceh-14 may only partly affect the thermosensory functions of the AFD neurons (Cassata et al., 2000; this work). On the basis of these results, we suggest that the observed cryophilic behavior of the lin-11; ceh-14 ttx-3 triple mutant results from a strong loss of function of only the thermophilic arm, but not the cryophilic arm, of the circuit and that the AFD-AIY-AIZ circuit is likely the major circuit mediating the examined thermotaxis behavior in C. elegans. Mutations in ttx-1 May Reset the Internal Preferred or Selected Temperature A key feature of thermoregulation in endotherms is the maintenance of internal temperature homeostasis by the hypothalamus (Bligh, 1973; Hensel, 1973). Ectotherms may also maintain an internal body temperature (the preferred or selected temperature), primarily through behavioral modifications (Crawshaw, 1980; Crawford, 1982). In ttx-1 mutants, the arm of the circuit mediating responses to cold temperature is intact, whereas the arm mediating sensation of warm temperatures is affected. It is possible that when ttx-1 mutants are grown at a given temperature in the presence of food, the absence of signals from the AFD-AIY arm of the circuit results in resetting the animal’s food-associated preferred temperature to a lower value. Thus, upon being challenged with a thermal gradient, ttx-1 mutants move to lower temperatures and are therefore cryophilic.

ttx-1 Specifies Thermosensory Neuron Fate 953

Because several temperature-regulated processes are normal in ttx-1 mutants (Ailion and Thomas, 2000), multiple thermosensory circuits must function to sense temperature and to relay them to different behavioral, developmental, and endocrine circuits. Under the conditions examined, entry into the dauer stage is only weakly affected in ttx-1 mutants, whereas exit is severely affected (this work; Ailion and Thomas, 2000). The sensory neurons and pathways regulating entry and exit from the dauer stage are distinct (Bargmann and Horvitz, 1991; Schackwitz et al., 1996; Tissenbaum et al., 2000), and their outputs are likely regulated by distinct sources of sensory inputs. Recovery from (but not entry into) the dauer stage may involve cholinergic signaling via a muscarinic acetylcholine pathway (Tissenbaum et al., 2000). It has been proposed that temperature inputs via the AFD-AIY-AIZ circuit promote acetylcholine release from unknown neurons, which in turn promotes insulin signaling and consequent recovery from the dauer stage (Tissenbaum et al., 2000). This is analogous to the hypothalamus-mediated temperature control of metabolic rate. Although daf-7; ttx-1 and daf-7; ttx-3 mutants enter the dauer stage (this work; Hobert et al., 1997), the resetting of the preferred or selected temperature to lower values may result in the inappropriate production of high levels of acetylcholine and the consequent insulin signaling, resulting in rapid recovery. Consistent with this hypothesis, neither daf-2; ttx-1 nor daf-2; ttx-3 double mutants recover at high temperatures (this work; O. Hobert, personal communication), confirming that insulin signaling, but not TGF-␤ signaling, is crucial for dauer recovery (Tissenbaum et al., 2000). Taken together, the AFD-regulated thermosensory circuit appears to act similarly to the hypothalamus in regulating both motor behavior and endocrine signaling in response to temperature cues. Few molecules involved in thermosensation have been identified. A major goal for the future is the identification of downstream targets of TTX-1 because a subset of these is expected to be genes directly involved in thermosensation. Experimental Procedures Strains Worms were grown using standard methods (Brenner, 1974). Strains were obtained from the Caenorhabditis Genetics Center unless noted otherwise. ttx-3(mg158) was provided by O. Hobert. Strains carrying stably integrated fusion genes were provided as indicated: mgIs18 (ttx-3::GFP) (Altun-Gultekin et al., 2001); kyIs111 (tax-2::GFP), kyIs104 (str-1::GFP), and kyIs140 (str-2::GFP) (C. Bargmann) (Coburn and Bargmann, 1996; Troemel et al., 1997, 1999); oyIs7 (odr-3::GFP) (T. Sarafi-Reinach) (Roayaie et al., 1998); chIs513 (ceh-14::GFP) (Cassata et al., 2000); and nuIs11 (osm-10::GFP) (Hart et al., 1999). Strains carrying stably integrated gcy-8::GFP (Yu et al., 1997) and nhr-38::GFP (Miyabayashi et al., 1999) fusion genes, together with lin-15(⫹) (Huang and Erikson, 1994) as a coinjection marker, were generated by treating animals carrying extrachromosomal arrays with UV/psoralen. The following integrated strains were isolated: oyIs15 (nhr-38::GFP), oyIs17 (gcy-8::GFP), and oyIs18 (gcy-8::GFP). Strains were mapped and outcrossed a total of six times prior to characterization. Double-mutant strains were constructed using standard methods and confirmed by complementation tests or by sequencing. Details of double-mutant strain construction are available upon request. Isolation, Mapping, and Cloning of ttx-1 oyIs17 animals were mutagenized with EMS using standard protocols. F2 progeny of mutagenized animals were examined for re-

duced gcy-8::GFP expression using a Leica MZ12 dissection microscope equipped with epifluorescence. oy26 and oy29 alleles were backcrossed a total of three times prior to analysis for gene expression phenotypes and a total of six times prior to behavioral analysis. The p767, oy26, and oy29 alleles fail to complement each other for the defect in AFD-specific gene expression phenotype. ttx-1 was mapped to the unc-51-rol-9 interval on LG V using standard three-factor crosses and mapping with respect to deficiencies. The presence of ttx-1 mutations was followed by reduced expression of gcy-8::GFP in all mapping experiments. To fine map ttx-1, single nucleotide polymorphisms (SNPs) in the unc-51-rol-9 interval were identified between the Bristol N2 and the CB4856 strain essentially as described (Jakubowski and Kornfeld, 1999). CB4856 males were crossed to unc-51 ttx-1 rol-9; oyIs18 hermaphrodites, and Rol non Unc recombinant progeny were isolated. Recombinant progeny were then scored for the presence or absence of the ttx-1 mutation and each SNP. The primers and sequence alterations that correspond to these SNPs have been deposited in Wormbase (http://www.wormbase.org/). A 14 kb PCR product containing 3 kb of promoter sequences, all ttx-1 exons, and ⵑ1 kb of 3⬘ flanking sequences was subcloned into the TOPO2.1 vector (Invitrogen) and found to rescue the gcy8::GFP expression and thermosensory behavioral defects in three of three independent lines. To identify the molecular lesions in the ttx-1 alleles, the genomic coding sequence of ttx-1 was amplified from p767, oy26, and oy29 mutants. Three independent PCR reactions were combined and the products directly sequenced. Because we did not detect a molecular lesion in the oy29 allele, it is formally possible that this allele defines a nonallelic noncomplementing locus. Characterization of ttx-1 cDNAs Four EST clones (yk345e3, yk51g10, yk600c12, and yk74e5) corresponding to the ttx-1 locus were obtained from Yuji Kohara and sequenced to determine the primary structure of the ttx-1 cDNA. In all cases, we found that exon 8 is 94 bp longer than predicted, resulting in an alteration of the reading frame of exon 9. Two alternatively spliced forms were identified. yk345e3 lacks the fourth exon (GenBank accession no. AF381627), whereas yk51g10 and yk74e5 contain the fourth exon. For expression experiments, a modified yk345e3 cDNA was used, which contains an additional 24 bases at the 5⬘ end, including the predicted ATG start codon. Identification of Transcripts in ttx-1(p767) RT-PCR was performed using the 5⬘ RACE System (GIBCO-BRL). First-strand synthesis on 5 ␮g of total RNA from wild-type animals was performed using a primer specific for sequences in exon 7 of ttx-1. ttx-1-specific sequences were then amplified using primers corresponding to sequences in exon 1 and exon 7 of ttx-1. PCR products were cloned into the TOPO2.1 vector, and several independent isolates were sequenced. Expression Constructs and Generation of Transgenic Animals The following were obtained from the indicated sources: osm-6 promoter (P. Swoboda) (Collet et al., 1998), rat Otx1 cDNA (S. McConnell), and mouse Crx cDNA (C. Cepko). To generate ttx-1 expression constructs, 3 kb of promoter sequences and sequences encoding either the first 13 residues (ttx1Exon-1::GFP) or 89 residues (ttx-1Exon4::GFP) were amplified by PCR from genomic DNA and fused in frame to GFP. ttx-1::GFPtag was generated by amplifying 3 kb of promoter sequences and all genomic sequences of ttx-1 and fusing in frame to GFP just prior to the STOP codon. To generate ttx-1, Crx, or Otx1 expression constructs, cDNAs were amplified, sequenced, and fused in frame to ttx-1Exon1 sequences. The amplified Crx cDNA contains a single silent base change. ttx-1 misexpression constructs were generated by fusing the osm-6, odr-10, or str-1 promoters in frame to a ttx-1 cDNA (see above). Germline Transformations and Microscopy Germline transformations were carried out using standard protocols (Mello and Fire, 1995). Coinjection markers used were ofm-1::GFP

Neuron 954

(Miyabayashi et al., 1999) at 30 ng/␮l, pJM23 (Huang and Erikson, 1994) at 50 ng/␮l, and pRF4 rol-1(su1006) at 100 ng/␮l. All plasmids were injected at 7.5–30 ng/␮l. All microscopy was carried out as previously described (SarafiReinach and Sengupta, 2000). RNA Interference PCR fragments of unc-22 and ttx-1 cDNAs were used as templates to synthesize RNA using Megascript T7 (Ambion). dsRNA was injected into hermphrodite gonads at 2 mg/ml. Single-injected animals were allowed to lay eggs and were transferred every 4–12 hr postinjection for 31 hr. Progeny were observed at all stages for a visual phenotype, and the total brood size of each injected animal was counted. For unc-22(RNAi), 97% of the progeny of injected animals exhibited a twitcher phenotype (Fire et al., 1998) (n ⫽ ⵑ500; five independent injected animals). Dauer Assays Dauer formation and recovery assays were carried out as described (Hobert et al., 1997; Ailion and Thomas, 2000; Daniels et al., 2000). Six adults of a given genotype were allowed to lay eggs for 5–6 hr at 22⬚C on HB101 bacteria. Parents were then removed, and the plates were placed at an incubator at 25.5⬚C. After 48 hr (for dauer formation assays) or 72 hr (for dauer recovery assays), animals were scored visually as dauer or nondauer on the basis of morphological characteristics (Cassada and Russell, 1975). To examine dauer formation at 15⬚C, animals were treated as described but scored at 120 hr. All relevant comparisons included all strains assayed in parallel.

Acampora, D., Mazan, S., Lallemand, Y., Avantaggiato, V., Maury, M., Simeone, A., and 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. Ailion, M., and Thomas, J.H. (2000). Dauer formation induced by high temperatures in Caenorhabditis elegans. Genetics 156, 1047–1067. Altun-Gultekin, Z., Andachi, Y., Tsalik, E.L., Pilgrim, D., Kohara, Y., and Hobert, O. (2001). A regulatory cascade of three homeobox genes, ceh-10, ttx-3, and ceh-23 controls cell fate specification of a defined interneuron class in C. elegans. Development 128, 1951–1969. Barabino, S.M., Spada, F., Cotelli, F., and Boncinelli, E. (1997). Inactivation of the zebrafish homologue of Chx10 by antisense oligonucleotides causes eye malformations similar to the ocular retardation phenotype. Mech. Dev. 63, 133–143. Bargmann, C.I., Hartwieg, E., and Horvitz, H.R. (1993). Odorantselective genes and neurons mediate olfaction in C. elegans. Cell 74, 515–527. Bargmann, C.I., and Horvitz, H.R. (1991). Control of larval development by chemosensory neurons in Caenorhabditis elegans. Science 251, 1243–1246. Bargmann, C.I., Thomas, J.H., and Horvitz, H.R. (1990). Chemosensory cell function in the behavior and development of Caenorhabditis elegans. Cold Spring Harbor Symp. Quant. Biol. 55, 529–538. Bligh, J. (1973). Temperature Regulation in Mammals and Other Vertebrates (New York: American Elsevier Publishing Company). Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics 77, 71–94.

Behavioral Assays Olfactory and osmotic avoidance assays were performed essentially as described (Bargmann et al., 1990, 1993). Thermotaxis assays were carried out as described previously (Mori and Ohshima, 1995).

Burmeister, M., Novak, J., Liang, M.Y., Basu, S., Ploder, L., Hawes, N.L., Vidgen, D., Hoover, F., Goldman, D., Kalnins, V.I., et al. (1996). Ocular retardation mouse caused by Chx10 homeobox null allele: impaired retinal progenitor proliferation and bipolar cell differentiation. Nat. Genet. 12, 376–384.

Acknowledgments

Burr, A.H. (1985). The photomovement of Caenorhabditis elegans, a nematode which lacks ocelli. Proof that the response is to light not radiant heating. Photochem. Photobiol. 41, 577–582.

We thank Julia Thompson and Anne Lanjuin for experimental assistance, Marc Colosimo for the osm-6 misexpression construct, the Caenorhabditis Genetics Center for strains, Ed Dougherty for assistance with confocal microscopy, Andy Fire for plasmids, Connie Cepko for the Crx cDNA, Sue McConnell for the Otx1 cDNA, Oliver Hobert for the ttx-3(mg158) null allele, and Yuji Kohara for ttx-1 cDNAs. We are grateful to Leon Avery for identification of the ttx1-expressing marginal cells. We acknowledge helpful suggestions and advice throughout the course of this work from members of the Sengupta and Mori labs. We thank Oliver Hobert, Cori Bargmann and members of the Sengupta lab for critical comments on the manuscript. This work was supported in part by an NIH training grant (T32 NS07292) and an NRSA postdoctoral fellowship to J.S. (F32 GM20048), NIH grant to P.S. (RO1 GM56223), and funds from the Packard Foundation. I.M. was supported by grants from the Ministry of Education, Science, Sports, and Culture of Japan and from the Japan Science and Technology Corporation. Received May 21, 2001; revised July 5, 2001. References Acampora, D., Avantaggiato, V., Tuorto, F., Barone, P., Reichert, H., Finkelstein, R., and Simeone, A. (1998). Murine Otx1 and Drosophila otd genes share conserved genetic functions required in invertebrate and vertebrate brain development. Development 125, 1691–1702. Acampora, D., Avantaggiato, V., Tuorto, F., and Simeone, A. (1997). Genetic control of brain morphogenesis through Otx gene dosage requirement. Development 124, 3639–3650. Acampora, D., Gulisano, M., and Simeone, A. (2000). Genetic and molecular roles of Otx homeodomain proteins in head development. Gene 246, 23–35. Acampora, D., Mazan, S., Avantaggiato, V., Barone, P., Tuorto, F., Lallemand, Y., Brulet, P., and Simeone, A. (1996). Epilepsy and brain abnormalities in mice lacking the Otx1 gene. Nat. Genet. 14, 218–222.

Cassada, R.C., and Russell, R.L. (1975). The dauer larva, a postembryonic developmental variant of the nematode Caenorhabditis elegans. Dev. Biol. 46, 326–342. Cassata, G., Kagoshima, H., Andachi, Y., Kohara, Y., Durrenberger, M.B., Hall, D.H., and Burglin, T.R. (2000). The LIM homeobox gene ceh-14 confers thermosensory function to the AFD neurons in Caenorhabditis elegans. Neuron 25, 587–597. Caterina, M.J., Schumacher, M.A., Tominaga, M., Rosen, T.A., Levine, J.D., and Julius, D. (1997). The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389, 816–824. Caterina, M.J., Rosen, T.A., Tominaga, M., Brake, A.J., and Julius, D. (1999). A capsaicin-receptor homologue with a high threshold for noxious heat. Nature 398, 436–441. Chen, C.M., and Cepko, C.L. (2000). Expression of Chx10 and Chx10–1 in the developing chicken retina. Mech. Dev. 90, 293–297. Coburn, C., and Bargmann, C.I. (1996). A putative cyclic nucleotidegated channel is required for sensory development and function in C. elegans. Neuron 17, 695–706. Collet, J., Spike, C.A., Lundquist, E.A., Shaw, J.E., and Herman, R.K. (1998). Analysis of osm-6, a gene that affects sensory cilium structure and sensory neuron function in Caenorhabditis elegans. Genetics 148, 187–200. Crawford, E.C. (1982). Behavioral and autonomic thermoregulation in terrestrial ectotherms. In A Companion to Animal Physiology, C.R. Taylor, K. Johansen, and L. Bolis, eds. (New York: Cambridge University Press), pp. 198–215. Crawshaw, L.I. (1980). Temperature regulation in vertebrates. Ann. Rev. Physiol. 42, 473–491. Daniels, S.A., Ailion, M., Thomas, J.H., and Sengupta, P. (2000). egl-4 acts through a transforming growth factor-beta/SMAD pathway in Caenorhabditis elegans to regulate multiple neuronal circuits in response to sensory cues. Genetics 156, 123–141. Finkelstein, R., and Perrimon, N. (1990). The orthodenticle gene

ttx-1 Specifies Thermosensory Neuron Fate 955

is regulated by bicoid and torso and specifies Drosophila head development. Nature 346, 485–488. Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E., and Mello, C.C. (1998). Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811. Forrester, W.C., Perens, E., Zallen, J.A., and Garriga, G. (1998). Identification of Caenorhabditis elegans genes required for neuronal differentiation and migration. Genetics 148, 151–165. Freund, C.L., Gregory-Evans, C.Y., Furukawa, T., Papaioannou, M., Looser, J., Ploder, L., Bellingham, J., Ng, D., Herbrick, J.A., Duncan, A., et al. (1997). Cone-rod dystrophy due to mutations in a novel photoreceptor-specific homeobox gene (CRX) essential for maintenance of the photoreceptor. Cell 91, 543–553. Furukawa, T., Morrow, E.M., and Cepko, C.L. (1997). Crx, a novel otx-like homeobox gene, shows photoreceptor-specific expression and regulates photoreceptor differentiation. Cell 91, 531–541. Furukawa, T., Morrow, E.M., Li, T., Davis, F.C., and Cepko, C.L. (1999). Retinopathy and attenuated circadian entrainment in Crxdeficient mice. Nat. Genet. 23, 466–470. Galliot, B., de Vargas, C., and Miller, D. (1999). Evolution of homeobox genes: Q50 Paired-like genes founded the Paired class. Dev. Genes Evol. 209, 186–197. Golden, J.W., and Riddle, D.L. (1984a). The Caenorhabditis elegans dauer larva: developmental effects of pheromone, food, and temperature. Dev. Biol. 102, 368–378. Golden, J.W., and Riddle, D.L. (1984b). A pheromone-induced developmental switch in Caenorhabditis elegans: Temperature-sensitive mutants reveal a wild-type temperature-dependent process. Proc. Natl. Acad. Sci. USA 81, 819–823. Gomez, M., De Castro, E., Guarin, E., Sasakura, H., Kuhara, A., Mori, I., Bartfai, T., Bargmann, C.I., and Nef, P. (2001). Ca2⫹ signaling via the neuronal calcium sensor-1 regulates associative learning and memory in C. elegans. Neuron 30, 241–248. Gottlieb, S., and Ruvkun, G. (1994). daf-2, daf-16 and daf-23: genetically interacting genes controlling dauer formation in Caenorhabditis elegans. Genetics 137, 107–120. Hart, A.C., Kass, J., Shapiro, J.E., and Kaplan, J.M. (1999). Distinct signaling pathways mediate touch and osmosensory responses in a polymodal sensory neuron. J. Neurosci. 19, 1952–1958. Hedgecock, E.M., and Russell, R.L. (1975). Normal and mutant thermotaxis in the nematode Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 72, 4061–4065. Hensel, H. (1973). Neural processes in thermoregulation. Physiol. Rev. 53, 948–1017. Hirth, F., and Reichert, H. (1999). Conserved genetic programs in insect and mammalian brain development. Bioessays 21, 677–684. Hirth, F., Therianos, S., Loop, T., Gehring, W.J., Reichert, H., and Furukubo-Tokunaga, K. (1995). Developmental defects in brain segmentation caused by mutations of the homeobox genes orthodenticle and empty spiracles in Drosophila. Neuron 15, 769–778. Hobert, O., D’Alberti, T., Liu, Y., and Ruvkun, G. (1998). Control of neural development and function in a thermoregulatory network by the LIM homeobox gene lin-11. J. Neurosci. 18, 2084–2096. Hobert, O., Mori, I., Yamashita, Y., Honda, H., Ohshima, Y., Liu, Y., and Ruvkun, G. (1997). Regulation of interneuron function in the C. elegans thermoregulatory pathway by the ttx-3 LIM homeobox gene. Neuron 19, 345–357. Huang, W., and Erikson, R.L. (1994). Constitutive activation of Mek1 by mutation of serine phosphorylation sites. Proc. Natl. Acad. Sci. USA 91, 8960–8963.

Komatsu, H., Mori, I., and Ohshima, Y. (1996). Mutations in a cyclic nucleotide-gated channel lead to abnormal thermosensation and chemosensation in C. elegans. Neuron 17, 707–718. Kupfermann, I., Kandel, E.R., and Iversen, S. (2000). Motivational and addictive states. In Principles of Neural Science, E.R. Kandel, J.H. Schwarz, and T. M. Jessell, eds.(New York: McGraw-Hill), pp. 998–1013. Leuzinger, S., Hirth, F., Gerlich, D., Acampora, D., Simeone, A., Gehring, W.J., Finkelstein, R., Furukubo-Tokunaga, K., and Reichert, H. (1998). Equivalence of the fly orthodenticle gene and the human OTX genes in embryonic brain development of Drosophila. Development 125, 1703–1710. Levine, E.M., Hitchcock, P.F., Glasgow, E., and Schechter, N. (1994). Restricted expression of a new paired-class homeobox gene in normal and regenerating adult goldfish retina. J. Comp. Neurol. 348, 596–606. Liu, I.S., Chen, J.D., Ploder, L., Vidgen, D., van der Kooy, D., Kalnins, V.I., and McInnes, R.R. (1994). Developmental expression of a novel murine homeobox gene (Chx10): evidence for roles in determination of the neuroretina and inner nuclear layer. Neuron 13, 377–393. Matsuo, I., Kuratani, S., Kimura, C., Takeda, N., and Aizawa, S. (1995). Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev. 9, 2646–2658. Mello, C.C., and Fire, A. (1995). DNA transformation. In Caenorhabditis elegans: Modern Biological Analysis of an Organism, H.F. Epstein and D.C. Shakes, eds. (San Diego: Academic Press), pp. 452–480. Miyabayashi, T., Palfreyman, M.T., Sluder, A.E., Slack, F., and Sengupta, P. (1999). Expression and function of members of a divergent nuclear receptor family in Caenorhabditis elegans. Dev. Biol. 215, 314–331. Mori, I. (1999). Genetics of chemotaxis and thermotaxis in the nematode Caenorhabditis elegans. Annu. Rev. Genet. 33, 399–422. Mori, I., and Ohshima, Y. (1995). Neural regulation of thermotaxis in Caenorhabditis elegans. Nature 376, 344–348. Morsli, H., Tuorto, F., Choo, D., Postiglione, M.P., Simeone, A., and Wu, D.K. (1999). Otx1 and Otx2 activities are required for the normal development of the mouse inner ear. Development 126, 2335–2343. Nagao, T., Leuzinger, S., Acampora, D., Simeone, A., Finkelstein, R., Reichert, H., and Furukubo-Tokunaga, K. (1998). Developmental rescue of Drosophila cephalic defects by the human Otx genes. Proc. Natl. Acad. Sci. USA 95, 3737–3742. Perkins, L.A., Hedgecock, E.M., Thomson, J.N., and Culotti, J.G. (1986). Mutant sensory cilia in the nematode Caenorhabditis elegans. Dev. Biol. 117, 456–487. Reichling, D.B., and Levine, J.D. (2000). In hot pursuit of the elusive heat transducers. Neuron 26, 555–558. Riddle, D.L., and Albert, P.S. (1997). Genetic and environmental regulation of dauer larva development. In C. elegans II, D.S. Riddle, T. Blumenthal, B.J. Meyer, and J. R. Priess, eds. (Cold Spring Harbor, NY: Cold Spring Harbor Press), pp. 739–768. Roayaie, K., Crump, J.G., Sagasti, A., and Bargmann, C.I. (1998). The G␣ protein ODR-3 mediates olfactory and nociceptive function and controls cilium morphogenesis in C. elegans olfactory neurons. Neuron 20, 55–67. Royet, J., and Finkelstein, R. (1995). Pattern formation in Drosophila head development: the role of the orthodenticle homeobox gene. Development 121, 3561–3572. Ruvkun, G., and Hobert, O. (1998). The taxonomy of developmental control in Caenorhabditis elegans. Science 282, 2033–2041.

Iggo, A. (1990). Whither sensory specificity? In Thermoreception and Temperature regulation, J. Bligh and K. Voigt, eds. (Berlin: SpringerVerlag), pp. 9–18.

Sagasti, A., Hobert, O., Troemel, E.R., Ruvkun, G., and Bargmann, C.I. (1999). Alternative olfactory neuron fates are specified by the LIM homeobox gene lim-4. Genes Dev. 13, 1794–1806.

Jakubowski, J., and Kornfeld, K. (1999). A local, high-density, singlenucleotide polymorphism map used to clone Caenorhabditis elegans cdf-1. Genetics 153, 743–752.

Sarafi-Reinach, T.R., and Sengupta, P. (2000). The forkhead domain gene unc-130 generates chemosensory neuron diversity in C. elegans. Genes Dev. 14, 2472–2485.

Klein, W.H., and Li, X. (1999). Function and evolution of Otx proteins. Biochem. Biophys. Res. Commun. 258, 229–233.

Sarafi-Reinach, T.R., Melkman, T., Hobert, O., and Sengupta, P. (2001). The lin-11 LIM homeobox gene specifies olfactory and che-

Neuron 956

mosensory neuron fates in C. elegans. Development 128, 3269– 3281.

refinement of exuberant axonal projections to subcortical targets. Neuron 24, 819–831.

Schackwitz, W.S., Inoue, T., and Thomas, J.H. (1996). Chemosensory neurons function in parallel to mediate a pheromone response in C. elegans. Neuron 17, 719–728.

White, J. (1988). The anatomy. In The Nematode, Caenorhabditis elegans, W. Wood, ed. (Cold Spring Harbor, NY: Cold Spring Harbor Press), pp. 81–122.

Sengupta, P., Chou, J.H., and Bargmann, C.I. (1996). odr-10 encodes a seven transmembrane domain olfactory receptor required for responses to the odorant diacetyl. Cell 84, 899–909.

White, J.G., Southgate, E., Thomson, J.N., and Brenner, S. (1986). The structure of the nervous system of the nematode Caenorhabditis elegans. Phil. Transact. R. Soc. Lond. B 314, 1–340.

Simeone, A. (1998). Otx1 and Otx2 in the development and evolution of the mammalian brain. EMBO J. 17, 6790–6798.

Wittenburg, N., and Baumeister, R. (1999). Thermal avoidance in Caenorhabditis elegans: an approach to the study of nociception. Proc. Natl. Acad. Sci. USA 96, 10477–10482.

Simeone, A., Acampora, D., Gulisano, M., Stornaiuolo, A., and Boncinelli, E. (1992). Nested expression domains of four homeobox genes in developing rostral brain. Nature 358, 687–690. Simeone, A., Acampora, D., Mallamaci, A., Stornaiuolo, A., D’Apice, M.R., Nigro, V., and 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. Sohocki, M.M., Sullivan, L.S., Mintz-Hittner, H.A., Birch, D., Heckenlively, J.R., Freund, C.L., McInnes, R.R., and Daiger, S.P. (1998). A range of clinical phenotypes associated with mutations in CRX, a photoreceptor transcription-factor gene. Am. J. Hum. Genet. 63, 1307–1315. Spray, D.C. (1986). Cutaneous temperature receptors. Ann. Rev. Physiol. 48, 625–638. Starich, T.A., Herman, R.K., Kari, C.K., Yeh, W.-H., Schackwitz, W.S., Schuyler, M.W., Collet, J., Thomas, J.H., and Riddle, D.L. (1995). Mutations affecting the chemosensory neurons of Caenorhabditis elegans. Genetics 139, 171–188. Suda, Y., Matsuo, I., and Aizawa, S. (1997). Cooperation between Otx1 and Otx2 genes in developmental patterning of rostral brain. Mech. Dev. 69, 125–141. Sulston, J.E., Schierenberg, E., White, J.G., and Thomson, J.N. (1983). The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol. 100, 64–119. Svendsen, P.C., and McGhee, J.D. (1995). The C. elegans neuronally expressed homeobox gene ceh-10 is closely related to genes expressed in the vertebrate eye. Development 121, 1253–1262. Swain, P.K., Chen, S., Wang, Q.L., Affatigato, L.M., Coats, C.L., Brady, K.D., Fishman, G.A., Jacobson, S.G., Swaroop, A., Stone, E., et al. (1997). Mutations in the cone-rod homeobox gene are associated with the cone-rod dystrophy photoreceptor degeneration. Neuron 19, 1329–1336. Thomas, J.H., Birnby, D.A., and Vowels, J.J. (1993). Evidence for parallel processing of sensory information controlling dauer formation in Caenorhabditis elegans. Genetics 134, 1105–1117. Tissenbaum, H.A., Hawdon, J., Perregaux, M., Hotez, P., Guarente, L., and Ruvkun, G. (2000). A common muscarinic pathway for diapause recovery in the distantly related nematode species Caenorhabditis elegans and Ancylostoma caninum. Proc. Natl. Acad. Sci. USA 97, 460–465. Troemel, E.R., Kimmel, B.E., and Bargmann, C.I. (1997). Reprogramming chemotaxis responses: sensory neurons define olfactory preferences in C. elegans. Cell 91, 161–169. Troemel, E.R., Sagasti, A., and Bargmann, C.I. (1999). Lateral signaling mediated by axon contact and calcium entry regulates asymmetric odorant receptor expression in C. elegans. Cell 99, 387–398. Vandendries, E.R., Johnson, D., and Reinke, R. (1996). orthodenticle is required for photoreceptor cell development in the Drosophila eye. Dev. Biol. 173, 243–255. Vowels, J.J., and Thomas, J.H. (1992). Genetic analysis of chemosensory control of dauer formation in Caenorhabditis elegans. Genetics 130, 105–123. Ward, S., Thomson, N., White, J.G., and Brenner, S. (1975). Electron microscopical reconstruction of the anterior sensory anatomy of the nematode Caenorhabditis elegans. J. Comp. Neurol. 160, 313–337. Weimann, J.M., Zhang, Y.A., Levin, M.E., Devine, W.P., Brulet, P., and McConnell, S.K. (1999). Cortical neurons require Otx1 for the

Younossi-Hartenstein, A., Green, P., Liaw, G.J., Rudolph, K., Lengyel, J., and Hartenstein, V. (1997). Control of early neurogenesis of the Drosophila brain by the head gap genes tll, otd, ems, and btd. Dev. Biol. 182, 270–283. Yu, S., Avery, L., Baude, E., and Garbers, D.A. (1997). Guanylyl cyclase expression in specific sensory neurons: a new family of chemosensory receptors. Proc. Natl. Acad. Sci. USA 94, 3384–3387. Accession Numbers The GenBank accession number for the ttx-1 sequence (EST clone yk345e3) reported in this paper is AF381627.