The temporal and spatial expression pattern of the Drosophila melanogaster α2-tubulin gene (α2) has been investigated by examining the expression of an α2-lacZ fusion gene. When this fusion gene is introduced into the germ line by P-element mediated transformation, expression is only detected in chordotonal organs and testes. Chordotonal organs, which are sensory organs of the peripheral nervous system, express the gene from late embryonic through adult stages in both males and females. Testicular expression occurs from larval through adult stages and is limited to germ-line cells, the primary and secondary spermatocytes and perhaps the early spermatids.

Microtubules are major structural components of eukaryotic cells and are essential for a variety of functions ranging from cell division and motility to secretion and intracellular transport. The structural subunit of microtubules is a heterodimer of α and β tubulin polypeptides, each of which is usually encoded by a small family of closely related genes. In most organisms, these genes are differentially expressed in a tissue- and temporalspecific manner, perhaps reflecting a requirement for differing quantities or isotypes of the tubulin polypeptides (reviewed by Raff, 1984; Cleveland & Sullivan, 1985; Cleveland, 1987).

The a-tubulins of Drosophila are encoded by a gene family with four members (α1, α2, α3, and a4) located at different chromosomal sites (84B3-6, 85E6-10, 84D48 and 67C4-6, respectively) (Sanchez et al. 1980; Kalfayan & Wensink, 1981; Mishke & Pardue, 1982; Natzle & McCarthy, 1984). The four genes have different temporal and tissue-specific transcription patterns and code for cr-tubulins that differ in amino acid sequence (Kalfayan & Wensink, 1982; Theurkauf et al. 1986; Matthews et al. 1989). The α1 and α3 genes are most similar and appear to be expressed in all tissues and at almost all developmental stages. The α1 and α3 tubulins differ at only two residues (one conservative and one non-conservative substitution) and are very similar to the abundant tubulins from many other species. In contrast, o4 has a highly specific expression pattern and codes for a tubulin with highly divergent sequence.

The pattern of α2 expression is the subject of this paper. This gene encodes an α tubulin that differs from al at 21 out of the 450 residues. These differences include non-conservative substitutions predicted to change the secondary structure of the protein (Theurkauf et al. 1986). Previous investigations used α2-specific hybridization probes to investigate the developmental expression pattern of the gene (Kalfayan & Wensink, 1982; Matthews et al. 1989). Transcripts first appeared in late embryos, increased thereafter, and with some fluctuations were present through adult stages. Adult transcripts were only found in males and appeared to be predominantly, and perhaps exclusively, in testes. However, low levels of α2 transcripts were detected before testes develop, raising the possibility that expression occurs in other organs. To investigate this possibility as well as details of the testicular expression pattern, we have transformed the Drosophila germ line with α2-lacZ fusion genes. In this paper, we report confirmation of testicular expression and localize it to germline cells. In addition, we demonstrate that transcription in chordotonal organs can account for the previously observed expression in late embryos. The chordotonal expression occurs in both sexes from late embryonic to early pupal stages when the organs degenerate. The fusion gene also allowed us to detect low levels of transcription in the chordotonal organs of male and female adults.

Since chordotonal organs are not commonly known, we describe them here as background for our results. These organs are part of the peripheral nervous system (Hertweck, 1931; Miller, 1965; Campos-Ortega, 1982; Campos-Ortega & Hartenstein, 1985; McIver, 1985; Ghysen et al. 1986; Bodmer et al. 1987; Dambly-Chaudière & Ghysen, 1986). They are elongate, internal sensory organs that lie near the body surface and have their ends attached to that surface. The organ is thought to sense pressure, vibration, muscle tension, and changes in the shape of the body surface (Wiggles-worth, 1965). The anatomy of chordotonal organs suggests that they have specialized microtubular structures (Campos-Ortega, 1982; McIver, 1985; Campos-Ortega & Hartenstein, 1985). An organ consists of 1 to 7 parallel filamentous structures (scolopidia), each with several different cells including a sensory neuron. The neuron has an internal ciliary structure that undoubtedly contains tubulin. However, the major concentration of tubulin appears to be in a sheath cell that surrounds the neuronal dendrite. On the inner surface of a sheath cell and stretching along the axis of the dendrite, there are 6 to 20 structures that have dense, extensive arrays of microtubules. These structures suggest that specialized α2 expression in chordotonal organs may be needed to construct scolopidia either by providing an increased quantity or a different form of α tubulin.

General nucleic acid methodology and DNA constructions

Unless otherwise specified, nucleic acid manipulations were as described by Schleif & Wensink (1981) and Maniatis et al. (1982).

The α2-specific hybridization probe was a ‘2P-labelled, antisense RNA produced by in vitro transcription of α2 DNA between −223 to +448 (nucleotide positions are relative to the cap site, +1). This α2 DNA extends 6 nucleotides into the second exon and includes the first three codons. It was inserted into the polylinker of pSP64-forming plasmid pSP64-Tα2-SSX, which was transcribed with the Riboprobe system (Promega Biotec; Melton et al. 1984).

The α2-lacZ fusions were made as follows. DNA from pDmTα2.1 (Kalfayan & Wensink, 1981) was recloned into pUC9 along with 1-95 kb of additional upstream α2 DNA (Sa/I-EcoRI fragment from ADmTα2’ [Kalfayan et al. 1982]) inserted in its normal position and orientation relative to α2, producing plasmid pTo2-S. The unique Ns il site (nucleotide +448 in codon 4) of α2 in this plasmid was changed to a Sall site by linker tailing (Lathe et al. 1984) with New England Biolabs linker #1027 producing plasmid pTα2-SS. This altered α2 gene was then fused in frame to the hsp70-IacZ gene from pSPl.l (Lis et al. 1983) that we had inserted into a P-element cloning vector, CP20.1 (Simon et al. 1985). The product was pTα2-lac.l (Fig. 1). In an amino- to carboxyterminal direction, pTα2-lac.l has α2 DNA to codon 4, then hsp70 DNA from codon 313 (a Pstl site converted to a Sall site by linker tailing as described above and subsequently joined to the Sall site at +448 in pTα2-SS) to 336, and finally lacZ DNA from codon 9. The pTα2-lac.2 was made by deleting a portion of the α2 gene (Fig. 1) in pTα2-S and inserting the product into the same lacZ-CP20.1 structure.

Fig. 1.

The α2 gene and constructs. The wild-type (Canton S) α2 gene is diagrammed with an arrow indicating the transcript and direction of transcription (Kalfayan et al. 1982; Theurkauf et al. 1986; Bo, 1989). Carets within the arrow locate the two introns. Numbers below the diagram indicate nucleotide positions relative to the transcription initiation site (+1). The precision of the endpoints (−2200 and +2700) is ±100 nucleotides. Below ±2 are diagrams of constructs used in germ line transformation. Parentheses locate the deleted DNA in α2.3 and the shaded boxes symbolize lacZ DNA which has a short hsp70 coding region at its 5’ end (see Materials and methods). Only α2 DNA is drawn to scale. Nucleotide positions of α2 breakpoints in the constructs are shown in the upper diagram.

Fig. 1.

The α2 gene and constructs. The wild-type (Canton S) α2 gene is diagrammed with an arrow indicating the transcript and direction of transcription (Kalfayan et al. 1982; Theurkauf et al. 1986; Bo, 1989). Carets within the arrow locate the two introns. Numbers below the diagram indicate nucleotide positions relative to the transcription initiation site (+1). The precision of the endpoints (−2200 and +2700) is ±100 nucleotides. Below ±2 are diagrams of constructs used in germ line transformation. Parentheses locate the deleted DNA in α2.3 and the shaded boxes symbolize lacZ DNA which has a short hsp70 coding region at its 5’ end (see Materials and methods). Only α2 DNA is drawn to scale. Nucleotide positions of α2 breakpoints in the constructs are shown in the upper diagram.

The α2.3 gene was constructed from pTα2-S by deleting a 318 bp Stul fragment (+1405 to +1723), digesting with Hin- dlll producing tails that were then partially filled with dATP and dGTP (Hung & Wensink, 1984), and joined to Cp20.1 that had been digested with Xbal and partially filled with dCTP and dTTP.

Germ line transformation and β-galactosidase assays

The transformation method of Spradling & Rubin (1982) was used with modifications described by Garabedian et al. (1985).

The intensity of staining was highly dependent on treatments that are likely to increase penetration of dye into the organism. Dechorionation and devitellination of embryos, mechanical destruction of larval, pupal, or adult cuticle, and the extent of dissection were important variables.

The preparation of embryos for β-galactosidase staining assays was carried out according to the general methods described previously (Karr et al. 1985; Freeman et al. 1986; Lis et al. 1983; Garabedian et al. 1986) with the following modifications. Embryos (1 to 24 h after fertilization, 25°C) were dechorionated by soaking in 3% (v/v) hypochlorite, 0·7 % (w/v) NaCl, 04 % (v/v) Triton X-100 at 25°C for 5 min, and then rinsed with 0·7 % NaCl and 0·4 % Triton X-100. The dechorionated embryos were then devitellinated and fixed with 1:1 heptane and buffer A (17mm-K2HPO4, 75mm-KCl, 25mm-NaCl, 3mm-MgCl2, 6% formaldehyde) for 10 min at 25°C with shaking. The product was washed several times with 0·7 % NaCl and 0·4 % Triton X-100, 15 min with shaking and then soaked in X-gal staining buffer at 25°C for 4 to 48 h. Staining buffer was: 10mm-Nα2HPO4 (pH 7·0, adjusted with phosphoric acid), 0·87mm-NaCl, lmm-MgCl2, 3·3 mm-K4Fe(CN)6.3H2O, 0·2% (w/v) 5-bromo-4-chloro-3-indolyl-1-D-galactopyranoside (X-gal; first dissolved at 200 mg ml-1 in N,N-dimethylformamide). The stained embryos were washed with water and put on slides in 75 % glycerin under a cover glass. Fujicolor Super HR 100 film was used for photography.

Postembryonic Drosophila were prepared for staining in a different manner. Whole adult flies were anesthetized under CO2 and transferred to dissecting tray wells containing 200 μl 1 % formaldehyde. The abdomens were opened with microscissors and the heads, thoraxes, and legs were gently punctured with dissecting needles. Fixation in formaldehyde continued for 5 min, 25°C. The samples were then washed in water, soaked in the staining buffer for 4 to 64h, 25°C, washed in water, and further dissected. Larvae and early pupae were opened with microscissors making an abdominal ventral-sagittal cut and then fixed and stained as described for adults. Late pupae were first soaked in Ringer’s buffer 10 min to soften the cuticle. The abdomens were then cut as described for larvae. After 10min fixation, they were washed and stained as described for adults.

In order to examine the temporal and tissue specificities of α2 expression during development, the gene and 5’ flanking DNA were fused to the E. coli lacZ gene and then introduced into the germ line of Drosophila. Fusion to the lacZ gene permits expression to be localized and sensitively detected by histological staining methods. Preliminary transformation experiments demonstrated that the structural gene and 2-2 kb of upstream DNA (n2.3; Fig. 1) were sufficient to produce the normal ratios between α2 transcripts of embryos, larvae, and adult males and females (Kalfayan & Wensink, 1982; Matthews et al. 1989; Bo, 1989). For this reason, our first construction fused 2-2 kb of upstream DNA and the first 448 bp of α2 to lacZ (α2-lac.l; Fig. 1). The α2 fusion site in this construct is located 6 bp downstream from the start of the second exon and is within the fourth codon of α2. A second construct (α2-lac.2; Fig. 1) used the same fusion joint, but included only 223 bp of upstream α2 DNA. These were introduced into the germ line yielding two independently transformed lines of flies for each construct. Southern blot hybridization analysis indicated that each line contains a single, unrearranged copy of the fusion gene (data not shown). The α2-lac.2 construct, which contains only 223 bp of upstream DNA, did not express in either line. In contrast, both of the α2-lac.l lines expressed and gave identical staining patterns. This result indicates that DNA sequences between -2-2 kb and -223bp are necessary for α2 expression. The staining results described below indicate that the o2 DNA in the α2-lac.l construct is sufficient for most and perhaps all of the α2 developmental specificities.

The expression pattern of the α2-lac.l fusion gene was examined by staining with X-gal, a colorless compound which produces a blue product when digested by the lacZ gene product, α-galactosidase. The staining pattern at all stages of development was reproducible and usually gave strong contrast between stained and unstained cells. The staining indicated that the hybrid gene is expressed in two organs, the testes and the chordotonal organs. The staining patterns are described in the following sections.

Embryonic staining pattern

Transformant embryos first stain at stage 12 of embryonic development (staging system of Bownes, 1982; approximately stage 15 in the system of Campos-Ortega & Hartenstein, 1985). At this time intense staining appears synchronously at many sites producing the pattern shown in Fig. 2. Lateral (Fig. 2A) and ventral (Fig. 2C) views reveal a set of stained cells in each of the twelve thoracic and abdominal segments. In the lateral view, a slightly S-shaped pattern of large, intensely stained spots traces the anterior-posterior axis along the surface of the embryo. These spots mark the three thoracic and the first seven abdominal segments. Large spots marking the eighth and ninth abdominal segments are difficult to interpret in this view and can be distinguished more easily in Fig. 2C as a pair of spots on each side of the embryonic posterior. This ventral view also shows that the staining pattern is bilaterally symmetric. The bilateral symmetry also can be seen in the lateral view of Fig. 2A, where out of focus spots of analogous structures on the far side of the embryo can be seen above most of the in-focus spots on the near side.

Fig. 2.

Embryonic expression of α2-lac.l. Lateral (A & B) and ventral views (C) of 15 h whole-mount embryos stained for αgalactosidase activity are shown in bright-field micrographs. Drawings of neural tissues (modified from Campos-Ortega & Hartenstein, 1985 [panels A & C]; and Hartenstein & Campos-Ortega, 1986 [panel B]) show the positions of chordotonal organs (blue). Symbols are: A1-A8, abdominal segments; A9, telson; a, anterior; ch, chordotonal organ; d, dorsal; 1, lateral; T1-T3, thoracic segments; and v, ventral. The letter preceding ch indicates the organ position and the number following indicates the number of scolopidia in the organ.

Fig. 2.

Embryonic expression of α2-lac.l. Lateral (A & B) and ventral views (C) of 15 h whole-mount embryos stained for αgalactosidase activity are shown in bright-field micrographs. Drawings of neural tissues (modified from Campos-Ortega & Hartenstein, 1985 [panels A & C]; and Hartenstein & Campos-Ortega, 1986 [panel B]) show the positions of chordotonal organs (blue). Symbols are: A1-A8, abdominal segments; A9, telson; a, anterior; ch, chordotonal organ; d, dorsal; 1, lateral; T1-T3, thoracic segments; and v, ventral. The letter preceding ch indicates the organ position and the number following indicates the number of scolopidia in the organ.

The details of embryonic staining indicate that the stained cells are in the chordotonal organs of the peripheral nervous system. This interpretation is based on previous descriptions of the peripheral nervous system (Hertweck, 1931; Campos-Ortega, 1982; Campos-Ortega & Hartenstein, 1985; Ghysen et al. 1986; Bodmer et al. 1987). Staining is coincident with the visible differentiation of chordotonal organs and increases in intensity throughout the remainder of embryogenesis. A photograph and diagram of staining in thoracic segments can be seen in Fig. 2A. Large heavily stained spots corresponding to the dorsal chordotonal triscolopidium (dch3, an organ with three neurons and three scolopidia) of each thoracic segment are visible in every stained embryo. Two additional, more ventral spots are slightly out of the plane of focus in this photograph. The larger, more anterior spot, corresponds to the lateral triscolopidium (lch3) and the smaller, more posterior spot, corresponds to the lateral monoscopidium (Ichl, an organ with one neuron). Both are known to occur in the posterior of segment Tl, but not in segments T2 and T3. Very faint spots are visible in more ventral portions of each thoracic segment and in the gnathal buds (Fig. 2A). Some of these can be seen in the ventral view shown in Fig. 2C. These additional spots are of approximately the correct quantity and position to account for the expected additional monoscopidial chordotonal organs near the ventral midline of each thoracic segment and in various parts of the gnathal buds. However, these additional organs have been less extensively described in the literature and the stained spots we observe are difficult to localize, so an unambiguous correlation cannot be made.

Correspondence between chordotonal organs and staining in the abdominal segments is shown in Fig. 2B. The organs are labelled in the A4 segment of the photograph and in the diagram. The visibility of faint spots is sensitive to the position of the focal plane so all cannot be seen in high-magnification photographs like that shown in Fig. 2B. The two lightly stained ventral chordotonal monoscolopidial organs (vchl), the heavily stained lateral chordotonal pentascolopidium (lch5) and the lightly stained lateral chordotonal monoscopidium (Ichl) are common to all abdominal segments and have been described by Campos-Ortega & Hartenstein (1985) and by Bodmer et al. (1987). A fifth spot stains lightly and is not visible in all segments of all stained embryos. Its position corresponds to the ventral chordotonal monoscolopidium (v’chl) which appears at the same time as other abdominal chordotonal organs but has been reported to migrate dorsally where, once it is dorsal to lch5 (Ghysen et al. 1986), it is termed Ichl (Campos-Ortega & Hartenstein, 1985; Bodmer et al. 1987). Since both v’chl and Ichl are visible in the same segment (Fig. 2B), our identification of v’chl is tentative. Staining in the eighth and ninth abdominal segments and a diagram of the corresponding chordotonal organs are shown in Fig. 2C.

Larval staining pattern

Larval staining specific to the transformed lines is limited to testes and what we interpret to be chordotonal organs. No staining was seen in testes or chordotonal organs of untransformed larvae or in the gonads of female transformants. Background staining does occur in the intestines of transformed and non-transformed larvae.

A strong indication that chordotonal organs are expressing this fusion gene is that the non-testicular larval staining follows the developmental path of chordotonal organs. The embryonic chordotonal organs elongate and mature to form larval chordotonal organs. The number and relative positions of these organs is maintained throughout this transition (Hertweck, 1931; Dambly-Chaudière & Ghysen, 1986). This same pattern is observed in the staining pattern. Larval staining occurs in the same positions as in embryos and increases in intensity and elongates as the organs develop through the first instar. Staining then diminishes, but remains strong through the third instar.

The staining pattern corresponding to chordotonal organs of a second instar larva is shown in Fig. 3A. The thoracic and abdominal patterns are segmentally reiterated. The photograph in Fig. 3A shows the correspondence to chordotonal organs found in the embryonic abdominal segments (Fig. 2B). Further indication that these stained structures are chordotonal organs comes from the five stranded structure that becomes visible when an organ we identify as the lateral chordotonal pentascolopidium (lch5) is squashed (Fig. 3B, upper photograph). The five visible strands correspond to the five scolopidia. The same chordotonal structure and staining pattern occurs in both male and female larvae (Fig. 3C).

Fig. 3.

Larval expression of α2-lac.l. Bright-field micrographs show dissected second instar larvae stained with X-gal. (A) The upper photograph shows the pattern of staining in the whole body wall. The lower photograph shows details of the pattern and its correspondence to chordotonal organs. (B) A higher magnification micrograph and a schematic diagram (from Miller, 1965) of a lateral pentascolopidial chordotonal organ (lch5) from a second instar larvae. (C) Male and female transformant larvae. Symbols are the same as in Fig. 2 with the addition of: C, chordotonal organ; Ca, cap cell; Cu, cuticle; F, female; M, male; N, nucleus; Se, sensory cell; Sh, sheath cell; Sn, segmental nerve; T, testis.

Fig. 3.

Larval expression of α2-lac.l. Bright-field micrographs show dissected second instar larvae stained with X-gal. (A) The upper photograph shows the pattern of staining in the whole body wall. The lower photograph shows details of the pattern and its correspondence to chordotonal organs. (B) A higher magnification micrograph and a schematic diagram (from Miller, 1965) of a lateral pentascolopidial chordotonal organ (lch5) from a second instar larvae. (C) Male and female transformant larvae. Symbols are the same as in Fig. 2 with the addition of: C, chordotonal organ; Ca, cap cell; Cu, cuticle; F, female; M, male; N, nucleus; Se, sensory cell; Sh, sheath cell; Sn, segmental nerve; T, testis.

Testicular staining first occurs in the second instar (Fig. 3C) and continues through the third instar. As expected, stained larval testes are spherical and in the fifth abdominal segment.

Pupal staining pattern

In pupae, staining caused by the α2-lac.l fusion gene occurs in the chordotonal organs of both sexes and in testes (Fig. 4A & B). Testicular staining persists throughout pupal development, but chordotonal staining gradually diminishes, becoming nearly undetectable by 12 h after pupariation (Fig. 4B) and undetectable thereafter. Except for a light background staining in the intestines, no staining was visible in untransformed pupae.

Fig. 4.

Pupal expression pattern of α2-lac.l. Whole mounts of dissected male and female pupae stained with X-gal are shown. (A) Early pupae (white pupae, before 12thh). (B) Late pupae (after 12th h). Symbols are as for Fig. 3.

Fig. 4.

Pupal expression pattern of α2-lac.l. Whole mounts of dissected male and female pupae stained with X-gal are shown. (A) Early pupae (white pupae, before 12thh). (B) Late pupae (after 12th h). Symbols are as for Fig. 3.

Adult staining pattern

Adult staining specific to transformed flies occurs in the testes (Fig. 5A & B) and in the chordotonal organs of both sexes. Both transformed and non-transformed adults showed background staining in intestines (Fig. 5 A & B ; Lis et al. 1983; Garabedian et al. 1986). In addition, after prolonged staining, male and female genitalia and rows of cells under the dorsal abdominal cuticle also showed background staining. While the non-testicular, transformant-specific staining generally corresponds to chordotonal organs previously described in the literature (reviewed by Miller, 1965), two sites expected to have these organs do not stain. In addition, some staining occurs in regions at which, to our knowledge, chordotonal organs have not been observed previously. Since chordotonal organs have been less extensively studied in adults than in embryos and larvae, we suspect that some adult chordotonal organs may not yet have been described. For this reason and for simplicity, we will tentatively identify all of the non-testicular, non-background staining as chordotonal staining. Note that adult chordotonal organs do not appear to be derived from larval-pupal chordotonal organs because chordotonal organs appear to degenerate approximately 12 h after pupation begins (Bodenstein, 1965).

Fig. 5.

Adult expression pattern of α2-lac.l. Stained, partially dissected transformant and control females (A) and males (B) are shown. A ventral view of a transformant male is shown in C. Details of staining: at the base of the wing (D) (staining of the wing radius and the dorsal region of the pteropleura is indicated by arrows); in prothoracic ganglion region (E); in the basi-proboscis (F) (two small rod-like stained regions are indicated by arrows); in the femur (G); and in the Wheeler’s organs (H). Symbols are: Co, untransformed control; Tr, transformant; I, intestines; O, ovary; H, haltere.

Fig. 5.

Adult expression pattern of α2-lac.l. Stained, partially dissected transformant and control females (A) and males (B) are shown. A ventral view of a transformant male is shown in C. Details of staining: at the base of the wing (D) (staining of the wing radius and the dorsal region of the pteropleura is indicated by arrows); in prothoracic ganglion region (E); in the basi-proboscis (F) (two small rod-like stained regions are indicated by arrows); in the femur (G); and in the Wheeler’s organs (H). Symbols are: Co, untransformed control; Tr, transformant; I, intestines; O, ovary; H, haltere.

The staining intensity in adult chordotonal organs is reproducible, but differs between organs at different sites. The photograph in Fig. 5C shows a stained adult male. The surrounding photographs (Fig. 5D to 5H) give more detailed views of some stained structures. One of the first chordotonal structures to stain is in the femur near the trochanter of each leg (Fig. 5G). Further staining reveals a second smaller structure in the femur near the tibia. Prolonged staining leads to merging of the two into a structure that stretches the entire length of the femur. Prolonged staining also causes staining in the coxa and tibia of the leg. Another chordotonal structure that stains strongly is a set of four spots that correspond to the Wheeler’s organ at the anterior margin of the second abdominal stemite (Fig. 5H). Staining in this organ has a striated structure similar to that observed in the larval chordotonal organs (Fig. 3B). Two sites at the base of the wing (arrows in Fig. 5D), the proximal radius of the wing blade and the dorsal region of the pteropleura, also stain. The radius stains more strongly than the pteropleura. A pair of oval spots stain weakly, one on each side of the prothoracic ganglion mass (Fig. 5E). Still weaker staining occurs in two pairs of small rod-like structures in the basi-proboscis region (arrows in Fig. 5F). Each is immediately above one of the maxillary palps. Prolonged staining reveals additional diffuse staining above these rods and localized staining in two small spots below the rods in the labrum.

The correspondence between adult chordotonal organs and staining is good, but not perfect. With the exception of the Wheeler’s organ, the pteropleura, and the tibia, all of the stained spots correspond to previously reported chordotonal organs (Miller, 1965; Campos-Ortega, 1982). Two sites have the organs but do not stain. The second antennal segment and the halteres do not stain despite prolonged staining treatment and destruction of some surrounding cuticle. The lack of staining in the second antennal segment may be due to the atypical character of antennal chordotonal organs. These organs are so unusual that they have been given a special name, the Johnston’s organ, and are suspected of having a function different from other chordotonal organs (Miller, 1965).

We conclude that the pattern of staining in adults indicates that most of the known adult chordotonal organs express the fusion gene and that several additional chordotonal organs may not yet have been described in the literature.

Details of the testicular staining pattern

The β-galactosidase activity was observed in male testes from larval to adult stages (Figs 3C, 4A, 4B, 5B & 6). Testicular staining first appears approximately 48 h after larval hatching. At this stage the testis contains spermatogonia clustered in a small volume at the anterior apex while most of the remaining testicular volume is occupied by spermatocytes. At this time approximately 15 spherical bodies stain in the posterior region (Fig. 6A). We interpret each to be a 16-cell cluster of primary spermatocytes. A higher magnification view of the posterior at a later stage (65 h) shows stained clusters of 16 primary spermatocytes in more detail (Fig. 6B). Our interpretation of their identity is based on previous observations that 16-cell clusters of primary spermatocytes first appear at approximately 48 h, that spermatocyte development proceeds from anterior to posterior, and that the latest germ-cell stage present in larval testes is the 16-cell cluster of primary spermatocytes (Bodenstein, 1965). This interpretation is strengthened by the pattern of staining found in adult testes.

Fig. 6.

Testicular staining pattern of α2-lac.l. A stained second instar larval testis (48 h) from a transformant is shown in A. The low contrast outer boundary of this testis is outlined with an ink line. A higher magnification view of the 16-cell clusters of primary spermatocytes in the posterior of a 65 h transformant larvae is shown in B. The stained male reproductive system of untransformed (C) and transformed (D) adults are shown. A stained testis is shown in E along with higher magnification views of the testicular apex (F), central region that has been slightly crushed (G), and basal region containing mature sperm (H). Symbols are: 1, testis; 2, testicular duct; 3, seminal vesicle; 4, ejaculatory duct; 5, accessory gland; 6, ejaculatory bulb; a, anterior; A, apex; B, base; MS, mature sperm; p, posterior; PS, 16-cell clusters of primary spermatocytes; SC, spermatocytes; ST, spermatids.

Fig. 6.

Testicular staining pattern of α2-lac.l. A stained second instar larval testis (48 h) from a transformant is shown in A. The low contrast outer boundary of this testis is outlined with an ink line. A higher magnification view of the 16-cell clusters of primary spermatocytes in the posterior of a 65 h transformant larvae is shown in B. The stained male reproductive system of untransformed (C) and transformed (D) adults are shown. A stained testis is shown in E along with higher magnification views of the testicular apex (F), central region that has been slightly crushed (G), and basal region containing mature sperm (H). Symbols are: 1, testis; 2, testicular duct; 3, seminal vesicle; 4, ejaculatory duct; 5, accessory gland; 6, ejaculatory bulb; a, anterior; A, apex; B, base; MS, mature sperm; p, posterior; PS, 16-cell clusters of primary spermatocytes; SC, spermatocytes; ST, spermatids.

Staining in adult testes occurs in primary and secondary spermatocytes and perhaps in early postmeiotic spermatids. A general view of the gonadal region shows staining is specific to the testis and does not occur in other parts of the adult male reproductive system (Fig. 6C & 6D). Spermatogenesis progresses from spermatogonia at the apex of the organ to mature bundles of 64 coiled sperm at the base of the organ (Lindsley & Tokuyasu, 1980; Bownes & Dale, 1982). Staining occurs in the central portion of the testis, but not at the apex or base (Fig. 6E). Staining first becomes intense at the site where 16-cell spermatocyte clusters first appear (Fig. 6F). In the remaining central portion of the organ, spermatocytes and elongated differentiating spermatid bundles are intermixed. When this central portion is slightly crushed to separate cell clusters, the stained spermatocyte clusters can be seen next to unstained, elongated sperm tail clusters (Fig. 6G). Our observation of such crushed preparations and of through-focal views of intact organs leads us to conclude that elongated spermatids do not stain. A view of the basal region shows that testicular staining has ceased at the site where mature sperm begin to coil (Fig. 6H).

We have described the developmental expression pattern in Drosophila of an α2 tubulin, β-galactosidase fusion gene. The principal observation is that this gene is expressed in only two organs: the testes and the chordotonal organs. Expression in testes is limited to a specific cell type, the germ line cells, and is confined to a specific stage of their development, spermatocytes, and perhaps early spermatids. Expression in chordotonal organs occurs in individual scolopidia (Fig. 3B) and may be in one or in all of the few component cells.

Interpretation of the staining pattern

The staining results reported in this paper are consistent with the expression pattern observed in previous studies of normal o2 gene transcription (Kalfayan & Wensink, 1982; Matthews et al. 1989). The previous work established that o2 transcripts occurred in adult testes and at low levels in unspecified organs from late embryonic to pupal stages. Our staining results account for this RNA pattern by localizing α2-lacZ expression to testes and chordotonal organs. The very low level of α2 transcripts detected before testicular expression occurs (that is, before second instar larvae) corresponds well with the amount of RNA likely to be produced by the small number of cells in embryonic and first instar chordotonal organs.

The sensitivity and precision of the staining assay allows us to localize expression more exactly and also to detect highly localized expression that was previously undetected because it represented such a small fraction of total fly RNA. Thus, the staining method allowed us to localize α2 expression to particular cells of the testes, the spermatocytes, and to individual chordotonal organs. It also allowed us to detect α2 expression where no transcripts had been detected by Northern blot analysis, namely, in adult females and in non-testicular tissue of adult males.

The fusion gene results indicate that the temporal and cell type specificities of α2 gene expression are highly restricted. These results demonstrate that the cloned α2 DNA (-2-2 kb to +448 bp) used in the fusion construction appears to determine all of the transcriptional specificities of the normal α2 gene. However, there are several reasons why this approach may not have detected other expression specificities of α2. First, it is possible that there are α2 expression control elements that lie outside the α2 region used in our construction. Second, it is possible that the assay is too insensitive to detect low level expression from one or a few cells. This latter possibility presumes that the expression observed in chordotonal organs is high. Third, it is also possible that the assay could be ineffective in some tissues, for example due to X-gal impermeability or to specific enzymic inhibition. This last possibility seems unlikely because previous studies with Drosophila indicated that α-galactosidase activity can be expressed and detected in virtually all tissues (Lis et al. 1983; Garabedian et al. 1986).

Although appearance of α-galactosidase activity seems to be a reliable assay for the initiation of gene activity, our limited knowledge of the half-life of the mRNA and protein in different tissues makes interpretation of continued /3-galactosidase activity difficult. With testicular expression of α2 this issue is not crucial because activity is only observed during a short period of spermatogenesis. In chordotonal organs this issue is more significant because activity persists from embryonic to pupal stages in cells that do not appear to divide. It is possible that all gene transcription occurs when chordotonal organs first appear and that continued and increased enzymic activity is due to continued translation of stable transcripts or to stable, but increasingly active enzyme. Staining in adult chordotonal organs is a positive indication of adult transcription because these organs are not derived from pupal chordotonal organs.

In summary, these experiments demonstrate that a segment of u2 DNA directs transcription in two tissues and appears to account for all of the known tissuespecific expression of the o2 gene. This α2 segment contains the cri-acting control sequences necessary for chordotonal and male germ line transcription.

Possible significance of the testicular and chordotonal expression of α2-tubulin

Two cr-tubulin genes from mice encode an isotype that appears to be expressed exclusively in testes (Villasante et al. 1986). A Drosophila α-tubulin gene also appears to encode an isotype expressed with testicular specificity (Kemphues et al. 1979; Raff, 1984; Michiels et al. 1987). It seems likely that these tubulins, which differ in structure from constitutively expressed tubulins, may have one of two roles. They may provide a special function unique to flagellar microtubules (Fulton & Simpson, 1976). Alternatively, they may function like constitutive tubulins, but are expressed in testes simply to provide the additional tubulin necessary for the microtubule intensive sperm structures (Raff, 1984). We know of no evidence to support either of these hypotheses (reviewed in Raff, 1984; Cleveland, 1987). However, it is interesting that the largest cluster of nonconservative substitutions in o2 (Theurkauf et al. 1986) is in the carboxy-terminal region likely to be involved in regulation of microtubule assembly and in interactions with microtubule-associated proteins (Serrano et al. 1984). This region may allow specialized functions for α2. needed in both flagellar microtubules and microtubules of chordotonal scolopidia.

We thank Michael Wormington, Kalpana White, Susan Hardin, Doug Jacoby, Susan Logan, and Kim O’Donnell for constructive criticism of this manuscript. Judith Black generously provided assistance in reproducing the color photographs. We also thank Claude Maina and Daniel Rosen for helpful discussions. This research was supported by a National Institutes of Health research grant (ROI GM31234) to P.C.W.

Bo
,
J.
(
1989
).
Ph.D. Thesis, Brandéis University
.
Bodenstein
,
D.
(
1965
).
The postembryonic development of Drosophila
. In
Biology of Drosophila,
(ed.
M.
Demerec
), pp.
275
367
.
New York
:
Hafner
.
Bodmer
,
R.
,
Barbel
,
S.
,
Sheperd
,
S.
,
Jack
,
J. W.
,
Jan
,
L. Y.
&
Jan
,
Y. N.
(
1987
).
Transformation of sensory organs by mutations of the cut locus of D. melanogaster
.
Cell
51
,
293
307
.
Bownes
,
M.
(
1982
).
Embryogenesis
. In
Handbook of Drosophila Development,
(ed.
Ransom
,
R.
), pp.
67
94
.
Amsterdam: Elsevier
.
Bownes
,
M.
&
Dale
,
L.
(
1982
).
Gametogenesis
. In
Handbook of Drosophila Development,
(ed.
Ransom
,
R.
), pp.
31
66
.
Amsterdam
:
Elsevier
.
Campos-Ortega
,
J. A.
(
1982
).
Development of the nervous system
. In
Handbook of Drosophila Development,
(ed.
Ransom
,
R.
), pp.
153
181
.
Amsterdam
:
Elsevier
.
Campos-Ortega
,
J. A.
&
Hartenstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster.
Berlin
:
SpringerVerlag
.
Cleveland
,
D. W.
(
1987
).
The multitubulin hypothesis revisited: What have we learned?
J. Cell Biol.
104
,
381
383
.
Cleveland
,
D. W.
&
Sullivan
,
K. F.
(
1985
).
Molecular biology and genetics of tubulin
.
A. Rev. Biochem.
54
,
331
365
.
Dambly-ChaudiÈre
,
C.
&
Ghysen
,
A.
(
1986
).
The sense organs in the Drosophila larva and their relation to the embryonic pattern of sensory neurons
.
Roux’s Arch, devl Biol.
195
,
222
228
.
Freeman
,
M.
,
NÜsslein-Volhard
,
C.
&
Glover
,
D. M.
(
1986
).
The dissociation of nuclear and centrosomal division in gnu, a mutation causing giant nuclei in Drosophila
.
Cell
46
,
457
468
.
Fulton
,
C.
&
Simpson
,
P. A.
(
1976
).
Selective synthesis and utilization of flagellar tubulin: the multi tubulin hypothesis
. In
Cell Motility,
(ed.
Goldman
,
R.
,
Pollard
,
T.
&
Rosenbaum
,
J.
), pp.
987
1005
.
Cold Spring Harbor, N.Y
.:
Cold Spring Harbor Laboratory
.
Garabedian
,
M. J.
,
Hung
,
M.-C.
&
Wensink
,
P. C.
(
1985
).
Independent control elements that determine yolk protein expression in alternative Drosophila tissues
.
Proc. natn. Acad. Sci. U.S.A.
82
,
1396
1400
.
Garabedian
,
M. J.
,
Shepherd
,
M. B.
&
Wensink
,
P. C.
(
1986
).
A tissue-specific transcription enhancer from the Drosophila yolk protein 1 gene
.
Cell
45
,
859
867
.
Ghysen
,
A.
,
Dambly-ChaudiÈre
,
C.
,
Aceves
,
E.
,
Jan
,
L. Y.
&
Jan
,
Y. N.
(
1986
).
Sensory neurons and peripheral pathways in Drosophila embryos
.
Roux’s Arch, devl Biol.
195
,
281
289
.
Hartenstein
,
V.
&
Campos-Ortega
,
A. J.
(
1986
).
The peripheral nervous system of mutants of early neurogenesis in Drosophila melanogaster
.
Roux’s Arch, devl Biol.
195
,
210
221
.
Hertweck
,
H.
(
1931
).
Anatomie und Variabilitat des nervensystems und der sinnesorgane von Drosophila melanogaster (Meigen)
Z. wiss. Zool.
139
,
559
663
.
Hung
,
M.-C.
&
Wensink
,
P. C.
(
1984
).
Different restriction enzyme-generated sticky DNA ends can be joined in vitro
.
Nucl. Acids Res.
12
,
1863
1874
.
Kalfayan
,
L.
,
Loewenberg
,
J.
&
Wensink
,
P. C.
(
1982
).
Drosophila a-tubulin genes and their transcription patterns
.
Cold Spring Harbor Symp. quant. Biol.
46
,
185
190
.
Kalfayan
,
L.
&
Wensink
,
P. C.
(
1981
).
cr-tubulin genes of Drosophila
.
Cell
24
,
97
106
.
Kalfayan
,
L.
&
Wensink
,
P. C.
(
1982
).
Developmental regulation of Drosophila or-tubulin genes
.
Cell
29
,
91
98
.
Karr
,
T. L.
,
Ali
,
Z.
,
Dregs
,
B.
&
Kornberg
,
T.
(
1985
).
The engrailed locus of D. melanogaster provides an essential zygotic function in precellular embryos
.
Cell
43
,
591
601
.
Kemphues
,
K. J.
,
Raff
,
R. A.
,
Kaufman
,
T. C.
&
Raff
,
E. C.
(
1979
).
Mutation in a structural gene for a fl tubulin specific to testis in Drosophila melanogaster
.
Proc. natn. Acad. Sci. U.S.A.
76
,
655
670
.
Lathe
,
R.
,
Kieny
,
M. P.
,
Skory
,
S.
&
Lecocq
,
J. P.
(
1984
).
Linker tailing: unphosphorylated linker oligonucleotides for joining DNA termini
.
DNA
3
,
173
182
.
Lindsley
,
D. L.
&
Tokuyasu
,
K. T.
(
1980
).
Spermatogenesis
. In
The Genetics and Biology of Drosophila, vol. 2D
, (ed. M. Ashburner &
T. R. F.
Wright
), pp.
225
294
.
New York
:
Academic Press
.
Lis
,
J. T.
,
Simon
,
J. A.
&
Sutton
,
C. A.
(
1983
).
New heat shock puffs and /3-galactosidase activity resulting from transformation of Drosophila with an hsp70-/acZ hybrid gene
.
Cell
35
,
403
410
.
Maniatis
,
T.
,
Fritsch
,
F. E.
&
Sambrook
,
J.
(
1982
).
Molecular Cloning. Cold Spring Harbor
,
New York
:
Cold Spring Harbor Laboratory
.
Matthews
,
K. A.
,
Miller
,
D. F. B.
&
Kaufman
,
T. C.
(
1989
).
Developmental distribution of RNA and protein products of the Drosophila "--tubulin gene family
.
Devl Biol.
132
,
45
61
.
Mciver
,
S. B.
(
1985
).
Mechanoreception
. In
Comprehensive Insect Physiology, Biochemistry, and Pharmacology,
vol.
6
, (ed. L.I. Gilbert &
G. A.
Kerkut
), pp.
71
125
.
New York
:
Pergamon Press
.
Melton
,
D. A.
,
Krieg
,
P. A.
,
Rebagliati
,
M. R.
,
Maniatis
,
T.
,
Zinn
,
K.
&
Green
,
R. M.
(
1984
).
Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter
.
Nucl. Acids Res.
12
,
7035
7056
.
Michiels
,
F.
,
Falkenburg
,
D.
,
MÜller
,
A. M.
,
Jinz
,
V.
,
Orro
,
V.
,
Bellmann
,
R.
,
GlÂtzer
,
H. K.
,
Brand
,
R.
,
Bialojan
,
S.
&
Renkawitz-Pohl
,
R.
(
1987
).
Testis-specific (P. tubulins are identical in Drosophila melanogaster and D. hydei but differ from the ubiquitous /?1 tubulin
.
Chromosoma
95
,
387
398
.
Miller
,
A.
(
1965
).
The internal anatomy and histology of the imago of Drosophila melanogaster.
In
Biology of Drosophila,
(ed.
M.
Demerec
), pp.
420
534
.
New York
:
Hafner
.
Mischke
,
D.
&
Pardue
,
M. L.
(
1982
).
Organization and expression of "-tubulin genes in Drosophila melanogaster
.
J. molec. Biol.
156
,
449
466
.
Natzle
,
J. E.
&
Mccarthy
,
B. J.
(
1984
).
Regulation of the Drosophila a- and (-tubulin genes during development
.
Devl Biol.
104
,
187
198
.
Raff
,
E. C.
(
1984
).
Genetics of microtubule systems
.
J. Cell Biol.
99
,
1
10
.
Sanchez
,
F. J. E.
,
Natzle
,
J. E.
,
Cleveland
,
D. W.
,
Kirschner
,
M. W.
&
Mccarthy
,
B. J.
(
1980
).
A dispersed multigene family encoding tubulin in Drosophila melanogaster
.
Cell
22
,
845
854
.
Schleif
,
R. F.
&
Wensink
,
P. C.
(
1981
).
Practical Methods in Molecular Biology.
New York
:
Springer-Verlag
.
Serrano
,
L.
,
De La Torre
,
J.
,
Maccioni
,
R. B.
&
Avila
,
J.
(
1984
).
Involvement of the carboxyl-terminal domain of tubulin in the regulation of its assembly
.
Proc. natn. Acad. Sci. U.S.A.
81
,
5989
5993
.
Simon
,
J. A.
,
Sutton
,
C. A.
,
Lobell
,
R. B.
,
Glaser
,
R. L.
&
Lis
,
J. T.
(
1985
).
Determinants of heat shock-induced chromosomes puffing
.
Cell
40
,
805
817
.
Spradling
,
A. C.
&
Rubin
,
G. M.
(
1982
).
Transposition of cloned P elements into Drosophila germ line chromosomes
.
Science
218
,
348
353
.
Theurkauf
,
W. E.
,
Baum
,
H. J.
,
Bo
,
J.
&
Wensink
,
P. C.
(
1986
).
Constitutive and developmentally regulated <r-tubulin genes in Drosophila melanogaster code for structurally distinct proteins
.
Proc. natn. Acad. Sci. U.S.A.
83
,
8477
8481
.
Villasante
,
A.
,
Wang
,
D.
,
Dobner
,
P.
,
Dolph
,
P.
,
Lewis
,
S. A.
&
Cowan
,
N. J.
(
1986
).
Six mouse tubulin mRNAs encode five distinct isotypes: testis specific expression of two sister genes
.
Molec. cell. Biol.
6
,
2409
2419
.
Wigglesworth
,
V. B.
(
1965
).
The Principles of Insect Physiology, pp. 229-230
.
London
:
Methuen
.