Embryonic Expression of the Divergent Drosophila ß3-Tubulin Isoform Is Required for Larval Behavior

Corresponding author: Elizabeth C. Raff, Department of Biology, Indiana University, Jordan Hall 142, 1001 E. 3rd St., Bloomington, IN 47405., eraff{at}bio.indiana.edu (E-mail)

Communicating editor: R. S. HAWLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have sought to define the developmental and cellular roles played by differential expression of distinct ß-tubulins. Drosophila ß3-tubulin (ß3) is a structurally divergent isoform transiently expressed during midembryogenesis. Severe ß3 mutations cause larval lethality resulting from failed gut function and consequent starvation. However, mutant larvae also display behavioral abnormalities consistent with defective sensory perception. We identified embryonic ß3 expression in several previously undefined sites, including different types of sensory organs. We conclude that abnormalities in foraging behavior and photoresponsiveness exhibited by prelethal mutant larvae reflect defective ß3 function in the embryo during development of chordotonal and other mechanosensory organs and of Bolwig's organ and nerve. We show that microtubule organization in the cap cells of chordotonal organs is altered in mutant larvae. Thus transient zygotic ß3 expression has permanent consequences for the architecture of the cap cell microtubule cytoskeleton in the larval sensilla, even when ß3 is no longer present. Our data provide a link between the microtubule cytoskeleton in embryogenesis and the behavioral phenotype manifested as defective proprioreception at the larval stage.

MICROTUBULES play diverse roles in development. For example, patterning of the Drosophila oocyte and early embryo depends on intricately choreographed reorganizations of the microtubule cytoskeleton (LEHMANN 1995 ; SULLIVAN and THEURKAUF 1995 ; DE CUEVAS and SPRADLING 1998 ; THEURKAUF and HAZELRIGG 1998 ; BRENDZA et al. 2000 ; FOE et al. 2000 ). Microtubules are dynamically assembled from - and ß-tubulin heterodimers. Tubulins are encoded in multiple gene families, each member of which is expressed in a specific tissue and temporal pattern (reviewed in RAFF 1994 ). The architecture and function of microtubule arrays are determined in part by the component tubulins from which they are assembled (HOYLE and RAFF 1990 ; MATTHEWS et al. 1993 ; HOYLE et al. 1995 ; HUTCHENS et al. 1997 ; RAFF et al. 1997 , RAFF et al. 2000 ; MATHE et al. 1998 ).

As one way to address the functional roles of different tubulin isoforms, we have studied Drosophila ß3-tubulin (ß3), a structurally divergent isoform expressed in a variety of tissues, first during midembryogenesis and then again during pupal development (RAFF et al. 1982 ; RUDOLPH et al. 1987 ; GASCH et al. 1988 , GASCH et al. 1989 ; LEISS et al. 1988 ; KIMBLE et al. 1989 , KIMBLE et al. 1990 ; HINZ et al. 1992 ; DETTMAN et al. 1996 ; SERRANO et al. 1997 ; DAMM et al. 1998 ; HOYLE et al. 2000 ). Although the pattern of ß3 expression is complex, its sites of expression have common features. ß3 is expressed primarily in nondividing cells undergoing final stages of differentiation, often in cells undergoing extensive shape changes, and during establishment of cell-cell interactions. ß3 expression is often transient, and in most cases ß3 comprises only part of the total cellular ß-tubulin pool. ß3 mutations result in multiple phenotypes, revealing that, although ß3 is a minor component of total tubulin usage in the fly, it has essential functions at each stage when it is expressed (KIMBLE et al. 1990 ; DETTMAN et al. 1996 ; HOYLE et al. 2000 ). The most severe ß3 mutations (class I) cause death during the first larval instar. Weaker alleles (class II) cause later larval or pupal lethality, or support viability of the fly but with a variety of developmental defects. Since ß3 is not expressed postzygotically until just prior to pupariation, and ß3 protein is no longer present after the end of embryogenesis, larval mutant phenotypes reflect required ß3 function in embryos.

Embryonic ß3 expression commences at stage 10 in the visceral mesoderm and continues in most mesodermally derived cells until early stage 17 (LEISS et al. 1988 ; GASCH et al. 1989 ; KIMBLE et al. 1989 ; HINZ et al. 1992 ; DETTMAN et al. 1996 ; DAMM et al. 1998 ). High levels of ß3 accumulate in all somatic and visceral muscles; by stage 16, the predominant ß3 accumulation is in the elaborate latticework of the body wall musculature (Fig 1A). A puzzle revealed by our genetic analysis is that ß3 is dispensable for myogenesis (DETTMAN et al. 1996 ). Rather, it is in the visceral mesoderm where ß3 is essential for viability. Class I ß3 mutant larvae hatch normally, feed, and can pass food through the gut. However, they die because they cannot absorb nutrients from the food (DETTMAN et al. 1996 ).

View larger version (103K): In this window In a new window Download PPT slide   Figure 1. ß3-tubulin accumulates in Bolwig's organ and other nonmuscle cell types in stage 15 and 16 embryos. Lateral optical focal planes of wild-type embryos stained by anti-ß3 are shown. Embryos are oriented anterior to the left, dorsal to the top. (A and B) Stage 16 embryo. (A) Focal plane close to the surface of the embryo showing the typical ß3-staining pattern in the body wall muscles. (B) Deeper plane of focus showing ß3 staining in the internal cephalic (cm), pharyngeal (phm), and visceral (vm) muscles. The somatic muscles (sm) are in focus on the periphery of the embryo; generalized background staining derives from out-of-focus staining throughout the body wall musculature. In this plane of focus, ß3 staining can also be observed in cells in the anteriormost region of the head (arrow) and in the left-side Bolwig's organ (bo). (Right-hand Bolwig's organ staining is obscured by muscle staining.) (C) Stage 15 embryo. High level ß3 staining is present throughout the musculature. The embryo is oriented so that ß3 staining in the typical flowerette-shaped photoreceptor organ and the associated Bolwig's nerve can be seen on the right-hand side of the embryo (the left-hand organ is also visible but out of focus and obscured by muscle staining). At this stage, Bolwig's organs have not yet moved to their final anterior-lateral position. Staining in the anterior sensory cells is also visible in this view (arrow). In addition, ß3 staining in the dorsal vessel (dv) and unidentified dorsal mesenchymal cells (arrowhead) can also be seen (slightly out of focus in this view). Bar in A, 50 µm.

The lethal failure of class I ß3 mutant larvae to grow can be phenocopied by starving wild-type larvae. However, the behavior of prelethal ß3 mutant larvae differs profoundly from wild type. Most striking is their constant foraging behavior, even in the presence of food. Starving wild-type larvae also forage, but if placed on food, they immediately cease foraging and remain on the food (thus sensibly escaping death by starvation). In contrast, ß3 mutants continue to forage, even though they make feeding movements and take food into the gut. This behavior is so distinctive that ß3 mutant larvae can be easily distinguished from their heterozygous siblings, as they leave the medium and crawl up the side of the vial (as do starving wild-type larvae in the absence of food). ß3 mutant larvae thus appear unable to sense the presence of food or that their guts are filled. This abnormal behavior cannot be explained by the gut dysfunction that causes death. Therefore, in addition to its role in gut differentiation, ß3 must also have other essential functions in the embryo. In support of this hypothesis, we show here that ß3 is expressed in several sensory organs, most notably the cap cells of mechanosensory organs and Bolwig's organ and nerve of the larval photosensory system. The behavioral deficits we observe in prelethal ß3 mutant larvae can be understood in terms of essential ß3 function in these novel sites of embryonic expression, which are required for function of the sensory organs later on during larval life.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks:
Cultures of Drosophila melanogaster and D. virilis were maintained at 25° on standard cornmeal/molasses/agar medium. Visible markers, deficiency chromosomes, and balancer chromosomes are described in LINDSLEY and ZIMM 1992 or FLYBASE 1994 .

Mutations in the ß3 gene (ßTub60D), designated as B3tⁿ, are described in KIMBLE et al. 1990 , DETTMAN et al. 1996 , and HOYLE et al. 2000 . None of the available ß3 mutations is a protein null (DETTMAN et al. 1996 ). However, all of the class I alleles behave genetically as severe hypomorphic alleles. Class I ß3 mutant homozygotes and hemizygotes [carrying a ß3 mutation in combination with Df(2R)Px², a deficiency chromosome that deletes the ß3 gene] display similar phenotypes, slightly more severe in hemizygotes (KIMBLE et al. 1990 ; DETTMAN et al. 1996 ). During this study, we sequenced the coding region of the class I allele B3t²; no lesions were found, and we therefore conclude that this allele is a regulatory mutation, in which sufficient levels of ß3 fail to be expressed in essential cell types. Homozygous or hemizygous ß3 mutant larvae were generated from parental stocks carried in a yellow (y) or yellow, white (yw) background with a second chromosome balancer carrying a copy of the wild-type y gene (CyO-y⁺, described by MARDAHL et al. 1993 ). Mutant larvae could be identified by the y phenotype, compared to their phenotypically wild-type y⁺ heterozygous sibs. (The other y⁺ class, CyO-y⁺ homozygotes, are morphologically abnormal and rarely hatch.)

The ß3-enhancer trap utilized in this work (designated P[ß3-lacZ]) is line A1-2-26 isolated by BIER et al. 1989 , and shown by HARTENSTEIN and JAN 1992 to map to the 60-D locus and to express ß-galactosidase in somatic muscle and the dorsal vessel in a pattern like that of ß3. We confirmed that P[ß3-lacZ] supports ß-galactosidase expression coincident with the known pattern of ß3 expression in stage 8–12 embryos. P[ß3-lacZ] homozygotes are viable and fertile and have no visible phenotypes. We cloned genomic sequences flanking the P element (BALLINGER and BENZER 1989 ) and determined that the P element in this line is inserted in the 5' region of the ß3 gene, very close to the start of transcription.

Immunohistochemistry:
Embryos staged according to CAMPOS-ORTEGA and HARTENSTEIN 1997 were fixed in 4% paraformaldehyde (KIMBLE et al. 1989 ) and stained using the following primary antibodies: (1) anti-ß3, an affinity-purified rabbit polyclonal antiserum specific to Drosophila ß3-tubulin (KIMBLE et al. 1989 ); (2) anti-ß-gal, a monoclonal antiserum specific to ß-galactosidase (Promega, Madison, WI); (3) anti-Mhc, a rabbit polyclonal antiserum specific to myosin heavy chain (YOUNG et al. 1991 ); (4) anti-85E, a rabbit polyclonal antiserum specific to Drosophila 85E-tubulin (MATTHEWS et al. 1990 ; HUTCHENS et al. 1997 ); and (5) mAb-22C10, a monoclonal antiserum specific to Drosophila sensory neurons that recognizes a MAP1B-like protein (FUJITA et al. 1982 ; ZIPURSKY et al. 1984 ; HARTENSTEIN 1988 ; HUMMEL et al. 2000 ). Primary antibody binding was detected using an appropriate horseradish peroxidase-conjugated secondary antiserum (Jackson Immunoresearch Laboratories, West Grove, PA) and a diaminobenzidine tetrahydrochloride color reaction (KIMBLE et al. 1989 ; MATTHEWS et al. 1990 ).

Electron microscopy:
For analysis of chordotonal organ ultrastructure, first instar larvae were fixed in 2% glutaraldehyde, 2% paraformaldehyde, 0.1 M cacodylate, 0.07% sucrose, permeablized to the fix solution by pricking the cuticle. Fixed larvae were stained with 2% osmium tetroxide and 0.5% uranyl acetate, dehydrated with ethanol and acetone, embedded in DER resin (Electron Microscopy Sciences, Fort Washington, PA), and prepared for transmission electron microscopy using standard methods.

Determination of larval photobehavior: the Darth Vader assay:
We tested photobehavior of foraging larvae with an assay similar to that described by SAWIN-MCCORMACK et al. 1995 . Larvae were tested at room temperature (22°–24°) on 5-cm petri dishes containing standard Drosophila medium; half of the dish and lid were covered with electrical tape or aluminum foil, rendering it light tight (the "dark side"). Each plate was illuminated directly overhead by one of two fiber optic cables from a single light source, giving a sharply demarcated light boundary. A total of 20–30 larvae were placed along the midline between the dark and light sides of the plate. Larvae on the dark and light sides of the plate were counted after 45–60 min. Both mutant and control foraging larvae moved far from the midline; most were near the periphery of the plate at the time of counting. Larvae that failed to leave the midline starting position (i.e., were not foraging) were not included in the count. Control larvae sometimes began feeding at the start site and more frequently remained at the start than ß3 mutant larvae, most of which carried on vigorous foraging during the photobehavior tests.

For the photobehavior tests shown in Table 1, mutant larvae were selected from crosses of heterozygous parents, identified by the y marker. Tests were done on larvae of different ages, from shortly after hatching to 48–60 hr posthatching. Control larvae were either the corresponding stock of the same genetic background as the ß3 mutants (y or yw but otherwise wild type), or the heterozygous y⁺ sibs of the mutant larvae. Since class I ß3 mutant larvae fail to grow or molt, all mutant larvae homozygous or hemizygous for B3t² or B3t¹⁰ were at the first larval instar stage and the size of newly hatched larvae, regardless of their age. Homozygotes and hemizygotes for B3t^SK, the class II allele tested, grow slowly (hemizygotes grow more slowly than homozygotes), but undergo larval molts; these larvae were thus at comparable developmental stage as sib controls but smaller in size. In tests in which wild-type larvae were used as controls, first instar control larvae were selected (i.e., for approximately same-sized controls). Starved wild-type larvae (yw), which like class I ß3 mutant larvae fail to grow or molt, also exhibited robust photonegativity when tested 48–60 hr posthatching (included in the control set). In tests in which sibs were used as controls, the sibs were the same age as the mutants but larger in size (e.g., ranged from first instar to late second instar depending on when the test was carried out relative to the time of hatching). Sibs of ß3 mutant hemizygotes used as controls consisted of a mix of parental genotypes [i.e., either the ß3 mutant allele or the Df(2R)Px² deficiency chromosome in combination with the CyO-y⁺ balancer chromosome]. To confirm that hemizygosity for other genes deleted by the Df(2R)Px² deficiency chromosome did not contribute to the loss of photonegativity exhibited by the ß3 mutant hemizygotes, first instar larvae of genotype Df(2R)Px²/CyO-y⁺ were also tested separately.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of nonmuscle sites of embryonic ß3-tubulin expression:
ß3 mutations result in multiple phenotypes (KIMBLE et al. 1990 ; DETTMAN et al. 1996 ; HOYLE et al. 2000 ). One of the most intriguing is the abnormal foraging behavior exhibited by prelethal class I ß3 mutant larvae. In this study we sought to identify the cellular basis for the behavioral phenotype. Aside from the visceral mesoderm, only the macrophage cells that appear on the ventral midline after stage 15 have been previously defined as nonmuscle sites of ß3 accumulation (LEISS et al. 1988 ). However, immunolocalization of ß3 in mutant embryos that fail to develop a mesodermal layer suggested that ß3 is also expressed in some nonmesodermal cells (LEISS et al. 1988 ). We used antibody staining to look for additional sites of embryonic ß3 expression. To distinguish nonmuscle ß3 staining from the high background of muscle-specific staining, we compared ß3 localization with that of myosin heavy chain as a muscle marker in wild-type embryos and with nuclear ß-galactosidase in embryos carrying P[ß3-lacZ] as another locator of ß3 expression (see MATERIALS AND METHODS). We detected ß3 staining in several different nonmuscle cell types in stage 15–17 embryos. As described below, ß3 staining at stage 15–16 in anterior sensory cells and Bolwig's organ, and at stage 17 in the cap cells of chordotonal organs, identified sites of ß3 expression that are essential for larval behavior and sensory perception. In addition, we also identified several other nonmuscle sites of ß3 expression, including the lymph and ring glands in stage 17 embryos (not shown), as well as unidentified dorsal mesenchymal cells at stage 15 (Fig 1C) and cells at the pharyngeal midline at stage 17 (not shown).

These observations expand our understanding of the contexts for ß3 expression. Because of its early accumulation in the visceral mesoderm and subsequent high level accumulation throughout the musculature, ß3 has been valuable as an early mesodermal marker in the embryo. However, a strict mesodermal paradigm for ß3 expression is not supported by our results. For example, the cells of the chordotonal organs and Bolwig's organ and nerve are ectodermally derived (SCHMUCKER et al. 1992 ; CAMPOS et al. 1995 ; OKABE and OKANO 1997 ). In addition, we did not detect ß3 in every embryonic cell type identified as mesoderm derived, such as in dorsal median cells (GORCZYCA et al. 1994 ; LUER et al. 1997 ) or cells of the clypeolabrum.

ß3-tubulin is expressed in Bolwig's organ in stage 15–16 embryos:
Fig 1B and Fig C, shows ß3 staining in Bolwig's organs, the paired larval photosensory organs that form at the anterior of later stage embryos. We think it likely that Bolwig's organ is also the identity of ß3-staining neuronal bundles observed by LEISS et al. 1988 in head segments of dorsalized embryos lacking a mesoderm. As described below, some of the behavioral abnormalities associated with ß3 mutant larvae can be explained by ß3 function essential for larval photoreception.

ß3 is also present in cells at the anteriormost tip of the embryo (Fig 1B and Fig C). On the basis of their position, we have identified these anterior cells as part of the antenna-maxillary complex, a larval sensory structure that includes two monoscolopidial chordotonal organs (CAMPOS-ORTEGA and HARTENSTEIN 1997 ). We also observed comparable ß3 accumulation in anterior cells in the head of D. virilis embryos (not shown). We previously showed that the ß3 muscle expression pattern is conserved in other Drosophila species (DETTMAN et al. 1996 ). These data show that nonmuscle, and, indeed, nonmesodermal, sites of ß3 expression are also conserved.

ß3-tubulin is expressed in cap cells of mechanosensory organs in stage 17 embryos:
Nonmuscle ß3 expression in stage 15–16 embryos demonstrated a potential role for ß3 in differentiation of sensory organs. Further underscoring this conclusion is the finding by HINZ et al. 1992 that sequences within the large first intron of the ß3 gene were capable of driving expression of a LacZ reporter in cap cells of chordotonal organs in late stage 16 embryos. We examined ß3 staining in late stage embryos and found that ß3 accumulates in the cap cells of chordotonal and other mechanosensory organs during stage 17, the final stage of embryogenesis (Fig 2). The Drosophila embryo forms several types of mechanosensory organs, including external campaniform sensilla and the internal chordotonal organs. These organs serve as proprioreceptors through which the larva can sense strain or deformation of the cuticle or internal body structures and movements or stretching of the viscera and muscles (HARTENSTEIN 1988 ; JAN and JAN 1993 ). Chordotonal sensilla consist of a bipolar neuron surrounded by three nonneural support cells, the scolopale, ligament, and cap or attachment cells (MOULINS 1976 ; HARTENSTEIN 1993 ; CAMPOS-ORTEGA and HARTENSTEIN 1997 ; CARLSON et al. 1997 ; see Fig 4A, below). Different classes of chordotonal organs contain one or more four-celled sensilla.

View larger version (112K): In this window In a new window Download PPT slide   Figure 2. ß3 is present in the cap cells of chordotonal organs in late stage embryos. Optical focal planes near the surface of stage 17 embryos stained with anti-ß3 are shown. Embryos are oriented anterior to the left, dorsal to the top. (A) Ventrolateral view of an embryo showing the overall ß3-staining pattern in late stage embryos. General ß3 staining in the musculature is still present, but greatly diminished in intensity; by the time of hatching, ß3 will no longer be present in the larva. At this stage, the highest level ß3 staining is in the mechanosensory organs. In this view of the embryo, ß3 staining can be seen in the lateral pentascolopidial organs in segments A1–A7 (large arrowheads) and in triscolopidial organs in segments T2, T3, and A8 (small arrowheads). There is also staining in the ventral campaniform sensilla, largely out of focus in this view (best seen in segment A4; indicated by arrow). The dark structures in the anterior are the larval mouth hooks (mh), slightly out of focus in this surface view. Their color is the natural pigmentation and does not represent ß3 staining. The darkly pigmented ventral denticle belts (db) are also visible. (B) ß3 staining in abdominal monoscolopidial (ch 1) and pentascolopidial (ch 5) chordotonal organs. Staining is present only in the cap cells (cap). Arrowheads indicate sites of insertion of the scolopale cells into the ß3-staining cap cells. (C) ß3 staining in the cap cells of an anterior ventral campaniform sensilla. Arrowhead indicates site of insertion of the scolopale cells (sc) into the ß3-staining cap cells. (D and E) Lateral views of two dorsal thoracic triscolopidial chordotonal organs in a single embryo. Focal planes show insertion of the scolopale cells into the ß3-staining cap cells of the left-hand (anterior) organ in D and of the right-hand (posterior) organ in E. Bar in A, 50 µm; bar in B for B–E, 30 µm.

View larger version (137K): In this window In a new window Download PPT slide   Figure 3. Chordotonal organs and campaniform sensilla develop in embryos deficient for ß3 synthesis. Stage 17 embryos (before development of mouth hooks) were collected from heterozygous Df(2R)Px2 parents and stained with anti-ß3. Homozygous Df(2R)Px2 embryos, identified as immunonegative for anti-ß3, were selected and stained with either mAb22C10 or anti-85E-tubulin. Embryos are oriented anterior to the left, dorsal to the top. (A) Lateral view of a homozygous Df(2R)Px2 embryo stained with mAb22C10, showing neurons of the central and peripheral nervous systems. The ventral nerve chord (vnc) and nerve bundles of the maxillary cirri (mx) are indicated. Large arrowheads point to neurons of the lateral monoscolopidial and pentascolopidial chordotonal organs; small arrowheads indicate neurons of the dorsal campaniform sensilla. (B) Lateral view of a homozygous Df(2R)Px2 embryo stained with anti-85E, showing the dorsal triscolopidial organs (dch3), the lateral pentascolopidial organs (lch5 and large arrowheads), and monoscolopidial organs (small arrowheads). 85E is present in its normal domain of expression in both the ligament (lig) and cap cells (cap). Bar in A, 50 µm.

View larger version (100K): In this window In a new window Download PPT slide   Figure 4. Microtubule organization in cells of chordotonal organs. (A) Schematic drawing of an insect surface sensillum indicating the organization of the support cells around the neuron. Abbreviations: Cap, cap cell; cd, ciliary dilation; Sco, scolopale cell; Lig, ligament cell; R, rootlet. Microtubules are indicated as vertical lines. (B–E) Electron micrographs of chordotonal organs in hatched first instar larvae. (B) Longitudinal section through a monoscolopidial chordotonal organ in a wild-type larva, showing the cell types diagrammed in A. Arrows indicate the relative positions of cross sections of chordotonal organs shown in C–E. (C) Cross section near the distal tip of the cilium in a monoscolopidial chordotonal organ in a wild-type embryo, showing the cilium (ci) and ciliary dilation within the neuron, the scolopale cell, and the cap cell. The cap cell contains many microtubules arrayed parallel to the longitudinal axis of the organelle. The boxed region of the cap cell cytoplasm is shown at higher magnification in Fig 5B. (D) Cross section near the distal tip of the cilium in a monoscolopidial chordotonal organ in a larva of genotype B3t2/Df (2R)Px2. This micrograph is at a similar position to that in C, but at a slightly more oblique angle. The overall morphology of the neuron, scolopale cell, and cap cell is wild type. In this view the nine doublet microtubules of the cilium can be seen. As in wild type, the cytoplasm of the cap cell has many microtubules organized parallel to the longitudinal axis of the organelle. The boxed region of the cap cell cytoplasm is shown at higher magnification in Fig 5C. (E) Cross section near the base of a monoscolopidial chordotonal organ in a wild-type larva, showing the ligament cell and the rootlet of the neuron. Both the ligament cell and the neuron contain densely packed, highly ordered arrays of parallel microtubules. Bars in B–E, 500 nm.

We observed ß3 accumulation in cap cells of many types of mechanosensory organs. Fig 2 shows examples of ß3 staining in cap cells of the lateral abdominal pentascolopidial chordotonal organs (Fig 2A and Fig B), lateral abdominal monoscolopidial chordotonal organs (Fig 2B), dorsal triscolopidial chordotonal organs (Fig 2A, Fig D, and Fig E), and ventral campaniform sensilla (Fig 2C). Morphogenesis of the mechanosensory organs of the larva begins prior to stage 13, after which morphologically mature organs first appear (MATTHEWS et al. 1990 ; CARLSON et al. 1997 ). Elongation of cap cells begins during stage 14 and continues throughout larval development. We did not detect ß3 staining in cap cells until stage 17. However, because ß3 staining in muscles is so strong, we thought it possible that accumulation of ß3 in chordotonal organs might be obscured in earlier stage embryos. Double antibody staining with anti-ß3 and mAb-22C10, a marker for sensory neurons (ZIPURSKY et al. 1984 ; HARTENSTEIN 1988 ; HARTENSTEIN and POSAKONY 1989 ; HUMMEL et al. 2000 ), confirmed that ß3 was not present in cap cells before stage 17 (not shown). Thus the ß3 isoform is present only during final stages of differentiation of the sensilla.

The predominant embryonic tubulins are ß1-tubulin and 84B-tubulin, which are both maternally and zygotically expressed. ß3 is only transiently expressed from the zygotic genome; the ß3 transcript cannot be detected in larvae until late third instar larvae nearing pupariation (ANDRES et al. 1993 ). ß3 protein disappears from its prominent sites of expression by the end of embryogenesis. Another minor isoform, 85E-tubulin (85E), is also zygotically expressed in a temporal and tissue-specific pattern similar, although not identical, to that of ß3 (MATTHEWS et al. 1990 ). In embryos, 85E accumulates in the somatic and visceral musculature and in the support cells of mechanosensory organs, beginning in stage 13 and persisting throughout larval development. To ascertain whether the ß3 protein similarly persists in mechanosensory organs after the completion of embryogenesis, we stained filleted third instar larval pelts to determine if ß3 is present in cap cells in larvae. Specific staining in imaginal discs with anti-ß3 served as our positive control (KIMBLE et al. 1989 ). We did not observe any ß3 staining in cap cells in larvae (not shown) and therefore conclude that ß3 synthesis in cap cells, as in the other embryonic tissues in which it is expressed, is restricted to the embryo and is transient.

To understand the potential function of ß3 in cap cells we wanted to determine whether ß3 is the sole ß-tubulin in cap cells during stage 17. We utilized 85E as a marker for a stable tubulin pool in chordotonal support cells. -tubulins are unstable in cells that do not synthesize any ß-tubulins (KEMPHUES et al. 1982 ; HOYLE et al. 1995 ). We therefore reasoned that if ß3 is the only isoform in cap cells, then 85E would not accumulate in cap cells of embryos deficient for ß3 [homozygous for Df(2R)Px²]. Although Df(2R)Px² homozygotes do not complete development to the larval stage, both the central and peripheral nervous systems form in these animals, including the mechanosensory organs (Fig 3A). Fig 3B shows that 85E accumulated normally in support cells of Df(2R)Px² embryos, demonstrating that cap cells possess a stable tubulin pool even in the absence of ß3. These data show that ß3 is not required for the morphogenesis of mechanosensory organs and that ß3 comprises only part of the total ß-tubulin pool in cap cells. The additional ß-tubulin in the cap cells must be ß1, since chordotonal organ morphogenesis begins during stage 13, at a time when only ß1 and ß3 are being expressed in the embryo. This is consistent with other sites of ß3 expression, in which ß3 is present together with ß1 (KIMBLE et al. 1989 ; DETTMAN et al. 1996 ; HOYLE et al. 2000 ).

Organization of the microtubule cytoskeleton is altered in the cap cells of chordotonal organs in ß3 mutant larvae:
To gain clues to the possible function of ß3 in the cap cells, we examined ultrastructure and microtubule organization in cells of chordotonal organs in wild-type larvae and larvae homozygous or hemizygous for B3t², the most severe class I ß3 mutation. Fig 4A shows a schematic drawing of an insect mechanosensory organ. Fig 4B, Fig C, and Fig E, shows the corresponding ultrastructure of chordotonal organs in wild-type Drosophila first instar larvae. Mature mechanosensors extend an axon from their proximal end, adjacent to the ligament cell, and a sensory dendrite toward their distal end to contact the cap cell. Within the dendrite there is a nonmotile sensory cilium. The dendrite and cell body are ensheathed within a single scolopale cell. Electron-dense scolopale rods are postulated to provide rigidity to the membrane of the scolopale cell facing the dendrite. The space between the scolopale and dendritic cell membranes is filled with lymph, which appears lucent in micrographs.

We observed many microtubules in the neuron, cap, and ligament cells, but few microtubules in scolopale cells (Fig 4, C–E). In cap cells, the site of ß3 expression, microtubules fill the cytoplasm (Fig 4C and Fig D), arrayed in parallel register with the axis of the dendrite, but not in any specific organization. In contrast, the arrays of microtubules in the neuron and in the ligament cells are much more highly ordered. In ligament cells, closely packed parallel microtubules surround the cell membrane of the neuron in the vicinity of the rootlet (Fig 4E). Within the neuron, parallel microtubules form a ring surrounding the rootlet (Fig 4E), with less organized clusters of microtubules at the periphery of the cell. In the ligament and neuron, a striking feature of the microtubule arrays is that many microtubules are crosslinked via electron-dense bridges, either side by side, or in clusters of three to four microtubules. In the cytoplasm of cap cells, microtubule crosslinking was less frequent.

The overall morphology of chordotonal organs was normal in ß3 mutant larvae (Fig 4D), consistent with the timing of ß3 expression late in differentiation. The general features of microtubule organization in cap cells were similar in wild-type and ß3 mutant larvae. However, as shown in Fig 5, we found that the occurrence of crosslinked microtubules was greater in cap cells from B3t²/B3t² homozygotes and B3t²/Df(2R)Px² hemizygotes than in wild type. On average, 16% of the microtubules in cap cells were crosslinked in wildtype larvae, 24% in B3t²/B3t² larvae, and 42% in B3t²/Df(2R)Px² larvae. Thus microtubule crosslinking was greater in B3t²/B3t² larvae than in wild type, and more than twice as frequent as wild type in B3t²/Df(2R)Px² larvae, which is significantly different from both B3t²/B3t² and wild type. The ß3 gene dose dependence of microtubule crosslinking is concordant with our conclusion that B3t² is a partial loss-of-function allele with a more severe phenotype as a hemizygote than as a homozygote (DETTMAN et al. 1996 ; see MATERIALS AND METHODS).

View larger version (90K): In this window In a new window Download PPT slide   Figure 5. Microtubule crosslinking in cap cells of chordotonal organs is more frequent in ß3 mutants than in wild type. (A) Microtubules were scored as not touching or touching (i.e., crosslinked by electron-dense material joining adjacent microtubules) in electron micrographs of cap cells of chordotonal organs in wild-type, B3t2/B3t2, and B3t2/Df(2R)Px2 larvae. Only microtubules that were in sufficiently good cross section as to be unambiguously scored were counted. The percentage of crosslinked, or touching, microtubules in cap cells was calculated for each cell and averaged for each genotype. Bars show the mean percentage of crosslinked microtubules in cap cells of each genotype, with standard deviation calculated with StatView. One-way analysis of variance (ANOVA) was done to compare the variance of the three populations (percentage of linked microtubules); P value was 0.0032. (B) Crosslinked microtubules (arrowheads) in the cap cell of a monoscolopidial chordotonal organ in a wild-type first instar larva; view shown is the boxed area of the cap cell shown in Fig 4C. (C–E) Examples of crosslinked microtubules in cap cells of chordotonal organs in first instar larvae of genotype B3t2/Df(2R)Px2. The view shown in C is the boxed area of the cap cell shown in Fig 4D. Note that many microtubules in cap cells of chordotonal organs in both wild-type and ß3 mutant larvae are associated with electron-dense cap cell matrix in the cytoplasm (* in D and E). Bar, 125 nm for B–D.

We conclude that in wild-type embryos, the transient presence of ß3 in the tubulin pool during late stages in differentiation of the sensory organs decreases the capacity of microtubules to form crosslinks with other microtubules. Transient ß3 expression thus confers permanent features of microtubule organization that persist in the fully differentiated cell, even after ß3 is no longer present. The sensilla in which ß3 is expressed function as stretch receptors and mechanosensors that allow the larva to sense the state of the cuticle and the viscera. The continuous foraging behavior of ß3 mutant larvae indicates that they are incapable of sensing when they are in the presence of food, or when the gut is filled. Our data support the hypothesis that the misorganization of microtubules in the cap cells of the chordotonal organs causes a functional defect in these sensilla, contributing to the sensory deficits exhibited by ß3 mutant larvae.

Severe ß3 mutants exhibit defective photosensitivity:
In pupae, transient expression of the ß3 isoform in the photoreceptor neurons of the compound eye and neurons within the optic lobe is required for neuronal patterning and connectivity in the developing adult visual system (HOYLE et al. 2000 ). Our observation of zygotic ß3 expression in Bolwig's organ and nerve suggested that ß3 might also be essential for development of the larval photoreceptors.

We examined the morphology of Bolwig's organ and Bolwig's nerve at the light microscope level in newly hatched wild-type and ß3 mutant larvae utilizing transgenic lines in which green fluorescent protein (GFP) is expressed under control of Kruppel gene regulatory elements (CASSO et al. 1999 ). Kruppel is expressed throughout development of Bolwig's organ and nerve (SCHMUCKER et al. 1992 ); the transgenic animals accumulate GFP in Bolwig's organ and nerve, allowing detection by fluorescence confocal microscopy. Consistent with the late timing of the onset of ß3 expression in the developing larval photoreceptor system, the overall morphology of Bolwig's organ and targeting of Bolwig's nerve to the brain were normal in larvae that were homozygous or heterozygous for the class I ß3 mutant alleles B3t² and B3t¹⁰ and the class II allele, B3t^SK (not shown). Thus, similarly to the situation with the chordotonal organ cap cells, we conclude that ß3 is not required for morphogenesis of the larval photosensory organs.

We next examined whether ß3 is required for function of the larval photosensory system. SAWIN-MCCORMACK et al. 1995 showed that from the second half of the first instar until late in the third instar, foraging wild-type Drosophila larvae exhibit a strong tendency toward photonegativity, such that 75% avoid light. However, glass mutant larvae, which lack larval photoreceptors, were photoneutral. The larval photoresponse is thus dependent on the larval photoreceptors. As shown in Table 1, tests of the photobehavior of ß3 mutant larvae revealed that a feature of the abnormal foraging behavior of ß3 mutant larvae is that they do not avoid light as do wild-type larvae. We found that foraging wild-type first instar larvae displayed variable photobehavior when newly hatched, but, as observed by SAWIN-MCCORMACK et al. 1995 , became photonegative by 12 hr after hatching. Homozygous ß3 mutant larvae were also photonegative, but hemizygous ß3 mutant larvae were photoneutral. Control experiments showed that normal photonegativity was exhibited both by starving wild-type larvae and by larvae heterozygous for the Df(2R)Px2 chromosome in combination with a second chromosome wild type at the ß3 locus. Thus the loss of normal larval photonegativity in the ß3 mutant hemizygotes is not attributable either to the inability of the mutant larvae to grow and molt or to hemizygosity for any of the other genes deleted by Df(2R)Px². We therefore conclude that the defective photoresponse displayed by hemizygous mutant animals results from defective function of the larval photoreceptors caused by insufficient ß3 function during differentiation of Bolwig's organ and Bolwig's nerve.

The ß3 dose-dependent response resembles the ß3 dose-dependent microtubule phenotype in chordotonal organs (Fig 5). Similarly, decreasing the ß3 dose affects some but not all ß3-dependent functions in the developing adult visual system (HOYLE et al. 2000 ). Together, our observations suggest that different cell types may have differing requirements either for the absolute amount of ß3 or for a given ratio of ß3 to ß1, the predominant Drosophila ß-tubulin isoform present with ß3. The three light-sensing systems in the fly—Bolwig's organ and Bolwig's nerve in larvae and the compound eyes and ocelli of the adult—are based on different developmental mechanisms, but all utilize ß3 during differentiation (this work; HOYLE et al. 2000 ). An intriguing question is whether a ß3-containing microtubule cytoskeleton provides a common function for differentiation of distinct photosensory cells.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

As TULLY 1996 has pointed out, mutations that affect behavior exert their effects through a hierarchy of processes from genes to molecules to cells to neural circuits to behavior. Previous genetic analysis has identified many components involved in Drosophila larval behavior, including numerous signal transduction systems (see discussions in VARNAM et al. 1996 ; OSBORNE et al. 1997 ; WANG et al. 1997 ; IYENGAR et al. 1999 ; SAROV-BLAT et al. 2000 ; YANG et al. 2000 ). Our results reveal a link between cytoskeletal organization and behavior.

Identification of a permanent alteration in microtubule organization in chordotonal organ cap cells provides the cellular basis for one aspect of the ß3 mutant behavioral phenotype. In wild-type larvae, we observed extensive crosslinking in the neuron and ligament cells of chordotonal organs, but much less microtubule crosslinking in the ß3-expressing cap cells. Crosslinking is increased in ß3 mutants; we therefore conclude that in wild-type animals incorporation of ß3 into cap cell microtubules acts to depress crosslinking.

Microtubule crosslinking in the neuron and ligament cells most likely contributes to rigidity necessary for the proper function of the mechanosensory organ. Other electron-dense structures in the neuron, scolopale, and ligament cells are also hypothesized to contribute to rigidity of the sensilla (MOULINS 1976 ). A precedent for microtubule-mediated rigidity in mechanosensory organs can be found in Caenorhabditis elegans mec-7 ß-tubulin. In touch neurons, mec-7 is required for assembly of crosslinked microtubule arrays, deformation of which mediates signal transduction required for touch perception (SAVAGE et al. 1989 ; TAVERNARAKIS and DRISCOLL 1997 ). The role of ß3 in cap cells of Drosophila chordotonal organs, however, appears to be opposite to that of mec-7 in C. elegans, in that microtubule crosslinking is reduced in cells in which ß3 is expressed. Thus the function of ß3 in the cap cells may be to render microtubule arrays in cap cells less rigid.

What are the possible functions of decreased rigidity in cap cells relative to other cells of the chordotonal organs? One possibility is that ß3 is synthesized in cap cells during a developmental stage when crosslinking of microtubules would interfere with the maturation of the organ. During the final stages of morphogenesis, the cap cell must elongate and attach to either the cuticle or viscera. Rigidity conferred by microtubule crosslinking might inhibit attachment or elongation of the organelle; for example, microtubule crosslinking might prevent adjacent microtubules from sliding past one another in the elongating cap cell. The second possibility, suggested by the fact that the permanent microtubule cytoskeleton in cap cells is less crosslinked than in the neuron and ligament cells, is that a greater degree of flexibility of the cap cells is required for function of the mature organ. Like the cells that attach the organ to other body structures, perhaps a more flexible cap cell imparts a greater sensitivity to subtle deformations, serving to amplify stretch signals transduced through the more rigid cells that comprise the basal structure of the organelle.

How does the transient presence of ß3 modulate microtubule crosslinking in the cap cells? The finding that microtubule organization is altered in chordotonal organs argues that ß3 provides a specialized function in these cells. Also supporting the "specialized function" hypothesis is the observation that we cannot rescue ß3 mutant phenotypes by increasing generalized expression of the predominant isoform, ß1-tubulin (E. RAFF, unpublished data). What specialized properties might ß3 confer on the microtubules into which it is incorporated? There are at least two mechanisms, not mutually exclusive, by which ß3 might alter the properties of the microtubules into which it is incorporated. First, many microtubule-associated proteins bind to microtubules via the carboxy terminus; thus one possibility is that the unique ß3-specific C-terminal domain is unable to associate with other proteins necessary to form crosslinks (i.e., proteins that would normally bind to the ß1-C terminus). Thus a mixed polymer assembled from a pool containing both ß3 and ß1 would generate less extensive crosslinking. This possibility is consistent with our demonstration of ß-tubulin C-terminal-specific functions in the male germ line (HOYLE and RAFF 1990 ; FACKENTHAL et al. 1993 ; HOYLE et al. 1995 ; RAFF et al. 2000 ). Another possibility is that ß3-containing microtubules may be more unstable than microtubules in the neuron and ligament cells. An overall increase in the dynamics of microtubules might well generate the less well-ordered microtubule cytoskeleton that we observe in cap cells, compared to the highly ordered microtubule arrays in the neuron and ligament cells. A unique structural feature of the ß3 molecule is an extra six-amino-acid "loop" in the internal variable domain (RUDOLPH et al. 1987 ). The recently solved structure of the tubulin heterodimer unexpectedly revealed that this region of the molecule faces the lumen of the microtubule (NOGALES et al. 1998 , NOGALES et al. 1999 ). Perhaps this or other structural features of ß3 alter the relative stability of the dimer within the microtubule protofilament, shifting the dynamics of the microtubules toward more frequent catastrophe.

We conclude that the structural defect in microtubule organization in the cap cells reduces the capacity of the mechanosensory organs in ß3 mutants to transduce information about the internal or external environment. Chordotonal organs act as stretch receptors, one function being to sense a filled gut. The unrelenting foraging activity of ß3 mutant larvae is consistent with a loss of perception of the state of the viscera. According to this hypothesis, class I ß3 mutant larvae feed but do not feel full, thus continuing to seek food even in its presence. Other ß3-expressing cell types may also be involved in the sensory failure that causes continuous foraging, since ß3 is expressed in other anterior sensory structures. Although we have not as yet defined a requirement for ß3 in these other sensilla, it is possible that inability to sense the presence of food may reflect loss of chemosensory or olfactory function in the mutant animals.

In summary, our data show that the structurally unique ß3 isoform is deployed in numerous cell types during the final steps in differentiation and that, at least in some of its sites of expression, the ß3 isoform confers specialized properties to the microtubules into which it is incorporated. Our data reveal how transient modulations of the microtubule cytoskeleton during development may determine the eventual functional capacity of a given tissue. In their study of development of chordotonal organs, CARLSON et al. 1997 remarked that "Overall, the chordotonal organs may be made in `haste' (about 3 h), but they are perfected at leisure during the remainder of embryonic life (about 8 h)" (p. 387). We conclude that the expression of ß3, and the resulting long-term change in microtubule cytoskeleton organization, is one of the "perfecting" processes by which not only are the mature functional mechanosensory organs generated, but many other cell types as well.

	FOOTNOTES

This manuscript is dedicated to the memory of our dear friend and colleague, Jeffrey A. Hutchens.
¹ Present address: Department of Pediatrics, Northwestern University, Ward 12-191, 303 E. Chicago Ave., Chicago, IL 60611. E-mail: r-dettman{at}northwestern.edu

	ACKNOWLEDGMENTS

We thank Bruce Diaz and Linda Brunick for their contributions to analysis of ß3-tubulin function; James Sliger for his contributions to analysis of Bolwig's organ development during an undergraduate research project; Huy Nguyen for sequence analysis of the site of insertion of the P element in the ß3 enhancer trap line, P[ß3-lacZ]; Mark Neilsen for his help in statistical analysis; and William Saxton and Rudolf Raff for critical reading of the manuscript. We thank D. Keihart for anti-Mhc antiserum and K. Matthews for anti-85E antiserum. We obtained the monoclonal antisera Mab22C10 from the Developmental Studies Hybridoma Bank developed under the auspices of the National Institute of Child Health and Human Development and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, Iowa, 52242. Our study was supported by a research grant to E.C.R. from the U.S. Public Health Service. R.W.D was supported in part as a predoctoral trainee under a Department of Health and Human Services National Research Service Training Grant awarded to the Indiana University Department of Biology.

Manuscript received August 24, 2000; Accepted for publication January 26, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ANDRES, A. J., J. C. FLETCHER, F. D. KARIM, and C. S. THUMMEL, 1993  Molecular analysis of the initiation of insect metamorphosis: a comparative study of Drosophila ecdysteroid-regulated transcription. Dev. Biol. 160:388-404[Medline].

BALLINGER, D. G. and S. BENZER, 1989  Targeted gene mutations in Drosophila. Proc. Natl. Acad. Sci. USA 86:9402-9406[Abstract/Free Full Text].

BIER, E., H. VASSIN, S. SHEPERD, K. LEE, and K. MCCALL et al., 1989  Searching for pattern and mutation in the Drosophila genome with a P-lacZ vector. Genes Dev. 3:1273-1287[Abstract/Free Full Text].

BRENDZA, R. P., L. R. SERBUS, J. B. DUFFY, and W. M. SAXTON, 2000  Kinesin I in oocyte polarity: posterior localization of oskar mRNA and Staufen protein. Science 289:2120-2122[Abstract/Free Full Text].

CAMPOS, A. R., K. J. LEE, and H. STELLER, 1995  Establishment of neuronal connectivity during development of the Drosophila larval visual system. J. Neurobiol. 28:313-329[Medline].

CAMPOS-ORTEGA, J. A., and V. HARTENSTEIN, 1997 The Embryonic Development of Drosophila melanogaster, Ed. 2. Springer, New York/Berlin.

CARLSON, S. D., S. L. HILGERS, and J. L. JUANG, 1997  Ultrastructure and blood-nerve barrier of chordotonal organs in the Drosophila embryo. J. Neurocytol. 26:377-388[Medline].

CASSO, D., F. RAMIREZ-WEBER, and T. B. KORNBERG, 1999  GFP-tagged balancer chromosomes for Drosophila melanogaster. Mech. Dev. 88:229-232[Medline].

DAMM, C., A. WOLK, D. BUTTGEREIT, K. LÖHER, and E. WAGNER et al., 1998  Independent regulatory elements in the upstream region of the Drosophila ß3 gene (ßTub60D) guide expression in the dorsal vessel and the somatic muscles. Dev. Biol. 199:138-149[Medline].

DE CUEVAS, M. and A. C. SPRADLING, 1998  Morphogenesis of the Drosophila fusome and its implications for oocyte specification. Development 125:2781-2789[Abstract].

DETTMAN, R. W., F. R. TURNER, and E. C. RAFF, 1996  Genetic analysis of the Drosophila ß3-tubulin gene demonstrates that the microtubule cytoskeleton in the cells of the visceral mesoderm is required for morphogenesis of the midgut endoderm. Dev. Biol. 177:117-135[Medline].

FACKENTHAL, J. D., F. R. TURNER, and E. C. RAFF, 1993  Tissue-specific microtubule functions in Drosophila spermatogenesis require the ß2-tubulin isotype-specific carboxy terminus. Dev. Biol. 158:213-227[Medline].

FLYBASE,, 1994  The Drosophila genetic database. Nucleic Acids Res. 22:3456-3458. [available from flybase.bio.indiana.edu][Abstract/Free Full Text].

FOE, V. E., C. M. FIELD, and G. M. ODELL, 2000  Microtubules and mitotic cycle phase modulate spatiotemporal distributions of F-actin and myosin II in Drosophila syncytial blastoderm embryos. Development 127:1767-1787[Abstract].

FUJITA, S. C., S. L. ZIPURSKY, S. BENZER, A. FERRUS, and S. L. SHOTWELL, 1982  Monoclonal antibodies against the Drosophila nervous system. Proc. Natl. Acad. Sci. USA 79:7929-7933[Abstract/Free Full Text].

GASCH, A., U. HINZ, D. LEISS, and R. RENKAWITZ-POHL, 1988  The expression of beta 1 and beta 3 tubulin genes of Drosophila melanogaster is spatially regulated during embryogenesis. Mol. Gen. Genet. 211:8-16[Medline].

GASCH, A., U. HINZ, and R. RENKAWITZ-POHL, 1989  Intron and upstream sequences regulate expression of the Drosophila beta 3-tubulin gene in the visceral and somatic musculature, respectively. Proc. Natl. Acad. Sci. USA 86:3215-3218[Abstract/Free Full Text].

GOLDSTEIN, A., 1964 Biostatistics. Macmillan, New York.

GORCZYCA, M. G., R. W. PHILLIS, and V. BUDNIK, 1994  The role of tinman, a mesodermal cell fate gene, in axon pathfinding during the development of the transverse nerve in Drosophila. Development 120:2143-2152[Abstract].

HARTENSTEIN, V., 1988  Development of Drosophila larval sensory organs: spatiotemporal pattern of sensory neurones, peripheral axonal pathways and sensilla differentiation. Development 102:869-886[Free Full Text].

HARTENSTEIN, V., 1993 Atlas of Drosophila Development. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

HARTENSTEIN, V. and Y. N. JAN, 1992  Studying Drosophila embryogenesis with P-lacZ enhancer trap lines. Roux's Arch. Dev. Biol. 201:194-220.

HARTENSTEIN, V. and J. W. POSAKONY, 1989  Development of adult sensilla on the wing and notum of Drosophila melanogaster. Development 107:389-405[Abstract].

HINZ, U., A. WOLK, and R. RENKAWITZ-POHL, 1992  Ultrabithorax is a regulator of ß3 tubulin expression in the Drosophila visceral mesoderm. Development 116:543-554[Abstract].

HOYLE, H. D. and E. C. RAFF, 1990  Two Drosophila beta tubulin isoforms are not functionally equivalent. J. Cell Biol. 111:1009-1026[Abstract/Free Full Text].

HOYLE, H. D., J. A. HUTCHENS, F. R. TURNER, and E. C. RAFF, 1995  Regulation of beta-tubulin function and expression in Drosophila spermatogenesis. Dev. Genet. 16:148-170[Medline].

HOYLE, H. D., F. R. TURNER, and E. C. RAFF, 2000  A transient specialization of the microtubule cytoskeleton is required for differentiation of the Drosophila visual system. Dev. Biol. 221:375-389[Medline].

HUMMEL, T., K. KRUKKERT, J. ROOS, G. DAVIS, and C. KLAMBT, 2000  Drosophila Futsch/22C10 is a MAP1B-like protein required for dendritic and axonal development. Neuron 26:357-370[Medline].

HUTCHENS, J. A., H. D. HOYLE, F. R. TURNER, and E. C. RAFF, 1997  Structurally similar Drosophila -tubulins are functionally distinct in vivo. Mol. Biol. Cell 8:481-500[Abstract].

IYENGAR, B., J. ROOTE, and A. R. CAMPOS, 1999  The tamas gene, identified as a mutation that disrupts larval behavior in Drosophila melanogaster, codes for the mitochondrial DNA polymerase catalytic subunit (DNApol-gamma125). Genetics 153:1809-1824[Abstract/Free Full Text].

JAN, Y. N., and L. Y. JAN, 1993 The peripheral nervous system, pp. 1207–1244 in The Development of Drosophila melanogaster, edited by M. BATE and A. M. ARIAS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

KEMPHUES, K. J., T. C. KAUFMAN, R. A. RAFF, and E. C. RAFF, 1982  The testis-specific ß-tubulin subunit in Drosophila melanogaster has multiple functions in spermatogenesis. Cell 31:655-670[Medline].

KIMBLE, M., J. P. INCARDONA, and E. C. RAFF, 1989  A variant ß-tubulin isoform of Drosophila melanogaster (ß3) is expressed primarily in tissues of mesodermal origin in embryos and pupae, and is utilized in populations of transient microtubules. Dev. Biol. 131:415-429[Medline].

KIMBLE, M., R. W. DETTMAN, and E. C. RAFF, 1990  The ß3-tubulin gene of Drosophila melanogaster is essential for viability and fertility. Genetics 126:991-1005[Abstract].

LEHMANN, R., 1995  Cell-cell signaling, microtubules, and the loss of symmetry in the Drosophila oocyte. Cell 83:353-356[Medline].

LEISS, D., U. HINZ, A. GASCH, R. MERTZ, and R. RENKAWITZ-POHL, 1988  ß3 tubulin expression characterizes the differential mesodermal germ layer during Drosophila embryogenesis. Development 104:525-532[Abstract/Free Full Text].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego.

LÜER, K., J. URBAN, C. KLAMBT, and G. M. TECHNAU, 1997  Induction of identified mesodermal cells by CNS midline progenitors in Drosophila. Development 124:2681-2690[Abstract].

MARDAHL, M., R. M. CRIPPS, R. R. RINEHART, S. I. BERNSTEIN, and G. L. HARRIS, 1993  Introduction of y⁺ onto a CyO chromosome. Dros. Inf. Serv. 72:141.

MATHE, E., I. BOROS, K. JOSVAY, K. LI, and J. PURO et al., 1998  The Tomaj mutant alleles of alpha Tubulin67C reveal a requirement for the encoded maternal specific tubulin isoform in the sperm aster, the cleavage spindle apparatus and neurogenesis during embryonic development in Drosophila. J. Cell Sci. 111:887-896[Abstract].

MATTHEWS, K. A., D. F. MILLER, and T. C. KAUFMAN, 1990  Functional implications of the unusual spatial distribution of a minor alpha-tubulin isotype in Drosophila: a common thread among chordotonal ligaments, developing muscle, and testis cyst cells. Dev. Biol. 137:171-183[Medline].

MATTHEWS, K. A., D. REES, and T. C. KAUFMAN, 1993  A functionally specialized alpha-tubulin is required for oocyte meiosis and cleavage mitoses in Drosophila. Development 117:977-991[Abstract].

MOULINS, M., 1976 Ultrastructure of chordotonal organs, pp. 387–418 in Structure and Function of Proprioceptors in the Invertebrates, edited by J. P. MILL. Chapman & Hall, London.

NOGALES, E., S. G. WOLF, and K. H. DOWNING, 1998  Structure of the alpha beta tubulin dimer by electron crystallography. Nature 391:199-203[Medline].

NOGALES, E., M. WHITTAKER, R. A. MILLIGAN, and K. H. DOWNING, 1999  High-resolution model of the microtubule. Cell 96:79-88[Medline].

OKABE, M. and H. OKANO, 1997  Two-step induction of chordotonal organ precursors in Drosophila embryogenesis. Development 124:1045-1053[Abstract].

OSBORNE, K. A., A. ROBICHON, E. BURGESS, S. BUTLAND, and R. A. SHAW et al., 1997  Natural behavior polymorphism due to a cGMP-dependent protein kinase of Drosophila. Science 277:834-836. [comment][Abstract/Free Full Text].

RAFF, E. C., 1994 The role of multiple tubulin isoforms in cellular microtubule function, pp. 85–110 in Microtubules, edited by J. S. HYAMS and C. W. LLOYD. John Wiley & Sons, New York.

RAFF, E. C., M. T. FULLER, T. C. KAUFMAN, K. J. KEMPHUES, and J. E. RUDOLPH et al., 1982  Regulation of tubulin gene expression during embryogenesis in Drosophila melanogaster. Cell 28:33-40[Medline].

RAFF, E. C., J. D. FACKENTHAL, J. A. HUTCHENS, H. D. HOYLE, and F. R