Germline silencing of UASt depends on the piRNA pathway

One of the most extensively used techniques in Drosophila is the Gal4/UAS binary system, which allows tissue-specific misexpression or knockdown of specific genes of interest. The original UAS vector, UASt, can only be activated for transgene expression in somatic tissues and not in the germline cells. Rørth (1998) generated UASp, a modified UAS vector that is responsive to Gal4 in both somatic and germline tissues, by replacing both the hsp70 promoter and the SV40 3’UTR with the P transposase promoter and the K10 3’UTR respectively. At present, the mechanisms by which UASt is silenced in germline cells are not fully understood. Here, we report that the piRNA pathway is involved in suppressing UASt expression in ovarian germline cells. Individually knocking down or mutating components of the piRNA biogenesis pathway (e.g., Piwi, AGO3, Aub, Spn-E, and Vasa) resulted in the expression of the UASt-reporter (GFP or RFP) in the germline. An RNA-seq analysis of small RNAs revealed that the hsp70 promoter of UASt is targeted by piRNAs, and in the aub mutant ovary, the amount of piRNAs targeting the hsp70 promoter is reduced by around 40 folds. In contrast, the SV40 3’UTR of the UASt, which happens to be targeted by the Nonsense-mediated RNA decay (NMD) pathway, is not responsible for germline UASt suppression, as UASt-reporters with NMD-insensitive 3’UTRs fail to show germline expression. Taken together, our studies reveal a crucial role of the piRNA pathway, potentially via the suppression of the hsp70 promoter, in germline UASt silencing in Drosophila ovaries.


46
The success of the fruit fly Drosophila melanogaster as a model organism is heavily attributed to the expansive range and multitude of genetic and molecular tools available to modify gene Fly stocks and genetics 92 The following fly stocks were used in this study: act>CD2>Gal4, UASt-RFPnls (BL30558);  Antibodies, immunofluorescence staining and confocal microscopy 101 Immunocytochemistry was carried out as described previously (Deng et al., 2001). The 102 following antibodies were used: rat anti-Vasa (1:300; Development Studies Hybridoma Bank), 103 rabbit anti-HA-tag (C29F4, 1:100; Cell Signaling). Secondary antibodies were stained with 104 Alexa Fluor® 546 or 488 and nuclear DNA was labeled with DAPI (Invitrogen). Images were 105 acquired with a Zeiss LSM-800 confocal microscope and assembled in Adobe Illustrator.

106
Quantitative RT-PCR analysis 107 Total RNA from two-day-old Drosophila ovaries was isolated using Trizol Reagent (Invitrogen) 108 according to the manufacturer's instructions and then treated with 2 U/µl of DNase I (Ambion) 109 for 30 minutes at 37˚C. One microgram of total RNA was reverse-transcribed in 20 µl of reaction 110 mixture containing Superscript II reverse transcriptase (Invitrogen) and oligo (dT)12-18 primer 111 according to the protocol for Superscript II first-strand cDNA synthesis system. One microliter cDNA (reverse transcribed from 50 ng of RNA) was subjected to quantitative real-time PCR (in 113 25 µl reaction volume) by using primers specific to a transposable element or RP49 (primer 114 sequences below) and cDNA templates were amplified using the Platinum SYBR Green qPCR 115 SuperMix UDG kit, according to the manufacturer's instructions (Invitrogen). PCR conditions 116 are: 95˚C for 10 minutes; 40 cycles of 95˚C for 30 seconds, 58˚C for 15 seconds, and 68˚C for 45 117 seconds. Real-time PCR was performed using the ABI 7500 Thermocycler (Applied Biosystems), 118 and results were analyzed using SDS version 2.1 software (Austin Biodiversity Web site gallery).

119
Data analysis was done using the 2-∆∆CT method for relative quantification. Calculated 120 expression values of cDNA samples were normalized to RP49.  Small RNA reads (>22nt) that passed quality control and the removal of rRNAs, snoRNAs and 131 tRNAs reads by bowtie (-a --best --strata -v 1 --un) (bowtie-bio.sourceforge.net/index.shtml) 132 were mapped to the transgene UASt nucleotide sequence (detail sequence obtained from 133 Addgene, Cambridge, MA) (-a -v 0 -m 1) by bowtie. 137 The bipartite Gal4/UAS system is one of the most important and widely used genetic tools in  (Fig 1A').

146
In a genome-wide in vivo RNAi screen to identify genes that potentially regulate Notch  Unexpectedly, we found that the RFP signal was detected in some nurse cells in mosaic egg 153 chambers with Su(var)2-10 (CG8086) knockdown ( Fig 1B). Consistently, when the CoinFLP 154 system (Bosch et. al., 2015), which also carries the Actin5C promoter to generate mosaic Gal4 155 expression, was used to drive Su(var)2-10-TRiP-RNAi expression, UASt-GFP expression was 156 also detected in the nurse cells ( Fig 1D). These results suggest that Su(var)2-10 plays a role in 157 suppressing UASt-transgene expression in the germline.  Table 1). 178 We then examined UASt-reporter expression in the germline of mutants defective in 179 piRNA production. We expressed act-Gal4 driven UASt-RFP (act>UASt-RFP) in a trans-180 heterozygous mutant of AGO3 (AGO3 t2/t3 ), and found that all AGO3 t2/t3 mutant ovaries expressed heterozygous egg chambers showed RFP expression only in the somatic follicle cells (Fig 4A). 183 We also tested the expression of UASt-RFP driven by mat-tub-Gal4, a germline specific Gal4, in 184 aub QC42/HN2 trans-heterozygous mutants. As expected, RFP expression was detected in both the 185 nurse cells and the oocyte during oogenesis (Fig 4D). Taken together, these results indicate that 186 the suppression of UASt-transgene expression in the germline depends on the piRNA biogenesis 187 pathway.

188
To determine whether UASt-RFP suppression is at the transcription level, we performed 189 quantitative RT-PCR (qRT-PCR) analyses, and found the RFP transcript level was 1.3-fold 190 higher in AGO3 t2/t3 trans-heterozygous ovaries than the AGO3 t2/+ heterozygous controls. As a 191 control, we examined the expression of Het-A, a transposable element that is targeted by piRNAs 192 (Brennecke et al., 2007), and found a 60-fold increase of the Het-A transcript in AGO3 t2/t3 193 mutant ovaries (Fig 4C'), indicating that piRNA production is indeed strongly suppressed in 194 these ovaries. Together, these results suggest that germline expression of UASt-RFP is 195 suppressed by piRNAs at the transcriptional level (Fig 4C).

196
The hsp70 promoter is a piRNA target 197 To determine exactly how piRNAs suppress UASt-transgene expression in the germline, we 198 performed a RNA-seq analysis of small RNAs from ovaries of aub QC42/HN2 and w 1118 flies. After 199 mapping the piRNAs from the RNA-seq analysis on to the pUASt sequence, we found that the 200 hsp70 promoter was heavily targeted by piRNAs in w 1118 flies (Fig 5A, B). This result is 201 consistent with the report that the hsp70 locus itself provides the substrates for high piRNA 202 production in transgenic lines (Olonikov et al., 2013). By contrast, aub QC42/HN2 ovaries had a 203 significantly reduced level (~40-fold lower) of piRNAs targeting the hsp70 promoter sequence compared with the w 1118 controls (Fig. 5B). These findings, along with the qRT-PCR results that 205 AGO3 t2/t3 mutant ovaries had elevated levels of the UASt-RFP transcript, suggest that the hsp70 206 promoter in the UASt vector is targeted by piRNAs from the hsp70 locus, thus suppressing 207 UASt-transgene transcription in the germline cells.

208
The 3'UTR is not the cause of UASt germline suppression 209 Besides the promoter, the other major difference between UASt and UASp is the 3'UTR tail. showed no ovarian germline expression of UASt-GFP (Fig 6A-B). We further tested another 217 modified UASt vector, the pGW construct (Bischof et al., 2013), which has tubulin 3' UTR 218 instead of the SV40 3'UTR. Using the Flp-out Gal4 system, we co-expressed UASt-RFP and 219 GW-HA in the ovary, and detected neither RFP nor HA proteins in germline cells (Fig 6C). 220 Additionally, we knocked down genes involved in the NMD pathway (Upf1, Upf2, Upf3, Smg1, 221 and Smg5), individually, and no germline UASt-reporter expression was detected in the ovary 222 (Table 2). Taken together, these results suggest that the SV40 3' UTR and NMD pathway are 223 not involved in germline UASt silencing.
The Gal4/UAS system has been used extensively in a variety of forms and adaptations for a wide 227 range of applications. However, the one downside is that the original UASt is not expressed by 228 germline cells of Drosophila ovaries, which limits the experimental power of germline cells.