MosSCI and Gateway Compatible Plasmid Toolkit for Constitutive and Inducible Expression of Transgenes in the C. elegans Germline

Here we describe a toolkit for the production of fluorescently tagged proteins in the C. elegans germline and early embryo using Mos1-mediated single copy insertion (MosSCI) transformation. We have generated promoter and 3′UTR fusions to sequences of different fluorescent proteins yielding constructs for germline expression that are compatible with MosSCI MultiSite Gateway vectors. These vectors allow tagged transgene constructs to be inserted as single copies into known sites in the C. elegans genome using MosSCI. We also show that two C. elegans heat shock promoters (Phsp-16.2 and Phsp-16.41) can be used to induce transgene expression in the germline when inserted via MosSCI transformation. This flexible set of new vectors, available to the research community in a plasmid repository, should facilitate research focused on the C. elegans germline and early embryo.


Introduction
Transgene silencing in the C. elegans germline has hampered research in this tissue and the early embryo. Such silencing is caused by repetitive transgene arrays that form upon injection of DNA in the gonad. The creation of more ''complex'' extrachromosomal arrays through inclusion of fragmented genomic DNA, and the use of microparticle bombardment for low copy number insertions, finally allowed germline expression of transgenes [1,2]. However, bombardment is labour intensive and complex extrachromosomal arrays are often still silenced. Furthermore, both methods frequently yield transformants with multiple transgene copies, which can have disadvantageous dosage related effects.
Recently, the Mos1 mediated Single Copy Insertion (MosSCI) method was developed to insert single copies of transgenes into defined sites in the genome of C. elegans [3]. Single copy insertion overcomes problems of variable gene dosage and silencing of extrachromosomal or integrated arrays in the germline. This technique is based on the MosTIC technique [4]. It makes use of C. elegans strains harbouring single Drosophila Mos1 transposon insertions at annotated sites in the genome. Following the heterologous expression of the Mos1 transposase, the transposon is excised from the genome, leaving a site-specific double strand break. If excision is carried out in the presence of a vector containing genomic DNA sequences that flank the Mos1 insertion site, template-directed repair can occur via homologous recombination, leading to integration of sequences cloned between the Mos1 flanking genomic DNA sequences. A library of strains containing Mos1 insertions was generated by the NEMAgenetag consortium, providing a large number of potential sites of integration [5]. Currently four Mos1 insertion strains with corresponding integration vectors have been validated for MosSCI and made available to the community [3,6].
The advantageous features of single copy insertion motivated us to explore the use of MosSCI generated transgenes for studies in the germline and early embryo. We designed a vector toolkit of germline compatible constructs compatible with the MultiSite Gateway system. MultiSite Gateway technology enables users to fuse up to four different sequences captured in Gateway recombination frames, via a one step reaction into a single fusion sequence. The system guarantees that the fragments fuse in a defined orientation and order designated by the recombination frames. Prior to the recombination reaction each of the sequences of interest are subcloned into the appropriate MultiSite Gateway vector yielding entry clones; these are then combined into a destination vector yielding an expression clone. From a collection of entry clones, different combinations of fragments can be chosen which is pivotal for the flexibility represented by the MultiSite Gateway system. The system has been widely adopted in the C. elegans community and several genome scale resources such as the promoterome [7], ORFeome [8] and 39UTRome [9] were generated that are compatible with MultiSite Gateway.
The plasmids of the toolkit are entry clones designed for the generation of expression clones using three sequences: a 59, a middle and a 39 fragment. The toolkit allows both N-terminal and C-terminal fluorescent protein tags; we provide promoter and promoter fusions as 59 fragments for N-terminal tagging and 39 UTR fusions as 39 fragments for C-terminal tagging. The middle fragment contains the ORF of interest, provided by the user. The destination vector has sites for recombination of these three elements flanked by genomic sequences adjacent to a Mos1 site of interest; our reagents are compatible with all published MosSCI sites [3,6]. Using an appropriate combination of 59 and 39 constructs with the ORF of one's choice and one of the available destination vectors, it is easy to generate a construct that will integrate at a target site in the genome and mediate constitutive expression of an N-or C-terminal fluorescently tagged recombinant protein in the germline or early embryo.

Results and Discussion
mex-5 promoter and tbb-2 39UTR constructs for constitutive expression in the germline As regulatory 59 element for constitutive transgene expression in the germline we chose the mex-5 promoter. A small 486 bp mex-5 promoter fragment had previously been shown to drive robust germline specific gene expression in strains made by microparticle bombardment [10]. We generated a set of 59 entry clones containing the mex-5 promoter fused to gfp (S65C), egfp (F64LS65T), citrine and mCherry ( Figure 1). In addition, we also generated a 59 entry clone containing the mex-5 promoter lacking a start codon to allow use of the start ATG in an ORF clone.
We based our 39 constructs on the tbb-2 39UTR, which had been shown to be permissive for expression in all cell stages of the germline and in embryos [10]. We fused the tbb-2 39UTR to sequences of gfp (S65C), egfp (F64LS65T), citrine and mCherry. An untagged tbb-2 39UTR clone (pCM1.36) is already available [10].
Expression of transgenes in C. elegans is promoted by the presence of introns or syntrons (artificial introns) [11]. The sequences that code for fluorescent proteins in the fusion constructs of the toolkit all contain syntrons, which should be advantageous for production of recombinant protein if a cDNA middle entry clone is used to generate the transgene. We also designed our constructs such that the linker (Gly) 5 Ala separates the fluorescent protein from its fusion partner in order to avoid possible negative steric interactions. The linker is additionally elongated by the sequence of the att recombination site that is generated in the MultiSite Gateway reaction.
Users of the toolbox can place a fluorescent fusion protein at the N-terminus using a mex-5 promoter/fluorescent protein gene fusion, the ORF of choice, and the tbb-2 39UTR. C-terminal fusions are created using the mex-5 promoter, the ORF of choice, and a fluorescent protein gene/tbb-2 39UTR fusion. The tbb-2 39UTR fusion constructs can also be combined with other (nongermline specific) promoters for expression of C-terminally tagged proteins in other tissues. Combining these sets with a MosSCI destination vector in a Gateway reaction generates a construct ready for injection into the appropriate Mos1 harbouring strain.

Germline expression of transgenes
In order to validate the 59 and 39 entry clones of the toolkit for germline expression, we generated and integrated a series of transgenes fusing GFP, EGFP, Citrine, or mCherry as N-terminal or C-terminal fusions (see methods); representative examples for the histone HIS-58 and a portion of the Golgi enzyme AMAN-2, are shown in Figure 2. All fusion proteins were visible in all regions of the hermaphrodite germline and in embryos (Figure 2 and data not shown). Fluorescence was high in early embryos and then declined in most cells during embryogenesis, presumably through degradation. In the hermaphrodite germline, fluorescence remained continuously high throughout development ( Figure 2G, H, I). We also observed mex-5 promoter driven transgene expression in the male germline (data not shown).
Heat shock induced expression in the germline driven by Phsp-16.2 and Phsp-16.41 The mex-5 promoter allows constitutive expression of transgenes in the germline. However, inducible expression is needed when proteins might have a toxic effect. The heat shock promoters Phsp-16.2 and Phsp-16.41 have been used extensively for ectopic induction of gene expression in somatic cells, but such transgenes have failed to drive observable fluorescent fusion protein expression in the germline [12]. A recent report used hsp-16.2 promoter fusions to generate germline phenotypes suggesting that this promoter is active in the germline, but did not characterize its activity [13].
To test the activity of heat shock promoters in the germline when present as single copy insertions, we generated constructs containing the hsp-16.2 or hsp-16.41 promoter and tbb-2 39UTR regulating the expression of gfp tagged his-58 and integrated them using MosSCI. Five strains were generated differing in promoter, tag sequences and its location and integration site ( Figure 3A). All transgenes were expressed in soma, germline and embryos following heat shock. Somatic expression was much stronger than that in the germline and we observed variation in the intensity of expression in the germline. Additionally, the signal from constructs made with EGFP fused to Phsp-16.41 (strains JA1533 and JA1541) was weaker than the signal from GFP constructs. We do not know the cause of this difference but others have reported that GFP S65C performs better in C. elegans than EGFP F64LS65T [14].
We examined the timing of appearance of transgene expression using the hsp-16.41 promoter strain EG5295. We subjected adult hermaphrodites to a one hour heat shock at 33uC followed by recovery at 20uC, and observed the animals and their progeny at one hour intervals. Immediately following the heat shock, onset of GFP fluorescence was visible only in the soma. After one hour of recovery, weak nuclear localised GFP signal could be seen in proximal germ cell nuclei near the loop region ( Figure 3C). GFP signal was visible in oocytes after two hours, and then in embryos after four hours ( Figure 3D and E). The intensity of the signal also grew stronger between one and three hours following recovery ( Figure 3B). After six hours, signal in the gonad began to diminish ( Figure 3B). Similar results were seen using the hsp-16.2 promoter (data not shown).
In summary, we have generated a flexible set of constructs to produce fluorescent fusions to an experimenter's protein of interest in the C. elegans germline, using MultiSite Gateway technology and MosSCI transgenesis. The toolbox constructs, available through Addgene (http://www.addgene.org) should be a valuable resource for studying germline and early embryo development.

Plasmid construction
Entry clones were generated using the MultiSite Gateway Three-Fragment Vector Construction Kit (Invitrogen). Inserts were amplified from genomic DNA or plasmid templates using the High Fidelity Phusion Polymerase (Finnzymes, Espoo, Finland). PCR products were recombined into pDONRP4-P1R, pDONR221 or pDONRP2R-P3 using the BP clonase (Invitrogen). Inserts were verified by sequencing. To generate the expression clones a set of entry clones were fused into either pCFJ150 or pCFJ201 using the LR clonase II (Invitrogen). Resulting plasmids were verified by restriction digest. Toolkit plasmids (see Figure 1) are available from Addgene (http://www. addgene.org). The promoter of mex-5 was amplified from genomic DNA, and fluorescent protein ORFs (containing syntrons) were from the following: gfp 65C from pPD95.02 (Fire Lab Vector Kit, June 1995), egfp 64L 65T from pPD104.53 (Fire Lab 1997Vector Supplement, February 1997, citrine 203Y 221K a kind gift from Stefan Eimer, (CMPB, ENI, Goettingen), mCherry a kind gift from Karen Oegema (Ludwig Institute for Cancer Research, La Jolla). Sequence encoding a (Gly) 5 Ala spacer was added 39 to the fluorescent protein sequence. The promoter and fluorescent protein sequences were fused via PCR stitching, with the outside primers containing attB4 and attB1 sites to allow recombination into pDONRP4-P1R.

Strains made or used in this study
See Table 1.

MosSCI transformation
MosSCI transformation was performed based on the protocol described in [3] (http://sites.google.com/site/jorgensenmossci/). The Mos1 insertion strains EG4322 or EG5003 were used for injection. Injection mixes contained pJL43.1 (50 ng/ml), pCJF90 (2.5 ng/ml), pCFJ104 (5 ng/ml), and the respective expression clone (50 ng/ml) in 20 mM potassium phosphate and 3 mM potassium citrate (pH 7.5). We note that although we were able to obtain transgenic strains expressing each of the constructs described, some apparent integration events did not result in detectable expression; we do not know the reason for this variability.

Heat shock induced germline expression
Worms were grown at 15uC to young adult stage and then heat shocked incubating sealed plates for 1 h in a water bath at 33uC. Subsequently the plates were incubated at 20uC and groups of worms were observed at 1 h intervals for fluorescence signals in the germline and embryonic progeny. After heat shock, the GFP signal strength in the germline was significantly lower than in somatic cells. Therefore, to observe germline and embryo GFP signals, worms were cut open to release the gonad and embryos. Observations were made using the 636 oil objective on a Zeiss Axioplan 2 fluorescence microscope. For the time course assessment data were collected in two independent experiments observing seven to thirteen samples of the different stages per time point with a total number ranging between 16 and 23. Observation started at 1 h of recovery for the loop region and oocytes and at 3 h of recovery for embryos. The weak germline signals were classified qualitatively into two categories: + (just detectable) and ++ (easily detectable). This qualification was translated into three shades of gray for the chart in Figure 3. The lightest shade of gray was assigned to time points when fewer than a third of observed signals were ++, and the darkest shade of gray when more than two thirds were ++ signals. Time points when ++ signals made up more than one third but less than two thirds of signals were coloured with the intermediate shade of grey.