Developmental and Transcriptional Consequences of Mutations in Drosophila TAF_II60

Abstract

In vitro, the TAF_II60 component of the TFIID complex contributes to RNA polymerase II transcription initiation by serving as a coactivator that interacts with specific activator proteins and possibly as a promoter selectivity factor that interacts with the downstream promoter element. In vivo roles for TAF_II60 in metazoan transcription are not as clear. Here we have investigated the developmental and transcriptional requirements for TAF_II60 by analyzing four independent Drosophila melanogaster TAF_II60 mutants. Loss-of-function mutations in Drosophila TAF_II60 result in lethality, indicating that TAF_II60 provides a nonredundant function in vivo. Molecular analysis of TAF_II60 alleles revealed that essential TAF_II60 functions are provided by two evolutionarily conserved regions located in the N-terminal half of the protein. TAF_II60 is required at all stages of Drosophila development, in both germ cells and somatic cells. Expression of TAF_II60 from a transgene rescued the lethality of TAF_II60 mutants and exposed requirements for TAF_II60 during imaginal development, spermatogenesis, and oogenesis. Phenotypes of rescued TAF_II60 mutant flies implicate TAF_II60 in transcriptional mechanisms that regulate cell growth and cell fate specification and suggest that TAF_II60 is a limiting component of the machinery that regulates the transcription of dosage-sensitive genes. Finally, TAF_II60 plays roles in developmental regulation of gene expression that are distinct from those of other TAF_II proteins.

Initiation of transcription by RNA polymerase II (Pol II) in eukaryotic organisms requires assembly of multiprotein complexes at the core promoter of genes (22, 36). Assembly of TFIID is thought to precede and nucleate assembly of the other initiation complexes (TFIIA, TFIIB, TFIIE, TFIIF, and TFIIH) and RNA Pol II. The TFIID complex consists of TATA binding protein (TBP) and 10 to 12 TBP-associated factors (TAF_IIs) (1, 5). Stability of the TFIID complex requires multiple TAF_II-TAF_II and TAF_II-TBP interactions. TAF_II60 binds TAF_II40 and TAF_II250, and elimination of TAF_II60 leads to degradation of other TFIID subunits, suggesting that TAF_II60 interactions in TFIID are important for integrity of TFIID (32, 54). Association of TAF_II60 with TAF_II40 involves histone fold motifs, similar to those of histones H4 and H3, respectively, that cocrystalize in a histone-like structure (57).

TFIID, but not TBP, can mediate activator-directed transcription in an in vitro RNA Pol II system, indicating that one function of TAF_IIs is to respond to enhancer-bound activators (12). TAF_II60 physically interacts with Dorsal, Bicoid, p53, and NF-κB activators in vitro, suggesting that TAF_II60 mediates transcriptional activation by recruiting TFIID to particular promoters (21, 37, 42, 47, 60). Consistent with this proposal, reducing TAF_II60 gene dose in the Drosophila melanogaster embryo alters the pattern of transcription of Dorsal gene targets, twist and snail (37, 60). Drosophila TAF_II60 can also be cross-linked to the downstream promoter element (DPE), a core promoter element located downstream of the transcription start site in many TATA-less promoters, suggesting that TAF_II60 may stabilize the interaction of TFIID with certain promoters, possibly in an activator-dependent manner (7, 8, 28).

The TAF_II60 protein is highly conserved at the primary sequence level in all eukaryotic organisms examined to date (2). In Saccharomyces cerevisiae, TAF_II60 is essential and is required for the transcription of most RNA Pol II genes (32). However, it is difficult to assess how broadly TFIID-bound TAF_II60 functions during transcription, since yeast TAF_II60 is also a component of the SAGA (SPT-ADA-GCN5-acetyltransferase) histone acetyltransferase complex that affects transcription by altering chromatin structure (5, 20). In humans, the homologous HAT complex, PCAF (p300/CREB-binding protein-associated factor), contains a distinct TAF_II60-like protein, PAF65α, and a similar situation may occur in Drosophila, which also encodes a TAF_II60-like protein, TAF_II60-2 (2, 5, 35). Thus, analysis of TAF_II60 in Drosophila may provide a clearer picture of the role TAF_II60 plays as a component of TFIID.

While significant progress has been made in understanding how TAF_II60 contributes to transcriptional activation in vitro, it remains to be determined whether these mechanisms are valid in vivo and whether TAF_II60 functions as a general regulator of transcription in multicellular eukaryotic organisms. To this end, we have examined the phenotypic and transcriptional consequences of mutating, reducing, or eliminating Drosophila TAF_II60 protein in the germ line, in somatic cells, and at various points in development.

MATERIALS AND METHODS

Drosophila stocks and crosses.

Flies were cultured at 25°C on standard medium, unless otherwise noted. Initial characterization of TAF_II60^XS922 has been described previously (25). TAF_II60¹, TAF_II60², and TAF_II60³ alleles were kindly provided by J. Kennison, and P[hsp70-eTBP] transgenic flies were kindly provided by T. Burke and J. Kadonaga (7, 15). The P[hsp70-eTAF60] construct was generated by inserting the TAF_II60 cDNA into the pCaSpeR-hs-FLAG vector. This construct expresses TAF_II60 protein with a FLAG epitope tag on the N terminus (eTAF60). pCaSpeR-hs-FLAG was constructed by inserting the oligonucleotide 5′-AATTCAAAACATGGACTACAAGGACGACGATGACAAGCATATGAATTCGTT-3′ into the EcoRI and HpaI sites of pCaSpeR-hs. P-element-mediated transformation was performed by the method of Rubin and Spradling (40).

Four independent P[hsp70-eTAF60] lines were obtained that expressed eTAF60 and rescued TAF_II60 mutants to adulthood. Rescued TAF_II60 animals were obtained by crossing P[hsp70-eTAF60]/P[hsp70-eTAF60];TAF_II60/TM6B Tb Hu males with P[hsp70-eTAF60]/P[hsp70-eTAF60];TAF_II60/TM6B Tb Hu females and scoring larvae, pupae, or adults for loss of Tb and/or Hu dominant markers or by crossing P[hsp70-eTAF60]/P[hsp70-eTAF60];TAF_II60 red bw/TM6B Tb Hu males with P[hsp70-eTAF60]/P[hsp70-eTAF60];TAF_II60 red bw/TM6B Tb Hu females and scoring larvae for red Malphigian tubules or adults for brown eyes. Approximately 15 males and 15 females were rescue crossed, and the progeny were cultured at 25°C in an air-phase incubator and heat shocked every 8 h at 37°C for at least 10 days. Mitotic clones in the eye were generated using the FLP-FRT system (58). Germ line clones were generated using the ovo^D and the FLP-FDS system (24).

CAT assays.

Chloramphenicol acetyltransferase (CAT) assay samples containing 25 5-day-old adult females of the indicated genotype were homogenized in 500 μl of 250 mM Tris (pH 7.9), freeze-thawed twice in liquid N₂, and incubated at 65°C for 10 min. Insoluble material was pelleted by centrifugation, and five 80-μl aliquots of the supernatant were each combined with 50 μl of 5 mM chloramphenicol, 68 μl of 250 mM Tris (pH 7.9), and 2 μl of ¹⁴C-labeled acetyl coenzyme (0.1 μCi). The mixture was overlaid with scintillation fluid, and the rate of CAT activity was determined by measuring eight time points for 1 min over a period of ∼8 h. CAT activity rates were averaged for the 5 samples. At least four experiments were performed for each genotype.

Western blot analysis.

Adult flies of the indicated genotype and heat shock treatment were homogenized in 1× Laemmli sample buffer (Bio-Rad), boiled for 3 min, and centrifuged for 15 min at 20,000 × g. Embryos of the indicated age were homogenized in buffer containing 50 mM Tris (pH 7.5), 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 μg of leupeptin per ml, 1 μg of pepstatin per ml, and 150 mM NaCl and centrifuged for 15 min at 20,000 × g. Extracts were electrophoresed on sodium dodecyl sulfate (SDS)–10% polyacrylamide gels. Proteins were transferred to Immobilon-P membrane (Millipore), according to standard procedures. Primary antibodies used were anti-FLAG monoclonal antibody M5 (diluted 1:1,000) (Sigma) and anti-TAF_II60 monoclonal antibody (diluted 1:10) (kindly provided by R. Tjian) and were detected using horseradish peroxidase-conjugated goat anti-mouse monoclonal secondary antibody (diluted 1:2,000) (Amersham) by the enhanced chemiluminescence method (Amersham).

Analysis of TAF_II60 mutants.

The coding region and introns of the TAF_II60 gene were sequences by the method of Schlag and Wassarman (43). Adult eyes were fixed and sectioned by the method of Tomlinson and Ready (49). Scanning electron microscopy of adult eyes was performed by the method of Kimmel et al. (27) Wings were mounted in 50% glycerol and imaged by phase-contrast microscopy. Testis squashes were performed by the method of Kemphues et al. (26). Eggs were imaged by dark-field microscopy.

Quantitation of total mRNA levels.

Total RNA was extracted from 30 3-day-old adult males or females of the indicated genotype with TRIzol (Life Technologies), according to the manufacturer's protocol. RNA (10 μg) was denatured in a solution containing 17% formamide, 6% formaldehyde, and 14× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate), split into two samples, and applied to a nitrocellulose membrane (Protran; Schleicher and Schuell) with a manifold apparatus (Manifold I; Schleicher and Schuell). Blots were washed twice with 10× SSC, baked at 80°C for 2 h, prehybridized in a solution containing 10% dextran sulfate, 1 M NaCl, 0.5% SDS, and 1 mg of sheared salmon sperm DNA per ml, and hybridized for 12 to 16 h with ³²P-5′-end-labeled oligo(dT)₂₀ (10⁶ to 10⁷cpm) or a 5S rRNA oligonucleotide (5′-GCCACCGACCATACCACGCTG-3′) probes. Signals were quantitated using a Storm system and ImageQuant program (Molecular Dynamics).

Quantitative PCR.

Total RNA was extracted from testes of the indicated genotype with TRIzol (Life Technologies). Quantitation of mRNA levels was performed using QuantumRNA 18S internal standard, according to the manufacturer's protocol (Ambion). Primers for specific mRNAs are as follows: Ste, 5′-TGCCCACGGTGTAAAAGCAAC-3′ and 5′-GCAGCAGCGAGAAGAAGATGTC-3′; Su(Ste), 5′-TCCCTATGCCTTGATGCCACTC-3′ and 5′-GCTTGGACCGAACACTTTGAAAC-3′; Ssl, 5′-TCCAGGACAAGTTCAATCTGACG-3′ and 5′-ATTCCAATGTGGGGTAGCGGGATG-3′; CK2 (beta subunit), 5′-ACCTGGTTCTGTGGACTTCGTG-3′ and 5′-AACTGATTAGTAGGACGCTTGGGAC-3′; and β2t, 5′-TCTAGATGGCGGCGATGAATAATAG-3′ and 5′-CTCGAG TCGTAACCCAGAAATCACAGC-3′. PCR products were resolved on 6% denaturing polyacrylamide gels and quantitated using a Storm system and ImageQuant program (Molecular Dynamics).

RESULTS

Sev transcription and Rh4 transcription are affected by TAF_II60 gene dose.

We previously identified an X-ray-induced mutant allele of TAF_II60, TAF_II60^XS922, as a dominant suppressor of the rough eye phenotype cause by a sev-Ras1^V12 transgene (25). The Drosophila compound eye is composed of approximately 800 identical units, called ommatidia (56). Each ommatidium contains 8 photoreceptor neurons (R1 to R8) and 12 nonneuronal cells (4 cone cells and 8 pigment cells). During ommatidial assembly, the R7 cell fate is determined by a signal transduction pathway that is initiated by activation of the Sevenless (Sev) receptor tyrosine kinase. Sev activation triggers a signaling cascade mediated by the Ras1 GTPase. Expression of a constitutively active form of Ras1, Ras1^V12, in R7 and cone cell precursors, under control of sev cis-regulatory sequences, bypasses the requirement for Sev activation and transforms cone cell precursors into R7 cells (3, 6, 16). In addition to TAF_II60^XS922, the Df(3L)kto2 deficiency that removes the TAF_II60 gene, dominantly suppresses the sev-Ras1^V12 rough eye phenotype, indicating that TAF_II60^XS922 is a loss-of-function allele (Fig. 1 and data not shown).

FIG. 1.

TAF_II60^XS922 suppresses the rough eye phenotype of sev-Ras1^V12. Scanning electron micrographs of wild-type (A), P[sev-Ras1^V12]/+ (B), and TAF_II60^XS922/P[sev-Ras1^V12] (C) eyes are shown (anterior to the right).

Suppression of the extra R7 cell phenotype by TAF_II60^XS922 was monitored by counting the number of R7 photoreceptors in apical tangential sections of adult fly retinae (Table 1). In the sev-Ras1^V12 background, heterozygous TAF_II60^XS922 flies had approximately twofold-less R7 cells per ommatidium compared to flies that had two wild-type copies of TAF_II60. Thus, mutating one copy of TAF_II60 in a diploid organism improves the regularity of the external eye morphology and reduces the number of extra R7 cells caused by sev-Ras1^V12. Since the rate of cone-to-R7 cell transformation is exquisitely sensitive to the level of Ras1^V12 expression, TAF_II60 mutations may suppress the rough eye phenotype by reducing the level of sev-Ras1^V12 transcription (30). This predicts that TAF_II60 mutations would suppress phenotypes caused by other sev-driven transgenes.

TABLE 1.

Effect of a TAF60 mutation on transcription in the eye^a

Genotype	Ommatidia in sectionsb		Relative CAT activityc
	No. of R7 cells per ommatidium	% wild-type ommatidia	Rh3-CAT/+	Rh4-CAT/+
+/+	3.1 (160)	7.7 (494)	100 (5)	100 (4)
TAF60XS922/+	1.5 (197)	29.7 (475)	94 ± 15 (5)	51 ± 8 (4)

^aFlies of the indicated genotypes were assayed for eye phenotypes or CAT activity 

^bTangential sections of adult retinae were scored for R7 photoreceptors in the sev-Ras1^V12/+ background or wild-type ommatidia in the sev-rough/+ background. The numbers of ommatidia scored are shown in parentheses. Wild-type flies contain one R7 cell per ommatidium. 

^cCAT assays were performed on extracts prepared from adult flies. The CAT activity of control extracts prepared from flies carrying an Rh3-CAT or Rh4-CAT transgene in an otherwise wild-type background was arbitrarily set at 100%, and the activity of flies heterozygous for TAF60^XS922 was normalized to the appropriate control. The number of independent assays is shown in parentheses. See text for details. 

Accordingly, TAF_II60^XS922 dominantly suppressed the R7 to outer photoreceptor cell transformation caused by sev-rough (Table 1) (4, 27). Rough is a homeodomain protein that is expressed in R2, R5, R3, and R4 outer photoreceptor precursor cells and specifies the fate of R2 and R5 (4, 27, 48). Ectopic expression of rough in the R7 precursor cell changes its fate to that of an outer photoreceptor. In the sev-rough background, heterozygous TAF_II60^XS922 flies had approximately fourfold-more wild-type ommatidia than flies that had two wild-type copies of TAF_II60. TAF_II60^XS922 also dominantly suppressed rough eye phenotypes caused by sev-phyl, sev-yan^ACT, and sev-Notch^act/nucl (data not shown) (9, 38, 52). Since sev-Ras1^V12, sev-phyl, sev-rough, sev-yan^ACT, and sev-Notch^act/nucl had different effects on eye development but were similarly affected by a mutation in TAF_II60, suppression may result from reduced levels of transcription from the sev-driven transgenes.

To directly examine the effect of the TAF_II60^XS922 mutation on transcription in the eye, we quantitated the activity of transgenes that express the reporter gene product (CAT) under the regulation of Rh3 or Rh4 enhancer and promoter sequences (17). Rh3 and Rh4 direct expression in nonoverlapping subsets of R7 cells in adult retinae. CAT activities for extracts prepared from Rh3-CAT/+;TAF_II60^XS922/+ flies were similar to those for control extracts prepared from Rh3-CAT/+;+/+ flies (Table 1). However, extracts prepared from Rh4-CAT/+;TAF_II60^XS922/+ flies had approximately twofold-less CAT activity than control extracts prepared from Rh4-CAT/+;+/+ flies. In summary, these data indicate that TAF_II60 is required for transcription of the sev and Rh4 genes but not the Rh3 gene and that the level of TAF_II60 expression is critical for the transcription of some genes in vivo.

Molecular lesions in TAF_II60 that cause lethality.

In addition to TAF_II60^XS922, three other recessive lethal TAF_II60 alleles have been described. TAF_II60¹, TAF_II60², and TAF_II60³ were isolated in a screen for lethal mutations in the 76B chromosomal region (15). To identify regions of the TAF_II60 protein that are important for its function in vivo, we sequenced the TAF_II60 gene from each allele. TAF_II60^XS922 contains an in-frame insertion of 6 bp (TACTAC) that encodes two tyrosine (Y) residues adjacent to position 207 of the protein (Fig. 2A). TAF_II60¹ contains a 29-bp deletion that causes a frameshift at amino acid 298 of the 593-amino-acid TAF_II60 protein. TAF_II60² contains a single missense mutation that changes tryptophan (W) 128 to arginine (R) (Fig. 2B). TAF_II60³ contains a G-to-A transition at the +1 position of the first intron 5′ splice site, which may affect the efficiency of splicing of the TAF_II60 pre-mRNA.

FIG. 2.

TAF_II60 mutations disrupt residues or domains that are conserved in metazoan family members. Sequences of a region surrounding the tryptophan (W)-to-arginine (R) mutation in TAF_II60² (A) and a region surrounding the two tyrosine (Y) insertion in TAF_II60^XS922 (B) from Drosophila melanogaster, Xenopus laevis, Schizosaccharomyces pombe, and Saccharomyces cerevisiae as well as TAF_II60-like proteins, Drosophila melanogaster CG10390 (also denoted TAF_II60-2), and human PAF65α are shown. Residues identical to those in TAF_II60 are indicated in bold type, and the percent sequence identity of each region to Drosophila TAF_II60 is indicated in parentheses to the right of the sequences. Amino acid positions of the ends of each sequence are shown to the left and right of each sequence. Dashes indicate gaps introduced to maximize sequence alignment.

Amino acid changes in TAF_II60^XS922 and TAF_II60² are located within regions that are conserved in metazoan TAF_II60 proteins but not in TAF_II60-like proteins, such as PAF65α and TAF_II60-2 (Fig. 2). This suggests that these regions carry out functions that are unique to TFIID. Furthermore, the TAF_II60^XS922 lesion defines a region that is not conserved in Schizosaccharomyces pombe or Saccharomyces cerevisiae TAF_II60 proteins, possibly defining a function that is particular to higher eukaryotes. Unfortunately, it is difficult to ascribe a function to these regions because they do not display sequence similarity to previously described protein motifs and have not been implicated in specific protein-protein interactions.

TAF_II60 is required throughout Drosophila development.

Embryos transheterozygous for TAF_II60 alleles die during the first or second larval instar stage, with the exception of TAF_II60^XS922/TAF_II60² flies which die late in the pupal stage (Table 2). Survival of TAF_II60 mutants through embryonic and early larval stages is probably dependent on the maternal contribution of TAF_II60 mRNA, and zygotic expression is necessary during the second larval instar. Maternal expression of TAF_II60 appears to be required for oogenesis as homozygous germ line clones of TAF_II60^XS922 failed to produce eggs. TAF_II60 is also required for cell proliferation and/or survival, since mitotic clones of TAF_II60^XS922 generated in the eye were not recovered, unlike the twin spot of wild-type tissue generated by the same recombination event. Thus, TAF_II60 is essential for the development of both germ cells and somatic cells. Furthermore, the essential nature of TAF_II60 suggests that TAF_II60 and TAF_II60-2, a TAF_II60 family member that is ubiquitously expressed in flies (N. Aoyagi and D. A. Wassarman, unpublished observation), have nonredundant functions.

TABLE 2.

Results of TAF60 complementation and lethal phase analyses

Genotype	% Viability by complementation analysisa												Lethal phase analysisb
	TAF602			TAF60XS922			TAF603			TAF601			TAF602			TAF60XS922			TAF603			TAF601
	A	B	C	A	B	C	A	B	C	A	B	C	A	B	C	A	B	C	A	B	C	A	B	C
TAF602	<1	0	<1										EL	EL	EL
TAF60XS922	0	20	23	0	0	6							LP	C	C	EL	EP	C
TAF603	0	0	0	0	19	22	0	0	<1				EL	EL	EL	EL	C	C	EL	EL	EL
TAF601	0	16	18	0	0	15	0	18	16	0	0	<1	EL	C	C	EL	LP	C	EL	C	C	EL	EL	EL

^aPercent viability was determined for each homozygous and transheterozygous combination of TAF60 alleles under the following conditions: A, not heat shocked and lacking the P[hsp70-eTAF60] transgene; B, not heat shocked and has two copies of the P[hsp70-eTAF60] transgene; C, heat shocked and has two copies of the P[hsp70-eTAF60] transgene (heat shock treatment consisted of 30 min at 37°C every 8 h). Percent viability is defined as the number of rescued homozygous or transheterozygous progeny divided by the total number of progeny (150 > n > 1,200). Fully viable combinations should be 33%. <1 denotes that flies were observed at a rate of less than 1%. 

^bThe lethal phase was determined for each homozygous and transheterozygous combination of TAF60 alleles under the same conditions as in footnote a (heat shock and the P[hsp70-eTAF60] transgene). Abbreviations: C, complement; EL, first- or second-instar larval stage; EP, white pupal stage; LP, dark pupal stage. 

The inability to eliminate the maternal contribution of TAF_II60 in the egg or to produce homozygous TAF_II60 mutant cells in adult tissues makes it difficult to determine the developmental and transcriptional requirements for TAF_II60 in the germ line, larvae, pupae, and adults. To circumvent this problem, we attempted to rescue TAF_II60 mutants with an inducible, ubiquitously expressed TAF_II60 transgene. We generated transgenic lines that expressed the TAF_II60 cDNA under control of the heat shock protein 70 (hsp70) promoter. Transgenic TAF_II60 protein (designated eTAF60) was tagged with the FLAG epitope so that it could be distinguished from endogenous TAF_II60 protein. In the absence of heat shock treatment, eTAF60 expression was not detected by Western blot analysis of adult or 12- to 24-h-old embryo extracts but was detected in 0- to 12-h-old embryo extracts, indicating higher levels of leaky expression early in embryonic development (Fig. 3A and C). After a 30-min heat shock at 37°C, eTAF60 was detected with either an anti-FLAG antibody or an anti-TAF_II60 antibody at all stages of development (Fig. 3A and data not shown). Expression of eTAF60 perdured for 8 to 12 h after heat shock treatment of adult flies (Fig. 3B).

FIG. 3.

Expression analysis of transgenic FLAG-tagged TAF_II60 protein (eTAF60). (A) Western blot analysis of eTAF60 from whole-cell extracts of heat-shocked (+) and non-heat-shocked (−) adult flies. Lanes 1, 2, 7, and 8, extracts from flies carrying the P[hsp70-eTAF60] transgene; lanes 3, 4, 9, and 10, extracts from flies carrying the P[hsp70-eTBP] transgene which expresses FLAG-tagged TBP; lanes 5, 6, 11, and 12, extracts from w¹¹¹⁸ flies that do not contain a transgene (7). Lanes 1 to 6 were probed with an antibody to the FLAG epitope (α-FLAG), and lanes 7 to 12 were probed with an antibody to TAF_II60 (α-TAF60). (B) Western blot analysis to examine the perdurance of eTAF60 expression after a 30-min heat shock treatment at 37°C. Extracts were prepared from flies at the times indicated after heat shock and then probed with an α-FLAG or α-TAF60 antibody. (C) Western blot analysis of eTAF60 expression in non-heat-shocked embryos. Blots were probed with an α-FLAG antibody. Lanes 1 to 3 contain increasing amounts of extract from 0- to 12-h-old embryos, and lanes 4 to 6 contain increasing amounts of extract from 12- to 24-h-old embryos.

It is important to note that the level of eTAF60 protein produced after heat shock induction is significantly higher than endogenous TAF_II60 protein level (Fig. 3A, compare lane 8 to lane 12). However, in wild-type flies, expression of eTAF60 throughout development did not appear to cause any abnormalities, suggesting that overexpression of TAF_II60 does not interfere with endogenous transcriptional mechanisms.

In the absence of heat shock treatment, some transheterozygous TAF_II60 mutants survived to adulthood, indicating that low levels of eTAF60 expression are sufficient to complement some mutants. By using a variety of heat shock regimens, we determined that heat shock treatment every 8 h throughout development was necessary to rescue homozygous TAF_II60 adult flies. However, even though this regimen provided a constant source of TAF_II60 protein, the normal number of viable adult flies was not obtained with any transheterozygous mutant combination, indicating that the level or timing of TAF_II60 expression is critical during development. This proposal is consistent with the finding that adult phenotypes are less severe in adult flies that are rescued under heat shock conditions (i.e., high TAF_II60 conditions) than those that are rescued at 25°C (i.e., low TAF_II60 conditions) (Table 2 and Fig. 4 and 5).

FIG. 4.

Phenotypes of rescued TAF_II60 mutant flies. All of the flies shown carry a copy of the P[hsp70-eTAF60] transgene on each of the second chromosomes. (A) A heterozygous heat-shocked TAF_II60^XS922 eye that is phenotypically wild type, (B) a homozygous heat-shocked TAF_II60^XS922 eye, (C) a tangential section of a heat-shocked TAF_II60^XS922 homozygous eye [arrowheads indicate representative photoreceptor rhabdomeres that are larger in mutant ommatidia (−1) than in wild-type (wt) ommatidia], (D) a non-heat-shocked TAF_II60^XS922/TAF_II60² male eye, (E) a heat-shocked TAF_II60^XS922/TAF_II60² male eye, (F) a non-heat-shocked TAF_II60^XS922/TAF_II60² female eye, (G) a heat-shocked TAF_II60^XS922/TAF_II60² female eye, (H) a heat-shocked TAF_II60^XS922 heterozygous wing, (I) a heat-shocked TAF_II60^XS922homozygous wing, (J) a non-heat-shocked TAF_II60^XS922/TAF_II60² male wing, (K) a heat-shocked TAF_II60^XS922/TAF_II60² male wing, (L) a non-heat-shocked TAF_II60^XS922/TAF_II60² female wing, (M) a heat-shocked TAF_II60^XS922/TAF_II60² female wing, and (N, O, and P) melanotic pseudotumors (arrowheads) in heat-shocked homozygous TAF_II60³ third-instar larvae (N and O) and an adult abdomen (P).

FIG. 5.

Testis and egg phenotypes of rescued TAF_II60 mutant flies. The flies shown in panels A to H carry a copy of the P[hsp70-eTAF60] transgene on each of the second chromosomes. The images in panels A to D, E and F, G and H, and I to L were photographed at the same magnification. (A) A heat-shocked TAF_II60^XS922/+ testis that is phenotypically wild type, (B) a heat-shocked TAF_II60^XS922 homozygous testis, (C) a non-heat-shocked TAF_II60^XS922/TAF_II60² testis, (D) a heat-shocked TAF_II60^XS922/TAF_II60² testis, (E) a high-magnification image of primary spermatocytes from heat-shocked TAF_II60^XS922/+ testis, (F) a high-magnification image of primary spermatocytes from heat-shocked TAF_II60^XS922 homozygous testis, (G) a high-magnification image of sperm tails from heat-shocked TAF_II60^XS922/+ testis, and (H) a high-magnification image of sperm tails from heat-shocked TAF_II60^XS922 homozygous testis. In panels A to D, thick white arrows indicate the apical tip of the testis, white lines indicate the region that contains primary spermatocytes, small white arrowheads indicate meiotic cells, large white arrowheads indicate postmeiotic cells, thick black arrows indicate degenerate spermatocytes, and small black arrowheads indicate sperm bundles. In panel F, white arrows indicate crystal structures. In panel H, black arrows indicate aberrant axonemes. (I) A wild-type egg shell. The dorsal pattern is identified by a pair of dorsal appendages separated by a narrow gap. (J to L) Rescued TAF_II60 eggshells that appear dorsalized (J), ventralized (K), and have a mixture of dorsalized and ventralized characteristics (L).

In the case of mutant combinations, such as TAF_II60^XS922/TAF_II60¹, that required heat shock treatment for adult rescue, cessation of heat shock treatment during larval or pupal stages precluded adult rescue, indicating that a high level of TAF_II60 expression is required during these developmental stages. However, cessation of heat shock treatment during the adult stage did not affect the life span of the flies. Thus, the requirement for TAF_II60 appears to be either reduced or dispensable for the postmitotic and fully differentiated cells of the adult fly.

TAF_II60 is required for a variety of developmental events.

Homozygous and transheterozygous TAF_II60 flies that were rescued to adulthood by the eTAF60 transgene, exhibited a common set of developmental phenotypes (Fig. 4 and 5). Abnormalities were exhibited in organs derived from imaginal tissues and in testes and ovaries. Development of these tissues may be more sensitive to the level or timing of TAF_II60 expression, or as anecdotal evidence suggests, the hsp70 promoter may not be efficiently induced in the germ line, which may explain why the P[hsp70-eTAF60] transgene is unable to rescue spermatogenesis or oogenesis. We most thoroughly characterized the phenotypes of rescued homozygous TAF_II60^XS922 and transheterozygous TAF_II60^XS922/TAF_II60² flies because they were the easiest to obtain.

TAF_II60 is required for cell fate specification in the eye.

Rescued TAF_II60 adults had an external rough eye phenotype (Fig. 4, compare panels A and B). The phenotype was less severe if the flies were raised under heat shock conditions (Fig. 4, compare panel D to panel E and panel F to panel G) and was less severe in females than males (Fig. 4, compare panel D to panel F). The rough eye phenotype was due to a change in the number and size (see below) of photoreceptor cells. In wild-type flies, apical tangential sections reveal a trapezoid pattern of photoreceptors with six large outer photoreceptors (R1 to R6) forming the perimeter of the trapezoid and one small inner photoreceptor (R7) in the center of the trapezoid (Fig. 4C) (56). In TAF_II60 rescued flies, ommatidia were either missing an inner or outer photoreceptor or had an extra inner or outer photoreceptor. In flies that had the strongest rough eye phenotype, approximately 25% of ommatidia contained an abnormal number of photoreceptors. Thus, TAF_II60 is involved in specifying photoreceptor cell identities.

TAF_II60 is required for wing development.

Rescued TAF_II60 adults had a notched wing phenotype (Fig. 4, compare panel H to panel I). Regions of the wing margin were most commonly missing at the tip, but wing veins and bristles appeared normal. As occurred in the eye, the wing phenotype was less severe if the flies were raised under heat shock conditions (Fig. 4, compare panels J and K and panels L and M) and was less severe in females than males (Fig. 4, compare panels J and L and panels K and M).

TAF_II60 may regulate cell growth.

Rescued homozygous TAF_II60^XS922, TAF_II60², and TAF_II60³ larvae and adults had dark black masses of cells phenotypically similar to melanotic pseudotumors described in flies mutant for genes that regulate cell growth, dE2F, dDP, S6 ribosomal protein (RpS6), and S21 ribosomal protein (RpS21) (Fig. 4N, O, and P) (39, 46, 50). Melanotic pseudotumors arise in larvae when groups of aberrant cells, often imaginal precursor cells, are recognized by plasmatocytes and lamellocytes in the immune system and are encapsulated in melanized cuticle (10). In TAF_II60 mutants, small pseudotumors were first observed in second-instar larvae and grew as the larvae developed.

In addition, in mutant ommatidia, most rhabdomeres, the light-sensitive membrane of photoreceptors, were larger than in wild-type ommatidia, however, the size of the eye did not appear to be larger (Fig. 4C). This phenotype resembles that of dE2F and Rbf mutants in which the size of individual cells are affected but the size of the organ is not altered (33).

TAF_II60 is required for spermatogenesis.

Rescued TAF_II60 adult males were sterile. To determine the defect that caused sterility, testes were dissected from heat shock rescued TAF_II60 males and analyzed by phase-contrast microscopy (Fig. 5). In wild-type males, the various stages of spermatogenesis are ordered from the tip to the base of the testis (Fig. 5A) (18). At the tip, stem cells divide to produce a spermatogonium cell that undergoes four rounds of mitotic division to produce a cyst of 16 primary spermatocyte cells. After the fourth division, spermatocytes cease mitosis and initiate the meiotic program which contains an extended G₂ phase, during which time cells grow considerably and transcription occurs at a high level. Upon completion of the growth phase, most transcription is shut down and primary spermatocytes undergo meiosis I and II, resulting in a cyst of 64 haploid spermatids. Spermatid bundles then migrate toward the base of the testis and differentiate, which is marked by a number of morphological changes; subcellular compartments are remodeled, a sperm tail containing the axoneme (a microtubule-based organelle for motility) is generated, DNA condenses, and nuclei change shape.

TAF_II60 mutant alleles displayed a range of spermatogenesis defects (Fig. 5). Heat-shocked TAF_II60^XS922/+ flies that carry two copies of the P[hsp70-eTAF60] transgene were phenotypically normal, indicating that expression of the transgene and inactivation of one copy of TAF_II60 does not interfere with spermatogenesis (Fig. 5A). In a more severe case, such as heat-shocked TAF_II60^XS922/TAF_II60² flies, postmeiotic stages, including mature spermatids, were observed but were less abundant than in wild-type flies (Fig. 5, compare panels A and D). Finally, in the most severe case, such as non-heat-shocked TAF_II60^XS922/TAF_II60² flies, proliferation of stem cells to primary spermatocytes was normal (Fig. 5B and C). Primary spermatocytes became mature in size and occupied an abnormally large portion of the testis but degenerated without initiating meiotic chromosome condensation (13). Subsequent stages, beginning with the growth phase, were absent or defective. Thus, transcriptional regulation by TAF_II60 is required during spermatogenesis for meiotic cell cycle progression and spermatid differentiation.

TAF_II60 is required for dorsoventral patterning of the egg.

Rescued TAF_II60 adult females were sterile and laid eggs that had polarity defects. In wild-type eggs, follicle cells on the dorsal side of the egg form two dorsal respiratory appendages called dorsal appendages that serve as a marker for dorsal identity (Fig. 5I) (51). By examination of dorsal appendages, eggs produced by rescued TAF_II60 females had phenotypes that ranged from strongly dorsalized (Fig. 5J), where dorsal appendage tissue was no longer localized to a portion of the eggshell but instead surrounded the egg, to strongly ventralized (Fig. 5K), where dorsal appendages were positioned closer together than in wild-type eggs. Finally, some eggs were small, had thin egg shells, and appeared both dorsalized and ventralized (Fig. 5L). Asymmetric distribution of gurken mRNA and protein to the dorsal side of the egg is critical to the establishment of dorsal follicle cell fates (34). Localized Gurken protein activates the epidermal growth factor receptor triggering a signal transduction pathway that specifies dorsal follicle cell fates (51). By in situ hybridization, we found that gurken mRNA levels and localization were normal in rescued TAF_II60 egg chambers, suggesting that TAF_II60 is required for a downstream event in the pathway (data not shown).

Transcription in TAF_II60 mutants.

Since heterozygous TAF_II60^XS922 mutants dominantly affected the transcription of sev- and Rh4-driven transgenes in the eye, we were interested in determining if TAF_II60^XS922 had a general effect on the transcription of endogenous mRNA-encoding genes. While heterozygous TAF_II60^XS922 mutants appeared phenotypically normal at all stages of development, small changes in transcription may have occurred but were not deleterious to development. As an initial step to address this question, we analyzed the level of poly(A)⁺ mRNA in TAF_II60^XS922/+ mutants versus w¹¹¹⁸ control flies. Total RNA was isolated from TAF_II60^XS922/+ and w¹¹¹⁸ flies and was hybridized with ³²P-labeled oligo(dT) and 5S rRNA probes. The ratio of oligo(dT) to 5S rRNA signal was used as a measure of mRNA level. 5S rRNA, which is transcribed by RNA Pol III and therefore should not be affected by TAF_II60 mutations, served as a control for total RNA levels. This analysis revealed that for both male and female flies, total mRNA levels were statistically similar between TAF_II60^XS922/+ and w¹¹¹⁸ strains (Fig. 6A). Total mRNA levels were also not affected in male or female heterozygous TAF_II60¹ flies. Thus, mRNA levels do not appear to be globally affected by reducing the dose of TAF_II60 twofold. However, the sensitivity of the assay does not exclude the possibility that transcription levels are affected less than twofold, as was observed for Rh4-CAT and genes expressed in testes (Table 1 and Fig. 6B).

FIG. 6.

Analysis of transcription in TAF_II60 mutant flies. (A) Analysis of total mRNA levels in heterozygous TAF_II60 mutants. The ratio of poly(A)⁺ RNA to 5S rRNA was determined for TAF_II60 mutant adult males and females and normalized to that of w¹¹¹⁸ males or females. (B) The transcription level of Ste, Su(Ste), Ssl, CK2, and β2t was determined by quantitative PCR from total mRNA extracted from homozygous TAF_II60^XS922 testes and normalized to w¹¹¹⁸ levels.

Further analysis of spermatogenic stages in rescued TAF_II60 flies revealed phenotypes that suggested specific transcriptional targets for TAF_II60. First, needle-like structures were observed in spermatocytes (compare Fig. 5E to F). The needles were similar in size and shape to those observed in X/O males lacking a Y chromosome and in recessive mutants of Stellate (Ste), Suppressor of Stellate [Su(Ste)], Stellate-interacting gene [sting (sti)], and hsp83 (also known as scratch) (19, 44, 59). Ste mRNA levels are 30- to 70-fold-more abundant in X/O males or males deficient for the Su(Ste) locus than in wild-type males. Overexpression of Ste genes results in self-aggregation of the Ste protein in needle-shaped crystals. The Ste protein is highly similar to the β subunit of Drosophila casein kinase 2 (CK2) and in vitro can interact with the α subunit of casein kinase 2 to form a complex with properties similar to an active α₂β₂ holoenzyme (44). The Su(Ste) and Suppressor of Stellate-like (Ssl) proteins are also highly similar to the β subunit of CK2. This suggests that crystals in TAF_II60 mutants may be due to reduced expression of a CK2 family member.

By quantitative PCR, we examined mRNA levels for Ste, Su(Ste), Ssl, and CK2 in testes from rescued TAF_II60^XS922/TAF_II60² flies relative to w¹¹¹⁸ control flies. Ste mRNA levels increased approximately threefold in TAF_II60 mutant flies, but the level of the other mRNAs was not affected (Fig. 6B). The second phenotype that suggested a gene target for TAF_II60 is that spermatid bundles remained spherical and did elongate (compare Fig. 5G to H). A similar phenotype is observed in mutants of a testis-specific β2-tubulin (β2t) that affect axonemal microtubules of the sperm tail (26). However, quantitative PCR revealed that β2t mRNA levels in TAF_II60^XS922/TAF_II60² flies were not significantly different than in control w¹¹¹⁸ flies. The most-straightforward interpretation of these observations is that reduced levels of TAF_II60 protein either cause minor changes in transcription of the genes we examined or affect other genes.

DISCUSSION

We have presented the following evidence in Drosophila. (i) Two evolutionarily conserved regions of TAF_II60 are critical for viability. (ii) TAF_II60 is essential for development of germ cells and somatic cells. (iii) TAF_II60 is required, to various degrees, during all stages of development. (iv) Imaginal disk development and gametogenesis are particularly sensitive to the level of TAF_II60 expression. (v) Males are more sensitive than females to the level of TAF_II60 expression. (vi) TAF_II60 is limiting in the eye for the transcription of the sev and Rh4 genes. (vii) A twofold reduction in the dose of TAF_II60 does not affect bulk mRNA transcription. (viii) TAF_II60 mutant phenotypes are consistent with roles for TAF_II60 in cell fate specification, cell growth, and cell proliferation.

Based on in vitro defined mechanistic roles for TAF_II60, as a structural component of TFIID, as a coactivator, and as a promoter-interacting factor, one can imagine that the developmental and transcriptional defects observed in TAF_II60 mutants are due to disassembly of TFIID, disruption of interactions with activators, or disruption of interactions with promoters. At present, it is difficult to discriminate between these potential mechanisms, but our data indicate that the defective mechanism can be compensated by overexpression of wild-type TAF_II60 protein. The degree of compensation is correlated to the level of TAF_II60 expression, suggesting that the affinity of genes for TAF_II60 varies greatly in vivo.

Interestingly, overexpression of TAF_II60 (i.e., eTAF60) does not affect development or presumably transcriptional activity in wild-type flies. These data indicate that phenotypes observed in rescued TAF_II60 mutants are not due to excessive levels of TAF_II60 and that TAF_II60 protein that exceeds normal endogenous TAF_II60 levels is either nonfunctional or induces minor changes in transcription that lead to nondetectable phenotypic alterations. These data contradict the observation that overexpression of TAF_II60 or other TAF_IIs in Drosophila or mammalian tissue culture cells modulates transcription directed by specific activators and hints at the existence of an in vivo mechanism that buffers the transcription level of TAF_II-regulated genes (11, 14, 31).

Mutant phenotypes identify dosage-sensitive TAF_II60 gene targets.

TAF_II60 is expressed ubiquitously and is essential for cell survival or proliferation, yet specific developmental pathways were disrupted when TAF_II60 levels were reduced but not eliminated, suggesting that for a subset of TAF_II60-regulated genes, a reduction in the level of transcription of less than twofold has phenotypic consequences. This hypothesis is supported by two lines of evidence. First, only minor changes in transcription levels were observed in TAF_II60 mutants. Suppression of the sev-Ras1^V12 eye phenotype by a TAF_II60 mutation is most likely due to a small change in expression of the sev-Ras1^V12 transgene, as we have previously shown that a twofold reduction in the number of R7 cells in sev-Ras1^V12 flies resulted from a less than twofold reduction in the level of sev-Ras1^V12 transcription (Table 1) (30). Furthermore, a less than twofold change was observed in Rh4 transcription and no change was observed in bulk mRNA transcription or the transcription of numerous genes expressed in testes (Table 1 and Fig. 6).

Second, TAF_II60 mutant males had more severe phenotypes than females, suggesting that TAF_II60 mutations affect the transcription of Y-chromosome genes or dosage-compensated genes in males. In Drosophila, transcription of most X-linked genes in males is increased approximately twofold to compensate for the presence of only a single X chromosome (29). Mutations in components of the dosage compensation machinery cause male-specific lethality. TAF_II60 mutations also cause male lethality. The male/female ratio for rescued homozygous TAF_II60^XS922 flies was 1:4.3 (n = 283). In addition, rescued TAF_II60 mutant males had stronger eye and wing phenotypes than rescue females (Fig. 4). This may be due to downregulation of the sev and Notch genes, which reside on the X chromosome. Sev mutations cause loss of R7 cells in the eye, and Notch mutations cause notches along the wing margin, phenotypes that were stronger in rescued TAF_II60 males than females (Fig. 4). These findings imply that TAF_II60 is required for the twofold upregulation of genes on the X chromosome in males and, more generally, that TAF_II60 is a limiting component of the machinery that regulates the transcription of dosage-sensitive genes.

Does a DPE specify the requirement for TAF_II60?

In the absence of a TATA box, the DPE functions in conjunction with the Inr element for binding of TFIID. Cross-linking experiments have shown that TAF_II60 is in intimate contact with the DPE, suggesting that TAF_II60 mutations would affect transcriptional activation of DPE-containing genes (7, 8, 28). We have found that in TAF_II60 mutant testes, the steady-state transcription levels of the DPE- and Inr-containing genes Su(Ste) and Ssl were not affected and Ste levels increased approximately threefold (Fig. 6B) (28) (http://www.biology.ucsd.edu/labs/Kadonaga/DCPD.html). We have also found that TAF_II60 mutations affect the sev and Rh4 genes but not the Rh3 or β2t genes (Table 1). The sev promoter contains an atypical TATA box (TTAAAA), a consensus Inr element and no DPE, both Rh4 and Rh3 contain consensus TATA boxes and Inr elements and no DPEs, and β2t contains an Inr element but no TATA box or DPE (3, 6, 17, 41). These results suggest that TAF_II60 is not absolutely required for the transcription of DPE-containing genes, a conclusion that was also reached upon analysis of transcriptional defects in TAF_II40 mutant flies (45). Alternatively, the presence or absence of a DPE within these genes may be incorrect because sequence similarity may not accurately predict functional DPEs or the characterized TAF_II60 mutations, which are probably not null mutations, may not affect DPE recognition. The latter alternative, that different TAF_II60 alleles affect different TAF_II60 functions, is consistent with TAF_II60 complementation analysis which showed that TAF_II60 alleles are not equivalent, some combinations can be complemented while others cannot.

Requirements for TAF_II60 overlap with but are distinct from other TAF_IIs.

Our results support the conclusion, drawn from gene expression studies in yeast and biochemical studies in reconstituted transcription systems, that genes differ in their requirement for TAF_IIs. Genetic studies presented here and elsewhere indicate that Drosophila TAF_II60, TAF_II110, and TAF_II250 participate in transcriptional activation of the sev, twist, and snail genes, but differences in phenotypes of TAF_II60 and TAF_II250 mutant flies suggest that the transcription of some genes requires TAF_II60 and TAF_II250 to different extents (37, 53, 60). (i) In TAF_II60 mutant flies, notches occur along the wing margin but wing veins appear normal, while in TAF_II250 mutant flies, deltas form at the distal end of wing veins but the wing margin appears normal. These phenotypes are similar to those of Notch and Delta mutants, respectively. Notch and Delta are components of the Notch signaling pathway that includes many of the relatively few haploinsufficient genes in Drosophila, including Notch and Delta. Thus, TAF_II60 and TAF_II250 may regulate different dose-sensitive genes in the Notch pathway. (ii) TAF_II60 and TAF_II250 are both required for cell fate specification in the eye, but only TAF_II60 mutations affect the size of photoreceptor cells, suggesting a TAF_II60-specific role in regulating genes involved in growth control. (iii) TAF_II60 and TAF_II250 mutants are sterile females, while TAF_II60 mutants are sterile males, suggesting TAF_II60-specific gene targets during spermatogenesis. Recently, a testis-specific isoform of TAF_II80, called Cannonball (Can), has been shown to be required for transcriptional regulation during spermatogenesis (23). can mutations, like TAF_II60 mutations, prevent the initiation of spermatid differentiation, resulting in male sterility (55). Phenotypic similarities between can and TAF_II60 mutants suggest that TAF_II60 is a component of an alternative TFIID complex that plays a role in male germ cell-specific gene expression.