Index INTRODUCTION MATERIALS AND METHODS Construction of transgenic fly lines Determination of viability and life span Quantification of mRNA RESULTS CG5216 (dSir2) CG5085 CG6284 DISCUSSION References

INTRODUCTION

Sir2, one of the components of the Sir complex that forms heterochromatin in S. cerevisiae, is an NAD-dependent histone deacetylase. Its functions include silencing at the mating loci and telomeres, repressing the recombination of rDNA, and regulating mitosis and meiosis (Guarente, 1999; Imai et al., 2000; Moazed, 2001; Garcia and Pillus, 1999). Sir2 also functions as an ADP-ribosyltransferase (Tanny et al., 1999). Furthermore, it has been reported that Sir2 is implicated in life span determination; the life span of a SIR2-deleted yeast mutant was found to be decreased by 50%, while that of a mutant with an additional copy of SIR2 was increased by approximately 30% (Kaeberlein et al., 1999). It has also been shown that Sir2 affects life span extension by calorie restriction (Lin et al., 2000).

Sir2 proteins are found in a range of organisms from archaebacteria to humans (Brachmann et al., 1995; Frye, 2000). The Sir2 family is classified into five groups, Classes I to IV and U, according to the amino acid sequences of the core domain, which are highly conserved (Frye, 2000). This domain is implicated in histone deacetylase activity (Imai et al., 2000). The function of regions other than the core domain has not yet been clarified. S. cerevisiae has four Sir2-like proteins from Class I. C. elegans has four Sir2-like proteins from Classes I, II, and IV. However, while a Class I gene in C. elegans, sir-2.1, which is most homologous to yeast Sir2, has been shown to extend its life span up to 50% when duplicated, other Sir2-like genes do not appear to extend life spans (Tissenbaum and Guarente, 2001). On the other hand, it has been reported that mammals have seven Sir2-like proteins. SIRT1 protein, a mouse homolog of yeast Sir2, has been shown to play an important role in embryonic development (McBurney et al., 2003; Cheng et al., 2003). It is interesting that Sir2 regulates the replicative life span of mother cells in yeast, the postmitotic life span in the soma of adult worms, and the immediate postnatal life span of mice.

D. melanogaster has been reported to have a Sir2 and four Sir2-like proteins, each belonging to a phylogenetically different class (Frye, 2000). They are CG5216 (dSir2, Class Ia), CG5085 (Class Ib), CG3187 (Class II), CG6284 (Class IVa), and CG11305 (Class IVb). It has been indicated that CG5216 (dSir2), the closest homolog of yeast Sir2, functions as an NAD-dependent histone deacetylase (Barlow et al., 2001). Also, dSir2 has been shown to be involved in position effect variegation and heterochromatic silencing (Rosenberg and Parkhurst, 2002), as well as Polycomb silencing (Furuyama et al., 2004). It has recently been reported that increased expression of dSir2 extends life span (Rogina and Helfand, 2004). Increased dSir2 transcription has been found in rpd3 (histone deacetylase) mutant flies and also in wild-type flies with calorie restriction, both of which are long-lived (Rogina et al., 2002). Similarly, it has been reported that the life span of dSir2-null mutants is decreased (Åström et al., 2003).

The genes that have been found to affect life spans in yeast, nematodes, and flies are all Class Ia genes. It has not been known whether Sir2-like genes of other classes also affect life spans. CG5216 (dSir2) shows 44% identity to yeast Sir2 in the conserved core domain, but CG6284 shows only 26% identity (Fig. 1). To investigate whether both Sir2 and Sir2-like genes affect life spans in Drosophila, we examined the life spans of flies in which Sir2 and Sir2-like gene expression was suppressed by RNA interference (RNAi) combined with the GAL4/UAS system, which has been called inducible RNAi (Fortier and Belote, 2000; Tavernarakis et al., 2000).

View Details

Fig. 1. Schematic representation of the structure of the Sir2 and Sir2-like genes. The conserved core domains are shaded; amino acid identities with the S. cerevisiae SIR2 core domain are indicated.

In this study, we established RNAi fly lines for three genes, CG5216 (dSir2), CG5085, and CG6284, and examined the viability and life spans for individuals in which the expression of these genes was suppressed. Our results indicate that the reduced expression of both Sir2 and Sir2-like genes affects life span and development.

MATERIALS AND METHODS

Construction of transgenic fly lines

Plasmid DNA for establishing RNA interference (RNAi) transgenic fly lines of CG5085, CG5216, and CG6284 were constructed by using modified pUAST-D13 or pUAST-R57 transformation vectors. These vectors differed in orientation on a part of the sequence. Part of the cDNA or genomic DNA of each gene was amplified by PCR; a 500-bp cDNA fragment of CG5216 (nucleotide position 468–967 of the mRNA sequence) was amplified from the cDNA clone LD07439 (BDGP Clone), and a 548-bp genomic DNA of CG5085 (nucleotide position 723–1271 of the mRNA sequence) and a 527-bp genomic DNA of CG6284 (nucleotide position 11–537 of the mRNA sequence) were amplified from total DNA. Extraction of total DNA was performed according to standard methods. PCR was carried out under the following conditions: for CG5085, 3 min at 94°C, then 1 min at 94°C, 1 min at 57°C, and 1 min at 72°C for 30 cycles, and then 7 min at 72°C; for CG6284, 3 min at 94°C, then 1 min at 94°C, 1 min at 61°C, and 30 s at 72°C for 30 cycles, and then 7 min at 72°C; for CG5216, 2 min at 98°C, then 30 s at 98°C, 30 s at 50°C, and 2 min at 72°C for 25 cycles, and then 2 min at 72°C. Each of the amplified fragments was inserted as an inverted repeat (IR) into the modified transformation vector as described (Pili-Floury et al., 2004). In all cases, the inverted repeats were constructed in a head-to-head orientation.

Transformation of Drosophila was carried out using the w¹¹¹⁸ strain, according to standard methods. Seven, nine, and 13 transgenic lines were generated for CG5216, CG5085, and CG6284, respectively. The transgenic lines were named UAS-CG number + IR + line number: e.g., UAS-CG5216IR1.

Determination of viability and life span

To induce RNAi, Act5C-GAL4 and elav-GAL4 lines were used as drivers. Transgenic flies were crossed to flies of the GAL4 line at either 25°C or 28°C, and the F₁ heterozygotes of Act5C-GAL4/UAS-IR or elav-GAL4/UAS-IR were used for the life span experiments. The lines in which the IR insertion homozygotes were viable were used to determine their life span, since their survival indicated that the insertion itself was not responsible for the F₁ lethality. Virgin females and males of the F₁ progeny were maintained separately on standard Drosophila medium at initial densities of 20 and 25 flies per standard vial, respectively, at 25°C or 28°C. Flies were passed into new vials every 2–3 days, and the number of dead flies was recorded until all flies were dead. Survivorships were plotted to construct survival curves, and mean life spans were calculated as the age in days required to reach 50% survivorship.

Quantification of mRNA

Total RNA was extracted from single F₁ adult flies within 24 hrs of eclosion (0 day) using TRIzol Reagent (Invitrogen). cDNAs were synthesized using the SuperScrip^TM First-Strand Synthesis System for RT-PCR (Invitrogen), and PCR was performed using LUX primers and Platinum Quantitative PCR SuperMix-UDG (Invitrogen) according to the manufacturer’s protocol. The primers used were as follows: 5’-gaaacctggcgtgtgcatggtttc-3’(labeled) and 5’-cggaatgccagcagatgtag-3’(unlabeled) for CG5085, and 5’-cacccgaaaacgacctcgaaatgggtg-3’(labeled) and 5’-aaagtggttccaagcgcaat-3’(unlabeled) for CG6284. The amount of rp49 mRNA amplified by the primers 5’-ctaagcgccatttgtgcgacagcttag-3’(labeled) and 5’-caggcccaagatcgtgaaga-3’(unlabeled) was used as the standard. Amplifications were performed under the following conditions: 2 min at 50°C, 2 min at 95°C, then 15 s at 95°C, 30 s at 67°C, and 30 s at 72°C for 45 cycles, using an ABI PRISM 7700 Sequence Detection System (Applied Biosystems).

RESULTS

To determine the effects of the loss or decrease of Sir2 or Sir2-like gene functions, UAS-IR transgenic flies were crossed with flies of GAL4 lines at 25°C or 28°C. Two kinds of GAL4 lines were used in the present study: Act5C-GAL4, which induces UAS-IR expression ubiquitously, and elav-GAL4, which induces UAS-IR expression only in neurons. For CG5216, CG5085, and CG6284, Act5C-GAL4/UAS-IR heterozygotes were examined for their viability, life span, and amounts of mRNA. For CG5216, elav-GAL4/UAS-IR heterozygotes were also examined for their viability and life span.

CG5216 (dSir2)

UAS-CG5216IR transgenic flies were crossed with Act5C-GAL4 flies. For all seven transgenic lines for CG5216 examined, the Act5C-GAL4/UAS-CG5216IR heterozygotes was lethal, such that eggs could not develop into larvae at 25°C (Table 1). When UAS-CG5216IR transgenic flies were crossed with elav-GAL4 flies at 28°C, elav-GAL4/UAS-CG5216IR heterozygotes were viable for all three lines examined (Table 1). For this reason, we used elav-GAL4 instead of Act5C-GAL4 for the following life span experiment.

View Details

Table 1. Viability of induced RNAi flies

For two lines, one with IR insertion into the chromosome III (UAS-CG5216IR1) and the other with IR insertion into the chromosome II (UAS-CG5216IR2), the F₁ progeny, elav-GAL4/UAS-CG5216IR, was further used to determine the life span at 28°C. In the first experiment, the mean life spans of elav-GAL4/UAS-CG5216IR1 females and males were 45.5 days and 28.9 days, respectively, while those of the control (elav-GAL4/+) were 48.4 days and 39.7 days, respectively (Fig. 2). The mean life span of elav-GAL4/UAS-CG5216IR1 flies showed a statistically significant decrease, to 94.1% and 72.9% of the control elav-GAL4/+ female and male flies, respectively. Similar decreases were observed in the second experiment and for the experiments using elav-GAL4/UAS-CG5216IR2 flies (data not shown).

View Details

Fig. 2. Survival curves of elav-GAL4/UAS-CG5216IR1 flies for (a) females and (b) males. Each survivorship curve represents data from over 450 flies. Solid lines indicate the results of the first experiment and dashed lines indicate those of the second experiment.

CG5085

UAS-CG5085IR flies were crossed with Act5C-GAL4 flies at 25°C. Of the eight transgenic lines for CG5085 examined at 25°C, the F₁ Act5C-GAL4/UAS-CG5085IR flies were viable in five (Table 1) and exhibited pupal lethality in three. On the other hand, when the F₁ flies were reared at 28°C, pupal lethality was observed for all nine lines examined (Table 1), as expected from the temperature dependency of the GAL4-UAS expression system (Fortier and Belote, 2000).

For the life span determination, Act5C-GAL4/UAS-CG5085IR1 and Act5C-GAL4/UAS-CG5085IR2 were used. The experiments were conducted at 25°C and 28°C. Because similar life span results were observed, the results at 28°C are shown in Fig. 3. Both UAS-CG5085IR1 and UAS-CG5085IR2 had an IR insertion in the second chromosome. In both the first and second experiments, the mean life span of Act5C-GAL4/UAS-CG5085IR1 females did not differ from those of the controls (Fig. 3a). For males, the survivorship decreased linearly in the first experiment, but the survival curve was nearly normal in the second experiment (Fig. 3b). For Act5C-GAL4/UAS-CG5085IR2 flies, lethality at the pupal stage was frequently observed (Table 1), especially in males. In the first experiment for life span, many Act5C-GAL4/UAS-CG5085IR2 flies died just after eclosion (Fig. 3c, 3d). In the second experiment, fewer Act5C-GAL4/UAS-CG5085IR2 flies died after eclosion when compared with the first experiment; however, the mean life span of the Act5C-GAL4/UAS-CG5085IR2 females was 38.4 days compared to 46.5 days for the control, a decrease to 82.6%. For the Act5C-GAL4/UAS-CG5085IR2 males, the survivorship decreased linearly, as shown in Fig. 3d.

View Details

Fig. 3. Survival curves and CG5085 mRNA levels of Act5C-GAL4/UAS-CG5085IR1 and Act5C-GAL4/UAS-CG5085IR2 flies at 28°C. Survival curves are indicated for (a, c) females and (b, d) males. Each survival curve represents data from over 200 flies, except that for Act5C-GAL4/UAS-CG5085IR2 males, which is based on approximately 100 flies. Solid lines indicate the results of the first experiment and dashed lines indicate those of the second experiment. (e) The results of real-time PCR analysis for the first (striped columns) and second (gray columns) experiments are shown. The level of CG5085 mRNA was normalized using a standard (rp49 mRNA), and the relative level is shown with the value of Act5C-GAL4/+ set at 100.

The levels of transcription of CG5085 at 25°C and 28°C were examined using real-time PCR. Five flies reared at 28ºC were used for the extraction of RNA in two experiments. The amounts of CG5085 mRNA examined for Act5C-GAL4/UAS-CG5085IR1 flies were 45% and 36% of the Act5C-GAL4/+ controls in females and 48% and 104% of the controls in males (Fig. 3e), in the first and second experiments, respectively. Similarly, the levels of transcription in Act5C-GAL4/UAS-CG5085IR2 flies were 32% and 34% of the controls in females and 40% and 60% in males in the first and second experiments, respectively. CG5085 gene expressions in the flies reared at 28°C were more decreased than those reared at 25°C (data not shown).

CG6284

For the 13 transgenic fly lines of CG6284, the F₁ of the crosses with the Act5C-GAL4 at 28°C were all viable (Table 1). Pupal lethality was observed for some individuals in only one line. CG6284IR1 with insertion into chromosome II and CG6284IR2 with insertion into chromosome III were used to examine life spans. In the first experiment, the mean life spans of the Act5C-GAL4/UAS-CG6284IR1 females and males were 36.9 days and 32.2 days, respectively, while the mean life spans of the control were 47.3 days and 47.4 days, respectively (Fig. 4a, 4b). The life spans of Act5C-GAL4/UAS-CG6284IR1 flies were significantly lower, 77.9% of the control in females and 68.0% in males. In the second experiment, similar decreases in the mean life span were observed. For Act5C-GAL4/UAS-CG6284IR2 flies, decreases in life span were also indicated (data not shown).

View Details

Fig. 4. Survival curves and CG6284 mRNA levels of Act5C-GAL4/UAS-CG6284IR1 flies at 28°C. Survival curves are indicated for (a) females and (b) males. Each survival curve represents data from over 350 flies. Solid lines indicate the results of the first experiment, and dashed lines indicate those of the second experiment. (c) The amount of CG6284 mRNA is shown as in Fig. 3e. The results of the first (white columns), second (striped columns), and third (gray columns) experiments are shown.

The level of transcription of CG6284 in the F₁ flies was also examined. Three to eight flies were used individually for the extraction of RNA. In three experiments on females, the amount of CG6284 mRNA found in Act5C-GAL4/UAS-CG6284IR1 flies was 43% to 71% of the Act5C-GAL4/+ controls; in three experiments on males, it was 21% to 38% of the control (Fig. 4c). Similarly, the levels of transcription in Act5C-GAL4/UAS-CG6284IR2 flies were 100% and 151% of the controls in females and 52% and 76% in males in the first and second experiments, respectively (data not shown).

DISCUSSION

In the present study, we have examined the effects of the loss or decrease of function of Sir2 and Sir2-like genes on the life span of Drosophila. Our results showed that the three genes examined might be involved in the regulation of life span. CG5216 protein is the closest fly homolog of the yeast Sir2 protein, and the CG5085 protein is the second closest fly homolog, while the CG6284 protein shows a relatively weak homology (Fig. 1). This suggests that Sir2 and Sir2-like genes have different functions, although Sir2 and Sir2-like proteins share a conserved core domain.

In the present study, individuals in which the expression of the Sir2 and Sir2-like genes decreased ubiquitously showed different viabilities and life spans. All the Act5C-GAL4/UAS-CG5216IR progeny reared at 25°C died, although some of the Act5C-GAL4/UAS-CG5085IR and Act5C-GAL4/UAS-CG6284IR progeny were viable at both 25°C and 28°C (Table 1). The differences in viability between Act5C-GAL4/UAS-CG5216IR, Act5C-GAL4/UAS-CG5085IR, and Act5C-GAL4/UAS-CG6284IR flies imply that the expression patterns of the Sir2 and Sir2-like genes during development differ from each other. Moreover, many Act5C-GAL4/UAS-CG5085IR flies died just after eclosion, and their survivorship decreased linearly (Fig. 3b–3d). Survivorship of the Act5C-GAL4/UAS-CG6284IR progeny exhibited normal curves (Fig. 4a, 4b). These results suggest that the suppression of Sir2 and Sir2-like functions has different effects on viability and life span.

Ubiquitous silencing of the CG5216 gene caused lethality during development, suggesting that CG5216 may play an important role in development as well as in life span. Because the silencing of CG5216 only in the neurons did not cause lethality (Table 1), CG5216 gene expression somewhere other than in the neurons might be critical to normal development in Drosophila. On the contrary, previous reports have indicated that the CG5216 gene is nonessential because mutants of null alleles are viable until adulthood (Åström et al., 2003; Newman et al., 2002). The reason for the difference in findings is not clear at present. We also observed that decreased expression of CG5216 in the neurons shortened life span (Fig. 2a, 2b). CG5216 protein has been detected in neurons (Newman et al., 2002; Rogina and Helfand, 2004), and it has been reported that neuronal CG5216 overexpression in Drosophila extends life span (Rogina and Helfand, 2004). The function of CG5216 in the neurons could be involved in determining life span.

The reduced expression of the CG5085 gene was shown in the present study to greatly shorten life span or to significantly lower viability (Fig. 3b–3d), except in female Act5C-GAL4/CG5085IR1 flies (Fig. 3a). The reduced expression of either the CG5216 or the CG6284 gene resulted in survival curves that were not greatly different from the controls in any experiment (Fig. 2a, 2b, Fig. 4a, 4b). Different shapes of survival curves were observed for the two lines of Act5C-GAL4/UAS-CG5085IR (Fig. 3b–3d). These survivorships greatly decreased just after eclosion. Despite using the same experimental conditions for the whole life span assay, only the results for this cross showed great decreases in the survivorship after eclosion. This finding suggests that this decrease in the survivorships is dependent on the function of CG5085 gene before and/or after eclosion. It might be possible that subtle and uncontrolled differences between experiments affected the function of CG5085, resulting in lethality at the pupal stage, or immediately after eclosion.

The rate of viability of males is generally lower than that of females for CG5085. This difference may be due to different function or expression of CG5085 gene between males and females. In fact, the amount of CG5085 mRNA in males was found to be five times that in females (data not shown). These results are consistent with those obtained in mammals; many SIRT1 knockout mice died soon after birth (McBurney et al., 2003; Cheng et al., 2003), suggesting that the CG5085 gene and SIRT1 have a similar function. CG5085 belongs to the same class (Class Ib) as yeast HST2 and human SIRT2 (Frye, 2000) among the Sir2 family. They both encode cytosolic NAD-dependent deacetylase (Perrod et al., 2001; North et al., 2003). It has been reported that HST2 modulates nucleolar and telomeric silencing (Perrod et al., 2001) and that human SIRT2, widely expressed in fetal and adult tissues, plays a role in the control of mitotic exit in the cell cycle (Dryden et al., 2003; Frye, 1999). Human SIRT3, another Class Ib protein, is an NAD-dependent deacetylase and is indicated to be localized in mitochondria (Onyango et al., 2002; Schwer et al., 2002). These facts suggest that the CG5085 protein may also be localized in the cytoplasm or mitochondria and may regulate mitosis in development.

We also observed that ubiquitously decreased expression of CG6284 shortens life span. CG6284 belongs to Class IV (Frye, 2000) of the Sir2 family. Mouse Sirt6, which belongs to the same class as CG6284, is enzymatically inactive as a histone deacetylase (North et al., 2003), but it shows nuclear mono-ADP-ribosyltransferase activity (Liszt et al., 2005). Mono-ADP-ribosylation is a post-translational modification of proteins leading to protein inactivation (Corda and Di Girolamo, 2003). SIRT6 is broadly expressed in adult mouse tissues and embryos (Liszt et al., 2005). As decreased CG6284 expression may cause life span reduction due to inactivation of ADP-ribosyltransferase, further study regarding this point should be useful.

To demonstrate a correlation between gene expression and life span, the extent of reduction of mRNA was examined for CG5085 and CG6284 in the F₁ progeny with induced RNAi using real-time PCR. The levels of transcription of CG5085 in Act5C-GAL4/CG5085IR flies reared at 25°C were 71% to 106% of the controls in females and 55% to 79% of the controls in males (data not shown). When the flies were reared at 28°C, CG5085 gene expression was more significantly decreased (Fig. 3e). Their viability was affected more severely at 28°C than at 25°C (Table 1), and the decreases in life span were larger at 28°C in most cases (Fig. 3b–3d). These results indicate that the reduced expression of CG5085 is correlated with a shortened life span or lower viability. As for females of Act5C-GAL4/CG5085IR1, the decrease in CG5085 mRNA might not be enough to shorten the life span. The level of transcription of CG6284 in both male and female Act5C-GAL4/CG6284IR1 flies decreased (Fig. 4c). The life span decreased in correspondence with the decrease of the expression of CG6284 (Fig. 4a, 4b). These results suggest that the expression of CG6284 may also be involved in life span determination.

In the present study, we showed that Sir2 and Sir2-like genes may be involved in the regulation of development as well as life span determination in Drosophila. The molecular mechanism by which the Sir2 protein affects life span has not been fully understood in other organisms. As Sir2 and Sir2-like genes are conserved from archaebacteria to humans, the Drosophila system that allows efficient manipulations of the function of genes with RNAi provides a useful model for the study of the role of these genes in life span determination. Further studies to analyze expression patterns of the Sir2 and Sir2-like genes during development could lead to a better understanding of the function of these genes.