Copyright2000 by the Genetics Society of America
A Novel Drosophila Alkaline Phosphatase Specific to the Ellipsoid Body
of the Adult Brain and the Lower Malpighian (Renal) Tubule
Ming Yao Yang, Zongsheng Wang, Matthew MacPherson, Julian A. T. Dow and Kim Kaiser
Division of Molecular Genetics, University of Glasgow, Glasgow G11 6NU, United Kingdom Manuscript received June 15, 1999
Accepted for publication September 20, 1999
ABSTRACT
Two independent Drosophila melanogaster P{GAL4} enhancer-trap lines revealed identical GAL4-directed expression patterns in the ellipsoid body of the brain and in the Malpighian (renal) tubules in the abdomen. Both P-element insertions mapped to the same chromosomal site (100B2). The genomic locus, as characterized by plasmid rescue of flanking DNA, restriction mapping, and DNA sequencing, revealed the two P{GAL4} elements to be inserted in opposite orientations, only 46 bp apart. Three genes flanking the insertions have been identified. Calcineurin A1 (previously mapped to 21E-F) lies to one side, and two very closely linked genes lie to the other. The nearer encodes Aph-4, the first Drosophila alkaline phosphatase gene to be identified; the more distant gene [l(3)96601] is novel, with a head-elevated expression, and with distant similarity to transcription regulatory elements. Both in situ hybridization with Aph-4 probes and direct histochemical determination of alkaline phosphatase activity precisely matches the enhancer-trap pattern reported by the original lines. Although the P-element insertions are not recessive lethals, they display tubule phenotypes in both heterozygotes and homozygotes. Rates of fluid
secretion in tubules from c507 homozygotes are reduced, both basally, and after stimulation by CAP2b,
cAMP, or Drosophila leucokinin. The P-element insertions also disrupt the expression of Aph-4, causing misexpression in the tubule main segment. This disruption extends to tubule pigmentation, with c507 homozygotes displaying white-like transparent main segments. These results suggest that Aph-4, while possessing a very narrow range of expression, nonetheless plays an important role in epithelial function.
A
LKALINE phosphatase (ALP) is a zinc and mag- strong1972, 1973;Parkashet al. 1993), but none has been previously characterized at the molecular genetic nesium-containing metalloenzyme (EC 3.1.3.1)level. that hydrolyzes phosphate esters with a high pH
opti-The enhancer-trap technique has proved valuable for mum. It is found in most species from bacteria to
hu-identification and isolation of genes with particular spa-mans. In Escherichia coli, ALP (encoded by the gene
tial or temporal patterns of expression (O’Kane and phoA) is found in the periplasmic space. In yeast, ALP
Gehring1987;Bieret al. 1989;Wilsonet al. 1989). We (encoded by the gene PHO8) is found in lysosome-like
have recently used the P{GAL4} enhancer-trap system vacuoles. And in mammals, it is a glycoprotein attached
(Brand and Perrimon 1993) to generate and screen to the membrane by a glycosylphosphatidylinositol
over 1400 independent Drosophila lines for patterns of (GPI) anchor (McCombet al. 1979; Trowsdaleet al.
GAL4-directedb-galactosidase expression in the adult 1990).
brain (Yanget al. 1995) and in the Malpighian (renal) In mammals, four different ALP isozymes are
cur-tubules (So¨ zen et al. 1997). Two lines with identical rently known, each encoded by at least one separate
expression characteristics (c507 and c232) were selected gene. Three are tissue specific: the placental,
placental-by both criteria and analyzed in detail. Expression in like (germ cell), and intestinal isozymes. The fourth
the brain is limited to only a small subset of neurons form, previously known as the liver/bone/kidney
iso-belonging to the ellipsoid body of the central complex, zyme, is tissue nonspecific and has physical and
bio-a structure implicbio-ated in the coordinbio-ation of motor chemical properties that clearly distinguish it from the
activity (Strauss and Heisenberg 1993; Ilius et al. other ALP isozymes (Harris1989;Maneset al. 1990).
1994). Expression in the Malpighian tubules is restricted There are also several independent ALP loci in
Dro-to a zone associated with fluid readsorption (O’Don-sophila melanogaster (Yao1950;BeckmanandJohnson
nellandMaddrell1995;So¨ zen et al. 1997). 1964; Schneiderman et al. 1966; Harper and
Arm-Here we report the characterization of the genomic region encompassing the P{GAL4} elements in lines c507 and c232. We find the expression patterns reported
Corresponding author: Kim Kaiser, Division of Molecular Genetics,
by their LacZ enhancer detectors to correspond to that
University of Glasgow, 56 Dumbarton Rd., Glasgow G11 6NU, United
Kingdom. E-mail: k.kaiser@bio.gla.ac.uk of a nearby ALP gene (Aph-4), implying a defined role
286 M. Y. Yang et al.
Double-stranded sequencing reactions using the dideoxy for this isozyme in specific aspects of neural and renal
chain termination method were carried out as described in function.
the Sequenase Version 2.0 manual (United States Biochemical Corp., Cleveland). The nucleotide sequence of both strands was obtained for all transcribed regions. Intron sequences
MATERIALS AND METHODS were sometimes read on one strand only. DNA sequences
were analyzed using MacVector and AssemblyLIGN (Sequence Drosophila strains: P{GAL4} lines described here were
iso-Analysis Software), BLAST and Prosite (NCBI), GCG
(Wiscon-lated in our laboratory (Yang et al. 1995). The secondary
sin), and SeqVu (Garvan Institute). reporter for GAL4 activity was a second chromosome insertion
Total Drosophila RNA was isolated using the acidic
guanidi-of UASG-lacZ (BrandandPerrimon1993). There is no visible
nium isothiocyanate method (Chomczynski and Sacchi
b-gal expression in the absence of GAL4. P{lacW} lines were
1987). Poly(A)1RNA was selected by oligo(dT)
chromatogra-kindly provided byDea´ket al. (1998) and P{PZ} line l(3)07028
phy, separated in 1.0% 3-[N-morpholino]propane-sulfonic
was kindly provided by A. Spradling’s laboratory. All genetics acid (MOPS)/formaldehyde denaturing gels, transferred to
markers used in the work were described by Lindsleyand nitrocellulose, and probed with radiolabeled cDNA fragments.
Grell(1968). Flies were maintained on standard cornmeal/ Hybridization was carried out at 428in 50% formamide, 53
yeast/agar medium at 258(Ashburner1989).
SSPE, 0.04% bovine serum albumin/0.04%
polyvinylpyrroli-Histochemistry:To obtain sections, recombinant flies
car-done/0.04% Ficoll/0.1% sodium dodecyl sulfate (SDS).
rying P{GAL4} and P{UASG-lacZ} were mounted in “fly collars” These filters were then stripped and reprobed with rp49
(modified from Heisenberg and Boehl 1979), soaked in (O’ConnellandRosbach1984) as a loading control.
OCT embedding medium (Miles, Kankakee, IL) for 10 min In situhybridization:The procedure for in situ hybridization
and then embedded in the OCT medium. Twelve-micrometer to third larval instar polytene chromosomes was essentially as
serial sections of head or body were cut in a cryostat (Anglia described by Pardue (1986). pBluescript and cDNAs were
Scientific) at2188. The sections were stained and mounted labeled with Bio-16-dUTP by nick-translation. Hybridization
as described byYanget al. (1995). Thereafter sections were was detected using 393-diaminobenzidine tetrahydrochloride
examined and photographed on a Nomarski optical micro- (DAB)/H
2O2. After hybridization, the slides were stained with
scope. Giemsa and mounted using DPX mountant.
For whole-mount preparations, brains or Malpighian tu- Nonradioactive in situ hybridization to tissues was carried
bules were dissected in PBS and fixed in 4% paraformaldehyde out essentially as described byNighornet al. (1991). Relevant
for 20 min. They were then washed three times for 20 min cDNA fragments were used as templates for the synthesis of
in PBS and stained with staining buffer and 2% X-gal for 1–2 DIG-labeled single-strand RNA probes according to the
manu-hr at 378. They were then washed for 20 min in PBS, cleared facturer’s instructions (Boehringer Mannheim). Heads and
overnight at 48with PBS/12.5% hydrogen peroxide, washed bodies of wild-type flies were cut into 12-mm sections using a
for 10 min with PBS, dehydrated through graded ethanol, cryostat (Yanget al. 1995) and hybridized with the relevant
and mounted in glycerol gelatin. probe.
For histochemical detection of alkaline phosphatase activity, Hybridization signals were detected by two different
meth-brains or Malpighian tubules were dissected in PBS and fixed ods. (1) Colorimetric detection with NBT and X-phosphate.
in 4% paraformaldehyde for 20 min. They were then washed For immunological detection of the hybridized probes, the
three times for 20 min in PBS and stained with NBT/X-phos in sections were washed in PBT and incubated with 200ml of
digoxigenin (DIG) detection buffer for up to 1 hr (Boehringer 5% (v/v) sheep serum in PAT for 2–3 hr. The sections were
Mannheim, Indianapolis). Finally, they were washed in PBS then incubated with 150–200ml of anti-digoxigenin antibody
and mounted in glycerol gelatin (Sigma, St. Louis). conjugated with alkaline phosphatase (Boehringer
Mann-Assay of tubule secretion phenotype:Tubules were dissected heim) diluted at 1:500 in PAT for 2–3 hr at room temperature
from adult (3–7-day-old) flies and placed in 10 ml drops of or overnight at 48. The sections were washed twice in PBT.
1:1 Schneider’s medium:Drosophila saline under paraffin oil. After extensive washes in 100 mmTris-HCl, pH 9.5, 100 mm
Fluid secretion experiments were performed as described pre- NaCl, 50 mmMgCl2, and 2 mmlevamisole, sections were placed
viously (Dowet al. 1994). Drosophila leucokinin was a kind with 200ml of diluted chromogenic substrate solutions (NBT
gift of J. Veenstra (Bordeaux); CAP2bwas a kind gift of N. J. and BCIP, X-phosphate) following the manufacturer’s
instruc-Tublitz (Oregon). tions (Boehringer Mannheim), and incubated in the dark at
Molecular methods:Genomic sequences flanking the P{GAL4} room temperature for 2–4 hr. The reaction was stopped by
element were cloned by plasmid rescue using standard tech- washing in PBT for 20 min and preparations were mounted
niques (Bieret al. 1989;Wilsonet al. 1989). Briefly, genomic in glycerol gelatin (Sigma). (2) Colorimetric detection with
DNA was digested with PstI to clone sequences “downstream” DAB and H2O2. Sections were washed in PBT and incubated
of the P{GAL4} element and with KpnI to clone sequences with 200 ml of 5% (v/v) sheep serum in PAT for 1–2 hr.
“upstream” of the P{GAL4} element, followed by ligation to They were then incubated with 150–200ml of
anti-digoxigenin-form a plasmid, which was propagated in E. coli NM621. The horseradish peroxidase (anti-DIG-HRP) Fab fragments
(Boeh-plasmids were used to screen a Drosophila genomic DNA ringer Mannheim), diluted at 1:500 in PAT for 2–3 hr at room
library in the vector EMBL3 (K. Kaiser,unpublished results), temperature or overnight at 48. After extensive washes, signal
and a head cDNA library in the vector NM1149 (Russell was detected with diaminobenzidine and hydrogen peroxide
1989). The 59rapid amplification of cDNA ends (RACE) pro- according to the manufacturer’s conditions (Boehringer
cedure was carried out using Superscript RT (Life Technolo- Mannheim). Reactions were stopped by washing in PBT for
gies) according to the conditions recommended by the manu- 20 min and preparations were mounted in glycerol gelatin
facturer. (Sigma).
Southern blotting was carried out essentially as described bySambrooket al. (1989). Hybridization was carried out at
648 in 63SSC, 53Denhardt’s reagent, 0.5% SDS, 100 mg/ RESULTS
ml denatured, fragmented salmon sperm DNA. Filters were
lacZexpression patterns inP{GAL4} lines c507, c232,
washed in 13SSC and 0.1% SDS for 15 min, and then in
Figure 1.—Expression patterns in en-hancer-trap lines and wild-type flies. (a and b) LacZ staining of ellipsoid body ring neu-rons in the adult brain of line c507. (c) Analogous section of a wild-type head show-ing Aph-4 mRNA expression detected by the standard anti-DIG method. (d and e) LacZ staining of the lower Malpighian tubule and ureter in line c507. (f) Analogous section of a wild-type body showing coexpression of Aph-4 RNA as detected by the standard anti-DIG method in the presence of levami-sole. The lower right inset shows the hybrid-ization pattern seen using the DAB/peroxi-dase detection system. (g) LacZ reporter staining of the lower Malpighian tubule and ureter in semilethal line l(3)07028. (h) Or-egon R, showing wild-type natural pigmen-tation pattern of opaque main segments and lower tubules. (i) c507 homozygotes, displaying transparent main segments
(remi-niscent of w2mutants) and opaque lower
tubules. (j) Wild-type tubule alkaline phos-phatase activity is concentrated exclusively in the ureter and lower tubule. (k) Alkaline phosphatase activity is concentrated in the main segment of the tubule in homozygous c507 flies. (l) Alkaline phosphatase activity is concentrated in the ureter, lower tubule, and main segment in heterozygous c507 flies. CB, ring neuron cell bodies; LTR, lat-eral triangle; EB, ellipsoid body; LT, lower Malpighian tubule; U, ureter; M, main seg-ment. Line c232 gives identical staining pat-terns.
identical, and extremely specific, expression patterns. Plasmid rescue of flanking genomic DNA from lines c507, c232, and l(3)96601:The enhancer-trap element GAL4-directedb-gal expression in the adult occurs in
just the two cellular domains shown in Figure 1, a and contains a bacterial plasmid replicon flanked by unique restriction sites, permitting rescue of genomic DNA b. Expression in the brain corresponds to ring neurons
of the ellipsoid body, and more specifically to the R3 from either side of the insertion by appropriate choice of enzyme (Wilsonet al. 1989). Genomic DNAs from and R4 morphological subtypes (Haneschet al. 1989;
Rennet al. 1999). Neurites arising from lateral cell bod- P{GAL4} lines c507 and c232 were digested with PstI, ligated, and rescued to produce downstream (i.e., 39to ies (CBs) extend dendritic arbors within a region known
as the lateral triangle (LTR) and send axonal process GAL4 transcription unit) plasmids pPC507 and pPC232, respectively. Similarly, KpnI was used to obtain the up-toward the midline where in combination they
contrib-ute to the characteristic ellipsoid body structure. In addi- stream (i.e., 59 to GAL4 transcription unit) plasmid pKC507 from line c507 (Figure 3). In similar fashion, tion, both GAL4 lines mark the lower section and ureter
of the Malpighian tubules (Figure 1, d and e), regions genomic DNA from P{lacW} line l(3)96601 was digested with EcoRI or PstI to produce downstream plasmid associated with reabsorption of fluid from the primary
urine (O’Donnell and Maddrell 1995; So¨ zen et al. pE96601 and upstream plasmid pP96601 (Figure 3). Although the P{GAL4} elements in lines c232 and 1997).
In P{PZ} line l(3)07028, b-gal expression is seen in c507 map to the same polytene band and give rise to identical expression patterns, the downstream rescued the nuclei of cells in the lower tubules and ureter
(Fig-ure 1g), as for the two GAL4 lines. No staining is seen fragments are of different size. To examine further the relationship between the two insertions, restriction map-in the adult bramap-in, however. By map-in situ hybridization to
polytene chromosomes, both P{GAL4} insertions were ping and DNA sequencing analysis was performed, re-vealing the two P{GAL4} elements to be inserted 46 bp found to reside at the same cytological location in the
proximal part of 100B (Figure 2), consistent with the apart and in opposite orientation (Figures 3 and 4). Rescued fragments from line c507 were used as localization of 100B2 documented for line l(3)07028
288 M. Y. Yang et al.
Figure2.—(Left) Polytene chromosome in situ hybridization. Arrows indicate hybridization signals in the proximal part of 100B on the third chromosome. (a) A set of chromosomes from line c507 probed with pBluescript, which hybridizes to the plasmid replicon component of the P{GAL4} element. (b–d) Wild-type (Canton-S) chromosomes probed with pMY51 (Aph-4), pMY8 (l(3)96601), and pMY4 (CanA1) cDNA, respectively. (Right) Southern blots of wild-type Drosophila genomic DNA (Canton-S), cleaved and probed as indicated.
These were arranged into a contig by restriction map- to none of the above on the basis of cytological location and thus has been designated Aph-4. Like most other ping and Southern blot analysis (Figure 3).
Isolation of cDNA clones:The rescued fragment from ALPs, Aph-4 has five potential N-linked glycosylation signals, the positions of which vary between species pPC507 contains a 0.9-kb BamHI fragment, which, on
the basis of “reverse Northern” analysis (data not (Maneset al. 1990;Itohet al. 1991). The cDNA does not contain an obvious polyadenylation site, and none shown), appeared to be transcribed. The 0.9-kb
frag-ment was thus used to probe a Drosophila head cDNA is seen in the genomic interval between Aph-4 and the 39end of the l(3)96601 transcription unit (see also Fig-library. Two independent (non-cross-hybridizing) classes
of cDNA were obtained, the longest versions of which ure 3).
Figure 4 also shows the sequence of a genomic interval were designated pMY51 and pMY8 (Figure 3). They
hybridize to nonoverlapping regions of the genome as containing the three P-element insertion sites of P{GAL4} c507, c232, and P{PZ} l(3)07028. Flies homozygous for shown in Figure 3. The head cDNA library was also
probed with a 4.5-kb HindIII fragment from pKC507. the insertions in lines c507 and c232 are viable, whereas homozygotes of l(3)07028 are described as having a The longest cDNA was designated pMY4 (Figure 3).
In situ hybridization to wild-type polytene chromo- semilethal phenotype (BDGP, accession no. AQ-073337). These closely spaced insertions presumably somes with each of the above three cDNAs showed the
respective transcription units to be contained within the disrupt the promoter of the Aph-4 gene to varying ex-tents.
100B2 region (Figure 2, left). Southern blotting further
confirms a single locus for each of the respective genes Among ALPs for which sequence data are available, Aph-4 is most closely related to vertebrate isozymes, frac-(Figure 2, right).
pMY51 encodes an alkaline phosphatase:Figure 4 shows tionally more so to tissue nonspecific varieties (Figures 5 and 6). The closest match was to a chicken nonspecific the nucleotide and deduced polypeptide sequences of
the pMY51 insert and of the 59RACE product derived isoform precursor (Crawford et al. 1995) with 43% identity and 51% similarity (determined using GAP). from pMY51 cDNA. The equivalent region of genomic
DNA was also completely sequenced, showing the gene The only other insect ALP sequence available is from the silkworm Bombyx mori (Itohet al. 1991; 38% identity, to be punctuated by three introns of 1332, 67, and 60
bp, respectively. The 1952 bp cDNA contains a long 47% similarity), presumably a paralogue. Metal and sub-strate phosphate binding sites determined for E. coli open reading frame (ORF) of 578 amino acids. Database
searching showed the deduced polypeptide to be clearly ALP by X-ray crystallography (Sowadski et al. 1985), and which are strongly conserved across phyla, are also a member of the ALP family (see Figures 5 and 6)
containing the highly conserved ALP signature present in Aph-4 (asterisks in Figure 5).
Aph-4expression: Aph-4 mRNA abundance is highest
“VPDSAGTAT.” Three Drosophila ALP loci have been
290 M. Y. Yang et al.
Figure4.—Nucleotide and deduced polypeptide sequences of Aph-4. Uppercase nucleotide sequences are from cDNA, lowercase
from genomic DNA. The 39end of the transcriptionally convergent l(3)96601 gene (pMY8 insert) is shown in italic type. The
range of insects (McCombet al. 1979). A dramatic in- transcription factors that normally specifies the lower tubule domain.
crease of Drosophila ALP activity in third instar larvae,
prior to the secretion of pupal cuticle, has been taken to Fluid secretion phenotypes: Tubules from c507 ho-mozygotes showed consistently lower basal rates of secre-suggest a role for the ALP in cuticle formation, possibly
regulated by the ring gland (Schneidermanet al. 1966). tion than wild-type tubules (Figure 8). They also showed reduced responses to a range of agonists: the cardioac-In situ hybridization to sectioned adult heads and
bodies revealed Aph-4 mRNA within the same cellular tive peptide 2b (CAP2b), which acts on principal cells
to stimulate the apical V-ATPase through intracellular domains as revealed by GAL4-directedb-gal expression,
i.e., the ellipsoid body (Figure 1, b and c) and the Mal- nitric oxide/cyclic GMP (Davieset al. 1997; Figure 8A); cyclic AMP, which acts on principal cells to stimulate pighian tubules (Figure 1, e and f). Since the detection
of DIG-labeled probes usually relies on an ALP reporter the apical V-ATPase (O’Donnellet al. 1996; Figure 8B); and Drosophila leucokinin, which acts on stellate cells enzyme coupled to anti-DIG antibodies, fresh levamisole
was used to block endogenous phosphatase expression to stimulate fluid production by activating the chloride shunt conductance through intracellular Ca21 (O’Don-(McCombet al. 1979), even though ALP activity in any
case appeared to be lost during hybridization (data not nellet al. 1998; Figure 8C). As the lower tubule does not participate in the fluid secretion assay, these results shown). An independent detection system, employing
diaminobenzidine/peroxidase staining to detect a per- must be interpreted in terms of the misexpression of Aph-4 in c507 homozygotes to the tubule main segment. oxidase/anti-DIG reporter gave the same result (Figure
1f, lower right inset). This expression may disrupt the tight coupling of metab-olism and transport at either the basal or apical
mem-The c507 insertion disrupts normal tubule
pigmenta-tion:Normal tubules in wild-type fly (Oregon-R) have branes.
pMY8 encodes an essential polypeptide of unknown
an opaque, yellowish color throughout the main
seg-ment and lower tubules (Figure 1h), caused by the accu- function: pMY8 corresponds to a single genetic locus in the 100B2 region (Figure 2). The pMY8 insert is 897 mulation of hydroxy-kynurenine and associated eye
pig-ments, through the action of the white gene product bp long, excluding the 18-bp polyA tail remnant (Figure 9). A partial cDNA sequence from the Berkeley Dro-and other organic solute transporters (Wessing and
Eichelberg1978). In c507 homozygotes, this pigmen- sophila Genome Project (GenBank accession no. AA696267) provided six additional residues at the 59 tation is strikingly suppressed in the main segment, but
not in the lower tubule (Figure 1i). A possible explana- end. Genomic DNA sequencing shows the gene to be punctuated by two introns of 87 bp and 265 bp, the tion is an effect of the Aph-4 promoter on expression
of the mini-white gene contained within the nearby en- first of which lies 38 bp upstream of the putative transla-tion start site. The sequence TATAAA is 84 bp upstream hancer-trap element (X chromosomes of c507 and c232
carry the disfunctional w1118 allele). A powerful lower of the beginning of the poly(A) tract. This is not the most common of poly(A) addition signals, but it is used tubule enhancer/main segment repressor element
might influence not just the lacZ reporter, but also mini- occasionally (Manley1988). The pMY8 cDNA contains an ORF of 223 amino acids. Database searches revealed white, in a lower-tubule-specific pattern.
The c507 insertion causes misexpression ofAph-4to no obviously related genes, with the possible exception of three mammalian expressed sequence tags (ESTs)
a genetically distinct domain:We found wild-type tubule
alkaline phosphatase activity to be concentrated exclu- revealed by a TBLASTN search (GenBank accession nos. AA097076, AA098325, and R86167). The two likeliest sively in the ureter and lower tubule (Figure 1j),
identi-cal to the three enhancer-trap patterns (Figure 1, d possibilities are a transcription factor and a neuropep-tide gene. The N-terminal part of the deduced protein and g). The Aph-4 pattern in c507 homozygotes was
profoundly different, however (Figure 1k). Expression contains a number of repeats (KPAA and similar) associ-ated with histone-like DNA regulatory proteins, which in the lower tubule is reduced compared to wild type,
while that in the main segment is greatly enhanced are found in a range of organisms. The protein also contains several dibasic repeats (at aa 33-4, 130-1, and (Figure 1, j and k). c507 heterozygotes have an
interme-diate pattern with expression concentrated in lower tu- 149-50) that could result in propeptide cleavage, with a downstream “GG” motif (at aa 93-4) normally associated bule and ureter, but with some expression in main
seg-ment (Figure 1l). The main segseg-ment is a genetic domain with N-terminal amidation.
Dea´ket al. (1998) have generated a large collection defined bySo¨ zenet al. (1997), which is known to
corre-spond to the fluid-secreting portion of the tubule of third chromosome recessive lethal lines by insertion of a P{lacW} enhancer-trap element. Using the plasmid (O’Donnell andMaddrell1995). This suggests that
the c507 insertion is impinging on a control region rescue method ofGuoet al. (1996), we found two lines (96601 and 35313) with P{lacW} insertions at the same that specifies tubule expression domains. The fact that
misexpression in the main segment corresponds so pre- nucleotide position and in the same orientation, just 59 to the pMY8 coding region (Figure 9). Both lines carry cisely to the domains described bySo¨ zenet al. (1997)
the pupal-to-pharate adult stage (Dea´ket al. 1998). One of the lines, l(3)96601, and its corresponding rescued plasmid, p96601 (Figure 3), were chosen for further characterization.
Excision of the P element caused reversion of the lethal phenotype, verifying insertion as the cause of lethality (rather than an unlinked event elsewhere on the third chromosome). Southern blotting of l(3)96601/wt DNA revealed the expected band shift due to P{lacW} insertion (not shown). Northern blotting of wild-type mRNA revealed an adult transcript ofz1 kb that is slightly elevated in the head (Figure 7, mid-dle). Northern blotting of l(3)96601/wt mRNA indi-cated a reduction of transcript abundance of 43% (nor-malized to rp49 loading control) compared with wild type (Figure 7, right). In situ hybridization to sections of the adult head and body revealed generalized expres-sion throughout the organism.
pMY4 encodes calcineurin A1:pMY4 corresponds to a single genetic locus in the 100B2 region (Figure 2). The sequence of the pMY4 insert revealed it to be a
Figure6.—Neighbor-joining plot showing phylogenetic re-partial (59-truncated) Drosophila calcineurin A1 (CanA1)
lationships between representative ALPs. Protein accession cDNA (Gueriniet al. 1992). This was surprising, since numbers are shown in parentheses.
CanA1 had been previously assigned to the cytogenetic location 21E-F (Guerini et al. 1992). There is no
evi-in situ hybridization to tissue sections and staevi-inevi-ing for dence from either Southern blotting or in situ
hybridiza-alkaline phosphatase activity show Aph-4 to have the tion to polytene chromosomes (Figure 2) that there are
same expression pattern as the two enhancer-trap lines. two such genes in the Drosophila genome. However,
Aph-4 is expressed strongly, if not exclusively, in the 21E-F and 100B are at similar positions to the telomeres
of 2L and 3R, respectively, and so we suggest that 100B2 is the true location. In support of this, two recently sequenced BDGP P1 clones (GenBank accession nos. AC007975 and AC008220) that have been mapped to 100A-C contain sequence identical to the published CanA1 cDNA.
DISCUSSION
We have identified a number of genes flanking the P{GAL4} elements in the enhancer-trap lines c507 and c232. Three different genes located at 100B were cloned and characterized. One of them is calcineurin A1, a Ca21/calmodulin-stimulated protein phosphatase
(Gue-rini et al. 1992), which is located upstream of the Figure 7.—Northern blot analysis. A total of 10 mg of
poly(A)1Canton-S RNA from the indicated stages and tissues
P{GAL4} element in line c507. Two other closely linked
was hybridized with the indicated cDNAs. As a control for genes are located downstream of the c507 element,
loading, stripped filters were reprobed with rp49 cDNA Aph-4 being closest to the site of transposon insertion. (O’Connell and Rosbach 1984). E, 0–24 hr embryo; L, Tightly linked to Aph-4, but in opposite orientation, is mixed larval instars; P, midpupal; H, adult head; B, adult body;
wt, wild type; lacW, line l(3)96601. the essential l(3)96601 locus of unknown function. Both
Figure 5.—Pileup of representative ALP protein sequences. Boxed residues are at least 50% conserved between the seven ALPs, shaded residues at least 80% similar. * indicates residues that participate in the active site of E. coli ALP. Drosophila Aph-4
(X98402); chicken nonspecific isozyme precursor (U19108;Crawfordet al. 1995); human nonspecific isozyme precursor (P05186;
Weisset al. 1986); human placental type 1 isozyme precursor (P05187;Knollet al. 1988); human placental-like isozyme precursor
(P10696;MillanandManes1988); human intestinal isozyme precursor (P09923;Henthornet al. 1987); silk moth
294 M. Y. Yang et al.
ellipsoid body of the brain and in the Malpighian (re- vertants were generated and analyzed. One line has a large deletion that removes part of the P element and nal) tubule.
The P{GAL4} insertions in lines c507 and c232 are flanking genomic DNA including the Aph-4, l(3)96601, and the nearby dbt (double-time, Kloss et al. 1998; see homozygous viable, and there are no other obvious
de-fects. To generate loss-of-function alleles, imprecise Figure 3). Although flies homozygous for this deletion P-element excision has been carried out. Over 130 re- die at the late embryo or early larval stage, both l(3)96601 and dbt are essential loci, so no conclusion can be drawn concerning Aph-4. However, a recently documented line, l(3)07028 (BDGP accession no. AQ073337; see Figure 4), in which P{PZ} is inserted 19 bp closer to the 59end of Aph-4 than the c232 insertion, has a semilethal phenotype. This may imply that Aph-4 is an essential, or at least an important, Drosophila gene. Reverse genetic analysis in most model organisms is complicated by the “phenotype gap,” a term that ac-knowledges the dearth of phenotypes for detailed func-tional analysis. This is particularly acute in Drosophila, where the small size of the organism militates against physiological study. However, the identification of cell-specific expression of Aph-4 in the Malpighian tubule (Skaer 1993) allows a more detailed examination of this enigmatic enzyme than would normally be possible and has provided novel insights into the tubule itself. First, the expression pattern delineated by the reporter genes in c507, c232, and l(3)07028 defines a lower tu-bule domain (So¨ zen et al. 1997) that had not been detected by classical analysis (WessingandEichelberg 1978), but which we now know corresponds with the reabsorptive, rather than secretory, portion of the tu-bule (O’Donnell and Maddrell 1995). Second, re-porter gene expression appears precisely to match the expression of Aph-4 mRNA (by in situ hybridization) and of alkaline phosphatase enzyme activity (by histo-chemistry) to a population of 2560.4 cells. It is rela-tively rare for an enhancer-trap pattern to display such close agreement with the gene into which it inserts, and this is particularly interesting in a region where the gene density is so high, with two other genes within a kilobase, each with very different expression patterns. Third, the gene is probably essential because one of three inser-tions in the upstream sequence is semilethal. Fourth, the phenotypes of these insertions are unusual; rather than producing nulls or hypomorphs, they appear to interrupt a control region that specifies the lower tubule domain in some combinatorial fashion. Although ex-pression of alkaline phosphatase in lower tubule is
re-Figure8.—Effect of homozygosity for P element c507 on fluid secretion. Secretion rates were measured in Oregon-R (empty square) or c507/c507 homozygous flies. Tubules were challenged at 30 min with maximal concentrations of agonists
known to act on principal or stellate cells. (A) 1025m
cardioac-tive peptide 2b (CAP2b); (B) 1023mcyclic AMP; (C) 1025m
Drosophila leucokinin. Data are presented as mean6SEM,
Figure 9.—Nucleotide sequence of the l(3)96601 gene. Uppercase nucleotide sequences are from cDNA, lowercase from genomic DNA. The hypothetical 223-amino-acid translation product is indicated. The putative polyadenylation signal sequence
is underlined. The 59end of the transcriptionally convergent dbt gene (Klosset al. 1998;B. Kloss,personal communication)
and the 39end of the transcriptionally convergent Aph-4 gene (pMY51 insert) are shown in italic. The cDNA sequence has been
deposited in the EMBL/GenBank/DDBJ database and assigned accession no. X94917.
duced, it is greatly enhanced in the main segment. Our phosphatase, which is usually a membrane-associated enzyme, may disrupt this order and so nonspecifically best explanation of this at present is that the boundary
between main and lower tubule domains is sharpened reduce the performance of the tissue, both of agonists that usually act on the prinicpal cell (CAP2band cAMP)
by interacting positive and negative regulatory elements
and that the P-element insertions separate these control and of those that act on the stellate cell (Drosophila leucokinin).
regions so that they can no longer interact. The more
distal, lower-tubule-specifying control region is thus What insights can we gain as to the role of alkaline phosphatase in tubule, or indeed in brain? In humans, moved 8 kb away from the more proximal, main
seg-ment control region and can no longer repress it. As there are multiple isozymes of ALP (McComb et al. 1979); the function of these is not well understood. there are a few kilobases between the CanA1 and Aph-4
coding regions (Figure 3), this might be a particularly However, ALP is a marker of osteoblast lineages, and hypophosphatasia, a recessive human disease with multi-rewarding area in which to perform a promoter analysis.
How can the ectopic expression of alkaline phospha- ple bone disorders, has been linked to mutations in ALPL, the non-tissue-specific isozyme of ALP (Henth-tase lead to a reduction in fluid secretion? In the fluid
secretion assay, the lower tubule sits outside the bathing ornandWhyte1992). The analogous gene in mouse, tissue-nonspecific alkaline phosphatase (TNAP), has drop, and so dies; it acts merely as a conduit for the
urine secreted by the main segment. The loss of alkaline been inactivated by homologous recombination (Way-mireet al. 1995). In this case, bone disorders were not phosphatase activity from the lower tubule is thus
un-likely to influence measured rates of fluid secretion, evident, but the homozygous mutant mice died from seizures, apparently from an inability to metabolize vita-and so overexpression in the main segment is more
likely to be responsible. The main segment is a highly min B6 (Waymireet al. 1995). This probably reflects a role for ALPL/TNAP as an ectoenzyme, anchored to ordered tissue, with semicrystalline arrays of V-ATPase
on the apical membrane of principal cells (reviewed the cell surface, that converts pyridoxal-59-phosphate to pyridoxal. The suggestion from the vertebrate literature byDow et al. 1998), while the basolateral membrane
296 M. Y. Yang et al.
tubules are sites for deposits of calcium and magnesium other phosphoric monoester hydrolase genes will be found near the 100B region.
phosphates, in the form of luminal concretion bodies,
or spherites (Wessing and Eichelberg 1978). These We thank Greg Stewart, Ping Li, Elaine Cleary, Ji Luo, Douglas bodies have traditionally been considered to be sites of Armstrong, Shireen Davies, and Jonathan Sheps for their help with this work. We also thank Peter Dea´k for P{lacW} lines, Nicole Mozden
“storage excretion” of phosphates of calcium,
magne-and Allan Spradling for line l(3)07028, Brian Kloss for dbt, Claude
sium, and other metals. As these spherites pack into the
Klee for CanA1. This work was supported by the U.K. Biotechnology
initial segments of the anterior tubules, it is possible and Biological Sciences Research Council and the Medical Research that the deposits are laid down in the initial segment Council, and by the Wellcome Trust.
and then pass though the tubule to be excreted in the urine. Alkaline phosphatase on the apical surface of the lower tubule might allow for the selective reclamation
LITERATURE CITED of calcium when needed. This might explain the
semile-Ashburner, M.,1989 Drosophila: A Laboratory Manual. Cold Spring
thal nature of the mutations; the severity of disruption
Harbor Laboratory Press, Cold Spring Harbor, NY.
of calcium metabolism might depend on the nature of Beckman, L.,andF. M. Johnson,1964 Variations in larval alkaline the diet. phosphatase controlled by Aph alleles in Drosophila melanogaster.
Genetics 49: 829–835.
Alkaline phosphatase activities in vertebrate brain are
Bier, E., H. Vaessin, S. Shepherd, K. Lee, K. McCallet al., 1989
generally lower than those in liver and kidney (McComb Searching for pattern and mutation in the Drosophila genome et al. 1979); however, in neuronal tissue, intense alkaline with a P-lacZ vector. Genes Dev. 3: 1273–1287.
Brand, A. H.,andN. Perrimon,1993 Targeted gene expression as a
phosphatase activity has been located in neuronal cell
means of altering cell fates and generating dominant phenotypes.
bodies and processes in some parts of central and pe- Development 118: 401–415.
ripheral nervous systems of various animals such as Chomczynski, P.,andN. Sacchi,1987 Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform
ex-mouse (Sood and Mulchandani 1977), rat (Nandy
traction. Anal. Biochem. 162: 156–159.
andBourne1963), guinea pig (Songet al. 1994), mon- Chung, H. J., C. ShafferandR. MacIntyre,1996 Molecular char-key (Friede 1966), and fish (Sood and Sinha 1983). acterization of the lysosomal acid phosphatase from Drosophila
melanogaster. Mol. Gen. Genet. 250: 635–646.
At the ultrastructural level, the enzyme has been
local-Crawford, K., H. Weissig, F. Binette, J. L. MillanandP. F.
Goe-ized on the outer surface of plasma membranes of nerve tinck, 1995 Tissue-nonspecific alkaline phosphatase
partici-pates in the establishment and growth of feather germs in
embry-cell bodies and dendrites (Mayaharaet al. 1967;Mori
onic chick skin cultures. Dev. Dyn. 204: 48–56.
and Nagano 1985) and postsynaptic membrane and
Davies, S. A., E. J. Stewart, G. R. Heusmann, N. J. V. Skaerand
synaptic vesicles (SugimuraandMizutani1979). Bio- S. H. P. Maddrell,1997 Neuropeptide stimulation of the nitric
oxide signalling pathway in Drosophila melanogaster Malpighian
chemical studies also suggest that alkaline phosphatase
tubules. Am. J. Physiol. 273: R823–R827.
is associated with synaptic vesicles isolated from bovine
Dea´k, P., M. M. Omar, R. D. C. Saunders, M. Pal, O. Komonyiet
cerebral cortex (ZisapelandHaklai1980). In view of al., 1998 P element insertion alleles of essential genes on the
third chromosome of Drosophila melanogaster: correlation of
physi-these facts that the ALP reactivity has been localized to
cal and cytogenetics maps in chromosomal region 86E-87F.
Ge-discrete subsets of neurons in different species, it is
netics 147: 1697–1722.
suggested that the ALP has a function related to particu- Dow, J. A. T., S. H. P. Maddrell, A. Gortz, N. J. V. Skaer, S. Brogan
et al., 1994 The Malpighian tubules of Drosophila melanogaster:
lar features of the reactive neurons. From the available
a novel phenotype for studies of fluid secretion and its control.
evidence, it appears that the ALP may play some role
J. Exp. Biol. 197: 421–428.
in transmembrane transport (possibly of calcium) and Dow, J. A. T., S. A. DaviesandM. A. So¨ zen,1998 Fluid secretion by the Drosophila melanogaster Malpighian tubule. Amer. Zool. 38:
cell differentiation or proliferation in the nervous tissue.
450–460.
Although the full size of the ALP gene family in
Dro-Friede, R. L.,1966 A quantitative mapping of alkaline phosphatase
sophila is not yet known, we can hypothesize that the in the brain of the rhesus monkey. J. Neurochem. 13: 197–203.
Frisardi, M. C.,andR. J. MacIntyre,1984 Position effect
variega-vertebrate neuronal and kidney roles, associated with
tion of an acid phosphatase gene in Drosophila melanogaster. Mol.
expression of the liver/bone/kidney (tissue nonspe- Gen. Genet. 197: 403–413.
cific) isozyme, are both played by transcripts of the Guerini, D., C. MontellandC. B. Klee,1992 Molecular cloning and characterization of the genes encoding the two subunits of
analogous Drosophila gene, Aph-4. It may also be of
Drosophila melanogaster calcineurin. J. Biol. Chem. 267: 22542–
interest that alkaline phosphatase (EC. 3.1.3.1) and pro- 22549.
tein phosphatase (also called calcineurin) (EC. Guo, Y., A. Gillan, T. To¨ ro¨ k, I. Kiss, A. T. J. Dowet al., 1996 Site-selected mutagenesis of the Drosophila second chromosome via
3.1.3.16) genes are located in the same chromosomal
plasmid rescue of lethal P-element insertions. Genome Res. 6:
region at 100B. Another kind of phosphatase, acid phos- 972–979.
phatase (EC. 3.1.3.2), is located at 99C5-7 of the same Hanesch, U., K.-F. FischbachandM. Heisenberg,1989 Neuronal architecture of the central complex in Drosophila melanogaster.
chromosome based on deletion mapping (Frisardiand
Cell Tissue Res. 257: 343–366.
MacIntyre1984;Chunget al. 1996). All the above three Harper, R. A.,andF. B. Armstrong,1972 Alkaline phosphatase phosphatases are classified as “phosphoric monoester of Drosophila melanogaster. I. Partial purification and
characteriza-tion. Biochem. Genet. 6: 75–82.
hydrolases” (International Union of Biochemistry
Harper, R. A.,andF. B. Armstrong,1973 Alkaline phosphatase
1979). A similar clustering has been observed for human of Drosophila melanogaster. II. Biochemical comparison among four
allelic forms. Biochem. Genet. 10: 29–38.
Harris, H.,1989 The human alkaline phosphatases: what we know and nuclei, pp. 111–137 in Drosophila: A Practical Approach, edited byD. B. Roberts.IRL Press, Oxford.
and what we don’t know. Clinica Chimica Acta 186: 133–150.
Parkash, R., I. SharmaandT. Neena,1993 Enzyme polymorphism
Heisenberg, M.,andK. Boehl,1979 Isolation of anatomical brain
in Drosophila melanogaster. Dros. Inf. Serv. 72: 156–157. mutants of Drosophila by histological means. Z. Naturforsch. 34:
Renn, S. C. P., J. D. Armstrong, M. Y. Yang, Z. Wang, X. Anet al.,
143–147.
1999 Genetic analysis of the Drosophila ellipsoid body neuropil:
Henthorn, P. S.,andM. P. Whyte,1992 Missense mutations of
organization and development of the central complex. J. Neuro-the tissue-nonspecific alkaline phosphatase gene in
hypophos-biol. 41: 189–207. phatasia. Clin. Chem. 38: 2501–2505.
Russell, S. R. H.,1989 Ph.D. Thesis, University of Glasgow.
Henthorn, P. S., M. Raducha, Y. H. Edwards, M. J. Weiss, C.
Sambrook, J., E. F. FritschandT. Maniatis,1989 Molecular
Clon-Slaughteret al., 1987 Nucleotide and amino acid sequences
ing: A Laboratory Manual. Cold Spring Harbor Laboratory Press,
of human intestinal alkaline phosphatase: close homology to
Cold Spring Harbor, NY. placental alkaline phosphatase. Proc. Natl. Acad. Sci. USA 84:
Schneiderman, H. A., W. J. YoungandB. Childs,1966 Patterns 1234–1238.
of alkaline phosphatase in developing Drosophila. Science 151:
Ilius, M., R. WolfandM. Heisenberg,1994 Central complex of
461–463.
Drosophila melanogaster is involved in flight control: studies on
Schubiger, M., Y. Feng, D. M. FambroughandJ. Palka,1994 A mutants and mosaics of the gene ellipsoid body open. J.
Neuro-mutation of the Drosophila sodium pump alpha subunit gene genet. 9: 189–206.
results in bang-sensitive paralysis. Neuron 12: 373–381.
International Union of Biochemistry,1979 Enzyme Nomenclature
Skaer, H.,1993 The alimentary canal, pp. 941–1012 in The
Develop-1978: Recommendations of the Nomenclature Committee of the
Interna-ment of Drosophila melanogaster, vol. 2, edited byM. BateandA.
tional Union of Biochemistry on the Nomenclature and Classification of
Martinez-Arias. Cold Spring Harbor Laboratory Press, Cold
Enzymes. Academic Press, New York.
Spring Harbor, NY.
Itoh, M., S. Takeda, H. Yamamoto, S. Izumi, S. Tominoet al., 1991
Song, Z-M., S. J. H. BrookesandM. Costa,1994 Characterization Cloning and sequence analysis of membrane-bound alkaline
of alkaline phosphatase-reactive neurons in the guinea-pig small phosphatase cDNA of the silkworm, Bombyx mori. Biochim.
Bio-intestine. Neuroscience 63: 1153–1167. phys. Acta 1129: 135–138.
Sood, P. P.,andM. Mulchandani,1977 A comparative
histoenzy-Kloss, B., J. L. Price, L. Saez, J. Blau, A. Rothenfluhet al., 1998
mological mapping on the distribution of acid and alkaline phos-The Drosophila clock gene double-time encodes a protein closely
phatase and succinic dehydrogenase in the spinal cord and me-related to human casein kinase Iepsilon. Cell 94: 97–107.
dulla oblongata of mouse. Acta Histochem. 60: 180–203.
Knoll, B. J., K. N. RothblumandM. Longley,1988 Nucleotide
Sood, P. P., andS. P. Sinha, 1983 A comparative histochemical sequence of the human placental alkaline phosphatase gene.
study of alkaline phosphatase and acetylcholinesterase in the Evolution of the 59flanking region by deletion/substitution. J.
hind-brain of Channa punctatus and Heteropneustes fossilis. Folia Biol. Chem. 263: 12020–12027.
Histochem. Cytochem. 21: 107–114.
Lindsley, D. L.,andE. H. Grell,1968 Genetic Variations in Drosophila
Sowadski, J. M., M. D. Handschumacher, H. M. Kirshna Murthy,
Melanogaster. Carnegie Institute of Washington, Washington, DC.
B. A. FosterandH. W. Wyckoff,1985 Refined structure of
Manes, T., K. Glade, C. A. ZiomekandJ. L. Millan,1990 Genomic
alkaline phosphatase from Escherichia coli at 2.8 A resolution. structure and comparision of mouse tissue-specific alkaline
phos-J. Mol. Biol. 186: 417–433. phatase genes. Genomics 8: 541–554.
So¨ zen, M. A., J. D. Armstrong, M. Y. Yang, K. KaiserandJ. A. T. Manley, J. L.,1988 Polyadenylation of mRNA precursors. Biochim.
Dow,1997 Functional domains are specified to single-cell reso-Biophys. Acta 950: 1–12.
lution in a Drosophila epithelium. Proc. Natl. Acad. Sci. USA 94:
Mayahara, H., H. Hirano, T. SaitoandK. Ogawa,1967 The new
5207–5212. lead citrate method for the ultracytochemical demonstration of
Strauss, R.,andM. Heisenberg,1993 A higher control center of activity of non-specific alkaline phosphatase. Histochemie 11:
locomotor behavior in the Drosophila brain. J. Neurosci. 13: 88–96.
1852–1861.
McComb, R. B., G. N. Bowers, Jr.and S. Posen, 1979 Alkaline
Sugimura K.,andA. Mizutani,1979 Histochemical and
cytochemi-Phosphatase. Plenum, New York.
cal studies of alkaline phosphatase activity in the synapses of rat
Millan, J. L.,andT. Manes,1988 Seminoma-derived Nagao isozyme
brain. Histochemistry 61: 123–129. is encoded by a germ-cell alkaline phosphatase gene. Proc. Natl.
Trowsdale, J., D. Martin, D. BicknellandI. Campbell,1990
Al-Acad. Sci. USA 85: 3024–3028. kaline phosphatases. Biochem. Soc. Trans. 18: 178–180.
Mori, S.,andM. Nagano,1985 Electron-microscopic cytochemistry Waymire, K. G., J. D. Mahuren, J. M. Jaje, T. R. Guilarte, S. P. of alkaline phosphatase activity in endothelium pericytes and
Coburn et al., 1995 Mice lacking tissue non-specific alkaline oligodendrocytes in the rat brain. Histochemistry 82: 225–231. phosphatase die from seizures due to defective metabolism of
Nandy, K.,andG. H. Bourne,1963 Histochemistry of alkaline phos- vitamin B-6. Nature Genet. 11: 45–51.
phatase in brain and spinal cord. Nature 200: 1216–1217. Weiss, M. J., P. S. Henthorn, M. A. Lafferty, C. Slaughter, M.
Nighorn, A., M. J. HealyandR. L. Davis,1991 The cyclic AMP Raduchaet al., 1986 Isolation and characterization of a cDNA phosphodiesterase encoded by the Drosophila dunce gene is encoding a human liver/bone/kidney-type alkaline phosphatase. concentrated in the mushroom body neuropil. Neuron 6: 455– Proc. Natl. Acad. Sci. USA 83: 7182–7186.
467. Wessing, A.,andD. Eichelberg,1978 Malpighian tubules, rectal
O’Connell, P.,andM. Rosbach, 1984 Sequence, structure and papillae and excretion, pp. 1–42 in The Genetics and Biology of codon preference of the Drosophila ribosomal protein 49 gene. Drosophila, edited byA. AshburnerandT. R. F. Wright.
Aca-Nucleic Acids Res. 12: 5495–5513. demic Press, London.
O’Donnell, M. J.,andS. H. P. Maddrell,1995 Fluid reabsorption Wilson, C., P. K. Pearson, H. J. Bellen, C. J. O’Kane, U. Gross-and ion transport by the lower Malpighian tubules of adult female niklaus et al., 1989 P-element-mediated enhancer detection: Drosophila. J. Exp. Biol. 198: 1647–1653. isolation and characterization of developmentally regulated
O’Donnell, M. J., J. A. T. Dow, G. R. Huesmann, N. J. Tublitzand genes in Drosophila. Genes Dev. 3: 1301–1313.
S. H. P. Maddrell,1996 Separate control of anion and cation Yang, M. Y., J. D. Armstrong, I. Vilinsky, N. J. StrausfeldandK. transport in Malpighian tubules of Drosophila melanogaster. J. Exp. Kaiser,1995 Subdivision of the Drosophila mushroom bodies
Biol. 199: 1163–1175. by enhancer-trap expression patterns. Neuron 15: 45–54.
O’Donnell, M. J., R. Rheault, S. A. Davies, P. Rosay, B. J. Harvey Yao, T.,1950 Localization of alkaline phosphatase during
post-em-et al., 1998 Hormonally-controlled chloride movement across bryonic development of Drosophila melanogaster. Q. J. Microsc. Sci. Drosophila tubules is via ion channel in stellate cells. Am. J. 91:89–105.
Physiol. 43: R1039–R1049. Zisapel, N.,andR. Haklai,1980 Localization of an alkaline
phos-O’Kane, C.,andW. Gehring,1987 Detection in situ of genomic phatase and other synaptic vesicle proteins. Neuroscience 5: regulatory elements in Drosophila. Proc. Natl. Acad. Sci. USA 2297–2303.
84:9123–9127.
