Development of a microinjection system for RNA interference in the water flea Daphnia pulex

Background The ubiquitous, freshwater microcrustacean Daphnia pulex provides a model system for both human health research and monitoring ecosystem integrity. It is the first crustacean to have a well annotated, reference genome assembly that revealed an unusually high gene count highlighted by a large gene orphanage,-i.e., previously uncharacterized genes. Daphnia are capable of either clonal or sexual reproduction, making them ideally suited for genetic manipulation, but the establishment of gene manipulation techniques is needed to accurately define gene functions. Although previous investigations developed an RNA interference (RNAi) system for one congener D. magna, these methods are not appropriate for D. pulex because of the smaller size of their early embryos. In these studies, we develop RNAi techniques for D. pulex by first determining the optimum culture conditions of their isolated embryos and then applying these conditions to the development of microinjection techniques and proof-of-principle RNAi experiments. Results We found that isolated embryos were best cultured on a 2% agar plate bathed in 60 mM sucrose dissolved in M4 media, providing optimal conditions for microinjections. Then, we injected double-stranded (ds)RNA specific to the Distal-less gene (Dll), which is a homeobox transcription factor essential for limb development in invertebrates and vertebrates. Injected embryos presented with defects in the second antenna and appendage development, and dsRNA induced the degradation of Dll mRNAs, indicating that this technique successfully inhibited transcription of the target gene. Conclusions We developed a microinjection system for RNAi studies in D. pulex. These techniques add to the growing genomic toolbox and enhance the genetic tractability of this important model for environmental, evolutionary, and developmental genomics.


Background
In both basic and human health research, there is a pressing need to understand how environmental conditions influence gene functions and, in turn, how individuals and populations cope with changing environments. Invertebrate species have emerged as models for experimental manipulation because of their unique biological attributes, life cycles, large numbers of offspring, easy maintenance, and sophisticated tools for high-throughput biology and open-source informatics [1]. However, the traits observed in laboratories are likely a small subset of the phenotypic variation that is expressed in natural ecosystems. This may partly explain why 15 to 50% or more genes are without experimentally determined functional annotations, even within the best-characterized genomes (e.g., yeast; [2]).
Daphnia possess several characteristics that make them valuable for environmental, evolutionary, and developmental genomics research-addressing the added complexity of genome-environment interactions. Daphnia are a ubiquitous, and ecologically important member of freshwater lakes and ponds, and have long been used as a sentinel of the integrity of these aquatic ecosystems. More recently with the release of the D. pulex genome [3], it now serves as a recognized surrogate model for human health research. The D. pulex genome possesses more genes than any previously sequenced animal genome (~31,000), due to a large orphanage of Daphnia genes that likely, allows the organism to respond to its environment [3]. In addition to their short generation time, large brood sizes, and ease of laboratory and field manipulation, Daphnia are capable of either clonal or sexual reproduction, making them ideally suited for genetic studies. At present, however, there are no effective methods for manipulating genes and characterizing gene function, which because of the large gene orphanage limits interspecies extrapolations.
RNA interference (RNAi) is an evolutionarily conserved post-transcriptional gene silencing mechanism, which is triggered by double-stranded (ds)RNA in a sequence specific manner [4,5]. Since RNAi was first reported in the nematode Caenorhabditis elegans by Fire et al. [6], it has been used as a powerful tool for the analysis of gene function in many organisms such as zebrafish Danio rerio [7], planarian Schmidtea meditteranea [8], cnidarian Hydra magnipapillata [9], fungus Neurospora crassa [10], fruit-fly Drosophila melanogaster [11] and mouse Mus musculus [12]. Microinjection is one method of introducing dsRNA into cells, and this method has been successfully developed for the daphnid species, D. magna [13]. In fact, microinjection techniques enabled the application of not only RNAi [13,14], but also overexpression of foreign genes [15] and the creation of transgenic individuals [16]. Establishing these techniques in D. pulex will extend the resources for environmental, evolutionary, and developmental genomics research for this species by providing needed tools to characterize gene function.
Distal-less (Dll) and its homologs Dlx genes, which function as homeodomain transcription factors, play one of the major roles in limb development throughout the animal kingdom [17]. Reduction of Dll activity caused defects of distal leg segments in arthropods including insects [18][19][20], crustaceans (Parhyale hawaiensis [21], D. magna [13]), the spider Cupiennius salei [22] and the spider mite Tetranychus urticae [23]. Because of its conserved role in limb development, defects in the expression of the Dll gene produce easily recognizable phenotype and ease of evaluation of its phenotype, which is why this endogenous developmental gene was selected as a target in proof-of-principle RNAi in D. pulex.
The goal of this study was to develop a microinjection system for RNAi in D. pulex. Culture requirements for isolated embryos were first determined and then microinjection techniques developed for conducting RNAi experiments using Dll-dsRNA.

Development of microinjection system
There are two major technical problems for microinjection in daphnid species. One is the rapid hardening of egg membrane [24] and the other is a considerable difference between the internal and external osmotic pressures of the egg. The former prevents egg membrane penetration by a needle, and the latter produces leaking of internal contents (yolk, oil droplets, and so on) when the needle is withdrawn. Kato et al. [13] determined the proper condition for microinjection in D. magna. Eggs were incubated on ice just after ovulation to inhibit the membrane hardening transiently and placed in plastic petri dish with 80 mM sucrose medium dissolved by M4 culture medium (M4) to increase external osmotic pressure. However, these conditions are not suitable for D. pulex possibly because of the difference in egg size and form (Additional file 1). To overcome these hurdles, we first examined culture conditions by varying sucrose concentration and culture media, and determining those conditions that allowed embryos isolated from the brood chamber to develop normally. Survival was greater in M4 culture media as compared to dechlorinated freshwater (FW) ( Table 1). A sucrose concentration of 40 mM on a 2% agar plate yielded the greatest viability, but this concentration of sucrose was not high enough to counter the internal osmolality and, therefore, not sufficient for injection (Table 1). Taken together, a 2% agar plate covered with 60 mM sucrose dissolved in M4 media provided the best conditions for microinjection of D. pulex early embryos, and was employed in subsequent experiments. These conditions allowed embryos to be injected within 30 to 60 min of isolation. It is critical that injections occur within the first hour following ovulation, because there is no cytokinesis during this period and dsRNA can easily diffuse throughout the egg as it remains a single cell [25].
Dll RNAi using microinjection in D. pulex The dsRNA was prepared from a 500 bp region derived from the Dll gene (Dll-dsRNA; Additional file 2), and dsRNA derived from a 729-bp region of the Escherichia coli malE gene (malE-dsRNA) was developed as a negative control. There is only one Dll gene in the D. pulex genome (gene ID: NCBI_GNO_194714, scaffold_121: 199759-203057). The Dll-dsRNA was synthesized from a region that included the homeodomain. Red fluorescence was used to judge whether microinjection were successful ( Figure 1). Results from RNAi experiments are summarized in Table 2. Only 57% of the embryos injected with Dll-dsRNA developed compared to 71.1% viability of those injected with of malE-dsRNA, indicating Dll-dsRNA produced embryonic lethal phenotypes. Embryonic lethality caused by loss of DLL functions was reported in insects [19,26].

Sequence-specific gene silencing of D. pulex Dll
We injected 1 μg/μl dsRNA with a fluorescent dye into each egg and measured the endogenous Dll mRNA levels using Q-PCR at 24 h after injection to determine if Dll-dsRNA triggers the degradation of endogenous Dll mRNAs. The quantity of Dll mRNA in Dll-dsRNA-injected embryos was decreased to 43.9% (±9.4%) of that in malE-dsRNA-injected control embryos ( Figure 2). These results were generally consistent with the result of D. magna where a 43.6% (±2.5%) reduction in Dll was observed [13].

Phenotype of Dll RNAi in D. pulex
We confirmed the spatial expression pattern of Dll in embryo using whole-mount in situ hybridization. Dll was strongly expressed in the distal portion of second antenna, appendages (from the first to fifth thoracic limbs) and labrum, and weakly expressed in first antenna, mandible and first maxilla between 25 to 30 h post ovulation (Figure 3). Dll-dsRNAi injection of isolated embryos produced defects in the Dll-expressing organs (second antennae, appendages, ocellus, abdominal claw, and abdominal setae) to various degrees. These defects included a truncation of the second antenna segments (Figure 4) that resembled a phenotype observed in D. magna Dll-RNAi experiments [13]. The first to fifth thoracic appendages, including each exopodite, were shortened ( Figure 4 and Additional file 3). In arthropods truncation of the distal portion of the appendages is one of the well-known phenotypes observed in Dll mutation/knockdown [13,[18][19][20][21][22][23]. In addition, a loss of the ocellus and abdominal setae, and a minimized abdominal claw (Additional file 3) were observed in Dll-dsRNA injected embryos, which are also concordant with D. magna [13]. Although expression of Dll in the ocellus   has not been reported in any organisms, DLL expression was observed in the prospective ocellus region of D. magna [27]. The injection of non-D. pulex malE-dsRNA did not induce any morphological abnormalities in D. pulex embryos (Figure 3, Additional file 3, and Table 2). These data strongly suggest that the observed phenotypes were specifically generated by Dll-dsRNA, mediating the disruption of endogenous Dll mRNAs, in D. pulex.

Conclusions
We developed a microinjection system for RNAi in D. pulex. The optimum conditions for microinjection were a 2% agar plate with 60 mM sucrose dissolved in M4 media. RNAi can be induced in early embryos by injecting dsRNA into eggs within 1 hour post-ovulation. The microinjection system is applicable to not only RNAi but also creation of transgenic animals, labeling specific target cells, and so on. Thus this technique will contribute D. pulex to become a more appropriate species as model organism for environmental, evolutionary, and developmental biology.

Daphnia strains and culture conditions
The D. pulex were obtained from the Center for Genomics and Bioinformatics (Indiana University, USA) and the National Institute for Environmental Studies (Tsukuba, Ibaraki, Japan). The strains were maintained in dechlorinated freshwater (FW), which was aerated and filtered through activated carbon for 2 weeks, at 18°C under artificial light conditions of 14 h light and 10 h dark to maintain reproduction. A 0.01-ml suspension of 10 9 cells/ml Chlorella vulgaris was added every day to each culture (20-25 individuals/L). For rearing embryos, M4 culture medium (M4) was prepared using MilliQ water [28].

Culture requirements for isolated eggs
Culture conditions were evaluated to determine the optimum conditions for microinjection. Eggs were dissected from brood chamber just after ovulation using forceps (Dumoxel #5 Biologie, Dumont, Montignez, Switzerland) and placed in a culture plate (glass petri dish, plastic petri dish) containing 2% agar and sucrose (160, 80, 60, 40 mM). Sucrose was dissolved in either FW or M4, as described in the Results and Discussion. These media were passed through a 0.45 μm filter (#431220; Corning, Steuben County, NY, USA) before use. Embryo viability (survived juveniles/total eggs) was evaluated during the first instar juvenile stage.

Cloning of Distal-less
Total RNA was extracted using an ISOGEN kit (NIP-PON GENE, Tokyo, Japan), and converted to cDNA using Superscript III and random primers (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's protocol. The Dll fragment was PCR amplified from the cDNA using a set of primers designed from the Dll sequence retrieved from wFleaBase http://wfleabase. org/ (Additional file 4). Subsequently, the cDNA fragments were cloned into the pGEM-T vector (Promega, Madison, WI, USA) according to the manufacturer's instructions. These were sequenced using Sanger techniques that included the Big Dye terminator kit on an ABI 3100 Avant DNA sequencer (Applied Biosystems Japan Ltd, Tokyo, Japan). Two separate clones were developed with same Dll region oriented in both directions for dsRNA preparation.

Whole-mount in situ hybridization
Dll transcripts were visualized by whole-mount in situ hybridization with a Dll antisense RNA probe. The 522 bp probes were synthesized using a set of primers (Additional file 4). Embryos at 25-30 h after oviposition were dissected from brood chamber. The samples were transferred into a 1:4 mixture of 4% paraformaldehyde in PBS and heptane for 20 min and rinsed 5 times in PBT (0.1% Tween 20 in PBS). These embryos were digested with proteinase K (4 mg/ml) for 15 min and then rinsed 2 times in PBT. After post-fixation in 4% paraformaldehyde for 20 min, the embryos were rinsed 5 times in PBT and once in a 1:1 mixture of hybridization buffer (HB) and PBST, then transferred to HB at 65°C for at least 2 h. The pre-hybridization buffer was replaced with RNA probe in HB (5 ng/mL) and incubated overnight at 60°C. The digoxigenin (DIG)-labeled RNA probe was prepared from the cloned partial sequences using DIG RNA Labeling Mix (Roche, Mannheim, Germany). Embryos were then rinsed in HB (

Preparation of double-stranded RNA
Double-stranded RNA was synthesized using the MEGAscript T7 RNAi Kit (Ambion, Austin, TX, USA). Two templates for the in vitro transcription were prepared by PCR using two separate clones (see Cloning of Distal-less section) with T7 primer and one side genespecific primer in each (Additional file 4). The reaction mixes were incubated for 4 h at 37°C, then synthesized sense and antisense RNA were annealed at 75°C for 5 min. This was followed by nuclease digestion to remove DNA and single-stranded RNA, and purification of dsRNA was performed according to manufacturer's recommended protocol. malE-dsRNA was prepared according to methods detailed in Kato et al. [13].

Injection of double-stranded RNA
The molt of female provides a reference time point of parthenogenetic cycle, which proceeds along a strict time course at 18°C [25]. The female begins to extrude eggs into the brood chamber at 13 min after molting. Eggs were obtained just after ovulation by two week old D. pulex and placed in ice-cold M4 medium containing 60 mM sucrose (M4-sucrose). M4-sucrose was passed through a 0.45 μm filter (#431220; Corning) before use. The synthesized dsRNA (2 μg/μl) was mixed with an equal amount of 2 mM Chromeo 494 fluorescent dye (Active Motif Chromeon GmbH, Tegernheim, Germany), which was used as a visible marker for injection. A glass needle was made from a glass capillary tube (GD-1; NAR-ISHIGE, Tokyo, Japan) by a Sutter Instrument (Novato, CA, USA), and the tip was manually cut off using a forceps (Dumoxel #5 Biologie) under a stereomicroscope. A glass petri dish was prepared by placing two cover glasses side by side with M4-sucrose. One egg was placed at the left side cover glass edge to immobilize and injected using an injector (Femtojet, Eppendorf, Hauppauge, NY, USA) and a micro-manipulator (M-152, MMO-220A, Narishige, Tokyo, Japan). The needle was withdrawn from the egg using the right side cover glass edge. Microinjections were carried out within 60 min after ovulation. Then, injected eggs were transferred into a 2% agar in 6well plate with M4-sucrose and incubated at 18°C. Firstinstar animals were examined under a stereomicroscope, and their phenotype and viability were recorded.
Quantitative real-time PCR (Q-PCR) The relative expression levels of Dll mRNAs associated with the two injected conditions (Dll-dsRNA and malE-dsRNA) were quantified and compared using Q-PCR. Thirteen embryos were collected 24 h following injections with Dll-dsRNA or malE-dsRNA. Total RNA was extracted and purified using RNAqueous-Micro (Ambion) and converted to cDNA with high capacity cDNA reverse transcription kit (Applied Biosystems) using OligodT primers according to the manufacturer's recommended protocol. PCR reactions were performed in an ABI Prism 7000 (Applied Biosystems), using SYBR Green I chemistry (Applied Biosystems) in the presence of appropriate primers. The expression stability and suitability of the eight candidate reference genes (GAPDH, aTub, Tbp, Stx16, Xbp1, MMP, CAPON and Actin) were validated employing geNorm and NormFinder based on Spanier et al. [29]. The actin gene was demonstrated as the most stable reference gene and used for normalization. Primers for both target and control genes were designed to amplify short PCR products of <150 bp (Additional file 4). Data acquisition and analysis were performed by ABI Prism 7000 SDS software ver. 1.1 (Applied Biosystems). The Ct (cycle threshold) was set automatically. Samples included three biological replicates were run in triplicate to capture technical variation, and mean and standard errors calculated by Microsoft Excel 2010. Significant differences between the expression levels of Dll and malE-dsRNA injected embryos was determined by the t-test.