Quantifying Fitness Costs in Transgenic Aedes aegypti Mosquitoes

Summary

The present protocol describes how to measure common life parameter data in Aedes aegypti mosquitoes, including fecundity, wing size, fertility, sex ratio, viability, development times, male contribution, and adult longevity. These measurements can be used to assess the fitness of transgenic mosquitoes.

Abstract

Transgenic mosquitoes often display fitness costs compared to their wild-type counterparts. In this regard, fitness cost studies involve collecting life parameter data from genetically modified mosquitoes and comparing them to mosquitoes lacking transgenes from the same genetic background. This manuscript illustrates how to measure common life history traits in the mosquito Aedes aegypti, including fecundity, wing size and shape, fertility, sex ratio, viability, development times, male contribution, and adult longevity. These parameters were chosen because they reflect reproductive success, are simple to measure, and are commonly reported in the literature. The representative results quantify fitness costs associated with either a gene knock-out or a single insertion of a gene drive element. Standardizing how life parameter data are collected is important because such data may be used to compare the health of transgenic mosquitoes generated across studies or to model the transgene fixation rate in a simulated wild-type mosquito population. Although this protocol is specific for transgenic Aedes aegypti, the protocol may also be used for other mosquito species or other experimental treatment conditions, with the caveat that certain biological contexts may require special adaptations.

Introduction

Survival of the fittest is the Darwinian idea that individuals who harbor genes best adapted to their environment will pass those genes to subsequent generations¹. This means that it is fitness that determines whether their genes will survive. This more than 150-year-old concept is perhaps the most significant determinant of engineering a successful gene drive in transgenic mosquitoes. Gene drives, or the super-Mendelian inheritance pattern of a selfish genetic element that allows it to spread through populations², are being explored for genetic pest management³. In the context of vector control, this strategy aims to either replace wild-type (WT) arthropods with those resistant to pathogens (population modification) or eliminate them all together (population suppression)⁴. However, transgenic mosquitoes often exhibit fitness costs (also called genetic load) in comparison to their WT counterparts, which means that the transgene will be lost in populations that can outcompete them. Coupling transgenes with a gene drive system is therefore necessary to offset any fitness costs and push the transgene through the population at levels greater than expected from typical Mendelian inheritance⁴.

Laboratory studies across mosquito vector species have shown that transgenes often display fitness costs^5,6,7. For example, Irvin et al. measured various life parameters in Aedes (Ae.) aegypti engineered to express enhanced green fluorescent protein (eGFP) or transposase genes under Drosophila actin 5C or synthetic 3XP3 promoters, and compared them to the wild-type Orlando strain (the same genetic background from which they were derived)⁵. Most notably, they found that all transgenic strains had significantly reduced reproductive fitness⁵. Dependent on the transgene, some Anopheles (An.) mosquitoes expressing transgenes that inhibited Plasmodium parasite development also exhibited fitness costs⁶. Specifically, An. stephensi expressing a bee venom phospholipase (PLA2) under constitutive or bloodmeal-inducible promoters laid significantly fewer eggs compared to controls⁶. Those authors also found that transgenic Anopheles expressing a different transgene, an SM1 dodecapeptide tetramer did not exhibit fitness costs, leading them to conclude that transgene-conferred fitness costs are dependent on, at least likely in part, the effect of the transgenic protein produced⁶. Indeed, fitness costs may be attributed to transgene products, positional effects, off-target effects, insertional mutagenesis, or inbreeding effects in laboratory-reared strains⁷. Gene drives must therefore be robust enough to offset these transgene-induced fitness costs while also avoiding the development of insertions and deletions (indels) that block the drive itself.

Developmental or reproductive fitness costs can be measured in laboratory or cage studies⁷, with a caveat being that unknown factors in the field may also have an impact. Nonetheless, controlled fitness studies are an important first step when planning or evaluating genetically modified mosquito releases, such as a gene drive program, to determine whether the transgenic line will persist over subsequent generations. For example, Hammond et al. evaluated An gambiae CRISPR/Cas9-based-gene drives meant to disrupt the genes necessary for female fertility⁸. Through cage studies maintained for at least 25 generations, the authors found that the mosquitoes incurred gene-drive resistant alleles that blocked CRISPR-targeted cleavage and restored female fertility⁸. Their modeling efforts suggested that fertility in females heterozygous for the gene drive had the most dramatic impacts on gene drive fixation in simulated conditions⁸. Along these same lines, Ae. aegypti (Higgs' White Eye strain, HWE) engineered to express autonomous gene drive cassettes under different promoters or at different intergenic loci exhibited different rates of gene drive fixation in simulated populations⁹. Using MGDrivE modeling¹⁰, combined with measured rates of indel formation, maternal effects, and laboratory-collected life parameter data, the authors found that mosquito fitness most strongly influenced the persistence of the gene drive in simulated conditions⁹. Fitness costs that are most likely to impair gene drive efficiency are attributable to somatic Cas9 (over) expression or from the gene that is targeted, particularly in heterozygotes, rather than the intrinsic drive itself^{11,12,13,14,15,16,17}.

Given its significance, fitness is an important factor for the ability of a transgenic line to persist over subsequent generations, and it can be used as an indicator for any physiological effects associated with a transgene. For example, off-target effects may be associated with fitness costs. In this case, backcrossing a transgenic line for several generations is recommended. Furthermore, crossing the transgenic line with one that is reflective of a field population may also be necessary to investigate how well the transgenic population can compete in the real world. To quantify fitness costs so that they can be comparable, this manuscript provides simple protocols for measuring common life history traits in Ae. aegypti mosquitoes, with a particular emphasis on fitness costs associated with transgenes, so that such studies can be more easily reproduced. Protocols include fecundity, fertility, sex ratio, viability, development times, male contribution, and adult longevity. Measuring wing length and area was also chosen as a fitness measurement as it correlates with thorax length^18,19 and body size measurements, which is directly linked to bloodmeal size, fecundity, and immunity^20,21. Although there are many ways to assess fitness, these parameters were chosen because they reflect reproductive success, are simple to measure, and are commonly reported in the literature.

Protocol

NOTE: This protocol is written for transgenic and wild-type Ae. aegypti lines that have been previously validated and established. For more information on generating transgenic Ae. aegypti, see Kistler et al.²² and Coates et al.²³. All experiments outlined below were performed under standard Ae. aegypti rearing procedures. Mosquitoes were maintained at 28 °C with 75%-80% relative humidity and a 12 h light/12 h dark cycle. A minimum of 100 individual egg papers per mosquito stain is highly recommended for statistical purposes. Comprehensive fitness studies take approximately 3 months to complete (Figure 1).

1. Measuring fecundity in female mosquitoes

1. Hatch transgenic and WT mosquitoes (of the same genetic background but lacking the transgene) by placing egg papers into sterile water overnight in the insectary, following standard rearing conditions. Allow the larvae to feed ad libitum by providing several drops (each about 50 µL) of ground fish food (Tetramin) slurry (roughly 30% w/v). The first instar larvae are recognizable 1 day post-hatching. Place larvae into larger containers so there is no crowding, generally 100-150 larvae can be reared in a 1 L volume of water with a surface area of ca. 250 cm². Continue to provide Tetramin slurry as a food source, allowing the larvae to feed ad libitum. Add the food slurry in discrete droplets so that sufficient food is always available to the larvae, and so that the water does not appear cloudy. The larvae will increase in size with each instar and will ultimately enter pupation in the larval pans.
2. Once the pupae begin emerging, ~5-6 days post-hatching, separate them by sex to ensure the adults are virgins once they emerge (for subsequent mating), as described below.
   1. To sort the pupae by sex, begin by picking the pupae using 3 mL plastic pipettors into small plastic cups containing clean tap water. Male pupae tend to be smaller and tend to emerge first compared to females. First, sort the pupae by size, and then confirm this initial screen by morphology.
   2. Move the pupa to a glass microscope slide and remove excess water with a tissue. This immobilizes the pupa. Under a stereoscope at 2x-4x magnification, position the pupa ventral side up using a finely tipped paintbrush.
   3. Visually analyze the tail for differences in genital lobe shape (at the end of the pupal abdominal segments behind the paddles; Figure 2). Females have pointed fins that extend past the genital lobes, roughly resembling a crown or vampire teeth, whereas male genital lobes are rounded and do not have these pointed structures. See Carvalho et al.²⁴ for additional figures.
3. Place male and female pupae in separate water cups and place them in proper containment, for example, 1.9 L cartons (1/2 gallon or 64 oz) with 0.2 m x 0.2 m (8 in x 8 in) white organdy fabric, each containing ~150 pupae.
4. Provide a sugar and water source on top of the cages (e.g., raisins and a water-filled cup covered with gauze, secured with two rubber bands). Cut a pen-sized hole into the cartons and cover it with a rubber stopper so that additional pupae can be added into the cages as they emerge.
5. At 3-5 days post-adult emergence, anesthetize the mosquitoes by placing the cartons at 4 °C for ~2-5 min. Although they may still twitch, the mosquitoes are sufficiently anesthetized once the majority fall to the bottom of the carton. Place the mosquitoes on a glass Petri dish on ice for ~30 min-1 h.
   NOTE: Leaving the mosquitoes at 4 °C for too long may significantly impact their health, although we have not found any impacts if they are left on ice for up to 1 h.
6. Cross WT or transgenic GD1 or GD2⁹ (CRISPR/Cas-9-based-gene drive) males en masse with virgin WT (control) females by adding anesthetized mosquitoes in ratios of five WT or transgenic males to one WT female into mosquito cartons, each containing ~150 mosquitoes. Be sure to label the cages by male type.
   NOTE: This crossing scheme is to determine the fitness for GD1 or GD2⁹ hemizygotes (i.e., mosquitoes harboring one copy of the transgene). To determine the fitness for homozygous transgenic lines, the crossing scheme can be adapted so that transgenic males and females are allowed to mate in comparison with crosses of WT males and females.
7. After 2-3 days, blood feed the females following standard rearing practices²⁵ by providing a blood source on top of the cage. Remove the sugar source 24 h pre-blood feeding and the water source ~2-3 h pre-blood feeding to increase feeding efficiency.
8. After blood feeding, anesthetize the mosquitoes by placing them at 4 °C, as in step 1.5.
9. Sort the blood fed (fully engorged) females from unfed females by visually screening their abdomens for engorgement. Discard the non-fed, partially fed, and males.
10. Place the engorged females back into their respective cages and provide a sugar and water source on top of the cages.
11. After 2-3 days, place individual females in 50 mL conical tubes lined with filter paper that has been pre-labeled in pencil (Figure 3). Cover the conical tubes with mesh and seal with two rubber bands.
12. Once females are awake, fill all the tubes with ~5 mL of tap water. Place a small piece of raisin or sugar-soaked cotton ball on top of each conical tube. Leave the females for 1-2 days in the insectary and allow them to oviposit.
13. After 1-2 days, drain each conical tube and anesthetize the mosquitoes by placing them at 4 °C. For quick processing, use the tip of a 3 mL plastic pipettor pressed against the mesh to drain the water from each conical tube. Rotate the tubes upside down and place them on top of a paper towel to drain. Place the 50 mL conical tubes in a tube holder and place the rack at 4 °C.
14. Collect the egg paper and count the number of eggs on each paper. Count the females that do not lay eggs as 0 and include them in the analysis. Calculate the average fecundity as the total number of eggs counted on each paper divided by total number of females screened.
    NOTE: Digital pictures of the individual egg papers may serve as a visual record for counting confirmation (Figure 4).
15. Dry the egg papers and place them back into the insectary. It is recommended that the egg papers dry for at least 3-5 days for optimum hatching²⁵.

2. Measuring wing length, area, and centroid size

1. Freeze at least 25 WT or transgenic age-matched blood fed females from the completed fecundity study by placing the mosquito cages at -20 °C for approximately 15 min.
   NOTE: Do not freeze for a longer period of time and do not shake or drop the container, as this may cause the wings to fall off the mosquitoes.
2. Dissect the left wing of each mosquito, excising at the joint of the wing and the thorax, removing the entirety of the wing, following the below protocol.
   1. In one hand, hold the mosquito with forceps, grasping the thorax region beneath the wings.
   2. Using another pair of forceps in the other hand, dissect the left wing of each mosquito, excising at the joint of the wing and the thorax using gentle pressure to pull the wing off, removing the entirety of the wing without ripping the wing veins.
3. Using forceps, place the wing on a glass slide, lying flat. Using a transfer pipette, add one drop from a 15 mL stock solution, comprised of 70% ethanol, and one drop of dish soap to the wing on the slide.
4. Add one drop of 80% glycerol to the coverslip and place on top of the wing in the ethanol-soap solution.
5. Photograph the wings mounted on the slides using a camera with a 65 mm lens (7D-65 mm-1x; zoom = 1, 200, 6.3, ISO = 100). Note the pixel/mm scale and ensure this remains the same for all images. If multiple wings are imaged together, crop the photos to individual wings and add scale bars to each photo. Save the images in TIFF format with LZW compression, with each image clearly labelled with all information for the individual wing.
   NOTE: The wings may be photographed using a different camera and magnification or with a camera attached to a microscope for other studies, if the settings remain the same and the pixel/mm scale remains the same for each photographed wing in the study. The pixel/mm ratio will be accounted for when determining measurements.
6. Download tpsUtil and tpsDig, which are free, morphometric image-processing softwares. See Rohlf^26,27 for more detailed information about the software.
7. Digitize the wings using tpsUtil to create a TPS file of the wing photographs for the next step of digitization.
   1. To do this, open tpsUtil, and under operation, select Build tps file from images. Click input, select the file containing the individual wing images, and then select any image to choose this folder. Click output and name the file that will be created in .tps format. Click save.
   2. Then, under actions, click setup. Select include all, click create and close the pop-up box. Under operation, now select Randomly order specimens. Click input and select the .tps file that was just created. Click output, rename the .tps file to indicate that this is the file version with the samples randomized, and click save. Then, click Create. Close the tpsUtil program.
      NOTE: To reduce measurement error, the wing photos may be duplicated and utilized twice for a technical replicate.
8. Use the tpsDig software to identify two-dimensional Cartesian coordinates for the landmarks. Under input, select the .tps file created in step 2.7. Set the scale to the scale determined by step 2.5; under options, select image tools, and then click measure.
9. Add landmarks to the image by clicking on the blue crosshair icon representing digitize landmarks, and then click on the wing joints indicated in Figure 5, in order from Landmark 1 to Landmark 14.
   NOTE: These landmarks were chosen because they have previously been used in wing landmark digitization and analysis¹⁹, are easy to locate at the joints of wing veins across many samples, and provide enough wing landmarks to capture the size and shape of the wing.
10. When marking landmarks, mark any missing or obscured landmarks (due to wing folding or damage) as missing in tpsDig by clicking the M icon to exclude them from analysis and retain the correct order of the landmarks. Click the red, right-facing arrow to repeat the landmark digitization for the next images until all images in the file have been completed. Save and close the file.
    NOTE: The landmarks must be added to the wing image in the correct order for each image for the calculations of wing length, area, and size to be correct.
11. Calculate the wing length and area using the R script provided in Supplementary File 1, modifying for the .tps file location and image scale as per the experiment.
12. Calculate the wing centroid size, a measure of size statistically independent from shape, defined as the square root of the sum of squared distances of each of the landmarks from the center point of the wing^28,29.
13. Use a generalized procrustes alignment to remove non-shape variation and plot mean landmark coordinates, using wings that have all the landmarks intact. For both wing centroid size and mean landmark determination, use the R script provided in Supplementary File 2, modifying for the .tps file location and image scale as per the experiment.

3. Assessing egg fertility

1. Hatch individual F₁ eggs from the papers collected from single females (collected in step 1) into polypropylene clear deli containers containing ~100 mL of freshly sterilized deionized water. Add two or three drops of ground fish food slurry for ad libitum feeding. Line the tops of the containers with mesh or organdy, and secure with a rubber band.
2. At 2-3 days post hatching, remove the egg papers from the water and allow them to dry overnight by placing them into a new container that is covered with a lid or netting and leaving them in the insectary.
3. Re-hatch the egg papers in the same containers in which they were initially hatched by placing them back into the water container.
   NOTE: Drying and re-hatching the egg papers greatly improves the hatch efficiency. This may not be necessary for other mosquito species.
4. At 3-5 days post-initial hatching, visually inspect the larvae (2^nd to 4^th instar) for a transgenic marker (if applicable; Figure 6). Inspect several larvae under a fluorescent microscope at the same time by placing them on a piece of filter paper placed on top of an organdy or mesh lined with paper towels. The paper towels absorb excess water, so that the filter paper can be used to screen and manipulate the immobilized larvae.
5. Record the number of positive or negative transgenic larva. Discard the negative larvae.
6. Clean the larval water to ensure optimal development times. To do this, pipette the larvae into a small cup, discard the dirty water, and replace it with fresh tap water Add the positive larvae and ensure enough ground fish food slurry is available.
7. Calculate fertility as the number of transgenic or control larvae divided by the total number of eggs.

4. Determination of sex ratio in the pupae

1. Up to 14 days post-hatching, continue collecting and recording the number of male and female pupae (as described in step 1.2.3). Females or males can be pooled, by sex, into small cups housed in 1.9 L containers.
2. To determine sex ratio, add the number of F₁ females or the number of F₁ males collected across the study timeline per individual adult female mosquito. Divide this number by the total number of pupae collected for the same mosquito line generated from a single female.

5. Determination of larvae viability

1. To determine larvae viability, add the total number of pupae collected across the study timeline per individual adult female mosquito.
2. Divide this number by the total number of larvae counted per individual adult female mosquito for that same line counted in step 3.

6. Determination of larvae-to-pupae development time

1. To determine larvae-to-pupae development, track the number of larvae and the number of pupae per line daily.
2. Calculate larvae-to-pupae development as the average time to pupae development post-hatching.

7. Determination of male contribution

1. Using age-matched males, anesthetize adult virgin transgenic male mosquitoes on ice, as well as adult virgin male and female control mosquitoes.
2. Cross either the transgenic or control male adults with control female adults by adding 50 transgenic or 50 control males (age-matched) with 100 control females into 64 oz cartons, so that there are a total of 50 males with 100 females (two control females:one control or transgenic male). Be sure to label which carton contains transgenic males and which carton contains control males.
3. Allow mating to occur for 2-3 days. After this time, offer mosquitoes a bloodmeal following standard rearing practices.
4. Three days after blood feeding, separate individual fully engorged females into 50 mL conical tubes lined with pre-labeled white filter paper, as described in step 1 (fecundity assay). Discard the non-fed females or offer them another bloodmeal the next day. Fill all the tubes with ~5 mL of tap water. Place a small piece of raisin on the mesh placed on top of each conical tube.
5. Leave females for 1-2 days in the insectary and allow them to oviposit. Visually inspect the egg papers for 4-5 days post-bloodmeal for eggs. Allow the females an additional day to oviposit to increase the egg-laying rates.
6. To collect the egg papers, drain the water and anesthetize the females by placing conical tubes at 4 °C, as previously described. Discard female mosquitoes by placing in 70% ethanol.
7. Leave the papers in tubes in the insectary for at least 5 days to dry. Use forceps to remove the papers and place them in polypropylene clear deli containers containing ~100 mL of freshly sterilized deionized water for hatching. Add 2-3 drops of ground fish food slurry for ad libitum feeding. Line the tops of the containers with mesh or organdy, secured with a rubber band.
8. At 2-3 days post-hatching, remove the egg papers from the water and allow them to dry overnight by placing them into a new container that is covered with a lid or netting and placing them back into the insectary.
9. Re-hatch the egg papers in the same containers by placing them back into the water container.
10. At 3-5 days post-initial hatching, visually inspect the larvae (2^nd to 4^th instar) for a transgenic marker (if applicable). Calculate the male contribution as the number of F₂ transgenic larvae divided by the number of total larvae per mosquito line.

8. Determination of mosquito longevity

1. Transfer 50 male and 50 female WT or transgenic age-matched non-blood fed mosquitoes into a proper enclosure. Use 64 oz cartons, with the top opening lined with panty hose, twisted and knotted shut, for easy access and manipulation to the mosquitoes as the experiment progresses (Figure 7).
   NOTE: It is recommended to do this in triplicate for statistical purposes.
2. Provide a sugar and water source. Here, raisins and a water-filled cup lined with gauze, secured with two rubber bands were used.
3. Track the number of dead male and female mosquitoes each day. Calculate longevity as the average number of days passed before mosquitoes die across the cages, omitting mosquitoes that die of non-informative causes (causes not due to mosquito lifespan, such as being squished or drowning in sugar) and omitting mosquitoes that are still alive at the end of the study.
   NOTE: Survival is defined as the probability of being alive up to time t, or the function S(t) = Pr(T>t). Median survival is the time at which the probability of survival is 0.5, or T_0.5 = S^-1 x 0.5.

Representative Results

Following the above protocol, the fitness of two mosquito lines were evaluated: (1) CRISPR/Cas9-mediated knock out of the Ae. aegypti D7L1 (AAEL006424) salivary protein and (2) Ae. aegypti lines expressing autonomous CRISPR/Cas9-mediated gene drives⁹. In the case of the former, a homozygous D7L1 knock-out line was established by exploiting the non-homologous end-joining pathway (NHEJ) to generate the disruption after microinjecting embryos with sgRNAs specific to the D7L1 gene. Heterozygous D7L1-KO mosquitoes were backcrossed for four generations to WT before generating a stable homozygous line to remove potential off-target effects. In the case of the latter, a gene drive cassette expressing a Cas9 ribonucleoprotein under either the nanos (AAEL012107) or the innexin-4 (AAEL006726) promoters (GD1 and GD2 in this protocol, respectively) was knocked into the Carb109 locus (Chr3:409699138) by microinjecting embryos with a CRISPR/Cas9 injection mix, as described previously⁹. Positive G₁ larvae were outcrossed to WT at least twice to establish the transgenic lines, and integration of the transgene was validated by PCR and Sanger sequencing of mosquito genomic DNA⁹.

Once the transgenic/mutant mosquito lines were established, fitness studies were performed following the protocol outlined above. The fitness of homozygous D7L1-KO Ae. aegypti was compared to WT Ae. aegypti Liverpool (LVP strain), while the fitness of hemizygote Ae. aegypti harboring a single copy of a gene drive cassette, generated by backcrossing the homozygous line to WT, was compared to the HWE strain. These comparison groups were chosen because they were of the same genetic background as their transgenic counterpart but lacking the transgene. At least 150-200 females per line were selected from the fecundity study (section 1) and allowed to oviposit for the subsequent sections to ensure at least 100 individual F₁ egg papers would be obtained for statistical purposes. The life parameter data described below are defined as the following: (1) fecundity: the number of eggs one female oviposits in the first gonotrophic cycle; (2) wing length, area, and centroid size: distance from wing end-to-end, area of the wing, and a measurement of wing size independent of shape variation; (3) fertility: total number of larvae that develop from eggs oviposited after one gonotrophic cycle divided by the total number of eggs oviposited after one gonotrophic cycle; (4) sex ratio: total number of female or male pupae divided by total number of pupae; (5) larvae viability: total number of pupae divided by total number of larvae; (6) larvae-to-pupae development: time to pupa development post-egg hatching; (7) male contribution: number of transgenic-positive F₂ larvae divided by total number of larvae; and (8) longevity: number of days adult mosquitoes survive while maintained on sugar.

Male mosquitoes lacking the D7L1 protein exhibited significantly delayed larvae-to-pupae and pupae-to-adult development times (Figure 8) and decreased survival times (Figure 9) as compared to LVP WT Ae. aegypti. Additionally, Ae. aegypti D7L1 knock-out mosquitoes had a significantly smaller wing area compared to their WT counterparts, but there were no differences in wing size or centroid size (Figure 10). This analysis suggests that the genetic modification did not alter mosquito body size but may have induced developmental stress that led to modified wing shape. On the other hand, mosquitoes harboring one copy of a gene drive cassette under expression of the nanos (GD1) or zpg (GD2) promoters did not have significantly different fitness when compared to the HWE line, with the notable exception of larva viability (Table 1). Ae. aegypti expressing a gene drive cassette under the nanos promoter exhibited a significantly lower percentage of larval survival (p = 0.035).

Figure 1: Fitness study timeline. Estimated time required to perform all fitness study protocols described in this manuscript. Each life history parameter is highlighted in red. Abbreviations: TG = transgenic; WT = wild-type; O/N = overnight. Figure made with Biorender.com. Please click here to view a larger version of this figure.

Figure 2: Distinguishing female versus male pupae by genital lobes. Female pupae have extended fins past the genital lobes (black arrows) not seen in males. Please click here to view a larger version of this figure.

Figure 3: The 50 mL conical tubes serve as individual egg chambers. The 50 mL conical tubes, lined with white filter paper (indicated by red arrow) that has been cut into rectangles, rolled, and pre-labeled in pencil, may serve as individual egg laying chambers. Mesh is applied to the top of the tube, secured with two rubber bands, where a sugar source may be applied. Nearly 5 mL of water is added into each tube, which serves as a water source. Please click here to view a larger version of this figure.

Figure 4: White filter paper serves as egg laying paper for individual female. White filter paper, cut into a rectangle, was removed from individual egg chambers and visually inspected for eggs, as seen here. Labeling the paper in pencil in the top left facilitates identifying the parental female. Please click here to view a larger version of this figure.

Figure 5: Aedes aegypti wing landmarks. A total of 14 landmarks (shown in red) are used for wing digitization in the tpsDig software to generate two-dimensional Cartesian coordinates. The order of the wing coordinates must be maintained throughout the process of digitizing coordinates for all images. Please click here to view a larger version of this figure.

Figure 6: Transgenic larvae screening. Transgenic larvae engineered to express mCherry under a photoreceptor-specific promoter (3XP3) can be screened for positive fluorescence in their eyes (left), not seen in non-transgenic larvae (right). A stereoscope was used with a magnification between 0.5x-4x. Please click here to view a larger version of this figure.

Figure 7: Mosquito containment for longevity assay. A 64 oz carton, with the top opening lined with panty hose, twisted, and knotted shut, allowing for easy access and manipulation to the mosquitoes as the experiment progresses, serves as convenient mosquito containment for the longevity assay. Please click here to view a larger version of this figure.

Figure 8: Male AeD7L1-knock out Ae. aegypti have delayed larvae-to-pupae and pupae-to-adult development times. Larvae-to-pupae and pupae-to-adult development times were assessed with a Kruskal-Wallis ANOVA and Dunn's post-hoc comparisons. WT male mosquitoes have a shorter time to pupation than WT females (Dunn's test, p < 0.05), while D7L1 males do not significantly differ in their pupation time relative to D7L1 females (Dunn's test, p > 0.05). D7L1 females take longer to reach pupation than WT females (Dunn's test, p < 0.05), as do D7L1 males relative to WT males (Dunn's test, p < 0.05). In terms of pupae-to-adult development time, WT and D7L1 females did not significantly differ. WT males had a faster time to pupation than WT females (p < 0.05) and a faster time to pupation than D7L1 males (p < 0.05). D7L1 males and females did not significantly differ (p > 0.05). However, D7L1 females had a slower time than WT females (p < 0.05). Abbreviations: WT = wild-type Liverpool strain; F = female; M = male; D7L1 KO = D7L1-knock out Ae. aegypti. N = 25 age-matched males and 25 females per mosquito line. * = p < 0.05, Dunn's post-hoc test, Kruskal-Wallis ANOVA. ns = not significant. Bars represent mean + standard error. Please click here to view a larger version of this figure.

Figure 9: WT (Liverpool strain) and D7L1-knock out Aedes aegypti males and females differed significantly in their survival. WT female mosquitoes never reached their median survival time during the study: p(survival) = 0.83 at 40 days. All other groups reached their median survival before 40 days, with a distinct temporal separation of the median survival times (WT males ≈ 33 days, D7L1 females ≈ 39 days, D7L1 males ≈ 22 days). Kaplan-Meier survival curves were generated to model the survival of each line over time and estimate the median survival times, while accounting for censoring and non-normal survival distributions. A Log-Rank test was performed to evaluate between-group differences. The Log-Rank test assumes non-informative censoring and proportional hazards. The Cox Proportional Hazards Model was then used to evaluate the hazard ratio, or instantaneous risk of death at time t (given it has not already occurred) for each group relative to the other. The models' proportional hazard assumption was met, verified by Schoenfeld's residuals (verified by the cox.zph function; p > 0.05 for sex and line). N = 50 male and 50 female age-matched, non-bloodfed mosquitoes per Ae. aegypti line. Please click here to view a larger version of this figure.

Figure 10: WT (Liverpool strain) and D7L1-knock out Ae. aegypti females differed significantly in wing area but not in wing length or centroid size. Wing length, area, and centroid size were measured after imaging the wings and digitizing the landmarks to assess for changes in size and shape. Wing area differed between WT and D7L1-knock outs, suggesting that the gene knock-out did not affect body size but may have induced developmental stress that altered wing shape. (A) Length: WT mean = 2.4 mm, σ = 0.21. D7L1 mean = 2.498 mm, σ = 0.35. Not significant (Mann-Whitney T-test, p = 0.6370). (B) Area: WT mean = 529,740 pixels/mm, σ = 58,852. D7L1 mean = 457,879 pixels/mm, σ = 143,633. Statistically significant difference (Mann-Whitney T-test, p=0.0158). C) Centroid Size: WT mean = 2321 pixels, σ = 61.08. D7L1 mean = 2,296 pixels, σ = 76.77, two outliers excluded. Not significant (Mann-Whitney T-test, p = 0.2993). Please click here to view a larger version of this figure.

Line	WT (HWE)	GD1	GD2
larva-to-pupa development (days)	9.8 ± 4.3	9.5 ± 5.6	9.5 ± 7.0
larva viability (% survival)	44.2 ± 7.3	30.9 ± 2.9*	44.8 ± 2.5
% female adults	45.4 ± 1.6	47.8 ± 1.5	48.2 ± 1.1
% male adults	54.6 ± 1.7	52.2 ± 1.5	51.8 ± 1.1
50% female survival (days)	34 ± 6.2	44.5	41.5
50% male survival (days)	20.7 ± 3.3	16 ± 1.6	18 ± 4.5
fecundity (# eggs)	62.6 ± 2.8	57.6 ± 2.6	64.6 ± 2.0
fertility (egg hatchability) (%)	49.2 ± 8.2	44.0 ± 3.6	58.3 ± 3.4
male contribution		35.3 ± 1.02 (1931/5447)	25.6 (1035/4049)

Table 1: Life parameter data for two Ae. aegypti lines (GD1 and GD2) hemizygous for gene drive cassettes compared to Higgs' White Eye (HWE) control line. Data derived from hemizygous males or HWE males that were allowed to mate with female HWE. * = 0.01 < p < 0.05 compared to HWE by one-tailed t-test. No stars indicate no significant difference by one-way ANOVA. Means + standard errors are provided. This table has been modified from⁹.

Supplementary File 1: R code to calculate wing length and area. Code is provided under the name: Supplemental_File_R_Winglengtharea.R. Please click here to download this File.

Supplementary File 2: R code to determine wing centroid size and mean landmark. Code is provided under the name: Supplemental_File_R_Wingcentroidsize_meanlandmarks.R. Please click here to download this File.

Discussion

Ae. aegypti fitness studies are often performed in the laboratory to assess fitness costs associated with transgenic cargo (e.g., gene drive elements) or gene knock outs, as discussed in this manuscript; however, these studies may be performed for a variety of purposes—any that aim to evaluate the health of an Ae. aegypti group, such as Wolbachia-infected^30,31, insecticide resistant^32,33, or arbovirus-infected³⁴. Therefore, because these experiments can be used for a wide array of purposes, the protocols were intended to be simple, efficient, and reproducible.

Investigators should use caution when extrapolating fitness data collected in the laboratory to mosquitoes in the field. In this protocol, the control group should be mosquitoes of the same genetic background but lacking the transgene (or treatment condition). However, common wild-type Ae. aegypti strains used in the laboratory have been inbred over many generations and thus may not be an accurate representation of those that are field-derived. For example, the common laboratory strain, Rockefeller (ROCK), is thought to have been established from Ae. aegypti collected in Cuba as early as 1881³⁵. Studies have shown that Rockefeller females survive significantly longer in the laboratory than field mosquitoes³⁶. Another commonly used laboratory strain, particularly in transgenic studies, is Higgs' White Eye (HWE), so named for Dr. Stephen Higgs, who selected these for their white eye mutant strain³⁷. Because these mosquitoes lack eye pigmentation, transgenic expression of fluorescent markers is easier to detect than in wild-type Ae. aegypti, that have black eyes. However, these mosquitoes are significantly less fit than those from the field³⁸. Laboratory fitness studies are also not an absolute representation of the conditions that mosquitoes endure in the field. Nonetheless, controlled laboratory fitness studies are a good indication of life parameters that may not otherwise be measurable. After performing fitness studies in the laboratory, larger cage studies may be necessary to evaluate fitness in semi-field conditions.

Although this manuscript is specific for Ae. aegypti, many of the protocols may be adapted for other species; however, we urge investigators to take into account the biology of their diverse species, which may differ in such a way that the protocols need slight modifications. For example, we found that hatching dried, individual egg papers twice—that is, hatching the egg papers once in water, drying the paper overnight, and re-hatching the same line again in water—was more efficient and a better representation of the hatching rates compared to those after a single hatch. However, it is important to note that Ae. aegypti eggs are resistant to desiccation^39,40, whereas other mosquito species may not be tolerant to drying, including most species of Anopheles.

Perhaps the most critical, and time consuming, aspect of these protocols is accurately sexing male and female pupae. Adults mate 36-48 h post-emergence⁴¹, so it is critical that pupae are sexed correctly to ensure they are virgins for subsequent crosses. Females are monogamous, but males are polygamous and can effectively inseminate at least five females^41,42,43. This means that a single male of an incorrect genotype can inseminate at least five females in a cage—rendering the cage relatively worthless for subsequent experiments—but a single female will not significantly impact a cage of males.

In summary, we have outlined basic protocols for measuring common life history traits in Ae. aegypti mosquitoes. Such studies are of paramount significance when assessing fitness costs associated with transgenes in genetically modified mosquitoes, as this data can be used in downstream applications such as modeling. For studies evaluating novel genetic control strategies, larger cage studies in semi-field conditions would then be necessary before deployment.

Materials

Name	Company	Catalog Number	Comments
1 oz. translucent plastic souffle cups	WebstaurantStore	301100PC
2 oz. translucent polystyrene souffle cups	WebstaurantStore	760P200N
3 mL plastic pipettors	Cornin	357524
50 mL conical tubes	Any brand
64 oz. white double poly-coated paper food cup	WebstaurantStore	999SOUP64WB	for mosquito enclosement
65 mm lens	Canon	MP-E 65mm f/2.8 1-5x Macro Photo	Canon Macro Photo MP-E 65mm, 7D-65mm-1X; zoom=1, 200, 6.3, ISO=100; for photographing wings or egg papers, although other cameras are likely sufficient
Aedes aegypti mosquitoes	BEI	multiple strains as eggs are available
Artifical membrane feeders	https://lillieglassblowers.com/	Meduim membrane feeder, Custom made, 33mm	Chemglass also offers, but sizes are wrong for us. Ours are about 3 cm?
ATP	MP Biomedical	ICN15026605	Any good quality ATP, 10mM filter sterile aliquots at -20
Autoclave			for sterilizing water for hatching
Canon EOS 7D camera	Canon	3814B004	for photographing wings or egg papers, although other cameras are likely sufficient
defibrinated sheep blood	Colorado Serum Co.	60 ml, every 2 weeks	https://colorado-serum-com.3dcartstores.com/sheep-defibrinated
Dual Gooseneck Microscope Illuminator	Dolan Jenner Fiber-Lite 180	181-1 System
Ethanol
Forceps	Dumont	5SF
Gauze		omnisorb ii	4" non-woven sponges
glass microscope slide	Fisher Scientific	12-544-2
Glass Petri dishes, 100 × 15 mm	VWR	75845-546	for anesthesizing/manipulating mosquitoes on ice
Hogs' gut	Any Deli		we buy in bulk, split, wash and store in small aliquots of ~4X12" at -20 in 50 ml conical
Ice
Ice bucket
Kimwipe	Fisher Scientific	06-666A
Leica GZ4 StereoZoom microscope			for screening transgenic mosquitoes (if applicable)
Paintbrush	AIT synthetic brush	size 10-0	for manipulating larvae/pupae (Amazon)
Panty hose	Walmart	L'eggs Everyday	Women's Nylon Plus Knee Highs Sheer Toe, 16 pairs (plus fits the carton)
Pencils	Any brand
Plastic containers for 2° storage of cartons	Walmart		Sterilite 58 Qt Storage Box Clear Base White Lid Set of 8
Plastic containers for growing larvae	Walmart		Sterilite 28 Qt. Storage Box Plastic, White, Set of 10
Plastic containers for hatching larvae	Walmart		Sterilite 6 Qt. Storage Box Plastic, White
polypropylene clear deli containers	WebstaurantStore	127DM12BULK	12 oz, or 16 oz if needed for bigger (127RD16BULK)
Rubber bands	Office Max	#100736/#909606 /#3777415	12", #64 and #10
Rubber stopper	VWR	217-0515	for mosquito enclosement
Sugar source, such as sugar cubes or raisins
Tetramin flake food	Tetramin	16106
tpsDig	Stony Brook Morphometrics		A free morphometric image-processing software distributed online available at https://www.sbmorphometrics.org/
tpsUtil	Stony Brook Morphometrics		A free morphometric image-processing software distributed online available at https://www.sbmorphometrics.org/
White organza fabric 8” × 8”	FabricWholesale.com	4491676	Joann Casa Collection Organza Fabric by Casa Collection
Whitman Grade 1 Qualitative Filter paper	Whitman	1001-824	for egg papers. The white color makes it easier to see the black eggs.