Menin Is a Regulator of the Stress Response in Drosophila melanogaster

Abstract

Menin, the product of the multiple endocrine neoplasia type I gene, has been implicated in several biological processes, including the control of gene expression and apoptosis, the modulation of mitogen-activated protein kinase pathways, and DNA damage sensing or repair. In this study, we have investigated the function of menin in the model organism Drosophila melanogaster. We show that Drosophila lines overexpressing menin or an RNA interference for this gene develop normally but are impaired in their response to several stresses, including heat shock, hypoxia, hyperosmolarity and oxidative stress. In the embryo subjected to heat shock, this impairment was characterized by a high degree of developmental arrest and lethality. The overexpression of menin enhanced the expression of HSP70 in embryos and interfered with its down-regulation during recovery at the normal temperature. In contrast, the inhibition of menin with RNA interference reduced the induction of HSP70 and blocked the activation of HSP23 upon heat shock, Menin was recruited to the Hsp70 promoter upon heat shock and menin overexpression stimulated the activity of this promoter in embryos. A 70-kDa inducible form of menin was expressed in response to heat shock, indicating that menin is also regulated in conditions of stress. The induction of HSP70 and HSP23 was markedly reduced or absent in mutant embryos harboring a deletion of the menin gene. These embryos, which did not express the heat shock-inducible form of menin, were also hypersensitive to various conditions of stress. These results suggest a novel role for menin in the control of the stress response and in processes associated with the maintenance of protein integrity.

Multiple endocrine neoplasia type I is an autosomal dominant cancer syndrome characterized by tumors of various glands or disperse endocrine cells (32). The MEN1 gene, identified by positional cloning in 1997, encodes a 610-amino-acid protein of 67 kDa called menin (3, 8, 28). Menin harbors two nuclear localization signals at the C terminus and it has been detected in the nucleus in different cell types (16, 21, 44).

Since menin showed little similarity to proteins of known function, several investigators sought to identify menin-interacting proteins by yeast two-hybrid screens or proteomics approaches. Agarwal and coworkers first identified JunD, but no other components of AP-1, as a menin-interacting protein (2). Menin blocks transcriptional activation without interfering with DNA binding by JunD, a process that may depend on the association of menin with an mSin3A-histone deacetylase complex (2, 26). Several factors of the NF-κB family, including p50 NF-κB1, p52 NF-κB2, and p65 RelA, are also inhibited by menin (18). In contrast, menin interacts with Smad3 and other members of the Smad family, enhancing DNA binding by the Smad3/Smad4 complex and promoting growth inhibition by the transforming growth factor beta pathway (25, 42, 43). Menin is also part of a histone methyltransferase complex involved in the maintenance of Hox gene expression (20, 52). Therefore, menin is an important regulator of gene expression. In agreement with this notion, Elledge and coworkers described the repression of human telomerase promoter activity by menin, suggesting a mechanism by which menin restricts the proliferation of tumor cells (29).

In the last few years, several studies have uncovered a number of other menin-interacting proteins or pathways that suggest a more pleiotropic mode of action for the menin tumor suppressor. For instance, menin associates with the candidate tumor metastasis suppressor Nm23 and defines a novel atypical GTPase for the menin-Nm23 complex (36, 51). Menin acts as a negative regulator of the ERK and JNK pathways and blocks the activation of AP-1 even in the absence of JunD binding (13). Finally, menin binds to the 32-kDa subunit of replication protein A (RPA2) and interacts with the product of the Fanconi anemia predisposition gene FANCD2, suggesting a role for menin in DNA repair or DNA surveillance (24, 46).

In this study, we took advantage of the model organism Drosophila melanogaster to investigate the function of the menin gene. Misexpression of menin or deletion of the menin gene had little effect on development but impaired the ability of Drosophila embryos, larvae, or flies to survive in response to several stresses, including heat shock, hypoxia, hyperosmolarity, and oxidative stress. Proper expression of HSP70 and HSP23 expression was dependent on menin, defining a new function for this protein and indicating that menin is a key regulator of the stress response in Drosophila melanogaster.

MATERIALS AND METHODS

Fly stocks.

The fly stocks used in these studies were maintained on standard yeast-agar media in a 25°C incubator, and include Oregon-R, nos-GAL4 (Bloomington 4442), arm-GAL4, GMR-GAL4, LE-GAL4, tub-GAL4 (Bloomington 5138), Hsp70-GAL4 (Bloomington 2077), Df(2L)spd^j2, wg^spd−j2/CyO (Bloomington 2414), and GE11370 (GenExcel, Inc.).

Lines UAS-Mnn-1, UAS-Human-MEN-1, and UAS-Mnn-1-RNAi were generated by P-element mediated transformation using plasmids described below. The results were confirmed with three and four independent lines for UAS-Mnn-1-RNAi and UAS-Mnn-1, respectively. Results obtained with UAS-Human-MEN-1 were confirmed on two independent lines. Lines Mnn1^e173 and Mnn1^e30, harboring a deletion within the Mnn1 locus, are described below. The hsp70-lacZ reporter line (Bg9L) used in this study was kindly provided by J. Lis (30).

Generation of Drosophila Mnn1 mutants.

A Drosophila line (GE11370, GenExcel, Inc.) with a P-element in exon 1 of Mnn1 was used to generate deletion mutants by imprecise excision of the P-element. From an initial pool of 189 Drosophila lines, screened by PCR amplification of genomic DNA fragments, two lines with a deletion within the Mnn1 locus were identified and characterized by sequencing analysis. Line E30 (renamed Mnn1^e30) harbors a 2,096-bp deletion beginning at the site of insertion of the P-element in exon 1 and extending to the end of exon 4 of Mnn1. A larger 4,216-bp deletion was found in line E173 (Mnn1^e173). In this line, the deletion begins in the 5′-flanking region of Mnn1, 661 bp upstream of the published Mnn1 cDNA sequence (31), and ends at the beginning of intron 4 (data not shown).

DNA constructs.

A DNA fragment containing exons 2 to 5 of the Mnn-1 gene (4411 nucleotides in length) was amplified from Drosophila melanogaster genomic DNA (Canton S) using oligonucleotides designated DMENIN-Gen-5′A (5′-TTA GGT ACC TGG AGT GGC AAT CAG AAG AAC CCC-3′) and DMENIN-Gen-3′A (5′-TTA TCT AGA CGA ACC ACC TTA GCT TGC AC-3′). Commercially available reagents (Expand Long Template PCR system, Roche Diagnostics) were employed for amplification, which was performed with a Perkin Elmer 9600 thermocycler according to the following cycling protocol: one cycle at 94°C for 3 min; 10 cycles at 94°C for 10 seconds, 60°C for 30 seconds, and 68°C for 2 min; 20 cycles at 94°C for 10 seconds, 60°C for 30 seconds, 68°C for 2 min plus an extra 20 seconds at each consecutive cycle; and, finally, one cycle of 72°C for 8 min.

The restriction sites included in these oligonucleotides allowed directional cloning of the resulting DNA fragment in the XhoI and XbaI sites of plasmid pUAST, thus generating plasmid pUAST-Mnn1. Plasmid pUAST-Human-MEN1 was generated by subcloning the EcoRI DNA fragment of plasmid pCMV Sport menin (containing the human MEN1 cDNA, generously provided by Sunita Agarwal) into the EcoRI site of the pUAST vector.

Two DNA fragments were amplified to generate plasmid pUAST-Mnn1-RNAi. First, a fragment corresponding to exon 2 and intron 2 of the Mnn1 gene (fragment C) was amplified from DNA of the Drosophila bacterial artificial chromosome BAC 39J17 (Bloomington Stock Center) using oligonucleotides Mnn1-A-5′ (5′ AATCTCGAGGGCAATCAGAAGAACCC 3′) and Mnn1-C-3′ (5′ TAAGGCCAATGAGGCCGGGACGCTGTTGCC 3′). Second, a fragment consisting of exon 2 only (fragment B) was amplified using oligonucleotides Mnn1-B-5′ (5′ AATTCTAGAGGCAATCAGAAGAACCC 3′) and Mnn1-B-3′ (5′ ATTGGCCACATTGGCCCTAATGTCACCTGCTGCG 3′). Fragments C and B were digested with XhoI and XbaI, respectively, and then with SfiI before directional cloning into the XhoI and XbaI sites of plasmid pUAST. Since fragments C and B include Sfi sites with cDNA overhangs at their 3′ end, the resulting DNA insert contains exon 2 and intron 2 followed by exon 2 of the Mnn1 gene in the reverse orientation. Therefore, it is expected that a double-stranded RNA molecule consisting of a 282-nucleotide stem, corresponding to exon 2, and a 30-nucleotide loop is generated upon splicing of intron 2.

Heat shock experiments.

Stage 6 to 10 embryos (3 to 5 h after egg laying [AEL]) were collected in food vials or grape juice/agar plates, subjected to a 1-h heat shock pulse at 37°C, allowed to recover at 25°C for various periods of time, and used to prepare samples for Western blotting analyses or phenotypical analyses as described below. Embryos collected in vials were allowed to develop until adulthood at 25°C in order to determine viability. Lethality was calculated based on the expected number of mutant progeny relative to control siblings from the same cross. All embryos were heterozygous for the transposable elements indicated in the graph. A minimum of three experimental crosses were carried out with a minimum of 50 progeny counted per cross.

Hypoxia experiments.

Equal numbers of stage 6 to 10 embryos (3 to 5 h AEL) were dechorionated, placed into a specially designed hypoxia chamber, and subjected to hypoxia (5% O₂) for a 2-hour period. At the end of the hypoxic exposure, embryos were allowed to develop and viability was assessed by counting the number of eclosed adults. The experiment was repeated four times with 100 embryos per genotype. The effect of chronic exposure to low oxygen on tracheal branching was carried out on third-instar larvae (L3) according to the protocol of Jarecki et al. (23). Briefly, larvae were grown until the late first-instar stage (L1) and the vials were then placed in sealed chambers containing 5% or 21% O₂ for 40 h at room temperature. The air-filled terminal branches emanating from a Tr3 dorsal terminal branch were viewed under bright-field optics and counted.

High-osmolarity experiments.

Parent flies were fed for 5 days on a diet containing 0.2 M NaCl (22). They were then placed in population cages for egg collection. Equal numbers of embryo progeny were collected from these cages and placed in culture medium containing 0.2 M NaCl. Embryos were then allowed to develop and viability was assessed by counting number of adults eclosed. This experiment was repeated four times with 200 embryos per cross.

Oxidative stress experiments.

Adult males (2 to 3 days old) kept in standard medium were starved for 6 h before being transferred to vials containing 2 ml of special medium consisting of 1% sucrose, 1.3% agar, and 5 mM paraquat (methyl viologen; Sigma M-2254) added when the temperature of the medium was 45°C to avoid loss of oxidative activity (34). Each vial contained 10 males and survival was scored every day for 6 days. This experiment was repeated three times with 100 adults per genotype.

Whole-mount immunocytochemistry.

For visualization of mitotic chromosomes, stage 6 to 10 embryos (3 to 5 h AEL) were subjected to a 1-hour heat shock treatment at 37°C and then allowed to recover for 1.5 h. Embryos not receiving this heat shock treatment or similarly heat shocked wild-type embryos were used as controls. The embryos were then prepared as described previously (45) and labeled with affinity-purified rabbit polyclonal anti-phosphorylated histone H3 antibody (Upstate Biotechnologies, dilution 1:1,000). Images were acquired using a Bio-Rad MRC 600 krypton/argon laser confocal microscope.

Acridine orange staining.

Stage 6 to 10 embryos (3 to 5 h AEL) were subjected to a 1-hour heat shock treatment at 37°C and then allowed to recover for 1.5 h. Embryos of the same genotype not receiving heat shock treatment or heat-stressed wild-type embryos were used as controls. The embryos were then prepared for acridine orange staining and viewed as described previously (1).

Immunohistology and quantitative β-galactosidase activity assays.

Stage 6 to 10 embryos (3 to 5 h AEL) of the hsp70-lacZ reporter strain were heat shocked at 37°C for 1 hour and allowed to recover for an additional 1.5 h before being harvested and stained with 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) (Sigma B-4252). Quantitative measurement of β-galactosidase activity in extracts was performed according to Fridell and Searles (12). Reactions were incubated at 37°C in the dark and substrate conversion was measured at 574 nm using a spectrophotometer 0, 0.5, 1, 1.5, and 2 h after addition of extract. Kinetic analyses of β-galactosidase activity were performed to ensure that the reaction was in the linear range. To control for endogenous β-galactosidase activity, extracts from the wild-type strain were prepared and this background reading was subtracted from readings obtained with each experimental line. The experiment was repeated four times independently.

Quantification of mitotic cells.

To count mitotic cells and metaphase, anaphase, and telophase figures, images of embryos were acquired through 40× and 60× objective lenses on a Bio-Rad MRC 600 krypton/argon laser confocal microscope. Mitotic cells and phases were counted by visual inspection of phosphorylated histone H3-positive labeled chromatin. Each pair of anaphase and telophase figures was counted as one.

Generation of Drosophila menin antibodies and Western blotting analysis.

Two Drosophila menin polyclonal antibodies were generated in rabbits. The first antibody was generated with a glutathione S-transferase (GST) fusion protein containing the conserved region of menin (amino acids 235 to 465). This antibody, numbered 116, is referred to as the core menin antibody. The second antibody was generated by Sigma Genosys using a synthetic peptide corresponding to the last predicted 15 amino acids of the 83-kDa Drosophila menin protein (GDSIAASRPKRTRRE). This antibody is designated the C-terminal peptide antibody and is numbered 5073.

Embryos between 3 and 5 h of development were heat shocked at 37°C for 1 hour and allowed to recover for 0, 1, 3, or 5 h. Embryos not receiving a heat shock treatment or wild-type embryos receiving the same heat shock treatment were used as controls. Embryos were collected and homogenized in 5× volume of SDS sample buffer [2% sodium dodecyl sulfate (SDS), 5% β-mercaptoethanol, 62.5 mM Tris-HCl (pH 6.8), 10% glycerol, 0.001% bromophenol blue, protease inhibitors (complete mini EDTA-free cocktail; Roche Diagnostics) and phosphatase inhibitors (1 mM sodium fluoride, 1 mM sodium vanadate)] and boiled for 5 min. The boiled samples were vortexed and spun at maximum speed in a microcentrifuge for 15 minutes at 4°C.

The supernatant was transferred to a new tube and the protein concentration was determined by the Bradford assay (Bio-Rad); 20 to 30 μg of protein was resolved on a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane (Schleicher & Schuell, BioScience). The membrane was blocked with TBS buffer (20 mM Tris-HCl pH 7.5, 136 mM NaCl) containing 5% nonfat milk for 1 h at room temperature, then incubated in TBST buffer (TBS with 0.1% Triton X-100) at 4°C overnight with anti-Drosophila menin rabbit polyclonal antibody 5073 (Sigma Genosys; 1:2,500 dilution) or 116 (1:2,000 dilution). Alternatively, the expression of heat shock proteins was examined using mouse monoclonal antibodies for HSP70 (7F6, 1:2,000 dilution [48]), HSP83 (3E6, 1:2,000 dilution), HSP23 (7B12, 1:2,000 dilution) and HSP26 (10D3, 1:2,000 dilution). Protein loading was monitored using antibodies for actin (MAB1501, Chemicon International; 1:2,000 dilution). After multiple washes in TBS buffer, the anti-rabbit immunoglobulin G horseradish peroxidase-conjugated secondary antibody (Horseradish Peroxidase; Cell Signaling Technology) or anti-mouse immunoglobulin G-horseradish peroxidase (BD Bioscience) was added at 1:5,000 dilution and incubated for 1 h at room temperature. The ECL Western blotting system (Amersham Pharmacia Biotech) was used to detect the immunoreactive proteins.

Chromatin immunoprecipitation assays.

Collection and formaldehyde cross-linking of embryos were done essentially as described by Cavalli and Paro (7). Dechorionated embryos were suspended in wash buffer 1 (PBST: 0.01% Triton X-100 in PBS) and the embryos were left to settle in a tube. The wash buffer was aspirated and washing was repeated one more time. To the embryos (0.25 g), 2 ml of fixing solution (50 mM HEPES, pH 7.6, 100 mM NaCl, 1 mM EDTA, 2% formaldehyde, 3% methanol, 0.5 mM EGTA) and 6 ml of heptane were added and mixed by shaking vigorously for 15 min. The embryos were collected by centrifugation at 5 000 × g for 5 min. Cross-linking was stopped by washing twice in stop solution (0.125 M glycine in PBST).

Fixed embryos were washed once in wash buffer 2 (50 mM HEPES, pH 7.6, 10 mM EDTA, 0.5 mM EGTA, 0.25% Triton X-100) and once in wash buffer 3 (50 mM HEPES, pH 7.6, 10 mM EDTA, 0.5 mM EGTA, 0.25% Triton X-100, and 200 mM NaCl). The embryos were then resuspended in 1 ml of radioimmunoprecipitation assay buffer (10 mM Tris-HCl, pH 8.0, 140 mM NaCl, 0.1% SDS, 0.1% deoxycholate, and 1.5% Triton X-100) and were sonicated (20-s pulse on and 30-s pulse off for 12 min) in a Heat Systems ultrasonicator with settings at maximum amplitude. The sonicated samples were clarified by centrifugation at 10,000 rpm for 10 min in a Beckman high-speed centrifuge. The supernatant was then diluted with buffer A (10 mM Tris-HCl, pH 8.0, 140 mM NaCl, 1 mM EDTA, and 1% Triton X-100) so that the final concentration of SDS and of deoxycholate was 0.025%.

The lysates were precleared by incubating with protein G-Sepharose for 2 h. The precleared lysates were then immunoprecipitated overnight at 4°C with anti-heat shock factor, anti-Drosophila menin (serum 5073) or the corresponding preimmune antibodies. The immune complexes were collected by incubation with protein G-Sepharose beads for 4 h at 4°C and followed by extensive washing (seven times in buffer B: 50 mM HEPES-KOH pH 7.5, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.5 mM phenylmethylsulfonyl fluoride; once in buffer E: 10 mM Tris-HCl pH 8.0, 0.25 mM lithium chloride, 1 mM EDTA, 0.5% NP-40, and 0.5% sodium deoxycholate; and once in TE: 10 mM Tris-HCl, pH 8.0, and 1 mM EDTA). The cross-linked protein-DNA immune complexes were eluted in 0.1 ml of TE buffer containing 1% SDS at 65°C for 15 min. This step was repeated one more time with 0.15 ml of elution buffer. The protein-DNA cross-links were reversed by incubation at 65°C overnight. DNA was then extracted twice with equal volumes of phenol and chloroform. The DNA was ethanol precipitated in the presence of 0.4 M LiCl and 20 μg of glycogen. The precipitated DNA was washed with ice-cold 70% ethanol, dried, and dissolved in 50 μl of TE buffer.

PCR analysis of immunoprecipitated chromatin samples was done as described by Andrulis and coinvestigators (4) with minor modifications. PCR primers are listed in pairs: HSP70-998 (P3): 5′ GCC CCT TTG AGT TCT AAC CAT CC 3′ and HSP70-572 (P4): 5′ GGT TTC TTT GCT TAA TTA AAC GC 3′; HSP70-365 (P1): 5′ GGC CTT TCT GGC GGA CAA CAT CC 3′; and HSP70+5 (P2): 5′ CGC TCC GTC GAC GAA GCG CC 3′. Typically, a PCR was carried out in a 25-μl reaction volume with 2 μl of the precipitated DNA and 28 cycles of amplification using the conditions of Andrulis and coinvestigators (4).

RESULTS

Misexpression of menin disrupts the Drosophila embryonic response to heat shock.

The function of menin was investigated by overexpressing or down-regulating menin expression in embryos, larvae, and adult flies. To this end, two different constructs of the menin gene (designated Mnn1 in Drosophila melanogaster) were generated with the pUAST vector. Construct UAS-Mnn1 includes 4,411 nucleotides of genomic DNA corresponding to exons 2 to 5 of the Mnn1 gene. This region includes all information coding for menin but does not contain the noncoding exon 1 (17, 31). Second, we generated a construct containing two copies of exon 2 inserted in opposite orientations and separated by intron 2 of the Mnn1 gene (UAS-Mnn1-RNAi). It is expected that a double-stranded RNA hairpin consisting of a 282-nucleotide stem, corresponding to exon 2, and a 30-nucleotide loop are generated upon splicing of intron 2.

These DNA constructs were used to generate transgenic lines by P-element-mediated insertion. These lines were crossed to others harboring the GAL4 trans-activator gene under the control of different promoters and enhancers to obtain the overexpression or down-regulation of menin. The predicted molecular size of Drosophila menin is 83 kDa (17, 31). This was confirmed when the expression of UAS-Mnn1 was activated by the Hsp70-GAL4 driver in embryos subjected to a 1-h heat shock treatment (Fig. 1A, lanes 5 to 8). In contrast, no accumulation of the 83-kDa menin was observed in response to heat shock in Hsp70-GAL4 embryos harboring both the UAS-Mnn1 and UAS-Mnn1-RNAi transgenes, indicating that the Mnn1 RNAi blocked the induction of menin in these conditions (Fig. 1A, lanes 9 to 12).

FIG. 1.

Effects of menin misexpression on the survival of Drosophila under various conditions of stress. (A) Western blotting analysis of menin expression in Hsp70-Gal4 strains harboring the UAS-Mnn1 or both the UAS-Mnn1 and UAS-Mnn1-RNAi transgenes and subjected to heat shock. Stage 6 to 10 embryos (3 to 5 h AEL) were heat shocked (+) or not (−) for 1 h and then returned to the normal temperature for increasing periods of time (Recovery). Upon heat shock and expression of GAL4, the 83-kDa menin protein accumulates in the UAS-Mnn1 strain but not in the strain harboring both the UAS-Mnn1 and UAS-Mnn1-RNAi transgenes. (B and C) Stage 6 to 10 embryos (3 to 5 h AEL), in which menin has been up- or down-regulated were subjected to a 1-h heat shock treatment (B) or 2 h hypoxic environment (C) and then allowed to develop until adulthood. Elevated temperature or oxygen deprivation caused lethality in embryos in which menin was misregulated but not in parental lines. (D) Embryos with abnormal levels of menin gene function were collected and grown in food containing 0.2 M NaCl. Stress caused by hyperosmolarity was lethal in conditions of menin misexpression. (E) Control adult males and males in which menin was up- or down-regulated were placed in vials containing 5 mM paraquat and survival was scored every day for 6 days. In panels B to E, Mnn1 and Mnn1-RNAi refer to strains harboring a UAS construct of these genes, while H-MEN1 corresponds to a UAS strain with the human menin cDNA transgene. Error bars represent the standard error of the mean.

Using the Hsp70-GAL4 driver, the overexpression (UAS-Mnn1) or inhibition of menin (UAS-Mnn1-RNAi) caused embryonic lethality following a 1-h heat shock treatment (Fig. 1B). The same high level of lethality was observed when embryos expressing human instead of Drosophila menin were subjected to a 1-h heat shock treatment (data not shown). Surprisingly, no lethality or disruption of development was observed when GAL4 expression was obtained from a number of constitutive or tissue- or development stage-specific drivers, including nanos (nos), armadillo (arm), tubulin (tub), leading edge (LE-GAL4), and glass (gl) (Fig. 1B and data not shown). Therefore, we reasoned that the combination of heat shock and menin misexpression was responsible for the lethality observed with the Hsp70-GAL4 driver.

In order to test this hypothesis, the UAS-Mnn1 and UAS-Mnn1-RNAi lines were crossed to lines harboring the nos-GAL4, arm-GAL4 or tub-GAL4 gene. Embryo progeny of these crosses, collected 3 to 5 h after egg laying, were subjected to a 1-h heat shock treatment at 37°C. Lethality was observed in a significant percentage of the embryos overexpressing the Mnn1 gene or its RNAi construct but only when coupled with heat stress (Fig. 1B and data not shown).

The lethality consequent to exposure of early embryos to heat shock was due to arrest of embryonic development soon after heat shock (Table 1). Developmental arrest was characterized by a dramatic increase in cell death as seen by acridine orange staining (Fig. 2A to C). Consistent with these findings is the observation that the stereotypical mitotic domains present at that stage, as revealed by anti-phospho-H3 antibody labeling, were severely reduced in heat shock treated embryos with menin misexpression (Fig. 2D to F, Table 2). Therefore, altering the level of menin interfered with the capacity of the embryos to mount a proper heat shock response.

TABLE 1.

Developmental arrest of embryos subjected to heat shock in conditions of menin misexpression^a

Parameter	Genotype
	Hsp70-GAL4; UAS-Mnn-1		Hsp70-GAL4/ UAS-Mnn-1-RNAi		Oregon R		nos-GAL4; UAS-Mnn-1		nos-GAL4/ UAS-Mnn-1-RNAi		nos-GAL4
	Early	Late	Early	Late	Early	Late	Early	Late	Early	Late	Early	Late
No. of embryos	93	51	67	29	0	91	28	49	39	58	0	52
% of arrest (n)	65 (144)		67 (96)		0 (91)		36 (77)		40 (97)		0 (52)

^aEmbryos 3 to 5 h old (stages 6 to 10) were subjected to a 1-h heat shock treatment at 37°C, allowed to develop until they were 15 to 17 h old (stages 16 to 17), and then examined under Nomarski. The term early refers to embryos between stages 6 and 11 (3 to 6 h old), while the term late refers to embryos between stages 16 and 17 (15 to 17 h old). Circa 36% of the embryos derived from females carrying both the nos-GAL4 and UAS-Mnn-1 transgenes and 40% of the embryos derived from females carrying the nos-GAL4 and UAS-Mnn-1-RNAi transgenes showed developmental arrest upon heat shock.

FIG. 2.

Lack of proper menin gene function during the heat shock response causes increased apoptosis and reduced cell proliferation. Stage 6 to 10 embryos of a standard wild-type strain (Oregon R) (A and D) and embryos in which menin was up-regulated (nos-GAL4; UAS-Mnn1) (B and E) or down-regulated (nos-GAL4; UAS-Mnn1-RNAi; panels C and F) were heat shocked for 1 h, allowed to recover for 1.5 h, and then processed for acridine orange staining (AO) (A to C) or anti-phosphorylated histone H3 immunolabeling (PH3; panels D to F). All photomicrographs are lateral views (AO) or ventral views (PH3) of a representative stage 8 embryo. At this stage in development, little acridine orange staining can be seen in wild-type embryos (A). Disruption in Mnn1 expression coupled to heat shock leads to an increase in apoptosis as seen by an increase in acridine orange labeling (B and C). In the control embryos (D), cells of mitotic domains 11 (arrows) and 14 (arrowheads) are seen at various stages of mitosis. In heat-stressed embryos, in which menin is overexpressed (E) or down-regulated (F), a marked reduction in the number of cells labeled was seen in different mitotic domains. Scale bar = 50 μm.

TABLE 2.

Quantitation of mitotic domains under menin misexpression^a

Domain	Parameter	Oregon-R		nos-GAL4; UAS-Mnn-1		nos-GAL4; UAS-Mnn-1-RNAi
		No heat shock	Heat shock	No heat shock	Heat shock	No heat shock	Heat shock
11 (lateral view, stage 8)	No. of labeled mitotic cells	81.4 ± 2.37	79.7 ± 2.48	77.9 ± 1.54	23.5 ± 2.32	79.6 ± 1.67	27.3 ± 2.17
	% Metaphase cells	11.0 ± 0.98	16.5 ± 1.52	12.4 ± 1.16	78.8 ± 3.21	12.5 ± 0.81	85.7 ± 2.3
	% Anaphase cells	46.7 ± 1.73	42.3 ± 1.64	41.9 ± 2.37	11.1 ± 2.45	45.2 ± 2.18	8.9 ± 2.47
	% Telophase cells	41.2 ± 1.58	41 ± 2.59	49 ± 2.70	10.1 ± 1.75	42.3 ± 1.90	5.5 ± 1.70
14 (ventral view, stage 8)	No. of labeled mitotic cells	68.8 ± 1.93	73.1 ± 1.66	73.1 ± 1.22	15.5 ± 1.54	71.4 ± 1.69	14.1 ± 1.43
	% Metaphase cells	12.5 ± 1.59	9.3 ± 0.92	11.5 ± 1.75	85.7 ± 2.74	12.4 ± 1.41	88.6 ± 2.9
	% Anaphase cells	40.1 ± 2.23	36 ± 2.19	45.2 ± 2.30	7.1 ± 1.95	38.1 ± 1.85	4.8 ± 1.43
	% Telophase cells	47.5 ± 2.36	54.8 ± 1.65	43.7 ± 2.74	7.2 ± 2.57	49.4 ± 2.13	6.6 ± 2.23
N (dorsal view, stage 9)	No. of labeled mitotic cells	65.8 ± 2.04	67.4 ± 1.17	64.4 ± 1.87	12.2 ± 1.68	65.1 ± 1.99	12 ± 1.8
	% Metaphase cells	49.2 ± 2.88	53.5 ± 2.33	49.7 ± 1.96	84.6 ± 3.65	52 ± 2.65	89.9 ± 2.43
	% Anaphase cells	37 ± 3.16	32.9 ± 2.01	34.6 ± 1.70	8.8 ± 3.8	34.1 ± 2.06	4.3 ± 1.9
	% Telophase cells	14.1 ± 1.07	13.9 ± 1.62	15.7 ± 2.67	6.8 ± 2.05	13.8 ± 1.69	5.9 ± 2.55

^aWild-type control embryos (Oregon R) and embryos with menin overexpression or down-regulation were heat shocked for 1 hour, allowed to recover at the normal temperature, and fixed and stained with anti-phospho-histone H3 antibody (PH3) to visualize mitotic chromosomes. Embryos that did not receive a heat shock treatment were used as controls. Three different mitotic domains were assessed. For each each data point, mitotic cells from 10 embryos from two separate experiments were counted. In embryos misexpressing menin and subjected to heat shock, cells of the same mitotic domains were reduced in number and appeared to be mostly in metaphase. Data are presented as means±standard errors of the mean.

Proper Mnn-1 gene function is required for the Drosophila response to various stresses.

To determine if the response to other stressors was also altered in conditions of menin overexpression or inhibition, nos-GAL4 and tub-GAL4 embryos harboring the UAS-Mnn1 or UAS-Mnn1-RNAi transgene were subjected to other conditions of stress and then allowed to develop to adulthood. A marked increase in lethality was observed when 3 to 5 h AEL embryos in which the Mnn1 gene was misexpressed, were subjected to hypoxic shock (2 h at 5% O₂) or were allowed to develop in food medium containing high concentrations of NaCl (0.2 M; Fig. 1C and D). To characterize the effects of menin misexpression in response to hypoxia, we examined the pattern of tracheal branching in larvae expressing the Mnn1 RNAi and subjected to continued exposure to low oxygen (23). The expression of the Mnn1 RNAi had no effect on the normal development of the trachea but the induction of lateral branching in larvae submitted to low oxygen was inhibited in conditions of menin down-regulation (Fig. 3 and Table 3).

FIG. 3.

Branching of the trachea in response to hypoxia is impaired in embryos expressing menin double-stranded RNAi. Representative photomicrographs of the dorsal branch (DB) of a Tr3 tracheal metamere are shown for third-instar larvae (L3) reared from the first instar (L1) in normoxic (21% O₂) (A and C) and hypoxic (5% O₂) (B and D) conditions for wild-type L3 (A and B) and L3 harboring a UAS-Mnn-1-RNAi transgene under the control of tub-GAL4 (C and D). Note the increased number and tortuosity of terminal branches in wild-type L3 kept in the hypoxic environment. Hypoxia induced tracheal branching is inhibited by down regulation of the Mnn-1 gene.

TABLE 3.

Quantitation of tracheal terminal branches in conditions of menin down-regulation^a

Genotype	No. of terminal branches (n)
	21% O2	5% O2
Wild-type yw	27.9 ± 0.865 (19)	36.8 ± 1.01 (14)
Wild-type Oregon R	33.1 ± 1.52 (15)	44.7 ± 1.19 (16)
tub-GAL4/UAS-Mnn-1-RNAi	26.1 ± 1.45* (10)	29.6 ± 0.948* (14)
UAS-Mnn-1-RNAi/TM3	30.9 ± 1.18 (10)	41.6 ± 1.43 (12)
yw/tub-GAL4 parental control	25.9 ± 0.929 (11)	32.8 ± 0.961 (11)
yw/TM3 parental control	28.7 ± 0.883 (10)	38.4 ± 1.22 (11)

^aThe number of tracheal terminal branches emanating from a Tr3 DB was determined in L3 submitted to continued exposure to normal (21% O₂) or low levels of oxygen (5% O₂) since late L1. The number of terminal branches (Tr3 DB) is significantly increased by continued exposure to low oxygen in all control genotypes (P < 0.001) except when the Mnn-1 gene was down-regulated by the ubiquitous expression of UAS-Mnn1-RNAi (tub-Gal4; UAS-Mnn1-RNAi, P = 0.062). Student's unpaired t test was used to determine the significance of differences between means. P < 0.05 is considered significant; * denotes not significantly different, P = 0.062.

In vertebrates, menin interacts with a number of proteins that play a role in DNA repair or replication (24, 46). Since the accumulation of reactive oxygen species is known to cause DNA damage, we looked at the effect of menin misexpression in the fly's response to oxidative stress. Sensitivity to the redox-cycling agent paraquat has been used as an indicator of oxygen defense status in a variety of model organisms, including Drosophila melanogaster (19, 33, 35). Therefore, we investigated the impact of paraquat on short-term survival of adult flies in which Mnn1 gene function was up- or down-regulated. Similar to what was found for other environmental stressors, post-eclosion exposure to paraquat dramatically reduced the viability of flies with abnormal expression of the Mnn1 gene (Fig. 1E). Taken together, these results support the notion that proper Mnn1 gene function is required for the response of Drosophila melanogaster to a variety of stressors at different stages of the life cycle.

Proper menin function is required for the expression of HSP70 and HSP23 during the heat shock response.

The response of animals and other organisms to various conditions of stress is characterized by the activation of a specific program of gene expression and the synthesis of heat shock proteins (HSPs). We took advantage of a Drosophila line harboring a copy of the lacZ gene under the control of the Hsp70 promoter (Hsp70-lacZ) (30) to characterize the heat shock response of embryos in conditions of menin misexpression. Embryos with the Hsp70-lacZ reporter gene had little β-galactosidase activity when maintained at the normal temperature (Fig. 4). As expected, a 1-hour heat shock treatment resulted in a marked increase in the level of β-galactosidase detected in these embryos. This level was enhanced by the overexpression of menin. Moreover, in these flies, β-galactosidase levels remained abnormally high during recovery even 3 hours after return to the normal temperature. Similar results were obtained when human menin was overexpressed in place of the Drosophila ortholog. In contrast, the inhibition of menin by RNAi antagonized the induction of β-galactosidase in response to heat shock, suggesting that menin is a regulator of HSP70 expression.

FIG. 4.

Regulation of Hsp70-lacZ transgene by heat shock and menin. Embryos carrying the Hsp70-lacZ and Mnn1 transgenes under the control of the nos-GAL4 driver were heat shocked for 1 h, fixed, and incubated with X-Gal. A representative stage 8 embryo overexpressing Drosophila Mnn1 (UAS-Mnn1) (B) or human MEN-1 (UAS-MEN1 [data not shown]) shows increased expression of the Hsp70 promoter during the heat shock response compared to wild-type control embryos (A). (C) Down-regulation of Mnn-1 results in the attenuation of Hsp70 promoter activity during heat shock (UAS-Mnn1-RNAi). (D) Extracts prepared from flies subjected to heat shock or flies that were not heat shocked (NHS) were used to measure β-galactosidase activity. The β-galactosidase activity was also measured in flies subjected to a 1-h heat shock treatment and then transferred to the normal temperature for 0, 1, or 3 h. Error bars represent standard errors of the mean.

To confirm these results, we looked at the expression of HSP70 and other heat shock proteins in embryos with menin misexpression (Fig. 5A). HSP70 accumulated rapidly in control embryos (nos-GAL4) subjected to a 1-hour heat shock treatment. However, in these embryos HSP70 levels declined during recovery and the protein was undetectable 3 h after return to the normal temperature (Fig. 5A, lanes 1 to 7). High levels of HSP70 were also detected in embryos overexpressing menin. However, HSP70 was not down-regulated in these embryos during recovery (Fig. 5A, lanes 8 to 14). The same applied to HSP23, although in this case full induction of the protein appeared to be attenuated by the overexpression of menin.

FIG. 5.

Regulation of heat shock proteins and menin by heat stress. (A) Role of menin in HSP expression. HSP expression was examined in control nos-GAL4 embryos and embryos with the UAS-Mnn1 or UAS-Mnn1-RNAi transgene. Embryos were heat shocked (+) or not (−) at 37°C for various times (Duration). Alternatively, embryos were heat shocked for 60 min and then returned to the normal temperature for 0 (−), 3, or 5 h (Recovery). The samples in lanes 1, 8, 15, and 22 are from non-heat-shocked embryos collected at the same time as embryos heat shocked for 5 min. The samples in lanes 2, 9, 16, and 23 were from non-heat-shocked embryos aged for 3 h before lysis. The expression of HSP70 and HSP23 is affected by the overexpression or down-regulation of menin. (B) Association of menin with the Hsp70 promoter upon heat shock. The association of menin and heat shock factor with the Hsp70 promoter was examined by chromatin immunoprecipitation. nos-GAL4;UAS-Mnn1 embryos were heat shocked (+) or not (−) and subjected to chromatin immunoprecipitation. Primers P1 and P2 were used to amplify a DNA fragment of circa 370 bp encompassing the heat shock response elements. (C) Characterization of the heat-inducible 70-kDa menin. The expression of the 70-kDa menin in control nos-GAL4 embryos and embryos harboring the UAS-Mnn1 or UAS-Mnn1-RNAi transgene was characterized by Western blotting analysis. Samples in lanes 1 and 2, 8 and 9, and 15 and 17 were from non-heat-shocked embryos lysed without aging (lanes 1, 8, and 15) or aged for 1 h (lane 16) or 3 h (lanes 2, 9, and 17). The expression of the heat-inducible 70-kDa menin is transient in control embryos but persists in embryos of the nos-GAL4; UAS-Mnn1 line upon heat shock. No expression of the 70-kDa menin is detected in the nos-GAL4; UAS-Mnn1-RNAi line.

Other heat shock proteins, such as HSP26 and HSP83, were present in early embryos even in the absence of heat shock and were not affected as drastically by menin (Fig. 5A) (53). These results suggest that menin is a positive regulator of HSP70 expression. This conclusion was supported by the analysis of heat shock proteins in embryos expressing the RNAi for menin. In these conditions, HSP70 was induced early in response to heat shock but did not accumulate to normal levels. Significantly, HSP70 expression was repressed before the end of the heat shock treatment. The induction of HSP23 appeared to be more dependent on menin since it was not detectable in embryos expressing the Mnn1 RNAi (Fig. 5A, lanes 22 to 28).

The regulation of Hsp70-lacZ in strains with menin misexpression suggests that menin modulates the activity of the Hsp70 promoter (Fig. 4). In agreement with this notion, menin was recruited to the Hsp70 promoter in response to heat shock (Fig. 5B). Menin recruitment mirrored that of heat shock factor, whose binding to the heat shock response element is rapidly induced in response to heat shock (37, 39, 41, 49, 54).

Menin expression is regulated by heat shock.

In the course of these studies, we detected the expression of an additional menin protein species of 70-kDa using two different polyclonal antisera generated against different regions of the menin protein (Materials and Methods). Interestingly, the expression of the 70-kDa protein species was altered in response to heat shock (Fig. 5C). Indeed, the 70-kDa menin accumulated in control embryos during heat shock (nos-GAL4) but declined in abundance during recovery at the normal temperature. Significantly, the 70-kDa species was more abundant in embryos overexpressing menin and remained abnormally high during recovery in these embryos. In contrast, no expression of the 70-kDa menin was detected in embryos expressing the Mnn1 RNAi in response to heat shock (Fig. 5C). Taken together these results demonstrate that the expression of menin is regulated during heat shock. Furthermore, they suggest that the 70-kDa menin species, targeted by the RNAi construct, plays a central role in the heat shock response.

Mnn1 mutants are hypersensitive to various conditions of stress.

To confirm the results obtained with the repression of menin by RNAi, we generated two mutant lines harboring a deletion of the Mnn1 locus (Fig. 6A). These lines, designated Mnn1^e30 and Mnn1^e173, were generated by imprecise excision of a P-element inserted in the noncoding exon 1 of Mnn1 in line GE11370 (GenExcel, Inc). Mnn1 mutant embryos developed normally, and the adult flies were fertile in the absence of stress (data not shown). In the experimental conditions used in these studies, embryos and adult flies of the parental GE11370 line behaved like control embryos and flies (Oregon R) and were not sensitive to the effects of heat shock. They also expressed normal levels of the heat-inducible 70-kDa menin at elevated temperatures (data not shown). In contrast, embryos of the two mutant lines were characterized by a high degree of lethality when subjected to heat shock, hypoxia, and oxidative stress, indicating that they were hypersensitive to various conditions of stress (Fig. 6B to D).

FIG. 6.

Mnn1 mutant embryos are hypersensitive to various conditions of stress. (A) Schematic representation of the CG13778 (Mnn1) region in mutant lines Mnn1^e173 and Mnn1^e30 harboring a deletion within Mnn1. The codon for the initiating methionine is located in exon II of Mnn1 (ATG). The position of the P-element in line CG11370 is indicated by the arrow-head in exon I. B and C- Stage 6 to 10 embryos (3 to 5 h AEL) were subjected to a 1-h heat shock treatment (B) or 2 h hypoxic environment (C) and then allowed to develop to adulthood. Elevated temperature or oxygen deprivation caused lethality in embryos harboring a Mnn1 deletion, but not in the wild-type strain (Oregon R). Line Df(2L)spd^j2, wg^spd-j2/CyO harbors a large deletion that includes the Mnn1 gene and it was used in trans with the Mnn-1 deletion mutants in order to confirm that the phenotype observed was due to disruption of the Mnn1 gene and not genetic background. (D) Wild-type (Oregon R) and Mnn1 deletion mutant adult males were placed in vials containing 5 mM paraquat, and survival was scored every day for 6 days. Error bars represent standard errors of the mean.

In order to confirm that this phenotype was in fact due to lack of Mnn1 gene function and not due to genetic background, embryos heterozygous for these deletions and a chromosomal deficiency encompassing the Mnn1 gene (Df[2L]spd^j2) were subjected to the same heat shock treatment and allowed to develop to adulthood. The percentage of lethality for these flies was not significantly different from that of the homozygous mutant carrying either one of the deletion alleles and the chromosomal deficiency (Fig. 6B). These results confirm that, in these strains, the phenotype of stress hypersensitivity maps to the Mnn1 gene. Moreover, these findings indicate that the Mnn1 deletions behave as null alleles.

As described above, the expression of heat shock proteins was characterized in embryos of the Mnn1^e30 and Mnn1^e173 lines. As shown for embryos expressing the Mnn1 RNAi, the expression of HSP70 and 23 was severely impaired in the Mnn1 mutant embryos subjected to heat shock (Fig. 7A and B). In these lines, HSP70 expression was induced early after heat shock but failed to reach the levels observed in control embryos (Fig. 7C). HSP23 induction was also markedly reduced in embryos of the two Mnn1 mutant lines. Consistent with the deletion of a large portion of the Mnn1 gene, embryos of the mutant lines did not express the 70-kDa inducible form of menin in response to heat shock. These results confirm the role of menin in the control of the stress response and the expression of HSP70 and HSP23.

FIG. 7.

Expression of heat shock proteins in mutant Mnn1 embryos. HSPs expression was examined in control Oregon-R embryos and embryos of the mutant Mnn1^e30 (A and C) and Mnn1^e173 (B) lines. Embryos were heat shocked (+) or not (−) at 37°C for various times (Duration). Alternatively, embryos were heat shocked for 60 min and then returned to the normal temperature for 0, 1, or 3 h (Recovery). In panels A and B, samples in lanes 1 and 7 are from non-heat-shocked embryos collected at the same time as embryos heat shocked for 1 h only, i.e., without recovery (0 h). The samples in lanes 2 and 8 were from non-heat-shocked embryos aged for 2 h before lysis, while the samples in lanes 3 and 9 are non-heat-shocked embryos aged for 4 h before lysis. In panel C, the samples in lanes 1 and 8 are of non-heat-shocked embryos of the same age as embryos heat shocked for 5 min, while the samples in lanes 2 and 9 are from non-heat-shocked embryos of the same age as embryos heat shocked for 1 h and transferred to the normal temperature for 3 h (Recovery). The expression of HSP70 and HSP23 is affected by the absence of menin expression. The heat-inducible 70-kDa menin is not expressed by embryos of the two mutant Mnn1 lines.

DISCUSSION

Menin is required for proper expression of heat shock proteins in Drosophila melanogaster.

The misexpression of menin impaired the response of Drosophila embryos, larvae, and flies to heat shock, hypoxia, hyperosmolarity, or oxidative stress (Fig. 1). In embryos subjected to heat shock, this impairment resulted in a high degree of developmental arrest and lethality (Table 1). The overexpression of menin enhanced the expression of HSP70 in response to heat shock and interfered with its repression during recovery at the normal temperature (Fig. 4 and 5). In contrast, the induction of HSP70 and HSP23 was markedly reduced or absent when menin was down-regulated by RNAi or when embryos of two Mnn1 mutant lines were subjected to heat shock (Fig. 5 and 7). These results indicate that menin is a positive regulator of heat shock protein expression in Drosophila melanogaster.

Using Drosophila lines harboring genomic deletions of Hsp70 genes (15), we asked whether a 50% reduction in Hsp70 gene copy number reduced the lethality of heat-shocked flies overexpressing menin but did not observe any changes in the rate of lethality (our unpublished results). These results suggest that Hsp70 genes are not the main target of menin or that menin is a more global regulator of the stress response, controlling the expression of several heat shock response genes, including Hsp70. The observation that HSP23 expression is also altered in conditions of menin misexpression supports the latter possibility (Fig. 5 and 7). It is also possible that the remaining copies of the Hsp70 genes were sufficient to provide levels of HSP70 above the threshold level for lethality caused by menin overexpression in conditions of stress. Therefore, whether or not the lethality observed in conditions of menin misexpression reflects predominantly the aberrant induction of HSP70 remains to be determined.

Menin was recruited to the Hsp70 promoter upon heat shock (Fig. 5B). Since our antibodies recognize both the 83- and 70-kDa forms of menin, the identity of the menin species recruited on the Hsp70 promoter is presently unknown. The results of immunoprecipitation assays indicated that menin does not interact directly with the heat shock factor (our unpublished data). Therefore, menin may be part of a multiprotein complex recruited to the Hsp70 promoter in response to stress. Other proteins, including DAXX, Spt5, Spt6, and elongin A, cooperate with heat shock factor in the regulation of heat shock proteins and, like menin, may play an important role in the survival of the organism facing adverse conditions (4, 5, 14). Since mammalian HSP70 is also part of a negative feedback loop controlling its own expression (40), we also asked whether menin interacts with HSP70 in Drosophila embryos but did not observe any interactions using immunoprecipitation and Western blotting analyses (our unpublished results).

Menin is a component and regulator of the stress response in Drosophila melanogaster.

A 70-kDa form of menin accumulated in response to heat shock, indicating that menin expression is itself regulated in conditions of stress (Fig. 5C). In contrast, the 83-kDa form of menin did not accumulate in control embryos subjected to heat shock (parental strains nos-GAL4 and Hsp70-GAL4) or in UAS-Mnn1 strains harboring the nos-GAL4 driver (Fig. 1A, lanes 1 to 4, and data not shown). Since the Mnn1 RNAi blocked the expression of the 70-kDa menin species and Mnn1 mutant embryos did not express this protein, full induction of HSP70, HSP23, and possibly other heat shock proteins may depend on this heat shock-inducible form of menin. Testing of this hypothesis will depend on the molecular characterization of the 70-kDa menin and awaits the development of reagents and Drosophila strains specific for single menin protein species.

Two alternative forms of menin, differing at the C terminus and encoding proteins 763 and 530 amino acids in length, are predicted by the sequence of existing cDNA clones for Mnn1 (http://flybase.bio.indiana.edu). Since the 70-kDa menin is recognized by antibodies generated against the central domain of the protein or the predicted C terminus of the 83-kDa form of menin (763 amino acids in length), it is not the 530-amino-acid menin protein encoded by the Mnn1-RB transcript. The expression of the 70-kDa menin is presently under investigation.

We did not investigate the expression or mechanism of action of menin in the response to hypoxia, hyperosmolarity, or oxidative stress. Since the expression of heat shock proteins is induced in response to several conditions of stress (11), it is probable that menin exerts a similar effect on the expression of HSP70 and HSP23 in response to these stresses. A recent study concluded that Drosophila larvae and embryos with a mutation of the menin gene are characterized by genome instability and are more susceptible to a variety of chemical mutagens (6). Interestingly, the effect of several of these mutagens was only observed at 29°C. Whether or not these observations reflect the impairment of the stress response caused by the absence of menin remains to be investigated.

The survival of organisms subjected to heat shock depends on the induction of stress-responsive genes. Interfering with the expression or activity of heat shock proteins results in increased lethality in organisms subjected to high temperatures (9, 38). However, precise regulation of these genes is also required for survival since forced expression of heat shock proteins, such as HSP70, is toxic (10, 27). In this report we demonstrate that, in Drosophila melanogaster, proper expression of heat shock proteins depends on menin gene function and that menin is itself regulated by heat shock. Consistent with our findings are recent reports describing a role for the BRCA1 tumor suppressor in the expression of HSP27 in mammals. In addition, BRCA1 was processed and inactivated at high temperatures, indicating that it is also regulated in response to heat stress (50). Similarly, Tid1, the mammalian homolog of the Drosophila tumor suppressor lethal (2) tumorous imaginal disks, is a cochaperone of HSP70 and therefore a candidate regulator of the stress response (47).

Our experiments did not address whether this novel function of menin is relevant for its role as a tumor suppressor in humans. However, they are consistent with the idea that processes involved in the maintenance of protein integrity may also be important for ensuring the integrity of the genome. These processes may depend on a number of common factors, such as menin, BRCA1, and possibly Tid1.