Novel method of cell-free in vitro synthesis of the human fibroblast growth factor 1 gene.

Recombinant DNA projects generally involve cell-based gene cloning. However, because template DNA is not always readily available, in vitro chemical synthesis of complete genes from DNA oligonucleotides is becoming the preferred method for cloning. This article describes a new, rapid procedure based on Taq polymerase for the precise assembly of DNA oligonucleotides to yield the complete human fibroblast growth factor 1 (FGF1) gene, which is 468 bp long and has a G+C content of 51.5%. The new method involved two steps: (1) the design of the DNA oligonucleotides to be assembled and (2) the assembly of multiple oligonucleotides by PCR to generate the whole FGF1 gene. The procedure lasted a total of only 2 days, compared with 2 weeks for the conventional procedure. This method of gene synthesis is expected to facilitate various kinds of complex genetic engineering projects that require rapid gene amplification, such as cell-free whole-DNA library construction, as well as the construction of new genes or genes that contain any mutation, restriction site, or DNA tag.

1. Introduction

Recombinant DNA projects generally involve cell-based gene cloning. However, because template DNA is not always readily available, in vitro chemical synthesis of complete genes from DNA oligonucleotides is becoming the preferred method for cloning. This approach is also simpler, more rapid, and more cost-effective. Recently, several methods of in vitro gene synthesis have been described, such as (1) ligation of overlapping single-stranded oligonucleotides comprising the entire gene followed by PCR amplification of the crude ligation mixture using two outer primers [1-3], (2) ligation of preformed duplexes of phosphorylated overlapping oligonucleotides [4, 5], (3) the Fok I method [6], (4) the LCR method [7], and (5) single-step gene assembly from multiple DNA oligonucleotides [8]. Although long oligomers of up to 300 to 600 bases can be directly synthesized, the yield is extremely low [9]. Hence, most of the oligomers in these methods are 25 to 60 nucleotides long [8, 10, 11].

The cloning of prokaryotic genes is relatively easier than that of eukaryotic genes because the former do not contain introns. In fact, eukaryotic genomic DNA cannot be cloned directly because of introns, and cloning procedures need to at least include (1) cell culture, (2) isolation of RNA, (3) reverse transcription from mRNA to cDNA, and (4) amplification of the cDNA. This conventional method has technological problems and is expensive and time-consuming, requiring about 2 weeks to perform [12]. For example, we previously used this approach to clone the open reading frame of the human fibroblast growth factor 1 (FGF1) gene, which is 468 bp long and has a G+C content of 51.5% [13]. The source of the FGF1 gene template was cultured HEK293 cells, from which total RNA was isolated and then reverse transcribed with the use of costly extraction kits. Furthermore, after the nonspecific transcription products were amplified, only one of the PCR products had the same molecular weight as FGF1, but it contained mutations. In other words, the conventional method is inefficient and not cost-effective. These shortcomings prompted us to consider developing a new PCR-directed gene synthesis method.

This new method is ligase-free, the oligonucleotides are designed to cover the complete sequence of both DNA strands, and the full-length molecule is generated in a single step by overlap extension PCR (OE-PCR). Our protocol is based on the method of Stemmer [8], which consists of synthesis of oligonucleotides, gene assembly, gene amplification, and cell-free cloning. Our method uses the same first and last stages. However, we have combined the middle two steps (gene assembly and amplification) into one step: equal volumes of each oligonucleotide solution are combined and used to generate the full-length template for gene amplification by PCR using the two outer primers.

2. Materials and Methods

2.1. Oligonucleotides

We designed and ordered 14 oligonucleotides, ranging from 50 to 70 nucleotides in length, which collectively encoded both strands of the FGF1 gene (Table 1 and Figure 1). The sense and antisense strands each consisted of 7 oligonucleotides. Complementary oligonucleotides were expected to overlap when the melting temperature (T) exceeds 60°C (Figure 1). We incorporated bolded sequences that overlapped with the parts of the sequence of the plasmid pcDNA3.1/V5-His-TOPO (Invitrogen, Carlsbad) to facilitate cloning into this plasmid [13]. Oligonucleotides and all reagents unless otherwise stated were purchased from Sigma-Aldrich (St Louis, MO, USA).

Table 1

List of oligonucleotides.

Oligonucleotide	Sequence (5′ to 3′)
pcFGF1F	gtccagtgtggtggaattgcccttaccatggctgaaggggaaatcaccacc
pcFGF1R12-71	TTG TAA TTC CCT GGA GGC AGA TTA AAC TTC TCG GTC AGG
	GCT GTG AAG GTG GTG ATT TCC
pcFGF1F42-101	gaagtttaatctgcctccagggaattacaagaagcccaaactcctctactgtagcaacgg
pcFGF1R76-135	TGT GCC ATC CGG AAG GAT CCT CAG GAA GTG GCC CCC
	GTT GCTACAGTAGAGGAGTTTGGG
pcFGF1F114-173	gaggatccttccggatggcacagtggatgggacaagggacaggagcgaccagcacattca
pcFGF1R151-210	ATA CAC CTC CCC CAC GCT TTC CGC ACT GAG CTG CAG CTG
	AAT GTGCTGGTCGCTCCTGTC
pcFGF1F188-247	cggaaagcgtgggggaggtgtatataaagagtaccgagactggccagtacttggccatgg
pcFGF1R225-284	ACA TTC CTC ATT TGG TGT CTG TGA GCC GTA TAA AAG CCC
	GTC GGTGTCCATGGCCAAGTACTGGCCAGTC
pcFGF1F268-327	ggctcacagacaccaaatgaggaatgtttgttcctggaaaggctggaggagaaccattac
pcFGF1R301-360	CTT CTC TGC ATG CTT CTT GGA TAT ATA GGT GTT GTA ATG
	GTT CTC CTCCAGCCTTTCCAG
pcFGF1F330-389	cacctatatatccaagaagcatgcagagaagaattggtttgttggcctcaagaagaatgg
pcFGF1R365-424	GGC CAT AGT GAG TCC GAG GAC CGC GTT TGC AGC TCC
	CAT TCT TCTTGAGGCCAACAAACC
pcFGF1F402-461	cggtcctcggactcactatggccagaaagcaatcttgtttctccccctgccagtctcttc
pcFGF1R	GCTGGATATCTGCAGAATTGCCCTT TTA ATC AGA AGA
	GACTGGCAGGGGG

Figure 1

Oligonucleotides used to synthesize FGF1. A total of 14 oligonucleotides, 50–70 nucleotides in length, collectively encoded both strands of the FGF1 gene.

2.2. Polymerase Chain Reaction

We synthesized the FGF1 gene using a single-step PCR method by setting up one reaction containing 14 oligonucleotides (4 pmol of each inner oligonucleotide and 20 pmol of each outer oligonucleotide), as shown in Figure 1. After the addition of 1 U AccuPrime Taq High Fidelity polymerase (Invitrogen, Carlsbad), the reaction was carried out in a PCR thermocycler (Takara Bio, Shiga, Japan) for 45 cycles at the following cycling profile settings: 95°C for 3 minutes, 95°C for 30 seconds, 49°C for 1 minute, and 68°C for 1 minute. The final cycle was followed by an additional 4 minutes at 68°C to ensure complete extension. The holding temperature was 4°C.

2.3. Assembly of Synthesized FGF1 Gene and pcDNA3.1/V5-His-TOPO Vector

Single-step DNA assembly and circularization was performed on nonpurified PCR amplification products by high-fidelity PCR, after the addition of Taq DNA polymerase [13]. Each 50-μL sample contained 5 μL 10 × AccuPrime PCR Buffer I, 0.5 μL of raw reaction product mix containing the linear vector backbone, 2 μL of raw reaction product mix including the FGF1 gene, and 0.2 μL AccuPrime Taq High Fidelity. Although the ratio of DNA insert to vector should change with the size of the fragments, maintaining the ratio at 4 : 1 for our insert range of 500 bp to 2000 bp gave satisfactory results. The PCR cycling profile (for 35 cycles) was as follows: 94°C for 2 minutes, 94°C for 30 seconds, 52°C for 30 seconds, and 68°C for 3 minutes; the holding temperature was 4°C. Because the primers included complementary directional overhangs, the correct head and tail sequences of the vector and gene were expected to anneal and assemble into the pcDNA3.1/V5-His-TOPO-FGF1 plasmid [13].

2.4. Transformation Procedure

One tube containing 50 μL of DH5αEscherichia coli cells was thawed on ice and 1 μL of raw reaction product mix from the combined gene assembly and plasmid insertion step was immediately added and mixed gently. The tube was then incubated on ice for 30 minutes, heated for 60 seconds in a 42°C water bath, and placed on ice for 2 minutes. The contents were added to 950 μL of prewarmed SOC medium (2% tryptone, 0.5% yeast extract, 0.05% NaCl, 2.5 mM KCl, and 20 mM glucose) in a 1.5-mL tube and shaken at 225 rpm at 37°C for 1 hour. Finally, 200 μL was spread onto a prewarmed LB agar plate containing 100 μg/mL ampicillin, and the plate was incubated overnight at 37°C.

2.5. Plasmid DNA Extraction

A colony of transformed E. coli cells was picked and grown in 5 mL LB medium at 37°C overnight. Plasmid DNA was extracted with the Plasmid Miniprep Kit (Sigma-Aldrich, St. Louis, MO, USA).

2.6. Restriction Enzyme Digestion

Double digestion was performed on a sample of extracted plasmid DNA with HindIII and XhoI (20,000 U/mL and 10,000 U/mL, resp.; New England Biolabs, Ipswich, MA, USA). The two restriction sites were located just beyond either side of the gene. Digestion products were subjected to agarose gel electrophoresis to determine the size of the inserted fragment.

2.7. Sequencing

The final plasmid construct was sequenced by the Genome Research Centre, The University of Hong Kong, with the ABI BigDye sequencing method, according to the manufacturer's instructions (Applied Biosystems, Foster City, CA, USA).

3. Results

3.1. Synthesis and Assembly of the FGF1 Gene Using PCR

To clone the FGF1 gene open reading frame (468 bp) into the pcDNA3.1/V5-His-TOPO vector, we added 27 bp of extra oligonucleotides to the 5′ end of gene and 25 bp extra oligonucleotides to the 3′ end, both with sequences identical to appropriate sites of the vector. The full-length of the FGF1 fragment obtained from double digestion of the PCR reaction was 520 bp (Figure 2). The total oligonucleotides used in the method of Stemmer [8] covered both strands of the DNA to be synthesized, without any intervening gaps between individual oligonucleotides. To minimize the cost of synthesizing oligonucleotides, we left gaps between each of the 50- to 70-mer oligonucleotides. This strategy allowed us to use only 14 oligonucleotides, compared with 18 used by Stemmer. Greater gaps and shorter overlaps resulted in a 22.22% (4/18) reduction in the cost of synthesizing the oligonucleotides.

Figure 2

PCR assembly of FGF1. Lane 1: DNA molecular weight markers. Lane 2: Assembled and amplified FGF1 gene (520 bp).

3.2. Cloning and Sequencing of the FGF1 Gene

After the crude PCR reaction mixtures that contained the product were directly cloned into the cloning vector, we readily obtained clones that contained the correct sequences. Double digestion of extracted plasmid DNA with HindIII and XhoI yielded two fragments on a 1% agarose gel, of sizes 5433 bp and 561 bp, which represented the linear vector pcDNA3.1/V5-His-TOPO and the FGF1 gene, respectively (Figure 3). Because the fragment yielded by HindIII and XhoI digestion that contained the gene also contained some base pairs from the vector, it was 31 bp longer than the corresponding earlier PCR fragment. The DNA sequence of the FGF1 gene in the complete pcDNA3.1/V5-His-TOPO-FGF1 construct was 100% in agreement with the sequences from the NCBI gene bank (data not shown).

Figure 3

Confirmation of assembled plasmid containing the FGF1 gene. The pcDNA3.1/V5-His-TOPO-FGF1 plasmid was cleaved by HindIII and XhoI. Lane 1: DNA molecular weight markers. Lane 2: Two digestion fragments of 5433 bp and 561 bp.

4. Discussion

Our goal of improving cloning technology was to reduce the cost and increase the efficiency of gene synthesis, compared with the conventional cloning method and the method of Stemmer [8]. On the basis of the data presented in this paper, we believe that our protocol of cell-free in vitro gene synthesis offers a simple, rapid, high-fidelity, and low-cost alternative. The essential feature of our approach is that the entire gene is synthesized in a single step by PCR. It does not require the production of a template before the PCR amplification step. Instead, we combined the two steps of template formation and full-length product amplification into one step. The single-step product did not require further purification and could be directly used for cloning by our previously published method [13].

Our new method would be very useful in the field of molecular biology, especially for the cloning of eukaryotic DNA, because the genes of many eukaryotes contain introns (several per gene) [14]; for example, FGF1 contains at least 5 introns [15]. We had previously tried to clone the FGF1 gene using cultured human HEK293 cells and human stem cells and compared the 2 methods [16]. Both cell types were cultured for at least 2 weeks. Total RNA [17] was isolated from cells with an RNeasy Mini Kit (Qiagen Sciences, Germantown, MD, USA). Reverse transcription [18] of the mRNA was then performed with oligo-(dT) primers and Moloney murine leukemia virus reverse transcriptase (Applied Biosystems, Foster City, CA, USA) to generate the cDNA pool. These procedures required a total of at least 2 weeks and were costly. Our new 2-day method would greatly reduce the cost of materials, personnel, and overheads, and increase the speed by which genes could be cloned.

The new method would also be useful to readily and quickly clone genes from viruses [19], bacteria [20] or prions [21], in a way that would reduce the risk of infection. In addition, gene expression and cloning in E. coli is a standard practice in biomedical research and production. Because each amino acid relates to more than one codon [22], gene sequences need to be optimized [23]. Our new method will help in codon optimization, because a series of degenerate oligonucleotide sequences can be used, and the amount of expression of foreign protein in E. coli can be maximised. Another potential application of our method is gene shuffling in protein research. Generally, one would have to choose a group of homologous genes, divide them into small pieces, recombine them, and select hybrids that yield improved protein products or that optimize production. However, it is very difficult to isolate DNA fragments that have fewer than 50 bp in conventional methods because most of the fragments are invisible on an agarose gel. Our cloning method would help in gene shuffling procedures because gene fragments of the desired size can be prepared and included reliably in the gene constructs.

In summary, we have developed a protocol that is a rapid and easy method for gene cloning. This method of gene synthesis is expected to facilitate various kinds of complex genetic engineering projects that require rapid gene amplification, such as cell-free whole-DNA library construction, as well as the construction of new genes or genes that contain any mutation, restriction site, or DNA tag.