Detection and persistence of environmental DNA (eDNA) of a vector mosquito, Culex pipiens pallens

Introduction: Preventing mosquito-borne infectious diseases requires that vector mosquitoes are monitored and controlled. Targeting immature mosquitoes (eggs, larvae, and pupae), which have less mobility than adults, is an effective management approach. However, conducting these surveys is often dicult due to the limitations of morphological classication and survey costs. The application of environmental DNA (eDNA) analysis can solve these issues because it allows easy estimation of species distribution and morphology-independent species identication. Although a few previous studies have reported mosquito eDNA detection, there is a gap in knowledge regarding the dynamics of mosquito eDNA during the developmental stages of immature mosquitoes. Methods: We used Culex pipiens pallens, a vector of West Nile fever, as a model species. First, we developed a species-specic detection assay and conrmed its specicity using in silico and in vitro tests. Next, we conducted laboratory experiments using breeding tanks. Water samples were collected at each developmental stage. In addition, water samples were collected daily until the seventh day after emergence from the pupae. We quantied eDNA using real-time PCR with the developed assay to investigate the dynamics of mosquito eDNA. Results: The specicity of the developed assay was conrmed by in silico and in vitro tests. Mosquito eDNA was detected at all developmental stages and detected up to seven days after emergence of pupae. In particular, high concentrations of eDNA were detected immediately after hatching from eggs and after emergence from pupae. Highly frequent positive eDNA signals were continuously detected between egg hatching and pupa hatching. Conclusions: Mosquito eDNA was detected immediately after the eggs were introduced, and eDNA-positive detections continued until pupae emergence, suggesting that eDNA analysis is useful for monitoring mosquito larvae. The results show that eDNA analysis provides valuable information about the water sources inhabited by immature mosquitoes in mosquito control. In the future, monitoring immature mosquitoes using eDNA analysis will aid in preventing mosquito-borne infectious diseases.


Introduction
Mosquitoes are insects of the family Culicidae, with various species acting as vectors of numerous human infectious diseases such as malaria, dengue fever, Zika fever, yellow fever, and West Nile fever [1][2][3]. These mosquito-borne infectious diseases are an important cause of mortality and morbidity in at least one million people each year [4]. Therefore, controlling mosquito populations is an effective method for preventing the spread of these infections, and insecticides have been widely used to control adult mosquito populations [1]. However, due to the long-term use of chemical insecticides, resistance to different classes of insecticides has been increasingly reported [5]. Regarding mosquito population control strategies, targeting the larval stage could be more effective because unlike adults, mosquito larvae have limited mobility and are not able to change their habitat, making them more susceptible to control methods [6,7]. Therefore, monitoring immature mosquito habitats is an important task for effective and ecological mosquito management [8]. As a strategy to target larvae, the use of an insect growth regulator (IGR) has been very effective for controlling dengue fever. The use of simple traps to monitor the egg-laying of vector mosquitoes has also been used as a monitoring tool for many years.
However, the vectors of other insect-borne diseases such as malaria are diverse, and the source of each larva is different. Thus, there is currently no facile method for monitoring larvae, such as in the case of dengue mosquitoes.
Ecological monitoring of the immature stages of Culicidae is usually performed by surveying and identifying eggs, larvae, or pupae [9]. However, because immature mosquitoes live in widely distributed and diverse water environments, such as pools (e.g., hoof prints and puddles), large and complex water bodies such as river or lake margins, and rice elds, conducting physical sampling of immature stages of mosquitoes requires signi cant human labor, which can be costly depending on the area to be surveyed [9]. In particular, targeting mosquitoes that prefer large water bodies makes conducting surveys of immature mosquitoes more di cult [9]. In addition, some mosquito larvae are sometimes di cult to morphologically identify [10]. Therefore, although mosquito monitoring is an important task for mosquito control, it is di cult to accurately identify the habitats of each species in a large area.
Evaluating environmental DNA (eDNA) is a relatively new approach in parasite ecoepidemiology that consists of detecting target DNA in environmental samples to determine the presence of a given species [11]. This ecological surveying approach has been used to determine the distribution of waterborne human parasite populations [12]. The eDNA analysis method has some advantages, such as the ability to detect organisms with high sensitivity without requiring directly collection and the ability to easily identify species without morphological knowledge [11,13]. In addition, because of the importance of mosquitoes as vectors of several infectious diseases [14], the Culicidae DNA database is extensive, facilitating the design of eDNA-based detection assays and analysis. Therefore, this method could be helpful for monitoring immature mosquito stages in ecological surveys.
To date, few studies have used eDNA analysis to detect mosquito species. Schneider et al. [15] rst reported that eDNA analysis could be applied to identify mosquito habitats. Odero et al. [16] revealed that eDNA from mosquito larvae can be detected six hours after releasing larvae to tanks under experimental conditions, and eDNA could be detected even at low larval densities. However, there is no knowledge of the detectable period of eDNA according to the developmental stage of mosquitoes. Here, we investigated when and how long Culex pipiens pallens eDNA could be detected from the release of eggs to the emerging adult stages under laboratory conditions. Our study provides new and important knowledge related to the eDNA detection period along mosquito life stages in an ecotope, which will be useful for surveying and monitoring water sources for immature stages of mosquitoes under natural conditions.

Materials And Methods
We set Culex pipiens pallens as a model species in this study because this species is a major mosquito in Japan, a vector of West Nile virus in the world, and a type of mosquito that inhabits large water bodies that are di cult to survey 3 . In addition, the management of the immature stages of this species is required because adults are di cult to eradicate because of their high resistance to pyrethroid insecticides 5 . We conducted a tank experiment to verify the detectable period of eDNA according to the developmental stage. The experiment was conducted twice at the Research and Development Laboratory of Dainihon Jochugiku Co., Ltd. over two periods: August 21-September 28, 2017 and July 3-29, 2018.
Per experimental period, two 6 L tanks were prepared with one egg patch of C. pipiens pallens released in each tank. Water samples were collected at each developmental stage of the mosquitoes, and water was added after each sampling to keep the amount of water in the tanks stable during the experiments.
A total of 14 water samples were collected from each tank in plastic bottles with a volume of 120 mL for each sample (for details, see Fig. 2). Benzalkonium chloride at a nal concentration of 0.1% was added to each bottle to prevent the degradation of eDNA 17 . The collected water samples were transported in frozen conditions to the dedicated eDNA laboratory and stored at -28 °C until further use.
After melting in a water bath at room temperature (ca. 20 °C), the water samples were ltered with a glass ber lter with a nominal pore size of 0.7 μm (GF/F; GE Healthcare Life Science). To monitor potential contamination during the ltration and eDNA extraction process, 120 mL of reverse osmosis membrane water was used as a negative control for each ltration day. eDNA on the lters was extracted using the Salivette (Sarstedt) and DNeasy Blood & Tissue Kit (QIAGEN Science, Hilden, Germany) and stored at −25 °C according to the methods described in a previous study 18 .
Sequences of the mitochondrial cytochrome c oxidase subunit I (CO1) gene of the target species C. pipiens pallens and closely related species (non-target species), C. bitaeniorhynchus, Aedes albopictus, and A. aegypti, potential sympatric species in Japan, were downloaded from the database of the National Center for Biotechnology Information (NCBI: https://www. ncbi.nlm.nih.gov) ( Table 1). Based on these sequences, we designed species-speci c primers satisfying three conditions: 1) melting temperature around 60 °C, 2) at least one each of species-speci c base within the ve bases at the 3' end of forward and reverse primers, and 3) a small target fragment (50-150 bp) 13,19 .
Potential cross-reactivity of the assay was checked in silico (i.e., Primer-BLAST was performed on all databases; https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?LINK_LOC=BlastHome). Then, an in vitro speci city test was performed by real-time PCR using target and non-target tissue DNA templates (C. bitaeniorhynchus, Aedes albopictus, and A. aegypti). Real-time PCR was carried out in triplicate using morphologically identi ed specimens extracted DNA from each species as a template. Each reaction mixture (20 μl nal volume) contained 900 nM primers and 125 nM TaqMan probe (Fig. 1)  water was used instead of DNA in three reaction mixtures (non-template negative controls). The limit of detection (LOD) and limit of quanti cation (LOQ) of the assay were 1 and 3 copies per reaction, respectively.
The concentration of eDNA in C. pipiens pallens was quanti ed by TaqMan real-time quantitative PCR (qPCR) targeting the CO1 region of C. pipiens pallen using primers and a probe (Fig. 1). The qPCR conditions were the same as above, including 5 µL of the DNA template. The qPCR also included a dilution series (30,000, 3,000, 300, and 30 copies) of standard DNA derived from PCR amplicons. All qPCRs for eDNA extracts, standards, and negative controls were performed in triplicate, and the eDNA concentrations were calculated by averaging the triplicates. PCR-negative replicates (indicating nondetection) and samples with less than one copy were regarded as containing zero copies 20 . The copy number detected from the negative control was subtracted from the corresponding samples according to the previously described method 13,21 .

Results
The assay with primers Cpi-pa-CO1-F (5-GTTTAGTAGAAAATGGAGCTGCGA -3) and Cpi-pa-CO1-R (5′-TGCACCTAAAATTGATGAAATTCCTG -3′) and a probe Cpi-pa-CO1-P (5-FAM-TGAACAGTGTATCCCCCTCT -MGB-3′) was developed (Fig. 1). The melting temperatures of Cpi-pa-Co1-F and Cpi-pa-Co1-R were 59.76°C and 60.79°C, respectively. The developed primers and probe each contain a base speci c to the target species on the 3' side of primers and the center of the probe (shown in Fig. 1). The result of Primer-BLAST showed ampli cation of only the target species. The in vitro test using tissue DNA revealed ampli cation of the target species only and no ampli cation of closely related species.
The results of the laboratory experiment showed that the uctuation in eDNA concentration along with changes in the mosquito developmental stage, and the range of eDNA concentrations detected was 1.0-754.9 copies/reaction. (Fig. 2; Table S1). In particular, high concentrations of eDNA were detected immediately after hatching from eggs and immediately after emergence from the pupae. Before installing mosquitoes (sample ID = 1), one sample showed eDNA ampli cation. The period when the eggs had not yet hatched after installation (sample ID = 3) showed little eDNA detection (only one tank was positive). From hatching eggs to just after emergence of pupae, many positive eDNA signals were detected (sample ID = 4 to 8). In particular, eDNA was detected in all tanks immediately after and up to 1 d after hatching (sample ID = 7 and 8). The mosquito eDNA could be detected up to seven days after the emergence of all pupae (Fig. S1).

Discussion
Our ndings revealed that eDNA can be detected for up to seven days after mosquito emergence and the organism disappears from the water body. eDNA detection with no living mosquitoes in the water after emergence from pupae is expected to facilitate the identi cation of inhabited water areas. Therefore, this nding will contribute to mosquito management. In the application of eDNA analysis to monitor the immature stages of mosquitoes, the period for which eDNA is detectable after a mosquito has own off as an adult is important [22]. In addition, this remaining period for which eDNA can be identi ed seems to be similar to the known eDNA dynamics of vertebrates [23]. In general, eDNA is affected by water state, temperature, sunlight (UV), and pH [24]. However, it is necessary to accumulate more basic knowledge related to mosquito eDNA because the dynamics of insect eDNA may not necessarily be comparable to that of vertebrates. The application of mosquito eDNA analysis to eld surveys requires careful consideration of the dynamics of mosquito eDNA. In addition, because the baseline information essential for determining the areas of larval control includes larval habitat preferences [25], eDNA analysis makes wide-range surveys easier and help to obtain such information, contributing to e cient mosquito control.
We developed a species-speci c real-time PCR assay for evaluating C. pipiens pallens. The speci city of our assay was con rmed through in silico and in vitro speci city checks. Although we considered only closely related species inhabiting Japan when developing this assay, it can be used not only in Japan but also worldwide based on the results of in silico tests. Therefore, our assay will contribute to monitoring the immature stages of C. pipiens pallens, which is a known vector of West Nile fever.
As a control management approach for targeting immature mosquito, the larval source management (LSM) method has been employed [26], which can potentially overcome problems of adult insecticide resistance [7]. In LSM, permanent or temporary reduction of the availability of larval habitats (habitat control) and adding substances to standing water that either kill or inhibit the development of larvae (larvae control) have been conducted [26]. In addition, IGRs have been applied to the management of larvae instead of insecticides because they affect other insects and the environment [27]. Considering the mechanism by which LSM functions and the application of IGRs, which requires that larval habitats are identi ed, eDNA analysis can contribute to LSM. Accumulation of basic information on the eDNA of immature mosquitoes is expected for future eDNA-based LSMs.
After eggs hatched and pupae emerged, the eDNA concentration was relatively high. Continuous sampling revealed that this strong eDNA signal may be useful for inferring the developmental stage of the target species [28]. Such eDNA signals would be useful for evaluating mosquito habitats. Although it is di cult to make a simple comparison because of the difference in taxonomic groups, some studies reported increased eDNA concentrations after eggs hatched [29]. However, eDNA is rarely detected during the egg period. In the future, experiments focused on each developmental stage should be conducted to better investigate the dynamics of mosquito eDNA.
In conclusion, eDNA analysis of mosquitoes will enable more e cient monitoring of the immature stages of mosquitoes. In addition, several mosquito species can be detected simultaneously using the eDNA metabarcoding assay reported in a previous study [15]. Such monitoring methods can be applied not only to this target species but also to other mosquito species, which are vectors of malaria and yellow fever, and we are planning to apply this approach to other mosquito species. We conducted experiments in the laboratory to examine the detectable period of eDNA in mosquitoes, and we will need to demonstrate whether this nding is similar in the eld for application to eld monitoring in the future. As an example, in areas that are di cult to survey such as tropical forests, it may be applicable to investigate mosquito  Tables   Table 1. Accession numbers of nucleotide sequences used for designing eDNA assays.  Figure 1 Alignment of the ampli ed region of mitochondrial CO1 gene of C. pipiens pallens and closely related, potentially sympatrically distributed species with C. pipiens pallens in Japan. Asterisks indicate C. pipiens pallens-speci c nucleotides. The length of the amplicons was 136 bp.