Calreticulin Mediates Anesthetic Sensitivity in Drosophila melanogaster 

The Drosophila  mutant hypersensitive to diethylether, eth  ^as311 , is named as an eth  er-a  nesthetic s  ensitive mutant, is found as a mutation on chromosome 3 , and is the 11  th mutant in our laboratory. The mutant that has a P-element in the crc  locus and the P-element excision–induced revertants, revertant-2 , -3 , -5 , and -44 , have been isolated previously. 3,CyO/Sp;SbP[ry  ⁺Δ2–3 99B]/TM6 (jump starter), w;+;PrDr/TM3  (balancer), Canton-S  (wild type), and w  (white ) were used in this study. 23 CyO , TM6 , and TM3  are balancers that inhibit the crossing over. P[ry  ⁺Δ2–3 99B] is a P-element construct to produce transposase and inserted in the third chromosome at 99B. The flies were raised on a standard medium (agar, yeast, glucose, corn grits) or on an egg-collecting medium (grape juice, whisky, agar) for collecting embryos at 25°C. The experimental protocol was approved by the Ethics Review Committee for Animal Experimentation of the participating institutions in Department of Earth and Life Sciences Osaka Prefecture University.

The general cloning techniques described by Sambrook et al.  24were followed. A 2.1-kb genomic DNA fragment (pr-clone in fig. 1A) was isolated by the plasmid rescue method 25and used as a probe to isolate cDNAs from the crc  locus. Using the pr-clone as a probe, wild-type cDNA clones were isolated from a Canton-S  adult male library in the phage λ gt10. 26We determined the complete nucleotide sequence of the cDNA clone on both strands. The cDNA insert was isolated from the phage vector, subcloned into the Bluescript KS⁺vector (Stratagene, Burlingame, CA), and sequenced by a dideoxynucleotide chain termination procedure using Dye Deoxy Terminator Cycle Sequencing Kit in a model 373A autosequencer (Perkin Elmer ABI, Foster City, CA). Analysis of sequence data was searched using BLAST programs provided by the NCBI server (National Institutes of Health, Bethesda, MD) and programs provided by the Berkeley Drosophila  Genome Project server (Berkeley, CA). All reagents used in this study were supplied by Nacalai tesque Inc. (Nara, Japan) unless otherwise mentioned.

Total RNA was isolated from the embryos, larvae, pupae and adult flies. Poly(A)⁺RNA was separated on oligo(dT)-cellulose columns (Amersham Life Science, Buckinghamshire, United Kingdom). Next, 5 μg per lane of poly(A)⁺RNA was loaded per lane of the formaldehyde-agarose gel. After electrophoresis, the RNA was transferred to a Hybond-N membrane (Amersham). Fluorescein labeling of the probes, hybridization, and detection of mRNA were performed according to the instructions provided by the supplier (Amersham). Furthermore, to determine the 5′ ends of the mRNA transcripts, 200 ng poly(A)⁺was applied for 5′ RACE using the RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) kit (Ambion Inc., Austin, TX). Briefly, the RNA is treated with calf intestinal phosphatase in a 20-μl reaction mixture to remove free 5′-phosphates and to remain the cap structure found on intact 5′ ends of mRNA. After phenol:chloroform extract, the RNA is then treated with Tobacco acid pyrophosphatase to remove the cap structure, and a 45-base RNA adapter oligonucleotide is ligated to the treated RNA using T4 RNA ligase. A random-primed reverse-transcription reaction and nested polymerase chain reaction then amplifies the 5′ end of crc  transcript: outer 5′RLM-RACE polymerase chain reaction using 5′ RACE outer primer and an outer crc  primer (N-2, 5′-gatattctgctcgtgcttcac-3′, 391–371 nt) and inner 5′RLM-RACE polymerase chain reaction using 5′ RACE inner primer and an inner crc  primer (5′-gttggagaagccatcaaact-3′, 340–321 nt).

To study the expression patterns of the crc  mRNA, we synthesized digoxigenin-UTP-labeled single-stranded DNA probes. Sense and antisense probes were synthesized by polymerase chain reaction using each N, P, and C domain primers of crc  cDNA subclone: sense primers, N-1 (5′-tctgaaggagaatttcgacaac-3′, 148–169 nt), P-1 (5′-aggacgactgggacttcctg-3′, 672–691 nt), and C-1 (5′-gttaagtccggcactatcttc-3′, 1043–1063 nt); and antisense primers, N-2, P-2 (5′-tttggctgccactcgccctt-3′, 915–896 nt), and C-2 (5′-caattcgtcgtgttcgctttga-3′, 1300–1279 nt). Probes were hybridized in situ  to whole mounts of embryos and developing brains according to the method of Tautz and Pfeifle. 27The preparation mounted in glycerol was viewed using an Olympus BX60 light microscope (Olympus Optical Co., Ltd., Osaka, Japan).

Recently, a lethal excision derivative, revertant-s20-1 , was isolated by dysgenic cross. 28Briefly, flies carrying the crc  with inserted P-element of eth  ^as311  (w;+;p ) were crossed to +;CyO/Sp;SbP[ry  ⁺Δ2–3 99B]/TM6. Dysgenic progeny carrying CyO;Sb  were crossed to w;+;PrDr/TM3  flies twice. w;+;p*/TM3  (p*  represents excised P-element) progeny were selected for lines. 3 revertant-s20-1  is a line in which no homo p*/p*  progeny emerged, i.e. , lethal.

Transgenic lines with the wild-type crc  gene were produced. A  new P-factor for driving genes behind the hsp70 promoter of the plasmid, pCaspeR-hs, was constructed with w  ⁺as the selectable marker. 29This vector, containing an Eco  RI fragment of the crc  cDNA, was injected into w  embryos. Chromosomal localization of the transgenes and the generation of homozygotes for the transgenes were performed by standard crosses. The crc  ⁺transgene resides on the second chromosome (crc-T1  and crc-T2 ) and mutated crc  gene on the third chromosome in crc-T1;eth  ^as311  and cr  c-T2;eth  ^as311 , respectively.

For the estimation of anesthetic concentration, each of 20 1-day-old female and male flies were administered a given concentration of anesthetics, and their responses were scored using avoidance reflex from stimulation by a fine brush as an endpoint. 3Diethylether, isoflurane (Forane; Abbott Laboratories Inc., Park, IL), and halothane (Fluothane; Takeda Chemical Industries, Osaka, Japan) were used. The dose–response curves from at least four to nine data points were linearized by log-probit transformation to estimate EC⁵⁰with 95% confidence limits. Comparison of EC⁵⁰s was conducted by analysis of covariance. 18 

eth  ^as311 , a hypersensitive mutant to diethylether anesthesia, was screened by determining EC⁵⁰of diethylether. It has a modified P-element between the eight base repeats (CTGACTGA) of 26–33-bp upstream from the initiation codon (ATG) within the first exon of the crc  locus. 3We obtained a 1.4-kb cDNA, sequenced it, and performed a 5′ RACE analysis (EMBL sequence accession No. AB000718). We identified a 1420-bp full-length cDNA including the 5′ and 3′ untranslated regions of crc , 82-bp and 117-bp, respectively (fig. 1A). The open reading frame and deduced amino acid sequence were similar to those predicted from the genomic sequencing reported by Smith. 30,Drosophila  calreticulin has high homology (∼ 70%) to calreticulins of other eukaryote species, indicating structural and functional similarities among them. The model of Drosophila  calreticulin shown in figure 1Bis built from a mammalian model of calreticulin. 20,21Northern blotting analysis showed that only a 1.4-kb transcript was found at all developmental stages in the wild type and eth  ^as311  (fig. 1C). In the wild type, the highest expression was observed at pupa stage; however, the mRNA level was much lower in the mutant than the wild type at all developmental stages, including pupal stage (fig. 1D). These results show that eth  ^as311  is a low-expression mutation in the crc  gene and support that eth  ^as311  is a fertile mutation. In fact, the endogenous calreticulins were detected in the wild type and the mutant (Sumiko Gamo, M.D., and Dai Keyakidani, M.S., Department of Earth and Sciences, Osaka Prefecture University, Sakai, Osaka, Japan, unpublished data, April 1999). However, another Drosophila  mutant, l(3)s114307 , and the knockout mice of calreticulin gene are lethal. 31,32 

crc  mRNA is expressed at all developmental stages in wild-type Drosophila . Whole-mount in situ  hybridization of the crc  transcripts was performed to identify the cells and tissues that are stained by the antisense cDNA of the probe at various embryonic stages. At the cellular blastoderm, where embryonic cells distribute on the egg surface, the perinuclear cytoplasm was stained, but no staining of the cell nucleus (fig. 2A-1) was found compared to a sense RNA probe (fig. 2A-2). The cells on the egg surface were stained, but the yolk granules were not stained at the central region of the egg. At embryonic stage 9, the crc  mRNA was expressed ubiquitously, especially at pericephalic and ventral neuroblast regions of the central nervous system (fig. 2B-1), but no expression was detected when a sense probe of crc  was used (fig. 2B-2). Strong expression was detected in the head region, ventral nervous systems, trachea, and hind gut at embryonic stage 15 in the wild type (fig. 2C-1), but disruption of the crc  expression was found in the mutant embryo (data not shown); the pattern was similar to that of the wild-type embryo when a sense RNA probe was used (fig. 2C-2).

We performed in situ  hybridization to find when and where the highest level of the crc  transcript appears. At the third instar larvae, the outer surface neurons in the brain of the wild type were stained most strongly (fig. 3A-1). In addition, the leg and wing disks and ring gland were stained but very weakly stained in the thoracic and abdominal ganglia (data not shown) as well as the ventral ganglion (fig. 3A-1). No strong crc  mRNA expression was observed when sense probes were used in wild-type brain (fig. 3A-2) and in the brain of eth  ^as311  by using antisense probes (fig. 3A-3). We focused on the crc  expression in the brains of pupae and adults. Unlike most other larval organs, the central nervous system persists into the adult stage in Drosophila . The staining intensity of newly emerged adult type neurons (outside of brain) was stronger than the larval type neurons (inside of brain) in the 19-h-old pupae (fig. 3B-1). The strongest expression was observed at 19 h after puparization in the wild type in comparison to the control straining (fig. 3B-2), and the level of crc  mRNA was markedly low in the eth  ^as311  brain even in 19-h-old pupae (fig. 3B-3). Subsequently, the expression decreased gradually in wild-type brain (fig. 3C-1) and was still clearer than the control of a sense probe (fig. 3C-2). In 48-h-old pupae, more differentiation of optic lobes occurred and the staining of neuronal cell bodies was stronger than the axons in the optic lobes, brains, and the subesophageal ganglia that develop from the larval part of the ventral ganglion and esophagus (fig. 3C-1). In 90-h-old pupae (figs. 3D-1 and -2), the brain was morphologically similar to the adult brain, and the staining was markedly weak (fig. 3D-1). In 1-day-old adult brains, staining was found at the limited small regions (fig. 3E-1), no staining was found in the control brain (fig. 3E-2), and little was found in the mutant brain (fig. 3E-3). During the formation of adult brain, the maximum expression is apparent in the earlier pupal stage, but it decreases gradually to minimal expression in the adult brain. Therefore, these staining patterns are due to expression of crc  mRNA and consistent with the results of Northern blot analysis of crc  (fig. 1D).

We demonstrated here that in the eth  ^as311 , the insertion of a P-element into the crc  gene results in a decrease in mRNA levels. To confirm that the inserted P-element causes the mutant phenotype (hypersensitivity to diethylether anesthesia), we determined EC⁵⁰s in the excised P-element revertants. The revertants of eth  ^as311  were isolated, and their genomic DNA were analyzed by Southern blot hybridization to show complete removal of the P-element from the eth  ^as311  mutants to revert wild type. 3In the revertant-2 , -3 , -5 , and -44 , EC⁵⁰s were of similar values to those of wild-type strains, Canton-S  and w , 1.94% and 1.95% atm in these female flies and 1.91% and 1.98% atm in the male flies, respectively (table 1). These results strongly indicate that the hypersensitivity to diethylether of the mutant phenotype is caused by the inserted P-element in the crc  locus, which induced a lower expression of the gene.

Furthermore, we prepared transgenic lines, crc-T1  and crc-T2 , by introducing wild-type crc  gene under a heat shock promoter, which expressed the gene by heat treatment and examined whether they could be rescued by heat shock from the mutant phenotype to wild type. Transgenic lines crossed to eth  ^as311  to emerge as double mutants, crc-T1;eth  ^as311  and crc-T2;eth  ^as311 . Their EC⁵⁰s for diethylether were compared to the wild-type and mutant strains (table 1). Double-mutant strains without heat shock had significantly lower EC⁵⁰s than those of the wild type, although their ED⁵⁰s were higher than those of eth  ^as311 . When exposed to heat (37°C), their EC⁵⁰s were not significantly different from those of the wild type. Thus, the transgenic flies were rescued from the mutation phenotypes to the wild type. The results also strongly indicate that crc  is the gene that causes the hypersensitivity to diethylether in eth  ^as311 . Three types of heat treatment were performed: heat^a, heat^b, and heat^cwere performed 1 h/day for the whole life and 4 days and 3 days before EC⁵⁰determination, respectively. However, their EC⁵⁰s were not different from one another. A double mutant with a crc  ⁺transgene plus a lethal revertant-s20-1  (crc-T1;s20-1 ) with heat treatment recovered and was still alive at adulthood in the homozygotes. The results indicate that a transgene could be expressed by heat treatment.

Previous genetic studies have identified several mutated genes that alter the sensitivity to various volatile anesthetics. 3–12In some cases, the mutant phenotypes to volatile anesthetics are not parallel, and some are different among themselves. 6,7,9Therefore, we compared the sensitivity to isoflurane and halothane between eth  ^as311  and the wild type. The dose–response curves for isoflurane were plotted to determine the EC⁵⁰s by using loss of avoidance response as an endpoint (fig. 4). The EC⁵⁰s of isoflurane were higher in eth  ^as311  than in the wild type in both sexes. eth  ^as311  was resistant to isoflurane but hypersensitive to diethylether. To confirm that the inserted P-element causes the isoflurane-resistant phenotype, we determined EC⁵⁰s in the revertant and transgenic strains as well as hypersensitivity to diethylether anesthesia. In the revertant-2 , -3 , -5 , and -44 , EC⁵⁰s were of similar values to those of wild-type strains, Canton-S  and w , 0.75% and 0.77% atm in these female flies and 0.80% and 0.79% atm in the male flies, respectively (table 2). crc-T1;eth  ^as311  without heat shock had significantly higher EC⁵⁰s than those of the wild type; their EC⁵⁰s with heat shock did not differ significantly from those of the wild type. Thus, the transgenic flies were rescued from the isoflurane-resistant phenotypes to the wild type. These results also strongly indicate that crc  is the gene that causes resistance to isoflurane anesthesia in eth  ^as311 . On the other hand, the EC⁵⁰s of halothane were not different in eth  ^as311  and wild type (table 2), suggesting that calreticulin is not involved in the halothane anesthesia pathway or halothane does not seem to affect the functional site of the protein.

We had collected Drosophila  mutants with altered sensitivity to diethylether anesthesia, which are transposon tagging strains to facilitate gene cloning. 3In the current study, we demonstrated that the 1.4 kb of crc  mRNA is expressed at all developmental stages and is widely distributed in tissues, including the central nervous system. In the cellular blastoderm, the crc  mRNA was already expressed in the peripheral nuclei in the cells. Increased expression in the ectoderm was observed especially in the central nervous system and then increased in mesoderm tissues with advancement of embryonic stage. crc  gene was expressed in the cell proliferation regions. At late larval and early pupal stages, its expression markedly increased in the brain; the highest expression of the mRNA was detected in newly developed adult neurons at 19 h after puparization and then decreased until adult stage. During the formation of adult brain and optic lobes, the mRNA expression was stronger in cell bodies of neurons than in axons. Considered together, our results suggest that calreticulin is required for all development stages of Drosophila  and especially for the formation of the brain and optic lobes at the late larval and early pupal stages (figs. 3A-1 and B-1).

We also showed that eth  ^as311 , which is the first fertile mutation in crc , has lower transcriptional levels of the crc  gene at all developmental stages and leads to the mutant phenotypes. The crc  mutants, revertant-s20-1  (this study) and l(3)s114307 , 31cause death during embryonic maturation and are probably null mutation. In addition, knockout mice for the calreticulin gene die during embryonic development. 32In this regard, calreticulin is important for the development, organization, and pathfinding of the peripheral nervous system in Drosophila  embryos 31and for formation of the heart, brain, and ventral body wall closure in mice. 32In fibroblasts, the proteins influence cell migration in calcium- and substrate-dependent manners. 33The crc  may play a role in cellular adhesion during cell proliferation, differentiation, movement, and synapse formation in the brain, similar to knockout mice and Aplysia . 32–34 

The eth  ^as311  shows different sensitivities to various volatile anesthetics, i.e. , hypersensitivity to diethylether, resistance to isoflurane, and wild-type sensitivity to halothane (tables 1 and 2). These findings support the possibility that diethylether and isoflurane interact with the different sites of calreticulin to induce anesthesia, but halothane does not interact with any functional site of the protein. In crc-T1;eth  ^as311 , a transgene of crc  controlled by a heat shock promoter is expressed only by heat shock. Transgenic flies that were not exposed to heat shock showed hypersensitivity to diethylether and resistance to isoflurane (tables 1 and 2). The transgenic flies that were exposed to heat shock were rescued to show the wild-type anesthetic responses. crc-T1;eth  ^as311  exposed to heat (37°C) for 1 h a day for 3 days just before anesthesia showed wild-type sensitivities to diethylether and isoflurane. This suggests strongly that the mutant phenotypes are caused by low expression of calreticulin in the adult flies.

At clinical concentrations, halothane and isoflurane cause a leak of calcium into the cytosol, deplete InsP³-sensitive calcium stores, and prevent the increase in cytosolic calcium concentration. 35Halothane reduces endoplasmic reticulum Ca²⁺content in airway smooth muscle cells via  increased Ca²⁺leak through both InsP³and ryanodine receptors, and it alters InsP³receptors directly and indirectly via  modulating InsP³levels. 36These results suggest that halothane targets as InsP³receptor, but it does not interact with calreticulin because lower levels of crc  expression cause diethylether hypersensitivity and isoflurane resistance but normal response to halothane in Drosophila . The reduced amount of calreticulin in the mutant may induce higher cellular Ca²⁺levels in resting cells and require a higher concentration of isoflurane to induce anesthesia, causing isoflurane resistance.

Diethylether might act on calreticulin directly; calreticulin easily loses its function in the mutant, leading to hypersensitivity to diethylether. A possible explanation of the hypersensitivity to diethylether and isoflurane resistance is that the two volatile anesthetics act on different regions of calreticulin to induce opposite effects. We speculated previously that isoflurane interacts with the P- and C-domains of calreticulin, which control intracellular Ca²⁺levels. Another function in the N-domain of calreticulin is as a modulator of both α-integrin adhesion and integrin-initiated Ca²⁺influx signaling in fibroblasts of knockout mice. 33Furthermore, the N-domain does not only decrease tyrosine phosphorylation of β-catenin, resulting in the modulation of cell adhesiveness, 37,38but it also is known as the receptor for thrombospondin, one of the de-adhesive matricellular proteins. 39The low level of calreticulin in the crc  mutant may lead to hypersensitivity to diethylether via  weakness of the adhesion complexes and the signaling pathway.

Last, the highest crc  expression was noted during development in the newly formed adult brain at the pupa stage of 19 h. In the insect, steroid hormone ecdysone induces metamorphosis. The marked reduction of calreticulin in the pupa stage may affect the formation of adult tissues by nuclear export of the glucocorticoid receptor, 22down-regulation of steroid hormone receptor functions, and chaperoning function to glycoproteins. 20,21These effects may be disregarded as the cause of sensitivity to diethylether and isoflurane directly, but because of their indirect effects they may not be ruled out. However, we cannot ignore the interaction between the anesthetics and other proteins, particularly in the case of halothane.