- Email: [email protected]
An efficient transgenic system by TA cloning vectors and RNAi for C. elegans
BBRC Biochemical and Biophysical Research Communications 349 (2006) 1345–1350 www.elsevier.com/locate/ybbrc
An efficient transgenic system by TA cloning vectors and RNAi for C. elegans Keiko Gengyo-Ando a
a,b
, Sawako Yoshina
a,b,c
, Hideshi Inoue c, Shohei Mitani
a,b,*
Department of Physiology, Tokyo Women’s Medical University School of Medicine, 8-1, Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan b CREST, JST, 4-1-8 Hon-cho, Kawaguchi, Saitama 332-0012, Japan c School of Life Sciences, Tokyo University of Pharmacy and Life Sciences, 1432-1, Horinouchi, Hachiouji, Tokyo 192-0392, Japan Received 29 August 2006 Available online 11 September 2006
Abstract In the nematode, transgenic analyses have been performed by microinjection of DNA from various sources into the syncytium gonad. To expedite these transgenic analyses, we solved two potential problems in this work. First, we constructed an efficient TA-cloning vector system which is useful for any promoter. By amplifying the genomic DNA fragments which contain regulatory sequences with or without the coding region, we could easily construct plasmids expressing fluorescent protein fusion without considering restriction sites. We could dissect motor neurons with three colors in a single animal. Second, we used feeding RNAi to isolate transgenic strains which express lag2::venus fusion gene. We found that the fusion protein is toxic when ectopically expressed in embryos but is functional to rescue a loss of function mutant in the lag-2 gene. Thus, the transgenic system described here should be useful to examine the protein function in the nematode. Ó 2006 Elsevier Inc. All rights reserved. Keywords: TA cloning vector; RNAi; Fluorescent protein; Transgenic
In the nematode Caenorhabditis elegans, transgenic analyses have been fruitful to analyze the expression patterns of genes of interest, rescuing mutants by introducing wild-type genomic fragments spanning the whole region of the genes of interest and so on [1]. These convenient transgenic analyses are based on the fact that DNA without obvious origin is able to be replicated in nematode nuclei [2]. The phenomenon has led to a number of useful transgenic protocols including injecting DNA into gonad from various biological sources such as plasmid, k phage, YAC, or cosmid clones, PCR fragments [1]. However, there are a number of potential problems to be solved to perform these experiments more efficiently. First, expression vectors most popularly used need appropriate restriction sites to subclone promoters [3]. Because promoter analyses need frequently about 5 kb of 5 0 flank*
Corresponding author. Fax: +81 3 5269 7362. E-mail address: [email protected] (S. Mitani).
0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.08.183
ing sequences so that genes of interest be expressed correctly. The longer a genome sequence is, the more difficult it is to choose an appropriate restriction site. The most convenient DNA appear to be plasmids, which are available in a large amount with high purity. Thus, it is helpful if PCR fragments are conveniently cloned into expression plasmids quickly, because the whole genome has been sequenced and all the genes have been deduced by a computer program and cDNA analyses [4–6]. TA-cloning vectors, which are designed to accept PCR fragments by overhanging deoxythymidines at both 3 0 ends of the cloning sites, are one of the easiest vectors to handle, because PCR fragments with any restriction sites can be subcloned [7]. Here, we designed a set of multicolor TA-cloning vectors, which are useful for expression analyses with or without coding sequences. Another problem is more elusive; we often have troubles that the transgenic animals fail to have any expected expression. A fraction of the cases may be caused by the
1346
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
toxicity of ectopically expressed products. To overcome this problem, we tried to knock down the products by feeding RNAi with a few generations of transgenic strains. After establishing transgenic strains, we could allow the transgene to express the fusion protein by transferring transgenic animals to the common food bacteria OP50-1. Combined together, transgenic analyses in C. elegans are expected to be done more quickly. Materials and methods Construction of reporters and rescue plasmids. To prepare reporter constructs used in this study, the PCR fragments amplified from N2 genome were directly cloned into the TA expression cloning vectors. The vectors were digested with XcmI to create the linearized vectors with 3 0 -T overhangs and used for inserting PCR fragments containing upstream sequences and the first several codons of the genes acr-5, unc-4, and unc-25 [10,11]. acr-5 was amplified using pyrobest polymerase (Takara), and other genes except acr-5 were amplified using platinum Taq high fidelity (Invitrogen), from wild-type (N2) genome. Primers used for amplification of each promoter region are as follows; 5 0 -GGGAAGCATGCTGAAA ATTG-3 0 and 5 0 -CGCAACGACGGCATTCACTT-3 0 for acr-5, 5 0 -CAG TGCACCGATCATTTTCA-3 0 and 5 0 -TCTGAGACCATGTTGATGG G-3 0 for unc-4, and 5 0 -GCATGCAAAAAACACCCACT-3 0 and 5 0 -CATT TTTGGCGGTGAACTGA-3 0 for unc-25. Approximate sizes of PCR fragments were 4.2, 2.8, and 1.8 kb for acr-5, unc-4, and unc-25, respectively. The expression plasmids having insert fragments were quickly verified by digesting with SpeI. If the plasmids cannot be cut with SpeI, they are unlikely to be cloned into the expression vector in-frame. We confirmed the junction sequence of cloning site using the appropriate primers. To construct functional lag-2::venus plasmid, a 5 kb promoter fragment obtained by amplifying N2 genome by primers 5 0 -TGGATCCCGT TTTCCGATTTGCCGGAA-3 0 and 5 0 -TGCGGCCGCCGATCATTTTC TGAAAAAAGGC-3 0 was subcloned between BamHI and NotI sites in the pFX_venusT giving rise to Plag-2::venus. A 1.5 kb of genomic fragment containing full length lag-2 coding sequence was amplified with 50 -AAATGATCGGCGGCCGCATGATCGCTTACTTCCTCTTACTCC-30 and 5 0 -GCTCACCATGCGGCCGCAGACATAGTGACAGGCTGGA ATAG-3 0 . The amplified DNA fragment was inserted into the NotI site of the Plag-2::venus using a recombinase in-Fusion (Clonetech) giving rise to lag-2::venus. Germline transformation. For triple staining, the Pacr-5::ECFP, Punc4:: EYFP, and Punc-25:: DsRedx were injected at 40 ng/ll into adult N2 hermaphrodites by standard germline transformation techniques [8]. pRF4 plasmid (40 ng/ll) containing the dominant roller marker rol-6 (su1006) was co-injected to identify transgenic animals. N2 and FX11226 (lag-2 (q411)/nT1 [qIs51]) mutant animals were injected with the rescue plasmid Plag-2::venus (10 ng/ll) and Pmyo-2::DsRed (10 ng/ll). qIs51 harbors an insertion of ccEx9747 with markers:myo-2::GFP, pes-10::GFP, and F22B7.9::GFP, resulting in expression of GFP in the pharynx, embryo, and intestine. Homozygous lag-2 mutant animals were identified by loss of GFP fluorescence linked to balanced chromosome nT1 (qIs51). Feeding RNAi. Feeding RNAi was performed as described [9]. Briefly, the venus cDNA was amplified by PCR using sense and antisense primers, 5 0 -CGCGGGTACCTGTACAGCTCGTCCATGCCG-3 0 and 5 0 -CGCG TCTAGAATGGTGAGCAAGGGCGAGGA-3 0 . The cDNA fragment was cloned into the L4440 (pPD129.36) vector and the cloned plasmid was transformed into Escherichia coli HT115 (DE3). Nematode growth medium (NGM) agar dishes with 1 mM IPTG and 50 lg/ml of ampicillin were seeded with the bacterial clone which expresses double-stranded RNA of venus. Animals of the L4 stage were pre-cultured on the venus RNAi dishes until adult stage and were picked up for microinjection of plasmids as described. After animals were microinjected with plasmid DNA, they were cultured on RNAi dishes. After picking up DsRedpositive transgenic F1 animals, we further cultured the animals until
isolation of F2 animals which kept DsRed-positive transgenic animals. To examine the toxicity of LAG-2/Venus protein, recipients were cultured on bacterial lawn of either OP50-1 or HT115 (DE5) carrying a plasmid expressing venus dsRNA. Transgenic lines (lag-2(q411); tmEx1432– tmEx1436[lag-2::venus + Pmyo-2::DsRed]) carrying stably transmitting extrachromosomal arrays were used for fertility assay. Microscopy. Animals were mounted on 5% agarose pads in M9 buffer containing 10 mM sodium azide and observed by differential interference contrast (DIC) or epifluorescence microscopy. Images were observed using a microscope system composed of the microscope (Olympus BX-50) and the CCD camera (Olympus sensys), and analyzed by the IPLab software (Scanalytics). We used filter and dichroic mirror set for each fluorescent protein as follows. ECFP: excitation band-pass 425–445, 450 dichroic, emission band-pass 460–510, EGFP: excitation band-pass 460–480, 485 dichroic, emission band-pass 495–540, EYFP, and venus: excitation bandpass 490–500, 505 dichroic, emission band-pass 515–560, and DsRedExpress-1: excitation band-pass 520–550, 565 dichroic, emission band-pass 580IF.
Results and discussion Design of expression plasmids All the plasmids constructed in this study share common Bluescript backbones (Stratagene, CA). We inserted a multiple cloning site followed by fluorescent proteins of various colors, and about 1 kb-long 3 0 untranslated region from the unc-86 gene to add a signal sequence for poly(A) addition by using restriction sites as described in Fig. 1A. To perform multicolor staining, we inserted cDNAs of ECFP, EGFP, EYFP, venus, and DsRedExpress1 for pFX_ECFPT, pFX_EGFPT, pFX_EYFPT, pFX_venusT, and pFX_DsRedxT, respectively. The cDNAs were purchased from Clonetech (ECFP, EGFP, EYFP, and DsRedExpress1) or kindly offered by Dr. A. Miyawaki (venus). The plasmids when digested by a restriction enzyme XcmI give rise to linear fragments with an overhanging deoxythymidine at each 3 0 cut end [7] (Fig. 1B). The PCR fragments, with or without coding region, can be subcloned in-frame by adjusting the 3 nucleotides of the 5 0 end of antisense primers coding for the amino acid(s) to the cDNA of fluorescent proteins. The reporter plasmids constructed this way once examined for the expression pattern could be converted to functional expression plasmids for either native nematode proteins of wild-type or any mutant forms, or proteins from, for example, human cDNAs. This could be easily performed by inserting a cDNA fragment amplified by PCR with a SpeI or NotI restriction site at the 5 0 end and BglII site at the 3 0 end (Fig. 1B). Dissection of neuron subtypes by a multicolor reporter system To address whether the multicolor reporter system works for the nematode expression analysis, we amplified and subcloned the promoter DNA fragment with the first methionine into expression plasmids as three color combination; ECFP for acr-5 promoter, EYFP for unc-4 promoter, and DsRed Express1 for unc-25 promoter. Three
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
1347
A
B
Fig. 1. Schematic representation of method for cloning PCR products into T vector pFX_EGFPT. (A) Schematic diagram of pFX_EGFPT. All vectors produced in this work have same sequences except for the coding sequence of each fluorescent protein. Sequence of multiple cloning site (MCS) is shown in (B). (B) PCR products containing promoter sequence with only first methionine or most of the coding region are cloned into pFX_EGFPT digested with XcmI. Expression constructs using this cloning vector have additional five amino acids derived from multiple cloning sequence between the examined coding sequence and the fluorescent protein coding sequence.
promoters are said to drive reporter expression in the neuronal subtypes, and motor neurons, which are exclusively with each other [10,11]. The fluorescence observed was consistent with the previous notion, indicating that the three color reporters can be easily distinguished with each other when co-injected to the germ cells (Fig. 2). To confirm that our expression system is also useful for other genes, we constructed the expression plasmids containing the regulatory sequences that control the expression of various tissues: myo-3 (body-wall muscle), myo-2 (pharyngeal muscle), ges-1 (intestine), dpy-7 (hypodermis), unc-119 (all neurons), unc-122 (coelomocytes), eft-3 (ubiquitous), and so on. These expression plasmids were injected
into N2 animals and the transgenic lines having the extrachromosomal arrays showed the expected expression patterns (data not shown). Efficient rescue of lag-2 mutant phenotype by the aid of feeding RNAi against venus Transgenic strains are generated by injecting plasmid constructs into syncytial gonads in C. elegans [8]. However, we and others often have difficulties to obtain transgenic strains expressing reporter genes in spite of correct marker expression [1]. These problems are probably due to multiple factors. We reasoned that at least a fraction of these
1348
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350 Table 1 Toxic ectopic expression causes the death of F1 animals of lag-2::venus transgenic animals Condition
With RNAi
Without RNAi
DsRed(+) F1 Hatched F1 DsRed(+) F2 lines LAG-2/Venus(+) lines
124 112 13 8
91 2 0 0
A mixture of plasmid DNAs consisting of Pmyo-2::DsRed and lag-2::venus (1:1 dilution) was injected to N2 animals treated with or without feeding RNAi against venus. Number of F1 animals were counted as DsRed (+) embryos. Some of the embryos without feeding RNAi were DsRed () venus (+), presumably because myo-2 gene expression is found only after organogenesis stage, hatching rate of RNAi () was a little overestimation by reducing the number of DsRed (+) F1 animals.
Fig. 2. Triple staining of motor neurons in ventral nerve cord. Transgenic strain carrying three expression plasmids, acr-5::ECFP, unc-4::EYFP, and unc-25::DsRedx, was constructed. Same L1 larva is shown in (A–E). (A) acr-5::ECFP expression. acr-5 encodes a acetylcholine receptor subunit expressed in B-type motor neuron (DB) in L1 larva ([10]). Some DB motor neurons could not be identified since out of focus. (B) unc-4::EYFP expression. unc-4 encodes a homeodomain transcription factor expressed in A-type motor neuron (DA) in L1 larva [15]. (C) unc-25::DsRedx expression. unc-25 encodes glutamic acid decarboxylase expressed in Dtype motor neuron (DD) in L1 larva ([11]). Merged image of (A–C) is shown in (D) and an image with Nomarski optics shown in (E).
cases are due to the toxicity of expressed products, and non-expressing transgenic strains tend to survive selectively. To examine this possibility and find out the solution, we
chose a gene lag-2, which is essential for fertility of the animal by regulating meiotic division of germ cells [12]. We found ectopically expressed LAG-2/Venus fusion protein is toxic to animals. When the lag-2::venus plasmid is injected into N2 animals and cultured on OP50-1 as a food, we could not isolate any stable lag-2::venus positive transgenic strains (Table 1). This failure appears partly due to ectopic expression of LAG-2/Venus fusion protein, because dead embryos are often venus-positive at variable cells (Fig. 3). It is known that expression of transgenes in germ cells is reduced over several generations of the transgenic strains by silencing [13]. However, during the first few generations, sometimes transgenes are only weakly silenced in the germ cells. We reasoned that if toxic products are, for example, expressed ectopically in embryonic cells of F1 animals, animals with transgenes selectively become unhealthy or dead. To examine this possibility, we used the feeding RNAi method to knock down venus, and thus the lag-2 gene expressed as a fused mRNA at the early generations [9]. When the plasmid is injected into N2 animals and cultured with feeding RNAi, which knocks down venus, we could isolate stable transgenic lines (Table 1). To address whether this transgenic method is useful for functional analysis of expressed proteins, we tried to express the same lag-2::venus fusion gene in the lag-2 mutant background. After a few generations of isolating transgenic establishment, we picked up homozygous lag-2 mutant animals without the balancer nT1[qIs51] but with the transgenic marker Pmyo-2::DsRed and transferred the animals to dishes with OP50-1 as a food source. The homozygous lag-2 mutant animals without transgene were all sterile, whereas homozygous lag-2 mutant animals with the lag-2::venus transgene were fertile (mean number of progeny = 50.8, n = 5). We also examined the expression pattern of the LAG-2/ Venus protein by fluorescence microscopy (Fig. 3). We found the LAG-2/Venus protein in DTC and also weakly in germ cells, suggesting that the fusion protein is secreted by DTC to proximal gonadal cells, consistent with the previous study [12].
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
1349
Fig. 3. Expression of lag-2::venus (A–B) Ectopic expression of lag-2::venus in a dying embryo of F1 animal fed with OP50-1 (A, DIC image; B, Venus fluorescence). (C–D) No ectopic expression of lag-2::venus in a normal embryo of F1 animals fed with RNAi against venus (C, DIC image; D, Venus fluorescence). (E–G) Functional expression of lag-2::venus in the homozygous lag-2 mutant background (E, DIC image; F, Venus fluorescence found in a Distal Tip Cell; G, merged). Scale bars = 10 lm for (A–C), 20 lm for (D–F).
We succeeded in isolating functional protein expressing transgenic lines by culturing the animals in an early few generations on feeding RNAi dishes. RNAi against fluorescent protein did not appear to knock down endogenous
lag-2 mRNA, which is important for fertility of the animals. This may be caused by the fact endogenous lag-2 mRNA is resistant to RNAi [14]. Another advantage is the use of the same fluorescent protein as an RNAi target,
1350
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
we are able to use the same construct for functional expression of many genes. It should be noted that any other fluorescent reporter like DsRed is not affected and can be used as a transgenic marker and so on. Conclusions We constructed a convenient TA-cloning vector system, which is useful for triple staining and functional gene expression. Because the whole genome sequences are available, the present TA-cloning vector system makes researchers possible to design many reporter constructs without restriction site analysis. The feeding RNAi method present here is useful to isolate reporter transgenic without considering ectopic expression of transgenes. In that context, the present transgenic system will facilitate the analysis of the genetics of this useful model organism. Acknowledgments We thank Drs. Yuji Kohara, Masatoshi Hagiwara, Yuichi Iino, Takeo Awaji, and Takeshi Fukuhara for discussion. We are grateful to Dr. Atsushi Miyawaki for a venus plasmid. Some nematode strains used in this work were provided by the Caenorhabditis Genetics Center funded by the National Institutes of Health National Center for Research Resources. This work was supported by grants from MEXT, Japan, and JST to S.M. and K.G.-A. References [1] C. Mello, A. Fire, DNA transformation, Methods Cell Biol. 48 (1995) 451–482. [2] D.T. Stinchcomb, J.E. Shaw, S.H. Carr, D. Hirsh, Extrachromosomal DNA transformation of Caenorhabditis elegans, Mol. Cell. Biol. 5 (1985) 3484–3496. [3] A. Fire, S.W. Harrison, D. Dixon, A modular set of lacZ fusion vectors for studying gene expression in Caenorhabditis elegans, Gene 93 (1990) 189–198.
[4] C. elegans, Sequencing Consortium, Genome sequence of the nematode C. elegans: a platform for investigating biology, Science 282 (1998) 2012–2018. [5] J. Reboul, P. Vaglio, N. Tzellas, N. Thierry-Mieg, T. Moore, C. Jackson, T. Shin-I, Y. Kohara, D. Thierry-Mieg, J. Thierry-Mieg, H. Lee, J. Hitti, L. Doucette-Stamm, J. Hartley, G. Temple, M. Brasch, J. Vandenhaute, P. Lamesch, D. Hill, M. Vidal, C. elegans ORFeome version 1.1: experimental verification of the genome annotation and resource for proteome-scale protein expression, Nat. Genet. 27 (2001) 332–336. [6] L. Stein, P. Sternberg, R. Durbin, J. Thierry-Mieg, J. Spieth, WormBase: network access to the genome and biology of Caenorhabditis elegans, Nucleic Acids Res. 29 (2001) 82–86. [7] A.Y. Borovkov, M.I. Rivkin, XcmI-containing vector for direct cloning of PCR products, BioTechniques 22 (1997) 812–814. [8] C.C. Mello, J.M. Kramer, D. Stinchcomb, V. Ambros, Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences, EMBO J. 10 (1991) 3959– 3970. [9] L. Timmons, A. Fire, Specific interference by ingested dsRNA, Nature 395 (1998) 854. [10] A.R. Winnier, J.Y.-J. Meir, J.M. Ross, J. Tavernarakis, M. Driscoll, T. Ishihara, I. Katsura, D.M. Miller, UNC-4/UNC-37dependent repression of motor neuron-specific genes controls synaptic choice in Caenorhabditis elegans, Genes Dev. 13 (1999) 2774–2786. [11] Y. Jin, E. Jorgensen, E. Hartwieg, R. Horvitz, The Caenorhabditis elegans gene unc-25 encodes glutamic acid decarboxylase and is required for synaptic transmission but not synaptic development, J. Neurosci. 19 (1999) 539–548. [12] E.J. Lambie, J.E. Kimble, Two homologous regulatory genes, lin-12 and glp-1, have overlapping functions, Development 112 (1991) 231– 240. [13] W.G. Kelly, S. Xu, M.K. Montgomery, A. Fire, Distinct requirements for somatic and germline expression of a generally expressed Caernorhabditis elegans gene, Genetics 146 (1997) 227– 238. [14] R.S. Kamath, A.G. Fraser, Y. Dong, G. Poulin, R. Durbin, M. Gotta, A. Kanapink, N. Le Bot, S. Moreno, M. Sohrmann, D.P. Welchman, P. Zipperlen, J. Ahringer, Systematic functional analysis of the Caenorhabditis elegans genome using RNAi, Nature 421 (2003) 231–237. [15] D.M. Miller, C.J. Niemeyer, Expression of the unc-4 homeoprotein in Caenorhabditis elegans motor neurons specifies presynaptic input, Development 121 (1995) 2877–2886.
An efficient transgenic system by TA cloning vectors and RNAi for C. elegans Keiko Gengyo-Ando a
a,b
, Sawako Yoshina
a,b,c
, Hideshi Inoue c, Shohei Mitani
a,b,*
Department of Physiology, Tokyo Women’s Medical University School of Medicine, 8-1, Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan b CREST, JST, 4-1-8 Hon-cho, Kawaguchi, Saitama 332-0012, Japan c School of Life Sciences, Tokyo University of Pharmacy and Life Sciences, 1432-1, Horinouchi, Hachiouji, Tokyo 192-0392, Japan Received 29 August 2006 Available online 11 September 2006
Abstract In the nematode, transgenic analyses have been performed by microinjection of DNA from various sources into the syncytium gonad. To expedite these transgenic analyses, we solved two potential problems in this work. First, we constructed an efficient TA-cloning vector system which is useful for any promoter. By amplifying the genomic DNA fragments which contain regulatory sequences with or without the coding region, we could easily construct plasmids expressing fluorescent protein fusion without considering restriction sites. We could dissect motor neurons with three colors in a single animal. Second, we used feeding RNAi to isolate transgenic strains which express lag2::venus fusion gene. We found that the fusion protein is toxic when ectopically expressed in embryos but is functional to rescue a loss of function mutant in the lag-2 gene. Thus, the transgenic system described here should be useful to examine the protein function in the nematode. Ó 2006 Elsevier Inc. All rights reserved. Keywords: TA cloning vector; RNAi; Fluorescent protein; Transgenic
In the nematode Caenorhabditis elegans, transgenic analyses have been fruitful to analyze the expression patterns of genes of interest, rescuing mutants by introducing wild-type genomic fragments spanning the whole region of the genes of interest and so on [1]. These convenient transgenic analyses are based on the fact that DNA without obvious origin is able to be replicated in nematode nuclei [2]. The phenomenon has led to a number of useful transgenic protocols including injecting DNA into gonad from various biological sources such as plasmid, k phage, YAC, or cosmid clones, PCR fragments [1]. However, there are a number of potential problems to be solved to perform these experiments more efficiently. First, expression vectors most popularly used need appropriate restriction sites to subclone promoters [3]. Because promoter analyses need frequently about 5 kb of 5 0 flank*
Corresponding author. Fax: +81 3 5269 7362. E-mail address: [email protected] (S. Mitani).
0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.08.183
ing sequences so that genes of interest be expressed correctly. The longer a genome sequence is, the more difficult it is to choose an appropriate restriction site. The most convenient DNA appear to be plasmids, which are available in a large amount with high purity. Thus, it is helpful if PCR fragments are conveniently cloned into expression plasmids quickly, because the whole genome has been sequenced and all the genes have been deduced by a computer program and cDNA analyses [4–6]. TA-cloning vectors, which are designed to accept PCR fragments by overhanging deoxythymidines at both 3 0 ends of the cloning sites, are one of the easiest vectors to handle, because PCR fragments with any restriction sites can be subcloned [7]. Here, we designed a set of multicolor TA-cloning vectors, which are useful for expression analyses with or without coding sequences. Another problem is more elusive; we often have troubles that the transgenic animals fail to have any expected expression. A fraction of the cases may be caused by the
1346
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
toxicity of ectopically expressed products. To overcome this problem, we tried to knock down the products by feeding RNAi with a few generations of transgenic strains. After establishing transgenic strains, we could allow the transgene to express the fusion protein by transferring transgenic animals to the common food bacteria OP50-1. Combined together, transgenic analyses in C. elegans are expected to be done more quickly. Materials and methods Construction of reporters and rescue plasmids. To prepare reporter constructs used in this study, the PCR fragments amplified from N2 genome were directly cloned into the TA expression cloning vectors. The vectors were digested with XcmI to create the linearized vectors with 3 0 -T overhangs and used for inserting PCR fragments containing upstream sequences and the first several codons of the genes acr-5, unc-4, and unc-25 [10,11]. acr-5 was amplified using pyrobest polymerase (Takara), and other genes except acr-5 were amplified using platinum Taq high fidelity (Invitrogen), from wild-type (N2) genome. Primers used for amplification of each promoter region are as follows; 5 0 -GGGAAGCATGCTGAAA ATTG-3 0 and 5 0 -CGCAACGACGGCATTCACTT-3 0 for acr-5, 5 0 -CAG TGCACCGATCATTTTCA-3 0 and 5 0 -TCTGAGACCATGTTGATGG G-3 0 for unc-4, and 5 0 -GCATGCAAAAAACACCCACT-3 0 and 5 0 -CATT TTTGGCGGTGAACTGA-3 0 for unc-25. Approximate sizes of PCR fragments were 4.2, 2.8, and 1.8 kb for acr-5, unc-4, and unc-25, respectively. The expression plasmids having insert fragments were quickly verified by digesting with SpeI. If the plasmids cannot be cut with SpeI, they are unlikely to be cloned into the expression vector in-frame. We confirmed the junction sequence of cloning site using the appropriate primers. To construct functional lag-2::venus plasmid, a 5 kb promoter fragment obtained by amplifying N2 genome by primers 5 0 -TGGATCCCGT TTTCCGATTTGCCGGAA-3 0 and 5 0 -TGCGGCCGCCGATCATTTTC TGAAAAAAGGC-3 0 was subcloned between BamHI and NotI sites in the pFX_venusT giving rise to Plag-2::venus. A 1.5 kb of genomic fragment containing full length lag-2 coding sequence was amplified with 50 -AAATGATCGGCGGCCGCATGATCGCTTACTTCCTCTTACTCC-30 and 5 0 -GCTCACCATGCGGCCGCAGACATAGTGACAGGCTGGA ATAG-3 0 . The amplified DNA fragment was inserted into the NotI site of the Plag-2::venus using a recombinase in-Fusion (Clonetech) giving rise to lag-2::venus. Germline transformation. For triple staining, the Pacr-5::ECFP, Punc4:: EYFP, and Punc-25:: DsRedx were injected at 40 ng/ll into adult N2 hermaphrodites by standard germline transformation techniques [8]. pRF4 plasmid (40 ng/ll) containing the dominant roller marker rol-6 (su1006) was co-injected to identify transgenic animals. N2 and FX11226 (lag-2 (q411)/nT1 [qIs51]) mutant animals were injected with the rescue plasmid Plag-2::venus (10 ng/ll) and Pmyo-2::DsRed (10 ng/ll). qIs51 harbors an insertion of ccEx9747 with markers:myo-2::GFP, pes-10::GFP, and F22B7.9::GFP, resulting in expression of GFP in the pharynx, embryo, and intestine. Homozygous lag-2 mutant animals were identified by loss of GFP fluorescence linked to balanced chromosome nT1 (qIs51). Feeding RNAi. Feeding RNAi was performed as described [9]. Briefly, the venus cDNA was amplified by PCR using sense and antisense primers, 5 0 -CGCGGGTACCTGTACAGCTCGTCCATGCCG-3 0 and 5 0 -CGCG TCTAGAATGGTGAGCAAGGGCGAGGA-3 0 . The cDNA fragment was cloned into the L4440 (pPD129.36) vector and the cloned plasmid was transformed into Escherichia coli HT115 (DE3). Nematode growth medium (NGM) agar dishes with 1 mM IPTG and 50 lg/ml of ampicillin were seeded with the bacterial clone which expresses double-stranded RNA of venus. Animals of the L4 stage were pre-cultured on the venus RNAi dishes until adult stage and were picked up for microinjection of plasmids as described. After animals were microinjected with plasmid DNA, they were cultured on RNAi dishes. After picking up DsRedpositive transgenic F1 animals, we further cultured the animals until
isolation of F2 animals which kept DsRed-positive transgenic animals. To examine the toxicity of LAG-2/Venus protein, recipients were cultured on bacterial lawn of either OP50-1 or HT115 (DE5) carrying a plasmid expressing venus dsRNA. Transgenic lines (lag-2(q411); tmEx1432– tmEx1436[lag-2::venus + Pmyo-2::DsRed]) carrying stably transmitting extrachromosomal arrays were used for fertility assay. Microscopy. Animals were mounted on 5% agarose pads in M9 buffer containing 10 mM sodium azide and observed by differential interference contrast (DIC) or epifluorescence microscopy. Images were observed using a microscope system composed of the microscope (Olympus BX-50) and the CCD camera (Olympus sensys), and analyzed by the IPLab software (Scanalytics). We used filter and dichroic mirror set for each fluorescent protein as follows. ECFP: excitation band-pass 425–445, 450 dichroic, emission band-pass 460–510, EGFP: excitation band-pass 460–480, 485 dichroic, emission band-pass 495–540, EYFP, and venus: excitation bandpass 490–500, 505 dichroic, emission band-pass 515–560, and DsRedExpress-1: excitation band-pass 520–550, 565 dichroic, emission band-pass 580IF.
Results and discussion Design of expression plasmids All the plasmids constructed in this study share common Bluescript backbones (Stratagene, CA). We inserted a multiple cloning site followed by fluorescent proteins of various colors, and about 1 kb-long 3 0 untranslated region from the unc-86 gene to add a signal sequence for poly(A) addition by using restriction sites as described in Fig. 1A. To perform multicolor staining, we inserted cDNAs of ECFP, EGFP, EYFP, venus, and DsRedExpress1 for pFX_ECFPT, pFX_EGFPT, pFX_EYFPT, pFX_venusT, and pFX_DsRedxT, respectively. The cDNAs were purchased from Clonetech (ECFP, EGFP, EYFP, and DsRedExpress1) or kindly offered by Dr. A. Miyawaki (venus). The plasmids when digested by a restriction enzyme XcmI give rise to linear fragments with an overhanging deoxythymidine at each 3 0 cut end [7] (Fig. 1B). The PCR fragments, with or without coding region, can be subcloned in-frame by adjusting the 3 nucleotides of the 5 0 end of antisense primers coding for the amino acid(s) to the cDNA of fluorescent proteins. The reporter plasmids constructed this way once examined for the expression pattern could be converted to functional expression plasmids for either native nematode proteins of wild-type or any mutant forms, or proteins from, for example, human cDNAs. This could be easily performed by inserting a cDNA fragment amplified by PCR with a SpeI or NotI restriction site at the 5 0 end and BglII site at the 3 0 end (Fig. 1B). Dissection of neuron subtypes by a multicolor reporter system To address whether the multicolor reporter system works for the nematode expression analysis, we amplified and subcloned the promoter DNA fragment with the first methionine into expression plasmids as three color combination; ECFP for acr-5 promoter, EYFP for unc-4 promoter, and DsRed Express1 for unc-25 promoter. Three
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
1347
A
B
Fig. 1. Schematic representation of method for cloning PCR products into T vector pFX_EGFPT. (A) Schematic diagram of pFX_EGFPT. All vectors produced in this work have same sequences except for the coding sequence of each fluorescent protein. Sequence of multiple cloning site (MCS) is shown in (B). (B) PCR products containing promoter sequence with only first methionine or most of the coding region are cloned into pFX_EGFPT digested with XcmI. Expression constructs using this cloning vector have additional five amino acids derived from multiple cloning sequence between the examined coding sequence and the fluorescent protein coding sequence.
promoters are said to drive reporter expression in the neuronal subtypes, and motor neurons, which are exclusively with each other [10,11]. The fluorescence observed was consistent with the previous notion, indicating that the three color reporters can be easily distinguished with each other when co-injected to the germ cells (Fig. 2). To confirm that our expression system is also useful for other genes, we constructed the expression plasmids containing the regulatory sequences that control the expression of various tissues: myo-3 (body-wall muscle), myo-2 (pharyngeal muscle), ges-1 (intestine), dpy-7 (hypodermis), unc-119 (all neurons), unc-122 (coelomocytes), eft-3 (ubiquitous), and so on. These expression plasmids were injected
into N2 animals and the transgenic lines having the extrachromosomal arrays showed the expected expression patterns (data not shown). Efficient rescue of lag-2 mutant phenotype by the aid of feeding RNAi against venus Transgenic strains are generated by injecting plasmid constructs into syncytial gonads in C. elegans [8]. However, we and others often have difficulties to obtain transgenic strains expressing reporter genes in spite of correct marker expression [1]. These problems are probably due to multiple factors. We reasoned that at least a fraction of these
1348
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350 Table 1 Toxic ectopic expression causes the death of F1 animals of lag-2::venus transgenic animals Condition
With RNAi
Without RNAi
DsRed(+) F1 Hatched F1 DsRed(+) F2 lines LAG-2/Venus(+) lines
124 112 13 8
91 2 0 0
A mixture of plasmid DNAs consisting of Pmyo-2::DsRed and lag-2::venus (1:1 dilution) was injected to N2 animals treated with or without feeding RNAi against venus. Number of F1 animals were counted as DsRed (+) embryos. Some of the embryos without feeding RNAi were DsRed () venus (+), presumably because myo-2 gene expression is found only after organogenesis stage, hatching rate of RNAi () was a little overestimation by reducing the number of DsRed (+) F1 animals.
Fig. 2. Triple staining of motor neurons in ventral nerve cord. Transgenic strain carrying three expression plasmids, acr-5::ECFP, unc-4::EYFP, and unc-25::DsRedx, was constructed. Same L1 larva is shown in (A–E). (A) acr-5::ECFP expression. acr-5 encodes a acetylcholine receptor subunit expressed in B-type motor neuron (DB) in L1 larva ([10]). Some DB motor neurons could not be identified since out of focus. (B) unc-4::EYFP expression. unc-4 encodes a homeodomain transcription factor expressed in A-type motor neuron (DA) in L1 larva [15]. (C) unc-25::DsRedx expression. unc-25 encodes glutamic acid decarboxylase expressed in Dtype motor neuron (DD) in L1 larva ([11]). Merged image of (A–C) is shown in (D) and an image with Nomarski optics shown in (E).
cases are due to the toxicity of expressed products, and non-expressing transgenic strains tend to survive selectively. To examine this possibility and find out the solution, we
chose a gene lag-2, which is essential for fertility of the animal by regulating meiotic division of germ cells [12]. We found ectopically expressed LAG-2/Venus fusion protein is toxic to animals. When the lag-2::venus plasmid is injected into N2 animals and cultured on OP50-1 as a food, we could not isolate any stable lag-2::venus positive transgenic strains (Table 1). This failure appears partly due to ectopic expression of LAG-2/Venus fusion protein, because dead embryos are often venus-positive at variable cells (Fig. 3). It is known that expression of transgenes in germ cells is reduced over several generations of the transgenic strains by silencing [13]. However, during the first few generations, sometimes transgenes are only weakly silenced in the germ cells. We reasoned that if toxic products are, for example, expressed ectopically in embryonic cells of F1 animals, animals with transgenes selectively become unhealthy or dead. To examine this possibility, we used the feeding RNAi method to knock down venus, and thus the lag-2 gene expressed as a fused mRNA at the early generations [9]. When the plasmid is injected into N2 animals and cultured with feeding RNAi, which knocks down venus, we could isolate stable transgenic lines (Table 1). To address whether this transgenic method is useful for functional analysis of expressed proteins, we tried to express the same lag-2::venus fusion gene in the lag-2 mutant background. After a few generations of isolating transgenic establishment, we picked up homozygous lag-2 mutant animals without the balancer nT1[qIs51] but with the transgenic marker Pmyo-2::DsRed and transferred the animals to dishes with OP50-1 as a food source. The homozygous lag-2 mutant animals without transgene were all sterile, whereas homozygous lag-2 mutant animals with the lag-2::venus transgene were fertile (mean number of progeny = 50.8, n = 5). We also examined the expression pattern of the LAG-2/ Venus protein by fluorescence microscopy (Fig. 3). We found the LAG-2/Venus protein in DTC and also weakly in germ cells, suggesting that the fusion protein is secreted by DTC to proximal gonadal cells, consistent with the previous study [12].
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
1349
Fig. 3. Expression of lag-2::venus (A–B) Ectopic expression of lag-2::venus in a dying embryo of F1 animal fed with OP50-1 (A, DIC image; B, Venus fluorescence). (C–D) No ectopic expression of lag-2::venus in a normal embryo of F1 animals fed with RNAi against venus (C, DIC image; D, Venus fluorescence). (E–G) Functional expression of lag-2::venus in the homozygous lag-2 mutant background (E, DIC image; F, Venus fluorescence found in a Distal Tip Cell; G, merged). Scale bars = 10 lm for (A–C), 20 lm for (D–F).
We succeeded in isolating functional protein expressing transgenic lines by culturing the animals in an early few generations on feeding RNAi dishes. RNAi against fluorescent protein did not appear to knock down endogenous
lag-2 mRNA, which is important for fertility of the animals. This may be caused by the fact endogenous lag-2 mRNA is resistant to RNAi [14]. Another advantage is the use of the same fluorescent protein as an RNAi target,
1350
K. Gengyo-Ando et al. / Biochemical and Biophysical Research Communications 349 (2006) 1345–1350
we are able to use the same construct for functional expression of many genes. It should be noted that any other fluorescent reporter like DsRed is not affected and can be used as a transgenic marker and so on. Conclusions We constructed a convenient TA-cloning vector system, which is useful for triple staining and functional gene expression. Because the whole genome sequences are available, the present TA-cloning vector system makes researchers possible to design many reporter constructs without restriction site analysis. The feeding RNAi method present here is useful to isolate reporter transgenic without considering ectopic expression of transgenes. In that context, the present transgenic system will facilitate the analysis of the genetics of this useful model organism. Acknowledgments We thank Drs. Yuji Kohara, Masatoshi Hagiwara, Yuichi Iino, Takeo Awaji, and Takeshi Fukuhara for discussion. We are grateful to Dr. Atsushi Miyawaki for a venus plasmid. Some nematode strains used in this work were provided by the Caenorhabditis Genetics Center funded by the National Institutes of Health National Center for Research Resources. This work was supported by grants from MEXT, Japan, and JST to S.M. and K.G.-A. References [1] C. Mello, A. Fire, DNA transformation, Methods Cell Biol. 48 (1995) 451–482. [2] D.T. Stinchcomb, J.E. Shaw, S.H. Carr, D. Hirsh, Extrachromosomal DNA transformation of Caenorhabditis elegans, Mol. Cell. Biol. 5 (1985) 3484–3496. [3] A. Fire, S.W. Harrison, D. Dixon, A modular set of lacZ fusion vectors for studying gene expression in Caenorhabditis elegans, Gene 93 (1990) 189–198.
[4] C. elegans, Sequencing Consortium, Genome sequence of the nematode C. elegans: a platform for investigating biology, Science 282 (1998) 2012–2018. [5] J. Reboul, P. Vaglio, N. Tzellas, N. Thierry-Mieg, T. Moore, C. Jackson, T. Shin-I, Y. Kohara, D. Thierry-Mieg, J. Thierry-Mieg, H. Lee, J. Hitti, L. Doucette-Stamm, J. Hartley, G. Temple, M. Brasch, J. Vandenhaute, P. Lamesch, D. Hill, M. Vidal, C. elegans ORFeome version 1.1: experimental verification of the genome annotation and resource for proteome-scale protein expression, Nat. Genet. 27 (2001) 332–336. [6] L. Stein, P. Sternberg, R. Durbin, J. Thierry-Mieg, J. Spieth, WormBase: network access to the genome and biology of Caenorhabditis elegans, Nucleic Acids Res. 29 (2001) 82–86. [7] A.Y. Borovkov, M.I. Rivkin, XcmI-containing vector for direct cloning of PCR products, BioTechniques 22 (1997) 812–814. [8] C.C. Mello, J.M. Kramer, D. Stinchcomb, V. Ambros, Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences, EMBO J. 10 (1991) 3959– 3970. [9] L. Timmons, A. Fire, Specific interference by ingested dsRNA, Nature 395 (1998) 854. [10] A.R. Winnier, J.Y.-J. Meir, J.M. Ross, J. Tavernarakis, M. Driscoll, T. Ishihara, I. Katsura, D.M. Miller, UNC-4/UNC-37dependent repression of motor neuron-specific genes controls synaptic choice in Caenorhabditis elegans, Genes Dev. 13 (1999) 2774–2786. [11] Y. Jin, E. Jorgensen, E. Hartwieg, R. Horvitz, The Caenorhabditis elegans gene unc-25 encodes glutamic acid decarboxylase and is required for synaptic transmission but not synaptic development, J. Neurosci. 19 (1999) 539–548. [12] E.J. Lambie, J.E. Kimble, Two homologous regulatory genes, lin-12 and glp-1, have overlapping functions, Development 112 (1991) 231– 240. [13] W.G. Kelly, S. Xu, M.K. Montgomery, A. Fire, Distinct requirements for somatic and germline expression of a generally expressed Caernorhabditis elegans gene, Genetics 146 (1997) 227– 238. [14] R.S. Kamath, A.G. Fraser, Y. Dong, G. Poulin, R. Durbin, M. Gotta, A. Kanapink, N. Le Bot, S. Moreno, M. Sohrmann, D.P. Welchman, P. Zipperlen, J. Ahringer, Systematic functional analysis of the Caenorhabditis elegans genome using RNAi, Nature 421 (2003) 231–237. [15] D.M. Miller, C.J. Niemeyer, Expression of the unc-4 homeoprotein in Caenorhabditis elegans motor neurons specifies presynaptic input, Development 121 (1995) 2877–2886.