A DREaMR system to simplify combining mutations with rescue transgenes in Aedes aegypti

Abstract

In most experimental animals, it is challenging to combine mutations and rescue transgenes and to use bipartite systems to assess gene expression. To circumvent the difficulties in combining multiple genetic elements, we developed the DREaMR (Drug-on, REporter, Mutant, Rescue) system. Using Drosophila white as the initial model, we demonstrated that introduction of a single insertion by CRISPR/Cas9 created a null mutation, a tagged rescue construct, which could be induced with doxycycline, and which allowed assessment of protein expression. To create a DREaMR in an organism in which combining multiple genetic elements is more problematic than in Drosophila, we tested the mosquito, Aedes aegypti—the insect vector for dengue, yellow fever, Zika, and other viral diseases. We generated a DREaMR allele in the kh gene, which permitted us to induce expression of the rescue construct, and detect expression of Kh. Thus, this system avoids the need to perform genetic crosses to introduce an inducible rescue transgene in a mutant background, or to combine driver and reporter lines to examine expression of the targeted protein. We propose that DREaMR provides a system that can be applied to additional mosquito vectors as well as other organisms in which CRISPR/Cas9 is effective.

Chen, Luo, Gurav et al. present the DREaMR (Drug-on, REporter, Mutant, Rescue) system for combining a mutation, a rescue transgene, and to assess gene expression using a single insertion in the endogenous gene. They demonstrate the effectiveness of the DREaMR system using the Drosophila white and Aedes kh genes. The authors suggest that the DREaMR system can be used in other organisms in which it is possible to employ CRISPR/Cas9.

Introduction

CRISPR/Cas9 has revolutionized the ability to introduce targeted mutations and transgenes in many animals in which gene manipulation was formerly very difficult to perform, or even inaccessible (Wright et al. 2016). These include animals ranging from the squid to an array of farm animals and many types of insect vectors that spread devastating diseases (Reardon 2019; Shaw and Catteruccia 2019; Menchaca et al. 2020). While CRISPR/Cas9 enables the introduction of genetic elements, in most animals combining them can be time consuming. One notable exception is the fruit fly, Drosophila melanogaster, which is endowed with balancer chromosomes, each of which consists of multiple inversions and a dominant marker, thereby simplifying genetic manipulations (Kaufman 2017; Miller et al. 2019). As such, there are a wide set of invaluable tools in Drosophila, such as bipartite expression systems, and transgenes to monitor and manipulate cellular activities (del Valle Rodriguez et al. 2012).

The DREaMR system is designed to accomplish three goals with a single CRISPR/Cas9 mediated insertion. First, it creates a mutation in a gene of interest. Second, it uses a tetracycline-inducible expression system (Bieschke et al. 1998; Stebbins et al. 2001; Lycett et al. 2004; Das et al. 2016) to introduce a transgene that restores function. Third, the rescue construct is tagged so that protein expression can be assessed. In addition to establishing the effectiveness of the DREaMR system in Drosophila, we tested the efficacy of the DREaMR system in Aedes (Ae.) aegypti. This mosquito spreads pathogens that cause diseases such as dengue that afflict between 100 and 400 million people annually (Messina et al. 2019; Shaw and Catteruccia 2019; Girard et al. 2020). Consequently, there have been efforts to develop methods to control Ae. aegypti populations using genetic approaches (Fu et al. 2010; Li et al. 2020, 2021; Navarro-Payá et al. 2020; Chae et al. 2021; Chen et al. 2021). However, it is difficult to combine multiple elements into one genetic background due in part to the absence of balancer chromosomes. In addition, in contrast to Drosophila melanogaster, which has a genome only 4.5% the size of humans, the Ae. aegypti genome is 40% as large as the human genome (Nene et al. 2007). We found that the DREaMR system is effective in Ae. aegypti, thereby circumventing some of the limitations that complicate combining genetic elements in this insect.

Materials and methods

Insect strains and rearing

The wild type Canton-S and the w¹¹¹⁸ (stock #5905) strain of D. melanogaster were obtained from the Bloomington Drosophila Stock Center, and maintained at 25°C on standard fly food prepared with corn flour and yeast.

The following Ae. aegypti strains were used in the study: wild-type Liverpool strain (gift from Omar S. Akbari), ubiL40-Cas9 and kh¹ (Li et al. 2017). Mosquito eggs were hatched, and larvae were grown in deionized water with fish food (TetraMin Tropical Granules, Tetra, 98531). The adult mosquitoes were maintained with 10% sucrose in 15 × 15 × 15 cm cages (BugDorm-4M1515 insect rearing cage, MegaView Science Co., Ltd., Taiwan), at 28°C, 80% humidity and under 14 h light/10 h dark cycles. Females were fed defibrinated sheep blood (HemoStat Laboratories, DSB250) at 37°C using the Hemotek membrane blood-feeding system (Hemotek, SP6W1-3) and eggs were collected on a damp filter paper 4–5-day post blood meal.

Molecular cloning

To aid the generation of DREaMR constructs, we created several backbone plasmids. The pDmU6-LgRNA-3xP3-GFP backbone plasmid (Supplementary Figure S1A) is similar to pAaU6-LgRNA-3xP3-GFP (Supplementary Figure S1B), which we previously reported for CRISPR/Cas9 editing in Ae. aegypti (Chen et al. 2021), except that it has a Drosophila U6:3 promoter to drive gRNA expression in Drosophila (Port et al. 2014).

We created several DREaMR generic constructs (pDREaMR-SC1; Supplementary Figure S1, D–G) that contain a DREaMR subcloning cassette flanked by unique KpnI and BmtI sites. The DREaMR subcloning cassettes contain the following: (1) a rtTA coding sequence synthesized by GenScript; (2) a 3xP3-DsRed or 3xP3-GFP transgenic marker that was PCR-amplified from the plasmid p3xP3-UASpBacFPN (DGRC#1287) or pHD-DsRed (Addgene #51434) and; (3) a 1xTRE or 3xTRE followed by the DSCP (Drosophila synthetic core promoter) (Pfeiffer et al. 2008) synthesized by GenScript. Ligations were performed using the In-Fusion HD Cloning Plus kit (Takara) to piece the sequences together. The four plasmids are referred to as: pDREaMR-SC1-DsRed-1xTRE, pDREaMR-SC1-DsRed-3xTRE, pDREaMR-SC1-GFP-1xTRE, and pDREaMR-SC1-GFP-3xTRE (Supplementary Figure S1, D–G).

To generate the final injection construct for creating w^DR1(pDm-w^DR1), we first introduced the gRNA sequence (GTTGACCAGCTGCCGCCATC) targeting the w gene into pDmU6-LgRNA-3xP3-GFP. To achieve this, we PCR-amplified the gRNA containing the 20-nucleotide target sequence and scaffold using pDmU6-LgRNA-3xP3-GFP as a template with the following primers (gRNA target sequence is underlined): forward: 5’AAGAGTAGTGAAATGGTTGACCAGCTGCCGCCATCGTCTTAGAGCTATGCTGGAAACAG-3′; reverse: 5′-GCCGCTCTAGAACTAGTGGATCCCCC-3′ (Figure 2A). We then subcloned this gRNA into pDmU6-LgRNA-3xP3-GFP by digesting it with KpnI and SpeI and using the In-Fusion Cloning system (Takara, In-Fusion HD Cloning Plus kit) (Figure 2, A and B). Next, we PCR-amplified two homologous arms from genomic DNA prepared from adult flies (left arm, 2.14 kb upstream of the target site; right arm, 2 kb downstream of the target site). We also synthesized a DNA sequence that contains (5’ to 3’) a self-cleavage peptide sequence P2A or P2A-T2A (Kim et al. 2011; Diao and White 2012; Liu et al. 2017), an adaptor DNA with a unique FseI restriction site flanked by sequences that correspond the 5’ and 3’ end sequence of the DREaMR cassette, and a recoded sequence (using synonymous codons) that includes the N-terminal 90 amino acids of the White protein (Supplementary Figure S2). The sequence encoding this N-terminus segment was recoded to reduce homology with the corresponding chromosomal region. The adaptor sequence containing the FseI site (underlined) is 5′-GCCACCATGAGTAGGTTGGACAAGTCCAAGGCCGGCCAACTTTGAATCACAAGACGCATACCAAACG-3′.

Figure 2.

Cloning flow chart for creation of the DREaMR constructs (pw^DR1, pw^DR2, and pkh^DR) for generation of the DREaMR alleles. All plasmids were grown in E. coli under ampicillin selection. (A) Introducing gRNA target into the pDmU6-LgRNA-3xP3-GFP or pAaU6-LgRNA-3xP3-GFP backbone plasmids (Supplementary Figure S1, A and B). The 20 nucleotide target sequence is introduced with the forward primer (see Materials and Methods for primer sequences used here) to amplify the gRNA with the target sequence. (B) Generating the intermediate plasmid, pDREaMR-Int (Supplementary Figure S1C), by linearizing the indicated plasmid with the target gRNA with PacI and NheI and subcloning: (1) the left homologous arm, (2) a synthesized DNA sequence, which contains a P2A or P2A-T2A cleavage peptide, an adaptor sequence with a unique FseI site (see Materials and Methods for the adaptor sequence), and the synonymously recoded CDS upstream of the knock-in site (can be untagged or tagged with a 3xHA at its N-terminus), and (3) the right homologous arm. (C) Step to generate the injection construct. Excise the DREaMR cassette from a pDREaMR-SC1 plasmid (Supplementary Figure S1, D–G) with BmtI and KpnI. Linearize the pDREaMR-Int plasmid with FseI and subclone the DREaMR cassette by performing an In-Fusion ligation. (D) Final plasmid for injections to generate DREaMR knock-in alleles. Note that there is no 3xHA tag in pw^DR1 construct. Abbreviations: AaU6, Aedes U6 promoter; DmU6, Drosophila U6:3 promoter; DSCP, Drosophila synthetic core promoter; DsRed, red fluorescent protein; GFP, green fluorescent protein; P2A, porcine teschovirus-1 2A self-cleavage peptide; T2A, Thosea asigna virus 2A; rtTA, reverse tetracycline-controlled transactivator; TRE, tetracycline responsive element.

We then digested the pDmU6-LgRNA-3xP3-GFP plasmid, which contained the gRNA targeting w, with PacI and NheI, and subcloned the left homologous arm, the synthesized DNA sequence mentioned above, and the right homologous arm to generate an intermediate plasmid (pDREaMR-Int, Figure 2, B and C and Supplementary Figure S1C) using the In-Fusion cloning system (Takara). Finally, we digested the resulting pDREaMR-Int plasmid with FseI (Figure 2C), and introduced the DREaMR-1xTRE cassette (digested with BmtI and KpnI from pDREaMR-SC1-DsRed-1xTRE; Figure 2C) using the In-Fusion cloning system to create pDm-w^DR1 (Figures 1A, 2C, and D). The creation of pDm-w^DR2 (Figure 3A) was similar, except that we added a 3xHA tag at the beginning of the coding sequence (CDS) that was recoded with synonymous codons distinct from the natural codons, and we used the DREaMR-3xTRE cloning cassette, which we obtained from the DREaMR-SC1-GFP-3xTRE (Supplementary Figure S1G) by digesting with BmtI and KpnI.

Figure 1.

DREaMR kno ck-in in the Drosophila w gene. (A) Schematic showing the DREaMR cassette knocked into the Drosophila white (w) gene to create the w^DR1 allele. The DREaMR construct includes: (1) P2A::rtTA, (2) DsRed expressed under control of the 3xP3 promoter, and (3) the 5’ end of the w coding region upstream of the knock-in site, which is recoded with silent substitutions (encoding amino acid residues 1–90; Supplementary Figure S2). This reconstructed w gene is expressed under the control of one copy of the TRE, which is activated by rtTA bound to doxycycline. The DNA encoding P2A::rtTA is inserted in-frame with the w coding region. F1 and R1 are the forward and reverse DNA primers used to perform the PCR shown in Figure 1B. Abbreviations: P2A, porcine teschovirus-1 2A self-cleavage peptide; rtTA, reverse tetracycline-controlled transactivator; TRE, tetracycline responsive element; 3xP3, three P3 binding elements and a minimal promoter from the Drosophila hsp70 gene; DsRed, red fluorescent protein. The entire knock-in is 2931 bp. (B) PCR genotyping of the Drosophila w^DR1 allele. The locations of the primers used (F1 and R1) are indicated by the arrows in Figure 1A. The expected PCR products are 367 base pairs in wild type and 3.3 kb in w^DR1. (C–H) Eye color phenotypes of males of the indicated genotypes and treatments. Scale bar, 200 µm. (I–N) Eye color phenotypes of females of the indicated genotypes and treatments. Flies were reared in food containing the indicated concentrations of doxycycline starting from the embryonic stage. Scale bar, 200 µm.

Figure 3.

DREaMR2 in the Drosophila w gene. (A) Schematic showing the DREaMR2 cassette with 3xTREs knocked into the w gene creating the w^DR2 allele. The recoded w sequence in the knock-in transgene included amino acid residues 1 to 90 fused to a 5’ sequence encoding an HA tag. The 3xP3 GFP eye marker is used as the transgenic marker. The knock-in fragment is 3623 base pairs. F1 and R1 are the forward and reverse DNA primers used to perform the PCR shown in Figure 3B. (B) PCR genotyping of the Drosophila w^DR2 allele. The locations of the F1 and R1 primers used for the PCR are presented in Figure 3A. The expected PCR products are 367 base pairs in wild type and 4.0 kb in w^DR2. (C–I) Eye color phenotypes exhibited by the control and w^DR2. The genotypes and doxycycline treatments are indicated. The flies were reared in food containing the indicated concentrations of doxycycline starting from the embryonic stage. Scale bar, 200 µm.

The generation of pAa-kh^DR was similar to the generation of pDm-w^DR2. We introduced the gRNA sequence (GCCCACGATCAGATCGGCCT) targeting the kh gene into pAaU6-LgRNA-3xP3-GFP (Chen et al. 2021) (Figure 2, A and B and Supplementary Figure S1B). We PCR-amplified two kh homologous arms from genomic DNA prepared from adult mosquitoes (left arm, 870 bp upstream of the target site; right arm, 1.2 kb downstream of the target site). We synthesized a DNA sequence that contains (5’–3’) a P2A-T2A tandem, an adaptor sequence that corresponds the 5’ and 3’ end sequence of the DREaMR cassette and a unique FseI restriction site, and included a recoded sequence (using synonymous codons) corresponding to the 5’ end of the Aa kh coding region (amino acid residues 1–170; Supplementary Figure S5) 5’ of the knock-in site (Figure 2B). The sequence encoding this N-terminus segment was recoded to reduce homology with the corresponding chromosomal region. We next digested the pAaU6-LgRNA-3xP3-GFP plasmid containing the kh gRNA target with PacI and NheI (Figure 2B), and subcloned the left homologous arm, the synthesized DNA sequence mentioned above, and the right homologous arm to generate an intermediate plasmid (pDREaMR-Int; Figure 2, B and C and Supplementary Figure S1C), using the In-Fusion cloning system (Takara). We then digested the resulting plasmid with FseI and introduced the DREaMR cloning cassette digested with BmtI and KpnI from pDREaMR-SC1-DsRed-1xTRE (Figure 2C and Supplementary Figure S1D), using the In-Fusion cloning system (Takara) to create the final construct for injection into embryos (Figure 2D).

Some constructs included the P2A-T2A rather than just the P2A cleavage peptide to test whether it increased the cleavage efficacy. However, we did not observe differences in the cleavage efficiency with the P2A-T2A vs the P2A.

Transgenic insects

To generate the Drosophila w^DR1, and w^DR2 lines, the plasmids were injected into nos-Cas9 (BDSC#78782) embryos (BestGene Inc) and transformants were identified on the basis of DsRed or GFP fluorescence in the eyes. To obtain the Ae. aegypti kh^DR strain, we injected the pAa-kh^DR plasmid at a concentration of 400 ng/µl into ∼1000 ubiL40-Cas9 embryos that express Cas9 (Li et al. 2017). Transformants were identified on the basis of DsRed fluorescence in the eyes. Approximately 50 G₀ mosquitoes survived. We then crossed individual G₀s to the wild-type (Liverpool) mosquitoes of the opposite sex and harvested the F₁ eggs. We hatched the eggs and screened the F₁ larvae for transformants on the basis of DsRed expression in the eye using a Zeiss stereo fluorescence microscope (Zeiss SteREO Discovery.V8). We obtained 4 positive F₁ progeny from one female G₀. We established four transgenic lines (line #1, #2, #3, and #4). We confirmed that all lines had the desired mutations by PCR and DNA sequencing. The primers used for the PCR were kh-F1, 5′-CTGGATGAGGGAGAGATGAGTTTCATTG-3′ and kh-R1, 5′-GGCCAAATGTGCAAATAGTTGTGAGGC-3′.

Since all four lines were identical, we arbitrarily selected line #1 to outcross to the wild-type Ae. aegypti Liverpool strain for four generations. We then performed inter se crosses and established a kh^DR homozygote stock by picking mosquitoes with darker pink pigmentation in their eyes due to the 3xP3-DsRed, which is expressed in the eyes. We confirmed that the line was homozygous by PCR. We used the resulting line for all experiments with kh^DR in this study. PCR genotyping was performed on ≥3 biological replicates.

Inducing expression of the rescue constructs with doxycycline

To rescue the phenotypes exhibited by the Drosophila DREaMR flies, the female parents were allowed to lay eggs on standard fly food containing the indicated concentration of doxycycline (Sigma), and the adult animals were also fed on doxycycline-containing food. The food with doxycycline was prepared by adding various amounts from a 10 mg/ml doxycycline stock solution to warm, melted fly food.

To induce expression of the Ae. aegypti DREaMR rescue transgenes, the eggs were hatched in deionized water with sprinkles of fish food (TetraMin Tropical Granules, Tetra, 98531). The next day, different amounts of 10 mg/ml doxycycline stock solution were added to the water to reach the indicated concentrations. The larvae and pupae were maintained in the doxycycline-treated water with fish food (TetraMin Tropical Granules, Tetra, 98531) until eclosion. The adult animals were fed 10% sucrose containing the same concentration of doxycycline that we used in the water during development.

Examination of eye colors

For visualization of the eye phenotypes, fruit flies (5–7-day old) and mosquitoes (5–7-day old) were frozen in a 20°C freezer for 2 h. They were then viewed using a Zeiss Axio Zoom.V16 stereomicroscope and the images were processed with Zeiss ZEN software. For each condition, ≥3 mosquitoes (biological replicates) were imaged, and representative images are shown in the Figures.

Western blotting

Whole Drosophila (5–7-day old) or Ae. aegypti (1–2-day old) heads were dissected and homogenized in lysis buffer (50 mM Tris-HCl pH 6.5, 0.1 M DTT, 2% SDS, 1.25 mM bromophenol blue, 1.1 M glycerol). Ten microliter lysis buffer per head was used to lyse 5–10 heads in each Drosophila sample, and 1.5 µl lysis buffer per head were used to lyse 20–40 heads in each mosquito sample. The equivalent of ∼1 head per lane was loaded for Drosophila samples and ∼7 heads per lane were loaded for the mosquito samples on Mini-PROTEAN Precast Gels (BIO-RAD). Proteins were fractionated by SDS-PAGE and transferred to 0.2 μm pore-size nitrocellulose membranes (BIO-RAD). The blots were blocked with LI-COR blocking buffer and then probed with primary antibodies: rabbit anti-HA antibody (1:3000, SG77, Invitrogen) and mouse anti-tubulin antibody (1:10000, Developmental Studies Hybridoma Bank). For secondary antibodies, IRDye800CW Goat anti-mouse and IRDye680LT Goat anti-rabbit antibodies (LI-COR Biosciences) were used at 1:20,000. Images of the blots were acquired using an LI-COR Odyssey system. ≥3 biological replicates were assayed for each condition.

Immunostaining of Drosophila and Aedes eyes

We used Drosophila (w^DR2) and Aedes (kh^DR1) adults for immunostaining. We cultured the doxycycline-treated Drosophila w^DR2 in cornmeal-yeast-agar medium supplemented with 20 µg/ml doxycycline. Control w^DR2 flies without doxycycline were cultured on cornmeal-yeast-agar medium only. To treat the Aedes kh^DR1 with doxycycline, we added the drug to the water at a concentration of 20 µg/ml. Control Aedes were cultured without doxycycline. ≥3 biological replicates were examined for each condition.

We fixed the insects (5–7-day old Drosophila or 1–2-day old Aedes) in PBS (9 g/L NaCl, 144 mg/L KH₂PO4 and 795 mg/L Na₂HPO4, pH 7.4) plus 4% paraformaldehyde on ice for 2 h, and then dissected out the eyes in PBS + 0.1% Triton X-100 right after fixation. We incubated the dissected eyes in blocking buffer (100 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% Triton X-100, 10% normal goat serum) at room temperature for 1 h. After blocking, we incubated the samples with primary antibodies diluted in blocking buffer at 4°C for 2 days. After washing the samples in PBS + 0.1% Triton X-100 for 20 min three times, we subsequently incubated the samples with secondary antibodies at 4°C overnight. We again washed the samples three times in PBS + 0.1% Triton X-100, and mounted them in Vectashield (Vector labs). We acquired images from whole mounted Drosophila or Aedes eyes with optical sections at 1 µm using a Zeiss LSM 700 confocal microscope with a Plan-Apochromat 63x objective.

The w^DR2 eyes were stained with mouse anti-HA (Sigma, H3663) at 1:100, rabbit anti-PDH (Wang et al. 2010) at 1:200, or rabbit anti-TRP at 1:500 (Chevesich et al. 1997) as primary antibodies. We then used goat anti-mouse Alexa Fluor 568 (ThermoFisher, A11004) and goat anti-rabbit Alexa Fluor 633 (ThermoFisher, A21070) both at 1:200 as secondary antibodies.

To stain the kh^DR1 eyes, we used rabbit anti-HA (ThermoFisher, 71-5500) at 1:100 as the primary antibodies, and used goat anti-rabbit Alexa Fluor 633 (ThermoFisher, A21070) at 1:200 for the secondary antibodies. To co-stain the DNA, we added DAPI (Sigma D9542, 1 mg/ml) at 1:1000 together with the secondary antibodies.

Assessment of doxycycline on the mosquito life cycle and lifespan

To assess whether the levels of doxycycline that rescued the eye phenotype in kh^DR animals impacted development, we examined the life cycle of the mosquitoes. We performed the analyses using wild-type Liverpool eggs under different doxycycline treatments. We used the same batch collection of eggs for all experimental groups. Three biological replicates were measured for each condition. In each group, 200 eggs that were 4-days post collection were hatched in one liter of deionized water supplemented with fish food under the indicated doxycycline concentrations. The number of 4$\mathrm{th}$ instar larvae, pupae and adults under the indicated doxycycline treatment were counted and the times for pupal and adult development were recorded.

To analyze whether doxycycline had an impact on survival of Ae. aegypti, we performed a lifespan assay. We cultured wild-type mosquitoes from the same batch of eggs with 0, 200, or 1000 ng/ml doxycycline. We introduced ∼30 newly eclosed males and ∼30 newly eclosed females in 15 × 15 × 15 cm cages. The adults were fed 10% sucrose supplemented with the same concentration of doxycycline used in the water during development. We then recorded the number of living males and females over time. For each doxycycline condition, 3 groups (cages) were assayed.

Statistical analysis

To determine whether doxycycline had an impact on development of Ae. aegypti (Life Table), we performed one-way ANOVA with Tukey’s multiple comparisons test. To determine whether doxycycline impacted on survival of adult mosquitoes, we analyzed the statistics using the Cox Proportional Hazards model using the “survival” R package.

Results

DREaMR in Drosophila white

To conduct a test of the DREaMR system, we introduced an insertion in the X-chromosome-linked white (w) gene of D. melanogaster, since null mutations in w result in an easily scorable phenotype—conversion of red to white eyes. To eliminate gene function, we designed the insertion cassette to interrupt the protein-coding sequence within the second exon (after residue 90), thereby creating the w^DR1 allele (Figure 1A). To obtain an inducible rescue transgene, we regenerated a full w coding sequence, by inserting amino acid residues 1–90 in frame with the rest of the w gene (Figure 1A). The sequence encoding this N-terminal segment in the DREaMR construct was recoded (replaced with synonymous codons) to reduce homology with the corresponding chromosomal region, and thereby prevent homology-directed repair with the rescue transgene (Supplementary Figure S2). The reconstructed w gene was placed under the control of a yeast promoter (tetracycline response element; TRE) that is activated by the reverse tetracycline-controlled transactivator (rtTA) (Das et al. 2016).

We introduced the gene encoding rtTA in frame near the beginning of the w coding sequence (after residue 90), so that it was expressed under control of the endogenous w promoter. The final construct, which was created through multiple steps (Figure 2 and “see Materials and Methods”) was injected into nos-Cas9 embryos (Ren et al. 2013) to generate the w^DR1 line. Following transcription and translation, rtTA would be released since it is flanked at the N-terminus with P2A (a 2A self-cleavage peptide) (Kim et al. 2011; Diao and White 2012). Phenotypic rescue would only occur upon addition of tetracycline or related analogs (i.e., doxycycline) because the rtTA must bind to the drug to associate with the TRE and activate transcription.

We confirmed the w^DR1 allele by PCR (Figure 1B) and DNA sequencing. Wild-type males have red eyes (Figure 1C). In the absence of doxycycline, the mutant males (w^DR1/Y) have white eyes indistinguishable from the w¹¹¹⁸ null mutant (Figure 1, D and E), indicating that w^DR1 is a null allele. To test if the phenotype could be rescued by addition of doxycycline, we allowed the animals to develop on food containing doxycycline. We found that w^DR1 males showed rescue of the phenotype upon addition of doxycycline. In the presence of 10 and 50 µg/ml doxycycline, the flies showed slight eye pigmentation, and orange eyes, respectively (Figure 1, F and G). When we increased the doxycycline levels to 100 µg/ml, the w^DR1 males exhibited red eyes comparable to wild type indicating full phenotypic rescue (Figure 1, C and H). Wild-type females also have dark red eyes (Figure 1I), while the w¹¹¹⁸ females have white eyes (Figure 1J). The w^DR1 mutant females exhibited a white-eye phenotype indistinguishable from the w¹¹¹⁸ (Figure 1, J and K). The pigment of the w^DR1 homozygote female flies increased in a dose-dependent manner, and was fully restored at 50 µg/ml doxycycline (Figure 1, L–N). The complete rescue in w^DR1 females required a lower concentration of doxycycline than in males (50 vs 100 µg/ml), which was surprising since transcription of the X-linked w gene would be expected to be dosage compensated in males due to an increase in the rate of elongation of RNA polymerase II (Lucchesi 2018). Dosage compensation occurs at the level of individual genes (Lucchesi 2018), suggesting that the insert of the transgenes in w might disrupt dosage compensation.

Modified DREaMR in Drosophila white

We created a second version of the DREaMR system (DR2), which included a third element—an epitope tag so that it would be possible to detect the cellular expression of the targeted protein. To do so, we added an N-terminal HA tag to the doxycycline-induced rescue transgene (Figure 3A). We also included three copies instead of just one copy of the TRE to test whether this would rescue the w phenotype using lower doxycycline concentrations (Figure 3A). We inserted the modified DREaMR cassette into the w gene at the same knock-in site as w^DR1 to generate w^DR2 (Figure 3A) and confirmed the creation of the w^DR2 allele by PCR (Figure 3B) and DNA sequencing.

We examined the phenotype of w^DR2 and found that both the males and females displayed white eyes (Figure 3, C and G). To test for phenotypic rescue, we added doxycycline to the larval food. At 2 µg/ml, the eye color of w^DR2 males was partially rescued to orange, while at 5 and 20 µg/ml the eye color was rescued completely (Figure 3, D–F). w^DR2 females showed a complete rescue at a concentration of 2 µg/ml doxycycline, which was lower than that was required for a complete rescue in males (Figure 3, E and I). Thus, the rescue with the 3xTRE version of the DREaMR was achieved at lower doxycycline levels than with the 1xTRE version of the DREaMR.

The DR2 rescue construct was tagged with HA at its N-terminus to facilitate detection of the protein targeted by the DREaMR insertion. Therefore, we fed the larvae doxycycline, and performed Western blots using adult head extracts. We found that protein expression was induced by doxycycline (Figure 4A and Supplementary Figure S3). The DREaMR also allows for examination of the spatial distribution of the W protein, since the rtTA that drives expression of the rescue transgene is under control of the endogenous w promoter. The W protein is known to be expressed in pigment cells (PCs) and the cell bodies of photoreceptor cells (Borycz et al. 2008). We found that in the absence of doxycycline treatment, w^DR2 showed only very low levels HA::W that were barely detectable (Supplementary Figure S4, A and D). We co-stained with anti-PDH, which stains PCs (Wang et al. 2010), to demonstrate that retinal tissue was present (Supplementary Figure S4, B, C, E, and F). Upon induction with 20 µg/ml doxycycline, we detected strong HA::W staining in many PCs, which were also labeled by PDH (Figure 4, B–D and H–J). The remaining cells expressing HA::W are photoreceptor cells, which include a microvillar structure, called the rhabdomeres that are stained by anti-TRP (Figure 4, F, G, L, and M), and extra-rhabdomeral cell bodies that were stained with anti-HA (Figure 4, E, G, K, and M). Thus, a single DREaMR insertion can be used to create a mutation, induce expression of a rescue construct to restore function, and detect the expression of the targeted protein.

Figure 4.

Doxycycline-induced protein expression in Drosophila DREaMR2 knocked into the w gene. (A) Western blot showing expression of the HA-tagged White protein in wild type (Canton S) and w^DR2 homozygous flies. Genotypes and doxycycline concentrations added to the food are indicated. Protein size markers (kD) are shown to the left. (B–M) Whole mount staining of w^DR2 eyes with either anti-HA (green) to detect the HA::W protein or anti-PDH (red), which is a pigment cell (PC) marker (Wang et al. 2010). The eyes were dissected from flies that were reared on food containing 20 µg/ml doxycycline. (B–D) Longitudinal view of a retina from a fly showing co-staining with anti-HA to detect HA::W and anti-PDH, which labels PCs. (B) Anti-HA. Scale bar, 10 µm. (C) Anti-PDH. (D) Merge of (B) and (C). (E–G) Longitudinal view of a retina from a fly showing co-staining with anti-HA to detect HA::W and anti-TRP, which labels photoreceptor cells. (E) Anti-HA. Scale bar, 10 µm. The arrow indicates an R8 photoreceptor cell stained with anti-HA. (F) Anti-TRP. (G) Merge of (E) and (F). The following retinal layers are indicated: cone cell (CC), R7 and R8. (H–J) Cross-sectional view of a retina at the R8 layer showing co-staining with anti-HA and anti-PDH. (H) Anti-HA. Scale bar, 5 µm. (I) Anti-PDH. (J) Merge of (H) and (I). (K–M) Cross-sectional view of the retina at the R8 layer showing co-staining with anti-HA and anti-TRP. (K) Anti-HA. Scale bar, 5 µm. (L) Anti-TRP. (M) Merge of (K) and (L). R, photoreceptor cell.

DREaMR in mosquito Aedes aegypti kh

To determine whether the DREaMR system is effective in an animal in which genetic crosses are more difficult to perform due to a lack of balancer chromosomes, we focused on Ae. aegypti. We targeted the kynurenine 3-monooxygenase (kh) gene, because null mutations in kh convert the black eyes to white (Aryan et al. 2013; Li et al. 2017), similar to the effects of the Drosophila w mutation. We designed the DREaMR cassette to insert in kh after residue 170 (Figure 5A; kh^DR). As with the Drosophila DREaMR, the rescue construct (covering amino acid residues 1–170) was recoded (Supplementary Figure S5) to prevent unintended recombination with this transgenic sequence and the chromosomal kh gene. We included an HA epitope tag at the N-terminus of the rescue construct (Figure 5A) to enable us to spatially localize the Kh protein. Due to the very low but detectable level of expression of the Drosophila w rescue transgene using 3xTRE in the absence of doxycycline (Figure 4A and Supplementary Figure S4, A and D), we used the 1xTRE to mitigate background expression of Kh.

Figure 5.

DREaMR knocked into Aedes kh. (A) Schematic showing the DREaMR cassette introduced in the kh gene to create the kh^DR allele. The kh coding sequence 5’ to the knock-in site is recoded with silent changes (encoding amino acid residues 1 to 170; Supplementary Figure S5), and introduced into the indicated transgene. This transgene includes a P2A-T2A::rtTA, which is inserted in-frame with the kh coding region. The 3xP3 DsRed eye marker is used as the transgenic marker. The entire knock-in fragment is 3291 base pairs. (B) PCR genotyping of the kh^DR allele. The locations of the primers used (F1 and R1) are presented in Figure 5A. The expected PCR products are 296 base pairs in wild type and 3.6 kb in kh^DR. (C–I) Eye color phenotypes in mosquitoes with the indicated genotypes and doxycycline treatments. kh¹ is an indel mutant and is included as a control that shows a white-eye phenotype. The pink eye in kh^DR without doxycycline is due to the DsRed marker that is expressed in the eye under control of the 3xP3 promoter. Shown are eyes from females. The mosquitoes were 5–7-days old. Larvae were reared in water containing the indicated concentration of doxycycline starting from 1-day post hatching. Following eclosion, the adult mosquitoes were offered 10% sucrose supplemented with the corresponding concentration of doxycycline.

We generated kh^DR mutant mosquitoes and found that the mutation abolished the black eye phenotype characteristic of wild-type Ae. aegypti (Figure 5, C and D). Unlike white-eyed kh¹, the kh^DR mosquitoes have pink eyes (Figure 5E), due to the DsRed transgenic marker, which is expressed at very high levels in Ae. aegypti eyes under control of the 3xP3 promoter (Figure 5A). We tested for rescue of the kh phenotype using animals fed doxycycline during development, by adding the drug to the water at levels ranging from 10 to 1000 ng/ml. The kh^DR mosquitoes showed increasing rescue of the eye color phenotype with higher drug concentrations (Figure 5, F–I). At doxycycline levels ≥200 ng/ml, the kh^DR animals had black eyes comparable to wild type (Figure 5, H and I). The levels of doxycycline required for rescuing the dark eye color were indistinguishable between male and females, but only females are shown due to the larger eyes.

To determine whether the concentrations of doxycycline required to obtain full rescue (200 ng/ml) impaired development, we examined whether there were any impacts of doxycycline on the number of animals that survived to adulthood, or on the number of days required for development. We found that rearing the animals on either 200 or 1000 ng/ml had no significant impact on the numbers of 4th instar larvae, pupae, and adults, or on the developmental times (Table 1). Thus, even 5 times more doxycycline (1000 ng/ml) than needed for full rescue (200 ng/ml) did not impact the survival and development times of the mosquitoes. Similarly, neither the male nor the female lifespan was reduced by 200 or 1000 ng/ml (Figure 6). However, females at all conditions had a longer lifespan than males, as the age at which 50% survived was ∼55–60 days for females and ∼30–35 days for males (Figure 6).

Table 1.

Life table for wild-type Ae. aegypti reared under different doxycycline concentrations

Developmental stage	Doxycycline (ng/ml)	Number	Developmental time (days)
Eggs	0	200	−
	200	200	−
	1000	200	−
4th instar larvae	0	164.0 ± 4.9	−
	200	158.0 ± 4.1 (P = 0.99)	−
	1000	154.0 ± 9.3 (P = 0.96)	−
	0	166.0 ± 4.0	8.2 ± 0.3
Pupae	200	156.0 ± 3.0 (P = 0.96)	8.7 ± 0.2 (P = 0.88)
	1000	153.0 ± 8.9 (P = 0.85)	8.8 ± 0.1 (P = 0.77)
	0	156.7 ± 3.7 (♀/♂ ratio: 0.91 ± 0.02)	10.4 ± 0.5
Adults	200	151.7 ± 5.0 (P = 0.99) ♀/♂ ratio: 0.90 ± 0.04)	10.5 ± 0.5 (P = 0.99)
	1000	151.7 ± 8.6 (P = 0.99) ♀/♂ ratio: 0.93 ± 0.03)	11.1 ± 0.1 (P = 0.58)

Two hundred eggs from wild-type were hatched from each group. Three groups were assayed for each doxycycline condition. The number of 4th instar larvae, pupae, and adults as well as the days of development from the time the eggs are placed in water are indicated. Means ±SEMs. Statistics were performed using one-way ANOVA with Tukey’s multiple comparisons test. There were no significant differences in the numbers of hatched animals, development times, adult female to male ratios among the 0, 200, and 1000 ng/ml doxycycline conditions. P-values were based on comparisons between the groups with the indicated doxycycline concentrations vs the groups without doxycycline treatment.

Figure 6.

Testing the impact of doxycycline on the lifespan of Ae. aegypti. Wild-type eggs were hatched and larvae and pupae were grown without doxycycline (Dox) or in water with either 200 or 1000 ng/ml doxycycline. Approximately 30 newly eclosed males and ∼30 newly eclosed females were combined in 15 × 15 × 15 cm cages and were fed 10% sucrose supplemented with the same concentration of doxycycline used in the water during development. The number of living males and females were recorded over time. Three cages were assayed for each doxycycline condition. The survival curves and data were analyzed using the Cox Proportional Hazards model. The curves were generated using the survminer R package. There were no significant effects of doxycycline on survival (HR < 1, P > 0.05), although the lifespan of the females was significantly longer (HR = ∼5, P < 0.0001).

Using anti-HA, we performed Western blots to assess the relationship between doxycycline levels and protein expression. In the absence of doxycycline, we did not detect HA::Kh (Figure 7, A–C and Supplementary Figure S6). We found that the level of HA::Kh increased in a concentration-dependent manner (Figure 7A and Supplementary Figure S6), with strong signals at 1000 and 5000 ng/ml dox (Figure 7A and Supplementary Figure S6).

Figure 7.

Doxycycline-induced expression of the HA::kh transgene in kh^DR mosquitoes. (A) Western blot showing HA::Kh protein expression in kh^DR homozygous mosquitoes. Doxycycline treatment concentrations are indicated. Protein size markers (kD) are shown to the left. (B–G) Eyes from the cone cell layer of whole-mount retinas from kh^DR mosquitoes that were either not exposed to doxycycline (− dox) or exposed to doxycycline (+ dox) as indicated. The whole mounts were stained with anti-HA (green) and DAPI (purple) to reveal nuclei. (B, C) No doxycycline treatment. (B) Anti-HA. Scale bar, 10 µm. (C) Merge of anti-HA and DAPI. (D–G) Doxycycline treatment by adding 20 µg/ml doxycycline to the water during development. (D) Anti-HA. Scale bar, 10 µm. (E) Merge of anti-HA and DAPI. (F) Higher magnification region from (D) (indicated by the dashed yellow box). Anti-HA. Scale bar, 5 µm. PC, pigment cell; CC, cone cell. (G) Higher magnification region from E (indicated by the dashed yellow box).

We also performed immunostaining to localize HA::Kh in the Ae. aegypti retina. The anti-HA was somewhat less sensitive for immunostaining at dox levels that were effective for the Western blots. We required treatment with 20 µg/ml doxycycline to obtain strong expression and found that the anti-HA staining localized to PC and cone cells (CC; Figure 7, D–G). Taken together, these findings demonstrate the utility of the DREaMR system in Ae. aegypti.

Discussion

We demonstrate that the three elements of the DREaMR system are effective in an organism such as Ae. aegypti, in which combining multiple genetic elements is much more laborious than in D. melanogaster. A single CRISPR/Cas9 mediated insertion resulted in the creation of a null mutation, a rescue transgene that can be induced to restore wild-type function, and provided the ability to detect the endogenous spatial distribution of the targeted protein. Moreover, the concentration of doxycycline needed to obtain full rescue had no obvious impact on development or lifespan. This is especially helpful in Ae. aegypti which do not have balancer chromosomes to stably maintain heterozygous strains.

While the DREaMR system was effective with the Drosophila w gene and the Ae. aegypti kh gene, we do not anticipate that the rescue would be effective for all genes. In some cases, an N-terminal tag epitope might disrupt protein function, especially with secreted proteins. Thus, it may be necessary to create versions of the system without an epitope tag, or introduce the tag in a position other than the N-terminus. If very high levels of expression of the rescue construct is needed, it might be necessary to engineer a version of the rescue construct with more than three copies of the TRE or use levels of doxycycline higher than 1000 ng/ml. Although 1000 ng/ml doxycycline did not impact on development or lifespan, if phenotypic rescue requires higher doxycycline levels, it would be necessary to determine if it causes any fitness effects.

The DREaMR system can be readily adapted to introduce many genetic tools in other organisms that are routinely used in Drosophila. The rtTA can be used to express a variety of genes introduced into the DREaMR cassette under control of the TRE. These include genes encoding proteins that permit investigators to monitor or manipulate the activities of cells such as calcium sensors, voltage sensors, optogenetic tools, cation channels, and many others. The DREaMR system would circumvent the need to perform genetic crosses, avoid having to first determine the target gene promoter, and allow for inducible expression of these tools. Finally, we suggest that the DREaMR system can be applied to accelerate research in many other organisms, and will be especially valuable in nontraditional animal models in which combining genetic elements is especially problematic.

Data availability

Strains, plasmids, and DNA sequences are available upon request. The authors affirm that all data necessary for confirming the conclusions of the article are present within the article, figures, and the table. The Supplementary materials include six figures.

Supplementary material is available at GENETICS online.