Accepted Manuscript Notes & tips Alternative methods for the efficient construction of shRNA expression vectors Kun Xu, Tingting Zhang, Lijun Guo, Ying Xin, Long Zhang, Zhiying Zhang PII: DOI: Reference:
S0003-2697(15)00103-7 http://dx.doi.org/10.1016/j.ab.2015.03.006 YABIO 12005
To appear in:
Analytical Biochemistry
Received Date: Revised Date: Accepted Date:
30 December 2014 28 February 2015 2 March 2015
Please cite this article as: K. Xu, T. Zhang, L. Guo, Y. Xin, L. Zhang, Z. Zhang, Alternative methods for the efficient construction of shRNA expression vectors, Analytical Biochemistry (2015), doi: http://dx.doi.org/10.1016/j.ab. 2015.03.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Alternative methods for the efficient construction of shRNA
2
expression vectors
3 Kun Xu#,1, Tingting Zhang#,1,2, Lijun Guo1, Ying Xin1, Long Zhang1, 3 , Zhiying Zhang*,1
4 5 6
1
7
2
Research Institute of Applied Biology, Shanxi University, Taiyuan, Shanxi, 030006, China
8
3
Frontier Institute of Science and Technology, Xi'an Jiaotong University, Xi’an, Shaanxi, 710061, China
9
*
College of Animal Science & Technology, Northwest A&F University, Yangling, Shaanxi, 712100, China
To whom correspondence should be addressed. Tel: +86-029-87092102; Fax: +86-029-87092164; Email:
10
[email protected]
11
#
Co-first authors
12 13 14
Subject category: DNA Recombinant Techniques and Nucleic Acids
15 16
Short title: Alternative methods for shRNA vector construction
17
1
1
Abstract
2
shRNA mediated RNA interference (RNAi) has become a basic technique in modern molecular biology
3
and biochemistry for studying gene function and biological pathways. Here, we report two alternative and
4
efficient methods to construct shRNA expression vectors，respectively based on multiple-step sequential
5
PCR (msPCR) and primer extension-homologous recombination (PE-HR). Both methods don’t require
6
synthesizing long oligonucleotides containing hairpin sequences as used in traditional approaches. The
7
hairpin sequences may produce mutations in oligo synthesizing, pose problems in annealing and lead to
8
inefficient cloning. The PE-HR method further provides rapid and economical construction of shRNA
9
expression vectors without needing the ligation procedure.
10 11
Keywords: RNAi; shRNA vector construction; multiple-step sequential PCR; primer extension;
12
homologous recombination
13
2
1
RNA interference (RNAi) has been widely used for gene function analysis and elucidation of cellular
2
signal transduction pathways [1]. The RNA interference can be achieved by the delivery of
3
double-stranded RNA (dsRNA) into mammalian cells, in which a single strand RNA (siRNAs, about 20 nt
4
in length) will be eventually generated for targeting mRNA [2]. Initially, RNAi was introduced by directly
5
transfecting cells with chemical synthesized or in vitro transcribed siRNA [2, 3]. Alternatively,
6
vector-based short hairpin RNA (shRNA) can be introduced and expressed within host cells as dsRNA
7
with a loop, which will further be processed into functional siRNA. Compared with the in vitro generated
8
siRNA, the construction of vector-based shRNA is much more economical.
9
The most popular technique used to generate shRNA expression vectors is oligonucleotide annealing,
10
which routinely requires the synthesis and annealing of two complementary oligonucleotides with each
11
oligonucleotide containing the whole shRNA hairpin sequence [4]. However, this method is often limited
12
by high mutation rate and low annealing efficiency [4, 5]. The synthesis of long oligonucleotides (more
13
than 40 nt) had been considered to be prone to introduce errors. But in our previous experiments, we have
14
synthesized several of long oligonucleotides (40~60 nt) and did not found any high rate of mutation [6-8],
15
to which the improved synthesis technology may contribute in recent years. Nevertheless, when we tried to
16
construct shRNA expression vectors by annealing oligonucleotides with hairpin sequences, we detected a
17
high number of mutations from positive clones (about 50%). Therefore we assumed that it is the hairpin
18
sequences rather than the oligo length that may impair the accuracy for the synthesis of long
19
oligonucleotides.
20
On the other hand, low annealing efficiency for shRNA construct fragments was considered to impair
21
the subsequent ligating reaction. We initially used pLL3.7 shRNA cloning vector (Addgene), which applies
22
HpaI/XhoI sites for shRNA construct cloning (Figure S1). We tried to construct four shRNA expression 3
1
vectors against human E2F5 gene by simply annealing oligonucleotides (Figure S2, Table S2), according
2
to the supplemented ‘pLL3.7 shRNA cloning’ protocol (https://www.addgene.org/11795/). Unfortunately,
3
after three rounds of cloning, we got only two correct E2F5-shRNA expression vectors as designed. The
4
inefficiency was firstly reasoned to the HpaI cloning site, which leaves blunt end that was thought to
5
impair the subsequent ligating reaction. Thus, we modified the pLL3.7 vector by introducing a type IIs
6
restriction enzyme BsmBI site, which was designed to cut just before the -1 position of the mU6 promoter
7
leaving cohesive ends. By applying the modified cloning strategy (Figure S3), although we succeeded to
8
construct three E2F5-shRNA vectors after two rounds of cloning, limited positive colonies were obtained
9
for screening correct clones. These experiment experiences indicated that the inefficiency, to some extent,
10
may be due to the low annealing efficiency for shRNA construct fragments, which may impair the
11
subsequent ligating reaction.
12
Besides, the cost for synthesis of oligonucleotides longer than 60 nt will be more than doubled, and
13
unfortunately if we choose shRNA target sites (Table S1) that are or more than 20 nt, the length for the
14
needed oligonucleotides will exceed 60 nt (Table S2).
15
To solve the problems as described above, we developed two alternative and efficient methods to
16
construct shRNA expression vectors, respectively based on multiple-step sequential PCR (msPCR) and
17
primer extension-homologous recombination (PE-HR). Both methods can be used to construct shRNA
18
expression vectors without needing to synthesize long oligonucleotides containing hairpin sequences.
19
For the msPCR method (Figure 1), the mU6 promoter sequence was firstly amplified by a pair of
20
primers mU6.F/Primer.P1. The Prmer.P1 contains the shRNA target antisense sequence and the loop
21
sequence.
22
cassette with primers mU6.F/Primer.P2. The Prmier.P2 consists of the loop sequence and the shRNA target
The first PCR product was used as template for amplification of the whole shRNA expression
4
1
sense sequence. For routine shRNA expression cassette, simply two sequential standard PCR steps will be
2
available with each primer limited under 60 nt in length (Table S2). The PCR amplified shRNA expression
3
cassette can be inserted into pLL3.7 between XbaI/XhoI sites replacing the former mU6 promoter sequence
4
by standard double digestion-ligation cloning method, which ensures to generate sufficient positive
5
colonies for screening correct clones in a single round of cloning. To demonstrate the feasibility of our
6
method, we initially constructed two E2F5-shRNA expression vectors. Double digestion screening yielded
7
80% (8 out of 10) of colonies harboring the desired shRNA expression plasmid clones. All of the positive
8
clones were further sequenced and demonstrated containing the correct shRNA expression cassettes.
9
Similarly, by employing an improved three-step sequential PCR, we successfully generated three complex
10
shRNA constructs flanked with Drosha site and miR30 sequence against mouse CD40 gene, which have
11
been demonstrated to function in both cell assay and in vivo targeted delivery study [6].
12
Compared with oligonucleotide annealing, the msPCR method promises the accuracy and efficiency by
13
avoiding to synthesize and to anneal oligonucleotides with hairpin sequences and guaranteeing to
14
generate sufficient positive colonies for further screening. According to the experience in our lab, the
15
msPCR with the long-tailed primers can be specifically and efficiently achieved by hot-start and
16
touch-down PCR. The length of the shRNA expression cassette is less than 600 bp, of which the accuracy
17
can be promised by standard PCR reactions using the mixture of Taq/Pfu (1:1) DNA polymerases
18
(TransGen Biotech, Beijing, China). It is noteworthy that PCR reactions with Taq DNA polymerase alone
19
usually cause point mutations in our experience. The msPCR method requires multiple PCR steps, routine
20
double digestion-ligation reactions. Although all of these procedures can be finished in one working day, it
21
seems still a little time-consuming.
22
There is an interesting method that firstly ligates the shRNA hairpin construct to a double-stranded DNA 5
1
end, then it uses nicking enzymes to open the hairpin construct to be a single-stranded inverted repeat
2
DNA, and finally with the help of the extension reaction, the single-stranded inverted repeat DNA will be
3
converted to be the whole double-stranded shRNA construct for shRNA expressing [9, 10]. Although this
4
method can be used to generate shRNA from cDNA , it requires a series of enzymatic reactions [9]. For the
5
construction of a given shRNA expression vector, it still requires the synthesis of long oligonucleotides
6
with hairpin sequences, two times of ligating reaction and one nicking reaction [10]. Another alternative
7
approach has been reported to generate the whole shRNA construct by primer extension, which could be
8
efficient to construct a miRNA library, but it also requires standard double digestion-ligation reactions to
9
generate corresponding vectors [11]. Besides, this method may require two times (after the extension and
10
digestion reactions respectively) of purification for the short shRNA construct fragment (~60 bp), which is
11
usually inefficient. Recent years, in vivo homologous recombination (HR) has been developed and widely
12
used for plasmid construction [12-15], and it performs well with the homologous sequence length to be
13
only 20 bp [15]. Therefore it is possible to rapidly and economically construct shRNA expression vectors
14
by combining the primer extension and homologous recombination strategies.
15
To validate the primer extension and homologous recombination (PE-HR) based method (Figure 2), five
16
shRNA target sites were chosen from human DYRK1A and mouse Igf2 genes (Table S1). The Klenow
17
fragment (TAKARA, Dalian, China) and Taq DNA polymerase (TransGen Biotech, Beijing, China) were
18
firstly compared for primer extension reactions. We found that Klenow fragment was better and 1 unit
19
Klenow for 10 min at 37 °C was sufficient for the extension reaction as shown in Figure S4. E.coli JM109
20
competent cells were co-transformed with the purified primer extension products and HpaI/XhoI-linearized
21
parental pLL3.7 plasmid DNA. Positive colonies were picked, and plasmids were prepared and checked by
22
double digestion for positive clones. 96% (24 out of 25) of these positive clones were sequenced to contain 6
1
the correct shRNA expression cassettes, and only one was detected with a point mutation. This result
2
indicated that our PE-HR method is efficient for the construction of shRNA expression vectors.
3
HR frequency in E.coli is crucial for the PE-HR method, and therefore we tested three different E.coli
4
stains (BJ5183, JM109 and DH5α) using different transformation methods. BJ5183 is used for doing HR
5
as employed in adenovirus vector construction [16]; DH5α and JM109 with high efficiency of
6
transformation are normally applied for plasmid construction and amplification. In theory, BJ5183 should
7
be a very appropriate strain for our method, and practically it did produce 3~5 times more colonies than
8
DH5α and JM109. But many clones (81 out of 96) were produced by non-homologous end joining (NHEJ)
9
and only a few (15 out of 96) were confirmed to be correct. On the other hand, we found more than 95%
10
clones of DH5α and JM109 were correctly generated. Since the RecA gene of DH5α and JM109 strains is
11
mutated, we speculated that the recombination occurred in DH5α and JM109 might be mediated by RecEF
12
[14]. In addition, high transformation efficiency is also important, which contributes to improve the
13
interaction possibility between homologous sequences. We tried electric and chemical transformations
14
(Inoue and calcium sodium methods), and observed that electro-transformation and Inoue method were
15
sufficient to produce more than 100 colonies per transformation, whereas only several colonies or none at
16
all for the calcium chloride method.
17
In conclusion, we showed two alternative and efficient methods to construct shRNA expression vectors
18
without needing to synthesize and to anneal long oligonucleotides with hairpin sequences, which were
19
considered to be responsible for the high mutation rate and low cloning efficiency. The msPCR method
20
employs routine PCR, double digestion-ligation reactions and guarantees to generate sufficient positive
21
colonies for further screening. The PE-HR method relies on homologous recombination and further
22
provides rapid and economical shRNA expression vector construction without needing the ligation 7
1
procedure. Compared with traditional oligonucleotide annealing as we tried, our alternative methods are
2
more efficient and cost effective for the construction of shRNA expression vectors.
3 4
Acknowledgments
5
This research was funded by National Natural Science Foundation of China (NSFC; No.30870119 and
6
31171186) and China's Ministry of Agriculture (948 Program; No.2013-Z27).
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
References [1] D.B. Stovall, M. Wan, Q. Zhang, P. Dubey, G. Sui, DNA vector-based RNA interference to study gene function in cancer, Journal of visualized experiments : JoVE, (2012) e4129. [2] S.M. Elbashir, J. Harborth, W. Lendeckel, A. Yalcin, K. Weber, T. Tuschl, Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells, Nature, 411 (2001) 494-498. [3] B. Luo, A.D. Heard, H.F. Lodish, Small interfering RNA production by enzymatic engineering of DNA (SPEED), Proceedings of the National Academy of Sciences of the United States of America, 101 (2004) 5494-5499. [4] M. Miyagishi, H. Sumimoto, H. Miyoshi, Y. Kawakami, K. Taira, Optimization of an siRNA-expression system with an improved hairpin and its significant suppressive effects in mammalian cells, The journal of gene medicine, 6 (2004) 715-723. [5] P.J. Paddison, J.M. Silva, D.S. Conklin, M. Schlabach, M. Li, S. Aruleba, V. Balija, A. O'Shaughnessy, L. Gnoj, K. Scobie, K. Chang, T. Westbrook, M. Cleary, R. Sachidanandam, W.R. McCombie, S.J. Elledge, G.J. Hannon, A resource for large-scale RNA-interference-based screens in mammals, Nature, 428 (2004) 427-431. [6] L. Zhang, T. Zhang, L. Wang, S. Shao, Z. Chen, Z. Zhang, In vivo targeted delivery of CD40 shRNA to mouse intestinal dendritic cells by oral administration of recombinant Sacchromyces cerevisiae, Gene therapy, 21 (2014) 709-714. [7] T. Zhang, H. Yang, R. Wang, K. Xu, Y. Xin, G. Ren, G. Zhou, C. Zhang, L. Wang, Z. Zhang, Oral administration of myostatin-specific whole recombinant yeast Saccharomyces cerevisiae vaccine increases body weight and muscle composition in mice, Vaccine, 29 (2011) 8412-8416. [8] K. Xu, C. Ren, Z. Liu, T. Zhang, T. Zhang, D. Li, L. Wang, Q. Yan, L. Guo, J. Shen, Z. Zhang, Efficient genome engineering in eukaryotes using Cas9 from Streptococcus thermophilus, Cellular and molecular life sciences : CMLS, (2014). [9] A. Dinh, Y.Y. Mo, Alternative approach to generate shRNA from cDNA, BioTechniques, 38 (2005) 629-632. [10] H. Tanaka, Construction of shRNA expression plasmids for silkworm cell lines using single-stranded DNA and Bst DNA polymerase, Methods in molecular biology, 942 (2013) 347-355. [11] D. Gou, H. Zhang, P.S. Baviskar, L. Liu, Primer extension-based method for the generation of a siRNA/miRNA expression vector, Physiological genomics, 31 (2007) 554-562. [12] J. Fu, X. Bian, S. Hu, H. Wang, F. Huang, P.M. Seibert, A. Plaza, L. Xia, R. Muller, A.F. Stewart, Y. Zhang, Full-length RecE enhances linear-linear homologous recombination and facilitates direct cloning for 8
1 2 3 4 5 6 7 8 9 10 11
bioprospecting, Nature biotechnology, 30 (2012) 440-446. [13] S.K. Sharan, L.C. Thomason, S.G. Kuznetsov, D.L. Court, Recombineering: a homologous recombination-based method of genetic engineering, Nature protocols, 4 (2009) 206-223. [14] Y. Zhang, J.P. Muyrers, G. Testa, A.F. Stewart, DNA cloning by homologous recombination in Escherichia coli, Nature biotechnology, 18 (2000) 1314-1317. [15] D. Zhu, X. Zhong, R. Tan, L. Chen, G. Huang, J. Li, X. Sun, L. Xu, J. Chen, Y. Ou, T. Zhang, D. Yuan, Z. Zhang, W. Shu, L. Ma, High-throughput cloning of human liver complete open reading frames using homologous recombination in Escherichia coli, Analytical biochemistry, 397 (2010) 162-167. [16] T.C. He, S. Zhou, L.T. da Costa, J. Yu, K.W. Kinzler, B. Vogelstein, A simplified system for generating recombinant adenoviruses, Proceedings of the National Academy of Sciences of the United States of America, 95 (1998) 2509-2514.
12 13
9
1
Figure Legends
2
Figure 1. pLL3.7 shRNA expression vector cloning by multiple-step sequential PCR (msPCR).
3
For amplification of the mU6-shRNA cassette, simply two sequential PCR steps will be available with
4
each primer limited under 60 nt in length. The PCR amplified shRNA expression cassette can be
5
inserted into pLL3.7 between XbaI/XhoI sites replacing the former mU6 promoter sequence by
6
standard double digestion-ligation cloning method. The -1 position (brown bold font) of the mU6
7
promoter should be reconstituted with nucleotide T and the +1 position (green bold font) with
8
nucleotide G is required for efficient RNA expression, according to the ‘pLL3.7 shRNA cloning’
9
protocol (https://www.addgene.org/11795/). The cutting sites or overhangs of restriction enzymes
10
HpaI and XhoI are indicated with red font. The orange and grey arrows represent respectively the
11
direct and inverted repeat sequences within the shRNA hairpin construct.
12 13
Figure 2. pLL3.7 shRNA expression vector cloning by primer extension and homologous
14
recombination (PE-HR). The shRNA construct flanked by homologous sequences (20 bp in length)
15
is generated by primer extension, with each primer limited under 60 nt in length. Corresponding
16
shRNA expression vector will be constructed by homologous recombination between the HpaI/XhoI
17
double-digested parental pLL3.7 backbone and the primer extension product. The -1 position (brown
18
bold font) of the mU6 promoter should be reconstituted with nucleotide T and the +1 position (green
19
bold font) with nucleotide G is required for efficient RNA expression, according to the ‘pLL3.7
20
shRNA cloning’ protocol (https://www.addgene.org/11795/). The cutting sites or residuals of
21
restriction enzymes HpaI and XhoI are indicated with red font. The orange and grey arrows represent
22
respectively the direct and inverted repeat sequences within the shRNA hairpin construct.
10
1
2 3 4
Fig. 1
5 6
11
1 2 3
Fig. 2
4 5
12
S0003-2697(15)00103-7 http://dx.doi.org/10.1016/j.ab.2015.03.006 YABIO 12005
To appear in:
Analytical Biochemistry
Received Date: Revised Date: Accepted Date:
30 December 2014 28 February 2015 2 March 2015
Please cite this article as: K. Xu, T. Zhang, L. Guo, Y. Xin, L. Zhang, Z. Zhang, Alternative methods for the efficient construction of shRNA expression vectors, Analytical Biochemistry (2015), doi: http://dx.doi.org/10.1016/j.ab. 2015.03.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Alternative methods for the efficient construction of shRNA
2
expression vectors
3 Kun Xu#,1, Tingting Zhang#,1,2, Lijun Guo1, Ying Xin1, Long Zhang1, 3 , Zhiying Zhang*,1
4 5 6
1
7
2
Research Institute of Applied Biology, Shanxi University, Taiyuan, Shanxi, 030006, China
8
3
Frontier Institute of Science and Technology, Xi'an Jiaotong University, Xi’an, Shaanxi, 710061, China
9
*
College of Animal Science & Technology, Northwest A&F University, Yangling, Shaanxi, 712100, China
To whom correspondence should be addressed. Tel: +86-029-87092102; Fax: +86-029-87092164; Email:
10
[email protected]
11
#
Co-first authors
12 13 14
Subject category: DNA Recombinant Techniques and Nucleic Acids
15 16
Short title: Alternative methods for shRNA vector construction
17
1
1
Abstract
2
shRNA mediated RNA interference (RNAi) has become a basic technique in modern molecular biology
3
and biochemistry for studying gene function and biological pathways. Here, we report two alternative and
4
efficient methods to construct shRNA expression vectors，respectively based on multiple-step sequential
5
PCR (msPCR) and primer extension-homologous recombination (PE-HR). Both methods don’t require
6
synthesizing long oligonucleotides containing hairpin sequences as used in traditional approaches. The
7
hairpin sequences may produce mutations in oligo synthesizing, pose problems in annealing and lead to
8
inefficient cloning. The PE-HR method further provides rapid and economical construction of shRNA
9
expression vectors without needing the ligation procedure.
10 11
Keywords: RNAi; shRNA vector construction; multiple-step sequential PCR; primer extension;
12
homologous recombination
13
2
1
RNA interference (RNAi) has been widely used for gene function analysis and elucidation of cellular
2
signal transduction pathways [1]. The RNA interference can be achieved by the delivery of
3
double-stranded RNA (dsRNA) into mammalian cells, in which a single strand RNA (siRNAs, about 20 nt
4
in length) will be eventually generated for targeting mRNA [2]. Initially, RNAi was introduced by directly
5
transfecting cells with chemical synthesized or in vitro transcribed siRNA [2, 3]. Alternatively,
6
vector-based short hairpin RNA (shRNA) can be introduced and expressed within host cells as dsRNA
7
with a loop, which will further be processed into functional siRNA. Compared with the in vitro generated
8
siRNA, the construction of vector-based shRNA is much more economical.
9
The most popular technique used to generate shRNA expression vectors is oligonucleotide annealing,
10
which routinely requires the synthesis and annealing of two complementary oligonucleotides with each
11
oligonucleotide containing the whole shRNA hairpin sequence [4]. However, this method is often limited
12
by high mutation rate and low annealing efficiency [4, 5]. The synthesis of long oligonucleotides (more
13
than 40 nt) had been considered to be prone to introduce errors. But in our previous experiments, we have
14
synthesized several of long oligonucleotides (40~60 nt) and did not found any high rate of mutation [6-8],
15
to which the improved synthesis technology may contribute in recent years. Nevertheless, when we tried to
16
construct shRNA expression vectors by annealing oligonucleotides with hairpin sequences, we detected a
17
high number of mutations from positive clones (about 50%). Therefore we assumed that it is the hairpin
18
sequences rather than the oligo length that may impair the accuracy for the synthesis of long
19
oligonucleotides.
20
On the other hand, low annealing efficiency for shRNA construct fragments was considered to impair
21
the subsequent ligating reaction. We initially used pLL3.7 shRNA cloning vector (Addgene), which applies
22
HpaI/XhoI sites for shRNA construct cloning (Figure S1). We tried to construct four shRNA expression 3
1
vectors against human E2F5 gene by simply annealing oligonucleotides (Figure S2, Table S2), according
2
to the supplemented ‘pLL3.7 shRNA cloning’ protocol (https://www.addgene.org/11795/). Unfortunately,
3
after three rounds of cloning, we got only two correct E2F5-shRNA expression vectors as designed. The
4
inefficiency was firstly reasoned to the HpaI cloning site, which leaves blunt end that was thought to
5
impair the subsequent ligating reaction. Thus, we modified the pLL3.7 vector by introducing a type IIs
6
restriction enzyme BsmBI site, which was designed to cut just before the -1 position of the mU6 promoter
7
leaving cohesive ends. By applying the modified cloning strategy (Figure S3), although we succeeded to
8
construct three E2F5-shRNA vectors after two rounds of cloning, limited positive colonies were obtained
9
for screening correct clones. These experiment experiences indicated that the inefficiency, to some extent,
10
may be due to the low annealing efficiency for shRNA construct fragments, which may impair the
11
subsequent ligating reaction.
12
Besides, the cost for synthesis of oligonucleotides longer than 60 nt will be more than doubled, and
13
unfortunately if we choose shRNA target sites (Table S1) that are or more than 20 nt, the length for the
14
needed oligonucleotides will exceed 60 nt (Table S2).
15
To solve the problems as described above, we developed two alternative and efficient methods to
16
construct shRNA expression vectors, respectively based on multiple-step sequential PCR (msPCR) and
17
primer extension-homologous recombination (PE-HR). Both methods can be used to construct shRNA
18
expression vectors without needing to synthesize long oligonucleotides containing hairpin sequences.
19
For the msPCR method (Figure 1), the mU6 promoter sequence was firstly amplified by a pair of
20
primers mU6.F/Primer.P1. The Prmer.P1 contains the shRNA target antisense sequence and the loop
21
sequence.
22
cassette with primers mU6.F/Primer.P2. The Prmier.P2 consists of the loop sequence and the shRNA target
The first PCR product was used as template for amplification of the whole shRNA expression
4
1
sense sequence. For routine shRNA expression cassette, simply two sequential standard PCR steps will be
2
available with each primer limited under 60 nt in length (Table S2). The PCR amplified shRNA expression
3
cassette can be inserted into pLL3.7 between XbaI/XhoI sites replacing the former mU6 promoter sequence
4
by standard double digestion-ligation cloning method, which ensures to generate sufficient positive
5
colonies for screening correct clones in a single round of cloning. To demonstrate the feasibility of our
6
method, we initially constructed two E2F5-shRNA expression vectors. Double digestion screening yielded
7
80% (8 out of 10) of colonies harboring the desired shRNA expression plasmid clones. All of the positive
8
clones were further sequenced and demonstrated containing the correct shRNA expression cassettes.
9
Similarly, by employing an improved three-step sequential PCR, we successfully generated three complex
10
shRNA constructs flanked with Drosha site and miR30 sequence against mouse CD40 gene, which have
11
been demonstrated to function in both cell assay and in vivo targeted delivery study [6].
12
Compared with oligonucleotide annealing, the msPCR method promises the accuracy and efficiency by
13
avoiding to synthesize and to anneal oligonucleotides with hairpin sequences and guaranteeing to
14
generate sufficient positive colonies for further screening. According to the experience in our lab, the
15
msPCR with the long-tailed primers can be specifically and efficiently achieved by hot-start and
16
touch-down PCR. The length of the shRNA expression cassette is less than 600 bp, of which the accuracy
17
can be promised by standard PCR reactions using the mixture of Taq/Pfu (1:1) DNA polymerases
18
(TransGen Biotech, Beijing, China). It is noteworthy that PCR reactions with Taq DNA polymerase alone
19
usually cause point mutations in our experience. The msPCR method requires multiple PCR steps, routine
20
double digestion-ligation reactions. Although all of these procedures can be finished in one working day, it
21
seems still a little time-consuming.
22
There is an interesting method that firstly ligates the shRNA hairpin construct to a double-stranded DNA 5
1
end, then it uses nicking enzymes to open the hairpin construct to be a single-stranded inverted repeat
2
DNA, and finally with the help of the extension reaction, the single-stranded inverted repeat DNA will be
3
converted to be the whole double-stranded shRNA construct for shRNA expressing [9, 10]. Although this
4
method can be used to generate shRNA from cDNA , it requires a series of enzymatic reactions [9]. For the
5
construction of a given shRNA expression vector, it still requires the synthesis of long oligonucleotides
6
with hairpin sequences, two times of ligating reaction and one nicking reaction [10]. Another alternative
7
approach has been reported to generate the whole shRNA construct by primer extension, which could be
8
efficient to construct a miRNA library, but it also requires standard double digestion-ligation reactions to
9
generate corresponding vectors [11]. Besides, this method may require two times (after the extension and
10
digestion reactions respectively) of purification for the short shRNA construct fragment (~60 bp), which is
11
usually inefficient. Recent years, in vivo homologous recombination (HR) has been developed and widely
12
used for plasmid construction [12-15], and it performs well with the homologous sequence length to be
13
only 20 bp [15]. Therefore it is possible to rapidly and economically construct shRNA expression vectors
14
by combining the primer extension and homologous recombination strategies.
15
To validate the primer extension and homologous recombination (PE-HR) based method (Figure 2), five
16
shRNA target sites were chosen from human DYRK1A and mouse Igf2 genes (Table S1). The Klenow
17
fragment (TAKARA, Dalian, China) and Taq DNA polymerase (TransGen Biotech, Beijing, China) were
18
firstly compared for primer extension reactions. We found that Klenow fragment was better and 1 unit
19
Klenow for 10 min at 37 °C was sufficient for the extension reaction as shown in Figure S4. E.coli JM109
20
competent cells were co-transformed with the purified primer extension products and HpaI/XhoI-linearized
21
parental pLL3.7 plasmid DNA. Positive colonies were picked, and plasmids were prepared and checked by
22
double digestion for positive clones. 96% (24 out of 25) of these positive clones were sequenced to contain 6
1
the correct shRNA expression cassettes, and only one was detected with a point mutation. This result
2
indicated that our PE-HR method is efficient for the construction of shRNA expression vectors.
3
HR frequency in E.coli is crucial for the PE-HR method, and therefore we tested three different E.coli
4
stains (BJ5183, JM109 and DH5α) using different transformation methods. BJ5183 is used for doing HR
5
as employed in adenovirus vector construction [16]; DH5α and JM109 with high efficiency of
6
transformation are normally applied for plasmid construction and amplification. In theory, BJ5183 should
7
be a very appropriate strain for our method, and practically it did produce 3~5 times more colonies than
8
DH5α and JM109. But many clones (81 out of 96) were produced by non-homologous end joining (NHEJ)
9
and only a few (15 out of 96) were confirmed to be correct. On the other hand, we found more than 95%
10
clones of DH5α and JM109 were correctly generated. Since the RecA gene of DH5α and JM109 strains is
11
mutated, we speculated that the recombination occurred in DH5α and JM109 might be mediated by RecEF
12
[14]. In addition, high transformation efficiency is also important, which contributes to improve the
13
interaction possibility between homologous sequences. We tried electric and chemical transformations
14
(Inoue and calcium sodium methods), and observed that electro-transformation and Inoue method were
15
sufficient to produce more than 100 colonies per transformation, whereas only several colonies or none at
16
all for the calcium chloride method.
17
In conclusion, we showed two alternative and efficient methods to construct shRNA expression vectors
18
without needing to synthesize and to anneal long oligonucleotides with hairpin sequences, which were
19
considered to be responsible for the high mutation rate and low cloning efficiency. The msPCR method
20
employs routine PCR, double digestion-ligation reactions and guarantees to generate sufficient positive
21
colonies for further screening. The PE-HR method relies on homologous recombination and further
22
provides rapid and economical shRNA expression vector construction without needing the ligation 7
1
procedure. Compared with traditional oligonucleotide annealing as we tried, our alternative methods are
2
more efficient and cost effective for the construction of shRNA expression vectors.
3 4
Acknowledgments
5
This research was funded by National Natural Science Foundation of China (NSFC; No.30870119 and
6
31171186) and China's Ministry of Agriculture (948 Program; No.2013-Z27).
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
References [1] D.B. Stovall, M. Wan, Q. Zhang, P. Dubey, G. Sui, DNA vector-based RNA interference to study gene function in cancer, Journal of visualized experiments : JoVE, (2012) e4129. [2] S.M. Elbashir, J. Harborth, W. Lendeckel, A. Yalcin, K. Weber, T. Tuschl, Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells, Nature, 411 (2001) 494-498. [3] B. Luo, A.D. Heard, H.F. Lodish, Small interfering RNA production by enzymatic engineering of DNA (SPEED), Proceedings of the National Academy of Sciences of the United States of America, 101 (2004) 5494-5499. [4] M. Miyagishi, H. Sumimoto, H. Miyoshi, Y. Kawakami, K. Taira, Optimization of an siRNA-expression system with an improved hairpin and its significant suppressive effects in mammalian cells, The journal of gene medicine, 6 (2004) 715-723. [5] P.J. Paddison, J.M. Silva, D.S. Conklin, M. Schlabach, M. Li, S. Aruleba, V. Balija, A. O'Shaughnessy, L. Gnoj, K. Scobie, K. Chang, T. Westbrook, M. Cleary, R. Sachidanandam, W.R. McCombie, S.J. Elledge, G.J. Hannon, A resource for large-scale RNA-interference-based screens in mammals, Nature, 428 (2004) 427-431. [6] L. Zhang, T. Zhang, L. Wang, S. Shao, Z. Chen, Z. Zhang, In vivo targeted delivery of CD40 shRNA to mouse intestinal dendritic cells by oral administration of recombinant Sacchromyces cerevisiae, Gene therapy, 21 (2014) 709-714. [7] T. Zhang, H. Yang, R. Wang, K. Xu, Y. Xin, G. Ren, G. Zhou, C. Zhang, L. Wang, Z. Zhang, Oral administration of myostatin-specific whole recombinant yeast Saccharomyces cerevisiae vaccine increases body weight and muscle composition in mice, Vaccine, 29 (2011) 8412-8416. [8] K. Xu, C. Ren, Z. Liu, T. Zhang, T. Zhang, D. Li, L. Wang, Q. Yan, L. Guo, J. Shen, Z. Zhang, Efficient genome engineering in eukaryotes using Cas9 from Streptococcus thermophilus, Cellular and molecular life sciences : CMLS, (2014). [9] A. Dinh, Y.Y. Mo, Alternative approach to generate shRNA from cDNA, BioTechniques, 38 (2005) 629-632. [10] H. Tanaka, Construction of shRNA expression plasmids for silkworm cell lines using single-stranded DNA and Bst DNA polymerase, Methods in molecular biology, 942 (2013) 347-355. [11] D. Gou, H. Zhang, P.S. Baviskar, L. Liu, Primer extension-based method for the generation of a siRNA/miRNA expression vector, Physiological genomics, 31 (2007) 554-562. [12] J. Fu, X. Bian, S. Hu, H. Wang, F. Huang, P.M. Seibert, A. Plaza, L. Xia, R. Muller, A.F. Stewart, Y. Zhang, Full-length RecE enhances linear-linear homologous recombination and facilitates direct cloning for 8
1 2 3 4 5 6 7 8 9 10 11
bioprospecting, Nature biotechnology, 30 (2012) 440-446. [13] S.K. Sharan, L.C. Thomason, S.G. Kuznetsov, D.L. Court, Recombineering: a homologous recombination-based method of genetic engineering, Nature protocols, 4 (2009) 206-223. [14] Y. Zhang, J.P. Muyrers, G. Testa, A.F. Stewart, DNA cloning by homologous recombination in Escherichia coli, Nature biotechnology, 18 (2000) 1314-1317. [15] D. Zhu, X. Zhong, R. Tan, L. Chen, G. Huang, J. Li, X. Sun, L. Xu, J. Chen, Y. Ou, T. Zhang, D. Yuan, Z. Zhang, W. Shu, L. Ma, High-throughput cloning of human liver complete open reading frames using homologous recombination in Escherichia coli, Analytical biochemistry, 397 (2010) 162-167. [16] T.C. He, S. Zhou, L.T. da Costa, J. Yu, K.W. Kinzler, B. Vogelstein, A simplified system for generating recombinant adenoviruses, Proceedings of the National Academy of Sciences of the United States of America, 95 (1998) 2509-2514.
12 13
9
1
Figure Legends
2
Figure 1. pLL3.7 shRNA expression vector cloning by multiple-step sequential PCR (msPCR).
3
For amplification of the mU6-shRNA cassette, simply two sequential PCR steps will be available with
4
each primer limited under 60 nt in length. The PCR amplified shRNA expression cassette can be
5
inserted into pLL3.7 between XbaI/XhoI sites replacing the former mU6 promoter sequence by
6
standard double digestion-ligation cloning method. The -1 position (brown bold font) of the mU6
7
promoter should be reconstituted with nucleotide T and the +1 position (green bold font) with
8
nucleotide G is required for efficient RNA expression, according to the ‘pLL3.7 shRNA cloning’
9
protocol (https://www.addgene.org/11795/). The cutting sites or overhangs of restriction enzymes
10
HpaI and XhoI are indicated with red font. The orange and grey arrows represent respectively the
11
direct and inverted repeat sequences within the shRNA hairpin construct.
12 13
Figure 2. pLL3.7 shRNA expression vector cloning by primer extension and homologous
14
recombination (PE-HR). The shRNA construct flanked by homologous sequences (20 bp in length)
15
is generated by primer extension, with each primer limited under 60 nt in length. Corresponding
16
shRNA expression vector will be constructed by homologous recombination between the HpaI/XhoI
17
double-digested parental pLL3.7 backbone and the primer extension product. The -1 position (brown
18
bold font) of the mU6 promoter should be reconstituted with nucleotide T and the +1 position (green
19
bold font) with nucleotide G is required for efficient RNA expression, according to the ‘pLL3.7
20
shRNA cloning’ protocol (https://www.addgene.org/11795/). The cutting sites or residuals of
21
restriction enzymes HpaI and XhoI are indicated with red font. The orange and grey arrows represent
22
respectively the direct and inverted repeat sequences within the shRNA hairpin construct.
10
1
2 3 4
Fig. 1
5 6
11
1 2 3
Fig. 2
4 5
12
