Drosophila TIF-IA is required for ribosome synthesis and cell growth and is regulated by the TOR pathway

Abstract

Synthesis of ribosomal RNA (rRNA) is a key step in ribosome biogenesis and is essential for cell growth. Few studies, however, have investigated rRNA synthesis regulation in vivo in multicellular organisms. Here, we present a genetic analysis of transcription initiation factor IA (TIF-IA), a conserved RNA polymerase I transcription factor. Drosophila melanogaster Tif-IA(-/-) mutants have reduced levels of rRNA synthesis and sustain a developmental arrest caused by a block in cellular growth. We find that the target of rapamycin (TOR) pathway regulates TIF-IA recruitment to rDNA. Furthermore, we show that the TOR pathway regulates rRNA synthesis in vivo and that TIF-IA overexpression can maintain rRNA transcription when TOR activity is reduced in developing larvae. We propose that TIF-IA acts in vivo as a downstream growth-regulatory target of the TOR pathway. Overexpression of TIF-IA also elevates levels of both 5S RNA and messenger RNAs encoding ribosomal proteins. Stimulation of rRNA synthesis by TIF-IA may therefore provide a feed-forward mechanism to coregulate the levels of other ribosome components.

Figure 1.

TIF-IA is required for cell and organismal growth. (A) Levels of TIF-IA mRNA and pre-rRNA were measured by quantitative RT-PCR, using RNA isolated from either wild-type or Tif-IA ^−/− mutant larvae. Data were corrected for levels of GPDH mRNA. Data are mean (± SEM) fold changes compared with wild type (n = 6). (B) Tif-IA ^−/− mutant larvae are growth arrested. Images of Tif-IA heterozygote (+/−) and Tif-IA homozygous mutant larvae (−/−) at different stages (48–120 h) of larval development are shown. (C) Loss of p53 has no effect on the growth arrest phenotype seen in TIF-IA mutant larvae. Images of TIF-IA ^+/− ; p53 ^−/− (top) or Tif-IA ^+/− ; p53 ^−/− (bottom) larvae at 120 h of development are shown. (D) The hsFlp–GAL4 system was used to generate mosaic expression of GFP-marked cells overexpressing TIF-IA (arrowheads) in the larval fat body of Tif-IA ^−/− mutant animals (red, phalloidin; blue, DAPI). Bar, 25 μm. (E) The hsFlp–GAL4 system was used to generate mosaic expression of both GFP and a TIF-IA RNAi construct in the polyploid cells of the larval fat body (arrowheads; green, GFP; red, phalloidin; blue, DAPI). Bar, 25 μm. (F) quantitative RT-PCR was used to measure levels of pre-rRNA in either control larvae or larvae overexpressing UAS–TIF-IA under the control of the da-GAL4 driver. Data were corrected for levels of GPDH mRNA. Data are mean ± SEM. *, P < 0.05 versus control (n = 7–8). (G) An en-GAl4 driver was used to express a TIF-IA cDNA in the posterior compartment of the developing larval wing imaginal disc. Levels of pre-rRNA were then measured in wandering L3 wing discs by in situ hybridization using a probe to the ETS region of the pre-rRNA precursor. Posterior is to the right. Bar, 50 μm.

Figure 2.

The D. melanogaster nutrient–TOR pathway regulates rRNA synthesis and ribosome biogenesis. (A) Wild-type larvae were starved and, at the indicated times, total RNA was isolated from equal numbers of larvae per time point, and then levels of rRNA were quantitated from ethidium bromide–stained agarose/formaldehyde gels. Data are mean (± SEM) percentage changes in rRNA levels relative to nonstarved animals (day 0 time point; n = 3–6 per time point). (B) Levels of pre-rRNA were measured in fed or 4-d starved larvae by in situ hybridization. A representative image of the gut is shown for both samples. Bar, 25 μm. (C and D) Levels of pre-rRNA were measured by quantitative RT-PCR using RNA isolated from either wild-type or starved larvae (C) or wild-type or tor ^−/− mutant larvae (D). Data are mean (± SEM) fold changes versus fed larvae (n = 3). Data were corrected for levels of dMyc mRNA. (E) The hsFlp–GAL4 system was used to overexpress Rheb transgene throughout developing larvae. Levels of pre-rRNA and GAL4 (as loading control) were measured by Northern blot. Reference DNA fragment sizes are indicated (kb). (F) An en-GAl4 driver was used to express a Rheb transgene in the posterior compartment of wing imaginal discs. Levels of pre-rRNA were then measured by in situ hybridization. Posterior is to the right. Bar, 50 μm. (G) The hsFlp–Gal4 system was used to generate cell clones coexpressing GFP and Rheb in developing wing imaginal discs. Discs were stained with an antibody to fibrillarin (red). The dashed line shows the clone outline. Bar, 20 μm.

Figure 3.

TIF-IA functions downstream of the TOR pathway. (A) The localization of TIF-IA to rDNA was measured using the DamID technique. D. melanogaster Kc cells were transfected with either Dam alone or a Dam–TIF-IA fusion. Cells were then treated with DMSO (control) or rapamycin for 16 h as indicated. Genomic DNA was isolated and digested with DpnII and identical amounts per sample were analyzed by quantitative PCR. Data represent mean (± SEM) fold changes (log scale) compared with Dam alone, the control cells (n = 3). (B and C) Quantitative RT-PCR was used to measure levels of pre-rRNA in either control larvae or larvae overexpressing UAS–TIF-IA under the control of the da-GAL4 driver. In B, larvae were either fed for 2 d or starved for an additional day. In C, larvae were fed for 2 d and then transferred to new vials for 24 h with either normal food containing 0.1% DMSO control or food supplemented with 10 μM rapamycin. Data are mean (± SEM) fold changes in pre-rRNA levels versus fed controls (n = 4). (B) *, P < 0.05 versus fed control; **, P < 0.05 versus starved control. (C) *, P < 0.05 versus DMSO control; **, P < 0.05 versus rapamycin control.

Figure 4.

D. melanogaster TIF-IA regulates rRNA synthesis and controls the levels of other components of the ribosome. (A–C) A da-GAL4 driver was used to express TIF-IA ubiquitously in developing larvae. Total RNA was extracted from control and TIF-IA–overexpressing larvae. Quantitative RT-PCR was used to measure levels of indicated transcripts. Data are presented as fold changes compared with wild type and represent the mean ± SEM (n = 7–8). Data were corrected for levels of GPDH mRNA. *, P < 0.05 versus control.

Figure 5.

Overexpression of TIF-IA is not sufficient to drive ribosome synthesis, mRNA translation, or cell growth. (A) 40S and 60S ribosome particles were separated by sucrose gradient centrifugation and their levels determined by measuring absorbance at 260 mm. (B) Using a da-GAL4 driver, levels of protein synthesis were measured in either control larvae or larvae in which TIF-IA was ubiquitously overexpressed (+ TIF-IA). Data represent the mean (± SEM) percentage change in radiolabeled amino acid incorporation compared with control larvae. (C) Flow cytometry profiles of wing imaginal disc cells. Cell cycle phasing (left) or cell size (right) comparisons of GFP-marked TIF-IA–overexpressing cells (gray traces) with surrounding nonmarked wild-type cells (black traces) are shown. (D) Data represent the mean (± SEM) percentage changes in clone area in either control or TIF-IA–overexpressing cell clones in the wing imaginal disc (n = 200). (E) Using a da-GAL4 driver, levels of protein synthesis were measured in control larvae and larvae in which a UAS– TIF-IA transgene was ubiquitously overexpressed (+ TIF-IA). In addition, either UAS-S6K (+S6K) or UAS-eIF4E (+eiF4E) transgenes were coexpressed. Data represent the mean (± SEM) percentage changes in radiolabeled amino acid incorporation compared with control larvae.