Validation of novel promoter sequences derived from two endogenous ubiquitin genes in transgenic Aedes aegypti

Abstract

To date, only a limited number of promoter sequences have been described to drive transgene expression in the disease vector Aedes aegypti. We sought to increase this repertoire by characterizing the ability of upstream sequences derived from the Ae. aegypti Ub_L40 and polyubiquitin genes to drive the expression of marker proteins. Both genomic fragments were able to drive robust expression of luciferase in cultured mosquito cells. Following Mos1-transformation, the Ub_L40 promoter drove strong expression of a fluorescent marker in early larvae and in ovaries, while the polyubiquitin promoter drove robust EGFP expression in all stages of development including constitutive expression throughout the midgut. These promoter fragments provide two new expression profiles for future Ae. aegypti genetic experiments.

INTRODUCTION

Aedes aegypti is the most important mosquito vector of viral disease agents worldwide, responsible for the transmission of viruses which cause dengue fever, dengue hemorrhagic fever, and yellow fever (Halstead 2007). Transposon-based transformation of this species with class II transposable elements is well established (Coates et al. 1998; Jasinskiene et al. 1998; Kokoza et al. 2001), and has allowed investigations into mosquito gene regulation and immunity (Adelman et al. 2008; Antonova et al. 2009; Bian et al. 2005) as well as establishing heritable virus-resistance phenotypes (Franz et al. 2006). These studies are enabled, and limited by, the catalog of promoter elements available to drive transgene expression.

Several endogenously-derived promoter fragments which drive the expression of transgenes in either the Ae. aegypti midgut or fat body have been described (Kokoza et al. 2000; Moreira et al. 2000). Each of these promoters is induced following a blood meal, and have been used to generate pathogen resistance phenotypes (Franz et al. 2006; Kokoza et al. 2001) or to study the mosquito innate immune response (Antonova et al. 2009; Bian et al. 2005). Salivary gland promoters derived from the maltase and apyrase genes have also been described in Ae. aegypti, but the utility of these promoters has been limited by their generally weak levels of expression (Coates et al. 1999). Promoters restricted to the ovary (Adelman et al. 2007; Cho et al. 2006) and testes (Smith et al. 2007) have also been described. We have recently used the synthetic 3xP3 promoter (Horn et al. 2000) to generate an RNAi sensor strain which can be used to identify genes involved in RNAi (Adelman et al. 2008). The use of tissue or timing-restricted promoters has been useful for the targeted expression or knockdown of genes. However, few tools are available to drive transgene expression in a constitutive, whole body manner which may be desirable for generating gene knockdown or pathogen resistant phenotypes prior to bloodfeeding or in multiple tissues. Thus far, the only whole-body constitutive promoter reported in transformed Ae. aegypti is the heterologous D. melanogaster Act5C promoter (Pinkerton et al. 2000). While successfully used as a marker for transformation, the activity of this promoter was found to be restricted to the fat body and gut of the developing larvae and to the gonads of the adult (Pinkerton et al. 2000).

Ubiquitin (Ub) is an ancient 76 amino acid protein which is almost perfectly conserved among all eukaryotes. While the best characterized role for ubiquitin is in tagging proteins for degradation by the proteosome, conjugation to ubiquitin has been observed to play a role in numerous other cellular functions such as DNA repair, immune activation and cell signaling (reviewed in Hochstrasser 2009). Due to its importance to cellular functions, ubiquitin genes have served as donors of control sequences to drive gene expression on a global basis. Insect polyubiquitin (PUb) gene promoters have been used to drive global transgene expression in D. melanogaster (Davis et al. 1995; Handler and Harrell 1999) as well as the red flour beetle Tribolium castaneum (Lorenzen et al. 2002). The D. melanogaster PUb promoter has also been used in non-drosophilids, successfully driving transgene expression in the carribean fruit fly, Anastrepha suspensa, the screwworm Cochliamyia hominivorax and the malaria vector Anopheles albimanus (Handler et al. 2009; Handler and Harrell 2001; Perera et al. 2002). While the PUb gene has been cloned from Anopheles gambiae (Beard et al. 1996), control sequences from this gene have not yet been characterized. Genes encoding ubiquitin or polyubiquitin have not been described as yet for the yellow fever mosquito, Ae. aegypti.

In this report we test and validate promoter fragments derived from the Ae. aegypti ubiquitin genes Ub_L40 and PUb. Both promoter fragments were found to drive robust transgene expression in mosquito cells. In transformed Ae. aegypti, the Ub_L40 promoter was found to be highly active in early larvae and ovaries, while the PUb promoter was active during all life stages including constitutive expression in the adult female midgut. These promoter fragments can now be used for future studies into pathogen transmission or gene function.

RESULTS and DISCUSSION

Characterization of two Ae. aegypti ubiquitin gene promoters in mosquito cells

Putative ubiquitin genes were identified from the Ae. aegypti genome through tBLASTn search using the D. melanogaster 76 a.a. ubiquitin monomer as a query. Five genes on four supercontigs were identified with an e value of less than 1e-30. Genes AAEL006511 (supercont1.209) and AAEL013536 (supercont1.864) both contained a single ubiquitin monomer and appeared to be homologues of the Drosophila Ub_L40 (Ub52) and Ub_S27 (Ub80) ribosomal fusion genes (Cabrera et al. 1992; Lee et al. 1988). These genes are both well conserved among eukaryotes, and consist of a ubiquitin monomer fused to the ribosomal proteins L40 or S27 (Finley et al. 1989). Genes AAEL003888 and AAEL003877 (supercont1.99) are arranged in tandem, and contain 2 and 14 consecutive ubiquitin repeats, respectively, while the last gene, AAEL000795 (supercontig 1.17), contains 8 repeats. Both the Ub_S27 and AAEL000795 genes appeared to contain intron sequences in the 5′UTR of greater than 10 kb based on the current annotation. As we sought to include all 5′UTR sequence, including any intron sequence, into potential transformation constructs, these genes were not pursued any further. Gene AAEL003888 only contained two ubiquitin repeats, one of which had an addition of 1 amino acid at the C-terminus. Due to the presence of the much longer ubiquitin gene immediately downstream (14 repeats), we did not pursue this gene any further. Therefore, we restricted our focus to genes AAEL006511 and AAEL003877, hereafter referred to as Ub_L40 and PUb, respectively. In order to verify the 5′ and 3′ start and stop of transcription for each gene, we performed Rapid Amplification of cDNA Ends (RACE). For both the Ub_L40 and PUb mRNAs we recovered a consistent 5′ end of the mRNA transcript (Fig S1A and C). The 3′ end of each transcript exhibited some minor variation in termination; of the seven sequenced PUb clones and eleven Ub_L40 sequenced clones, no 3′ termination event was recovered more than twice (Fig S1B and D). However, all seven PUb clones terminated within 4 nt of each other, and 7 out of the longest 11 Ub_L40 clones terminated within 11 nt of each other. The completed gene structures, using the longest obtained 3′UTRs, are shown in Figure 1 (A + B). Both Ub_L40 and PUb genes were found to contain a single small intron in the 5′UTR, with the 3′ splice site of each located just upstream of the start codon. To obtain putative promoter sequences from these two genes, primers were designed based on the completed Ae. aegypti genome sequence and amplified fragments were cloned upstream of the firefly luciferase open reading frame. The 114 bp upstream boundary selected for the Ub_L40 genomic fragment was due to the presence of a full-length copia retrotransposon further upstream. We selected 565 bp of upstream genomic sequence for the PUb gene, which represented the entire intergenic space separating PUb from the upstream ubiquitin gene AAEL003888. Downstream boundaries of each amplified genomic fragment included sequence corresponding to the entire 5′UTR of each gene, so that the luciferase ORF was placed exactly where ubiquitin translation would normally initiate (Fig1A, B). To determine whether each genomic fragment could drive gene expression in mosquito cells, we performed luciferase assays following plasmid transfection into C6/36 (Ae. albopictus) or Aag2 (Ae. aegypti) cells (Fig 1C, 1D). As a negative control we transfected a promoterless construct (pGL3), while the baculovirus IE-1 promoter was used for comparison. All samples were co-transfected with a Renilla luciferase-expressing plasmid to normalize for differences in transfection efficiencies between wells. As shown in Figure 1 (C + D), both the Ub_L40-derived and PUb-derived genomic fragments were able to drive strong luciferase expression in both cell types. While the activity of the Ub_L40 promoter was similar to that of IE-1, the PUb genomic fragment was found to be at least 10-fold stronger than IE-1 in all experiments conducted in both C6/36 (Fig. 1C, p≤0.02, Students t-test) and Aag2 (Fig. 1D, p<0.0001, Students t-test) cells.

Figure 1.

Ub_L40 and PUb promoters are highly active in mosquito cells. Schematic representation of Ub_L40 (A) and PUb (B) gene structure in Ae. aegypti. Exons (E) and introns (I) are indicated, with the predicted mRNA, including 5′ and 3′ untranslated regions (UTRs) and open reading frame (ORF) are shown below each gene. Numbers indicate length in base pairs. Gray boxes indicate the 76 amino acid ubiquitin monomer. Bars above each gene represent putative promoter regions utilized in (C, D) and all further experiments. (C, D) Relative luciferase activity from IE-1, Ub_L40 and PUb promoters compared to a no promoter control plasmid (pGL3) at 48hrs post-transfection in C6/36 (C) and Aag2 cells (D). Each experiment (Exp) was performed in triplicate. Statistical significance (*) was determined by Students t-test. Ub_L40 and PUb promoter/gene sequences have been deposited in Genbank (accession #s GU179017 and GU179018).

Generation of transgenic Ae. aegypti

As both promoter fragments appeared to be highly active in mosquito cells, we sought to determine their activity in stably transformed mosquitoes. Three Mos1-based donor constructs: pMos-3xP3DsRed-RH-Ub_L40-EGFP (Fig 2A, #1), pMos-3xP3DsRed-Ub_L40-EGFP-attP (Fig 2A, #2) or pMos-3xP3DsRed-PUb-EGFP-attP (Fig 2A, #3) were injected into Ae. aegypti embryos (kh^w or Liverpool strains) along with the pKhsp82M helper (Coates et al. 1998). G₁ larvae were screened for eye-specific expression of DsRED as early larvae, with putative transgenic individuals identified from each experiment (Table 1). Preliminary examination of early larvae revealed widespread EGFP fluorescence in Ub_L40-EGFP lines #18 and #P17A and in all PUb-EGFP derived lines, but not in Ub_L40-EGFP lines #19, P10, P13 and P17B. Southern analysis was used to verify each transgene insertion (Fig. 2B). With the exception of line #19, genomic DNA from each putative transgenic line was subject to restriction enzyme digestion (Fig 2A - EcoR I, Xba I or Nsi I) with no target recognition sequence within the transgene in order to reveal the number of Mos1 insertions (Fig. 2B). Two insertions were observed in lines #18 and P4, while the remaining lines appeared to contain a single Mos1 element (Fig. 2B). As not all Ub_L40-EGFP lines appeared to express EGFP, we sought to verify the integrity of each transgene. Ub_L40-EGFP-derived lines were subject to SalI digestion, expected to produce a common ~4 kb hybridization signal in lines #18 and #19, along with two junction fragments of variable length, or a common ~3kb signal in lines P10, P13A, P17A, P17B, along with two junction fragments expected to vary with the insertion site (Fig. 2A). These hybridization signals were indeed observed in all pools (Fig. 2B). Additionally, we performed PCR on genomic DNA isolated from lines Ub_L40-EGFP #P13A and #P17B and sequenced the resulting amplicons. No disruptions in the Ub_L40-EGFP cassette were found in either line (data not shown). This confirms that the entire gene cassette was integrated intact into the mosquito genome for all Ub_L40-EGFP transgenic lines and suggests that the site of Mos1 insertion has affected the activity of the Ub_L40 promoter in those lines which did not appear to express EGFP. Approximate transformation frequencies for the three experiments were calculated to be 3.2%, 3.2% and 5.5%, assuming a G₀ fertility rate of ~50%, which is similar to previous experiments with Mos1 (Adelman et al. 2008; Adelman et al. 2007; Coates et al. 1998). While lines #18 and P4 were found to contain two insertions each, we did not include these extra insertions in our determination of the transformation frequency, as they did not appear to segregate from each other following two generations of out-crossing to recipient strain mosquitoes (Table 1).

Figure 2.

Southern analysis of Ub_L40-EGFP and PUb-EGFP insertions in Mos1-transformed Ae. aegypti. (A) Schematic representation of hypothetical transgene insertions for each of the donor constructs used. Block arrows represent the right (R) and left (L) arms of the Mos1 transposon. Bar indicates the size of the entire insertion and the dashed line represents mosquito genomic DNA. Restriction enzyme sites EcoRI (E), SalI (S), XbaI (X) and NsiI (N) are indicated. (B) Genomic DNA from each of the families identified as DsRED⁺ was digested with the indicated enzymes and hybridized with a probe corresponding to the Mos1 arms as well as the DsRED-SV40 gene cassette. The recipient strain kh^w is included as a negative control for all hybridizations. Molecular weight markers are shown at left (in kbp).

Table 1.

Mos1-mediated transformation of Ae. aegypti with Ub_L40-EGFP or PUb-EGFP gene cassettes.

Mosl Donor	# embryos injected	# G0 survivors (%)	# Pools	# G1 progeny screened	Pools with DsRED+ progeny (#)	G2+/Total (% +)	G3+/Total (% +)
UbL40 (exp1)	1347	126 (9.4%)	24	7358	#18(3+)	5/18 (28%)	23/49 (47%)
					#19 (2+)	84/304 (28%)	338/626 (54%)
UbL40 (exp2)	1825	248 (13.6%)	17	13503	#P10 (3+)	18/51 (35%)	-
					#P13A (53+)	131/234 (56%)	-
					#P17A (7+)	232/342 (68%)	40/76 (53%)
					#P17B (5+)	9/74 (12%)	-
PUb	1902	218 (11.5%)	7	16391	#P2 (34+)	102/200 (51%)	205/354 (58%)
					#P3 (38+)	144/300 (48%)	309/591 (52%)
					#P4 (13+)	42/200 (21%)	135/246 (55%)
					#P5 (13+)	357/728 (49%)	157/305 (51%)
					#P6 (10+)	117/223 (52%)	172/329 (52%)
					#P7 (18+)	136/306 (44%)	131/259 (51%)

Expression of EGFP reporter in transgenic Ae. aegypti

As indicated above, early larvae (1^st-2^nd instar) from Ub_L40-EGFP line #P18 and PUb-EGFP line #P5 transgenic mosquitoes displayed high levels of green fluorescence throughout the entire body (Fig. 3A). Similar expression was observed in Ub_L40-EGFP #17A and in all five other PUb-EGFP transformed lines (not shown). While fluorescence remained strong throughout larval development and into pupation for all of the PUb-EGFP transformed lines, Ub_L40-EGFP mosquitoes became gradually dimmer, with little fluorescence remaining in pupae (Fig. 3B). Adult females from all PUb-EGFP lines remained highly fluorescent, with EGFP easily detectable through the cuticle in the head, thorax and abdomen, while no EGFP could be visualized in adult Ub_L40-EGFP mosquitoes (Fig. 3C). Northern blotting was used to further characterize the expression of EGFP mRNA from transgenic lines Ub_L40-EGFP #18 and P17A, as well as PUb-EGFP #P5. Total RNA was extracted from various life stages and tissues and hybridized to a probe corresponding to EGFP or Ub_L40 (Fig. 4A) or to the PUb 5′UTR present in both the endogenous PUb transcript as well as the introduced transgene (Fig. 4B). In both kh^w and Liverpool strains, Ub_L40 transcript was detected throughout development, with the strongest expression observed in L₁ larvae (Fig. 4A). Ub_L40 transcript abundance decreased during larval development but was still detectable in pupae. In adult tissues, Ub_L40 mRNA was detectable in the thorax and midgut of the adult mosquito, with the highest transcript amounts detected in the ovaries. Transgenic lines Ub _L40 -EGFP #18 (compare to kh^w) and P17A (compare to Lvp) displayed similar expression patterns, with high transcript expression in early larvae decreasing through development and adult expression strongest in the ovaries (Fig. 4A). The D. melanogaster Ub_L40 gene was also found to be expressed preferentially in ovaries (Cabrera et al. 1992). However, these researchers found that in the fly, Ub_L40 mRNA expression remained constant throughout development (Cabrera et al. 1992). Endogenous PUb mRNA expression was fairly uniform throughout development and in specific tissues (Fig. 4B). EGFP mRNA expression in line PUb-EGFP #P5 was similar (Fig. 4B, bottom), with the exception of ovary tissue, where EGFP mRNA was not readily detectable. EGFP mRNA was not detected in the salivary glands, though this could be due to insufficient RNA. Visual inspection of salivary glands from both Ub_L40-EGFP or PUb-EGFP mosquitoes confirmed that EGFP expression in these tissues was not robust (not shown). Interestingly, it appeared that in all but ovarian tissues, the expression of PUb-driven EGFP mRNA exceeded expression of the endogenous PUb mRNA. Though the reasons for this are not known, it is possible that the genomic context of the inserted Mos1 element is a more favorable location for PUb expression, either through insertion near neighboring activating elements or via the loss of neighboring repressive elements located in the endogenous PUb locus but not included in the transgenic constructs.

Figure 3.

Ub_L40 and PUb promoters drive EGFP expression in transformed Aedes aegypti. EGFP expression in L₁ larvae (A) and pupae (B) for PUb-EGFP #P5, kh^w and Ub_L40-EGFP #P17A. (C) White light, EGFP and DsRED images of adult female kh^w, Ub_L40-EGFP #18 and PUb-EGFP #P5.

Figure 4.

Northern analysis of Ub_L40-EGFP and PUb-EGFP expression in transformed Ae. aegypti. (A) Northern blot of endogenous Ub_L40 (kh^w, Lvp) mRNA and EGFP mRNA in transgenic lines Ub_L40-EGFP #18 and #P17A. Blots were probed with either the L40 portion of the Ub_L40 ORF (last 159bp) along with its 3′UTR or for EGFP. (B) Northern blot of PUb and PUb-EGFP expression in kh^w and PUb-EGFP #P5 transgenic mosquitoes. A probe corresponding to the PUb 5′UTR was used for hybridization to each blot, with ribosomal RNA shown as a loading control below. Individual lanes include: first through fourth instar larvae (L1-L4), male pupae (♂P) female pupae (♀P), adult whole body (WB), head (H), thorax without salivary glands (T), salivary glands (S), midgut (M), sugar-fed ovary (Os), and ovaries at 4 days post blood meal (Ob).

As both promoters appeared to have strong activity in the midgut, and the activity of the Ub_L40 promoter appeared strongest in the ovaries, we performed dissections to examine the expression pattern of EGFP visible in each of these tissues (Fig. 5). In ovaries taken 24 hr post bloodmeal, Ub_L40-EGFP expression was observed in the nurse cells (Fig. 5A, white arrow), with diffuse fluorescence present in the developing oocyte. By 48 hr the nurse cells had already disappeared and strong EGFP expression was observed throughout the oocyte. By 72 hr, however, EGFP-based fluorescence had largely disappeared from the oocyte, and was mostly visible in the undeveloped secondary and tertiary ovarioles (Fig. 5A, thin white arrows). While some EGFP expression was observed in the PUb-EGFP ovaries, this appeared to be primarily associated with accessory tissues (Fig. 5A, 24 + 72hrs), though some EGFP fluorescence was observed in oocytes (Fig 5A, 48 hr). This situation was reversed in midgut tissues, as robust EGFP expression could be seen throughout the midgut of PUb-EGFP mosquitoes (Fig. 5B). Most significantly, this expression did not require a bloodmeal, nor did it appear to diminish in older mosquitoes (not shown). While some EGFP expression was definitely observed in Ub_L40-EGFP midguts, expression was not robust and appeared patchy (Fig. 5B). No EGFP expression was observed in recipient strain (kh^w) mosquitoes in either ovaries or midguts under the same conditions (Fig. 5).

Figure 5.

Ub_L40 and PUb promoters drive robust transgene expression in Ae. aegypti ovary and midgut tissues. (A) White light and EGFP images of kh^w, Ub_L40-EGFP #18 and PUb-EGFP #P5 ovaries at 24 hr, 48 hr and 72 hrs post blood meal. Thick white arrow indicates concentrated EGFP expression in nurse cells, thin white arrows indicate undeveloped secondary and tertiary ovarioles. (B) White light and EGFP images of midgut tissue obtained from kh^w, Ub_L40-EGFP #18 and PUb-EGFP #P5 mosquitoes.

CONCLUSIONS

We have validated control sequences derived from the Ae. aegypti Ub_L40 and PUb genes capable of driving transgene expression in both cultured mosquito cells and in transgenic mosquitoes. Ub_L40-driven expression was strongest in early larvae and in ovaries, though weak to moderate expression was detected in other tissues and life stages. PUb-driven expression was robust in all life stages and most tissues, with strong expression observed in the midgut. These constructs provide two new expression patterns which will be of great use in future investigations of Ae. aegypti gene regulation and function as well as the generation of virus-resistant phenotypes. The PUb control sequence should also be quite useful as an alternative to 3xP3 in driving the expression of marker genes used for identifying new transformation events, as all six 3xP3-DsRED⁺ lines recovered also exhibited strong PUb-EGFP expression. As the transgenic strains described here were marked with both, it was readily apparent that the PUb-driven fluorescent proteins could be detected at a much lower magnification than required for detection of 3xP3-DsRED. Also, while 3xP3-driven fluorescent proteins are easily visible in white-eyed mutant strains, visualization in wild-type eyes is much more laborious.

Each of the transgene constructs we assembled and tested here also contain a number of features designed to increase the utility of the obtained transgenic strains, such as loxP, attP and homing endonuclease recognition sites. Excision of the Ub_L40-EGFP or PUb-EGFP gene cassettes with cre recombinase, a highly efficient process in Ae. aegypti (Jasinskiene et al. 2003), would allow the re-introduction of a Ub_L40- or PUb-driven gene cassette of interest into a genomic location now shown to be favorable for expression. This would be especially useful for Ub_L40-driven cassettes, as only 1 out of 3 lines appeared to escape repressive position effects. We have also recently shown that the homing endonuclease clusters which were incorporated into line Ub_L40-EGFP #18 and #P17A (also referred to as UUGFP) can be targeted for site specific double-stranded break formation in somatic tissues of Ae. aegypti (Traver et al. 2009). Thus, in addition to validating Ub_L40 and PUb promoter fragments, these transgenic lines will be of great value in studying the genetics of this important disease vector.

EXPERIMENTAL PROCEDURES

Plasmid construction

A promoterless EGFP expression construct was generated by polymerase chain reaction (PCR) of the EGFP-SV40 region from pMos3xP3-EGFP (Horn and Wimmer 2000) using primers F 5′-ctccaccATGGTGAGCAAGGGCGAGGAG-3′ and R 5′-ttttgaattcAAGATACATTGATGAGTTTGGACAAACCAC-3′, with underlined bases indicating restriction enzyme recognition sites added for cloning. This amplicon was digested with NcoI/EcoRI and cloned into the NcoI/EcoRI sites of pSLfa (Horn and Wimmer 2000), generating pSLfa-EGFP-SV40. Putative promoter regions were amplified from Ae. aegypti (Liverpool strain) using Platinum Pfx (Invitrogen; Carlsbad, CA) and Ub_L40 primers F 5′-ttttaagcttAGCGTGAACCTTGCATGCGTGC-3′ and R 5′-ataccatggCTTGGATCAGTCTGTGGAAAAGTCGG-3′ or PUb primers 5′-ttttacgcgtATCTTTACATGTAGCTTGTGCATTGAATCC-3′ and 5′-ttttccatgGTTGAAATCTCTGTTGAGCAGAAAAAGAAACGAGG-3′. PCR conditions were 94°C, 2 min; 94°C, 15 sec; 62°C, 30 sec; 68°C, 1 min; 35 cycles; 68°C, 10 min. Amplified genomic fragments were digested with their respective enzymes and cloned upstream of the EGFP ORF in pSLfa-EGFP-SV40. Two different modified pMos3xP3-DsRED transformation vectors, pMos-3xP3DsRed-RH and pMos-3xP3DsRed-attP were used to test Ub_L40 and PUb promoter-based gene cassettes in transformed mosquitoes. Plasmid pMos-3xP3DsRed-RH was generated by inserting a complete duplication of the Mos1 right hand (RH) downstream of the DsRED gene, as well as a cluster of homing endonuclease recognition sequences (Traver et al. 2009). Plasmid pMos-3xP3DsRed-attP was generated by inserting two clusters of homing endonuclease recognition sites (Traver et al. 2009) and loxP target sites flanking the multiple cloning site (MCS), as well as an attP recombination site downstream of the MCS. The entire Ub_L40-EGFP-SV40 or PUb-EGFP-SV40 gene cassettes were removed from pSLfa and inserted into the MCS (AscI/EcoRI) of the respective Mos1 vector, generating pMos-3xP3DsRed-RH-Ub_L40-EGFP, pMos-3xP3DsRed-Ub_L40-EGFP-attP, and pMos-3xP3DsRed-PUb-EGFP-attP (Fig 2A).

For luciferase experiments, Ub_L40 and PUb genomic fragments were excised from pMos3xP3-DsRED vectors using MluI/NcoI and ligated into the MluI/NcoI sites of pGL3 Basic (Promega; Madison, WI) upstream of the luciferase ORF. The baculovirus IE-1 promoter was amplified using Platinum Pfx (Invitrogen; Carlsbad, CA) from the pIE-Hyg vector (Adelman et al. 2002) using primers 5′-ttttggcgcgccCGCGTAAAACACAATCAAGT-3′ and 5′-ttttccatggGTCACTTGGTTGTTCACGAT-3′. Following PCR, the IE-1 amplicon was digested with AscI/NcoI and ligated into the MluI/NcoI sites of pGL3-Basic.

Cell culture expression assays

Aedes albopictus C6/36 cells and Aedes aegypti Aag2 cells were maintained at 28°C and 5% CO₂ in Dulbecco’s Modification of Eagle’s Medium (C6/36) or Schneider’s Drosophila Medium (Aag2), 10% FBS, 1% 200mM L-glutamine, 1% Penicillin (5,000 I.U./ml) streptomycin (5 mg/ml). C6/36 cells were transfected using Effectene (Qiagen; Valencia, CA) according to the manufacturer’s instructions. Aag2 cells were transfected using Fugene (Roche; Basel, Switzerland) using a ratio of 3μl Fugene reagent to 1μg of DNA (Meredith et al. 2006). Cells were seeded in 12-well plates and were transfected at ~50% cell confluency. C6/36 or Aag2 cells were co-transfected with 500ng of pGL3 construct (no promoter, IE1-luc, Ub_L40-luc or PUb-luc) and 25 ng pRL-CMV per well. At 48hrs post transfection, cells were washed twice with Phosphate Buffered Saline (PBS) and scraped with a rubber policeman in 1X Passive Lysis Buffer (PLB), freeze-thawed once and centrifuged briefly to remove cell debris. Lysates were diluted 1:100 in 1X PLB and luciferase activity was determined using the Dual-Luciferase Reporter Assay System on a GloMax 20/20 Luminometer according to the manufacturer’s instructions (Promega; Madison, WI).

Mos1-mediated transformation of Aedes aegypti

Aedes aegypti (kh^w, Liverpool strains) were maintained at 28°C and 60%-80% relative humidity on a photoperiod of 16 hr light/8 hr dark, as described previously (Adelman et al. 2008). Mosquitoes were maintained by bloodfeeding with defibrinated sheep blood (Colorado Serum Company, Denver, CO) using artificial feeders and parafilm membranes. For transgenic experiments, kh^w or Liverpool embryos were microinjected with 0.5μg/μl Donor plasmid (pMos-3xP3DsRed-RH-Ub_L40-EGFP, pMos-3xP3DsRed-Ub_L40-EGFP-attP, or pMos-3xP3DsRed-PUb-EGFP-attP) and 0.3μg/μl pKhsp82M in 1X Injection Buffer (Coates et al. 1998). Surviving G₀ females were merged into pools of 20-25 and mated to males of the recipient strain. G₀ males were mated individually to 10-15 recipient strain females and pooled prior to bloodfeeding and egg laying. All microinjections were performed using a Leica micromanipulator and a FemtoJet microinjector (Eppendorf; Westbury, NY). G₁ larvae were screened for DsRED⁺ eyes and/or GFP⁺ bodies using a Leica MZ-16FL microscope. All photos were taken using a Canon PowerShot S3 IS digital camera with a constant exposure time of 0.8 seconds.

Southern analysis

Mosquitoes were snap frozen in nitrogen and genomic DNA was isolated from 6 females or 10 males as described previously (Adelman et al. 2008). Genomic DNA was digested overnight with SalI, XbaI, EcoRI or NsiI, as indicated, followed by gel electrophoresis and capillary transfer to a nylon membrane. A probe corresponding to HindIII restriction fragments of the Mos1 construct was randomly primed and labeled with [α-³²P]-dATP, using the Amersham Megaprime DNA labeling System (GE Healthcare; Buckinghamshire, UK) and purified using an Illustra NICK column (GE Healthcare; Buckinghamshire, UK). Following hybridization overnight at 65°C, membranes were washed and exposed to Kodak BioMax maximum sensitivity film at −80°C.

RNA isolation, RACE and Northern analysis

Ae. aegypti adults, larvae and pupae were snap frozen in liquid nitrogen and stored at −80°C. Adult mosquitoes were dissected in PBS and heads, thorax (without salivary glands), salivary glands, midgut, ovaries (sugar-fed or 4 days post blood meal) were placed into eppendorf tubes and snap frozen in liquid nitrogen and stored at −80°C until RNA extractions were performed. Total RNA was extracted using Trizol reagent (Invitrogen; Carlsbad, CA) according to the manufacturer’s protocol. For 5′ and 3′ RACE, polyA⁺ RNA was purified with the PolyATract mRNA Isolation System (Promega; Madison, WI) according to the manufacturers’ instructions. RACE-ready cDNA was amplified using the SMART RACE cDNA Amplification Kit (Clontech; Mountain View, CA) and Superscript II RNase H-Reverse Transcriptase (Invitrogen; Carlsbad, CA). RACE PCR was performed using the Advantage 2 PCR Enzyme System (Clontech; Mountain View, CA) according to the SMART RACE cDNA Amplification Kit User Manual with one SMART primer and one gene specific primer. Gene specific primers for 5′RACE: 5′-GGATCAGTTCTTTCCCAACGTGTCTCTTCC-3′ (Ub_L40), 5′-GATTCCCTTCACCGAAATGCACTTTCAC-3′ and 5′-GGGTTTTGACGAAAATCTGCATTGTTGAAATCT-3′ (PUb) and 3′RACE 5′-ACTTCAATTTCCCCCCGTTCCAAGACCTGG-3′ (Ub_L40) and 5′-GAGAGGCGGACAGTAAAGATGATGTTGAGC-3′ (PUb). RACE amplicons were cloned using the TOPO TA Cloning Kit for Sequencing (Invitrogen; Carlsbad, CA).

For Northern analysis, three micrograms of total RNA was electrophoresed and transferred to nylon membrane. For probes, PCR amplicons derived from the 3′UTR of Ub_L40 (primers F 5′-TGGTGGTATTATTGAGCCATCTCTGCG-3′ and 5′-ATAGCAATGCATTGTGACAACTTCAATTC-3′); the 5′UTR of PUb (F 5′-AATCGAATACGTTTCCTAGT-3′ and R 5′-GGGTTTTGACGAAAATCTGCATTGTTGAAATCT-3′); or the EGFP coding region (F 5′-TTTTGCGGCCGCCATGGTGAGCAAGGGCGAGGAGC-3′ and R 5′-TTTTGGATCCCAAGATACATTGATGAGTTTGGACA-3′) were [α-³²P]-dATP labeled through random-priming. Specific activity ranged from 7.2×10⁹ to 5.2×10¹⁰ counts per minute for Ub_L40 and EGFP probes, 9.4×10¹⁰ to 2.0×10¹¹ counts per minute for PUb probes. Membranes were hybridized overnight at 65°C for EGFP/Ub_L40 probes or 55°C for PUb probe, washed and exposed to Kodak BioMax maximum sensitivity film.