Gibson assembly interposition improves amplification efficiency of long DNA and multifragment overlap extension PCR

Overlap extension PCR (OE-PCR) is a common molecular biology technique mainly used for DNA fragment fusion and gene site mutation [1,2]. Two DNA fragments with overlapping sequences are denatured, complementary annealed, extended to form fusion DNA and amplified exponentially [3]. Also, three or more DNA fragments synchronized for OE-PCR are used for multifragment fusion and gene multisite-specific mutagenesis [4,5]. Short DNA fragment OE-PCR is less difficult. However, the OE-PCR fusion of long DNA fragments is complex. When one fragment exceeds 2 kb, the target fusion DNA is not easily obtained in the second PCR. It is generally believed that long DNA interferes with the complementary end pairing of two DNA fragments [6]. Furthermore, there is little hope of the successful fusion of multiple DNA fragments in one round of OE-PCR. Various optimization strategies are available when encountering such problems. These include adjusting the overlapping sequence length, exploring the annealing temperature, adjusting the buffer composition of the reaction system and so on [7–9]. The OE-PCR program can also be run in two steps. In the first stage, the primer-free thermal cycle is carried out to overlap and completely extend DNA fragments. Then, the primed operation is carried out to exponentially amplify the fused DNA [10]. Additionally, various PCR additives can improve the performance and yield of complex PCR, including some nanomaterials and small molecules extracted from the nucleus, with limited effect [11]. Due to the lack of specific rules, repeated adjustment of reaction conditions cannot amplify the fusion DNA fragment. Therefore, researchers require a stable and effective OE-PCR solution.

Gibson and his colleagues first proposed Gibson assembly in 2009. DNA fragments with overlapping terminal sequences can be quickly spliced in a single tube at the same temperature using the cooperative effect of Taq DNA ligase, T5 exonuclease and Phusion DNA polymerase [12,13]. Comparing the OE-PCR fusion and Gibson assembly process, both methods have advantages. For some difficult OE-PCRs, such as long gene and multi-DNA fragment fusion, combining Gibson assembly with traditional OE-PCR could be an effective solution. This indicates joining the DNA fragments from the first PCR by Gibson assembly and amplifying it exponentially by the second PCR.

RB is a typical antioncogene with a length of 2787 bp encoding a 928 amino acid protein [14]. Rb protein contains multiple potential phosphorylation sites [15,16]. Thus, site-directed mutagenesis is a valuable method to search for their function. The single-point and double-site mutations of the RB gene were cloned to test the feasibility of this idea. This study combines Gibson assembly with OE-PCR to provide a feasible and effective gene fusion solution.

Materials & methods

Plasmids, reagents & primers

The vector pEGFP_C1 was obtained from Clontech (CA, USA). pEGFP_RB plasmid was constructed by the current authors. Pyrobest DNA polymerase was purchased from TaKaRa (Dalian, China). Restriction endonucleases Hind III, XhoI and NEBuilder Hifi DNA assembly MasterMix were purchased from NEB (Beijing, China). A Gel Extraction Kit was purchased from Bioer (Hangzhou, China). All primers were designed by Primer Premier 5 and commercially synthesized by AuGCT (Beijing, China). The primer sequences are listed in Table 1 and their stocking concentrations were 50 μmol/l.

DNA amplification

The DNA amplification method consists of two PCR reactions interposed by a Gibson assembly reaction. In the first PCR round, 1 ng pEGFP_C1-RB plasmid was added to 2 μl 10× buffer, 2 μl dNTPs mix (2.5 mmol/l each), 0.2 μl forward and 0.2 μl reverse primers, 0.2 μl (1 U) Pyrobest DNA polymerase and final replenishment with ddH₂O up to 20 μl. The initial denaturation was at 94°C for 5 min, followed by denaturation at 94°C for 30 s, annealing at 54°C for 30 s and extension at 72°C for 1∼3 min (30 cycles). The PCR products were run by 1% agarose gel electrophoresis at 150 V for 10 min, and the target band was cut and extracted to 20–50 μl ddH₂O to achieve a final concentration of 10 ng/μl. Double-fragment assembly was conducted by taking extractive upstream DNA and downstream DNA (2 μl), respectively, mixed with 4 μl 2× assembly MasterMix. Three-fragment assembly was conducted by taking 2 μl extractive DNA of F1, F2 and F, respectively and mixing with 6 μl 2× assembly MasterMix. Next, 2× assembly MasterMix was replaced with an equal amount of water as a negative control. The reactant was incubated at 50°C for 1 h and then denatured at 94°C for 5 min. The second round PCR was conducted by taking 0.5 μl assembly DNA or its negative control and 1 ng pEGFP_C1-RB plasmid as a template; the remaining components and steps were the same as the first PCR.

DNA identification & sequence analysis

PCR products were run by 1% agarose gel electrophoresis. The expected bands were cut and extracted with a Gel Extraction Kit to 20–50 μl ddH₂O to achieve a final concentration of 10 ng/μl. DNA detected by enzyme digestion was as follows: 7.5 μl extractive DNA, 2 μl 10× buffer, 0.5 μl (10 U) Hind III or XhoI with water added up to 20 μl, followed by incubation at 37°C for 2 h and analysis by agarose gel electrophoresis. DNA sequencing was entrusted to AuGCT (Beijing, China) and alignmented by DNAMAN 4.0.

Results & discussion

Inserting Gibson assembly improves amplification efficiency of long DNA OE-PCR

Cloning RB^S780E mutant by traditional OE-PCR was challenging due to its long sequence. This problem was overcome with a complicated procedure called double-fragment ligation [6]. To test whether the insertion of the Gibson assembly can improve the efficiency of OE-PCR amplification, cloning of the same mutant was performed. Three different gene fragments centered on RB_780S were prepared for comparative analysis to explore the fusion effect of this scheme on DNA fragments of different lengths (Figure 1A).

The first round of PCR was performed based on the conventional OE-PCR protocol. Three upstream target DNA with different lengths and one downstream target DNA fragment around the mutation site were successfully amplified (Figure 1B). Three different upstream gene fragments, Fa, Fb and Fc, were directly mixed with the downstream fragment Fd as the template for the second PCR round. Only the short Fc + Fd group could amplify the expected DNA fragment successfully. The amplification efficiency of the Fb + Fd group was low. However, no target gene was detected within the longer Fa + Fd group. The mixture of the three groups was assembled by the Gibson method and used as a template for the second PCR round. Thus, the expected DNA was successfully amplified (Figure 1C). The PCR product sequence alignment on the Fad assembly group revealed that it was an RB^S780E mutant (Figure 1D). These results showed that the upstream and downstream gene fragments assembled by the Gibson method can enhance the amplification efficiency of the second PCR round, which is evident in the long DNA fragment OE-PCR. It should be noted that the expected DNA product was difficult to obtain in the second PCR when the same terminal primers from the first round were used (results not shown) unless the primers at both ends were shifted inward.

Inserting Gibson assembly improves efficiency of multifragment OE-PCR

In many cases, multifragment OE-PCR is time efficient, but it is more challenging than long DNA OE-PCR. Based on the sequence characteristics of the RB gene, a unique two-site mutant was designed. The AAGCTG near 1295 bp was mutated into AAGCTT, creating a HindIII site. Meanwhile, the CTCCAG was mutated to CTCGAG near 1786 bp, forming an XhoI site (RB_HX). Two restriction sites were introduced for rapid analysis of the PCR products (Figure 2A). Three fragments, designated as F1, F2 and F3, were successfully amplified in the first PCR round (Figure 2B). After a Gibson assembly program, three more prominent DNA bands were detected using electrophoresis. Based on their migration rate, the biggest one (indicated by the red arrow) was presumed to be the full-length RB gene. In contrast, the other two were the F12 and F23 fusion products formed by the partial assembly (Figure 2C).

In the second PCR round, the simple mixture template of three DNA fragments could not amplify the full-length RB gene. However, it could amplify the target gene when the Gibson assembly mixture was directly used as the template without separation and purification (Figure 2D). After the corresponding bands were extracted and analyzed using Hind III or XhoI digestion, the wild-type RB from the positive control remained intact. In contrast, the PCR product from the assembly group was cut into two fragments of the expected size (Figure 2E), indicating the product to be the expected RB_HX double-site mutant. As mentioned, the matching sites of primers at both ends used in the second PCR round were inward shifted compared with the first ones.

Primer matching sites of second PCR round should be appropriately shifted inward

In the early attempt to explore our hypothesis, amplifying the target gene often was difficult when the same forward/reverse primers were used in two rounds of PCR. Based on Gibson's assembly principle, the T5 exonuclease in the assembly reagent partially digested the two ends of the assembly linear DNA. This led to invalid PCR templates, which were tested by amplifying RB_HX double-site mutant with three DNA fragments. During the second PCR round, we selected a series of forward primers whose matching position gradually shifted inward about 20 bp, and the reverse primer matched to 63 bp before the first one (Figure 3A). The electrophoresis of the PCR product showed that when the same forward primer as the first one (i.e., S1) was used, a weak target DNA band appeared. The amount of target product increased gradually with the inward shifting of the upstream primer. After shifting inward about 60 bp to the forward primer S4, there was no significant increase in the amount of the expected product (Figure 3B). There was almost no difference in the amount of target product amplified by the forward primers when pEGFP_C1-RB plasmid was set as the positive control (Figure 3C). These results indicate that the sequence difference between the six forward primers had little influence on the PCR amplification efficiency in this case. The gradient change of the target products should be due to the partial digestion of both ends of the template during the assembly reaction. Thus, it leads to low amplification efficiency of the primers with matching sites close to the template DNA end.

While exploring the entangled relationship of primers between the two PCR rounds, the DNA fragment concentration, the Gibson assembly reaction time and other factors slightly affected the subsequent results. It was believed that the primers' shift inward about 30–50 bp could guarantee the success of the second PCR amplification round while conducting a comprehensive analysis.

OE-PCR is often used for gene site-directed mutagenesis and DNA fusion. For single-point mutation, the most popular approach is rolling circle amplification, but this method has inherent defects. First, PCR amplification of 7–12 kb plasmids depends on a high fidelity and fast DNA polymerase, otherwise, the target product cannot be easily obtained or random mutations may be introduced unexpectedly. Second, if the DpnI enzyme fails to completely digest the template plasmid, it will seriously interfere with the target mutant. Third, rolling cycle amplification also poses a risk of mutation in the recombination plasmid other than the target sequence, as this part of the plasmid was amplified by PCR, but is generally not confirmed by sequencing [6]. Therefore, the Gibson assembly interposition protocol has value in terms of single-point mutation but has obvious advantages in multisite mutagenesis, which is hardly achieved at one time by rolling circle amplification or the mega-primer method.

When it refers to DNA fragment fusion, the Gibson assembly procedure is easier than OE-PCR, so it is a priority candidate if conditions permit [17,18]. However, the Gibson assembly method has many disadvantages, including strict sequence requirements, high cost and low efficiency. However, there are reports of successfully splicing multiple DNA fragments [19,20] with very few clones formed after joining four or more DNA fragments, which makes it challenging to screen positive clones. In contrast, the scheme proposed in this study could be used when direct Gibson assembly is inconvenient, such as TA cloning after splicing multiple exons; when it is challenging to prepare overlapping sequences in DNA fragments and linear vectors, making it impossible for Gibson assembly; and when the tremendously low Gibson assembly efficiency cannot screen out positive clones.

The cross-complementary and extension between two DNA fragments in conventional OE-PCR after high-temperature denaturation and annealing are critical factors in determining the efficiency of fusion gene amplification. As the gene length or the number of DNA fragments increases, the interference factors of complementary extension of overlapping sequences also increase. This leads to the gradual reduction of OE-PCR efficiency. Gibson assembly reacts at 50°C, and the dsDNA was not denatured due to the moderate temperature. Therefore, the length of the DNA fragment could not affect the assembly efficiency. The original fragment, partial and complete assembly DNA can simultaneously exist in the assembly mixture. Among them, the entire assembly of DNA can be a good template for the second PCR round (Figure 4), while the others did not affect DNA amplification. Therefore, the assembled product can be a template on the following PCR without electrophoresis or DNA purification. T5 exonuclease can affect the stability of DNA fragments, which will be denatured in the subsequent PCR high-temperature stage. Thus, interposing a Gibson assembly process to the OE-PCR will not make it cumbersome.

Gibson assembly reagent contains a T5 exonuclease component that can degrade linear DNA from 5′-end to 3′-end [21]. This leads to partial digestion of one DNA strand formed in the Gibson assembly. When the same primers are used in the previous and subsequent PCR, although they have a binding template, they cannot synthesize the matching sequence of another primer when it extends to the end, failing the exponential amplification of template DNA (Figure 4). Therefore, it is necessary for the second PCR round that the primers shift a little inward. There is no standard instruction for the inward shift distance of primers in the second PCR due to the variable digestibility of DNA molecular ends during the Gibson assembly process. It is affected by many factors, such as DNA fragment concentration, T5 exonuclease activity, incubation time and so on. The interposed Gibson assembly time is a “two-edged sword.” Extending the assembly time can increase the concentration of the expected fusion DNA template. However, the sequences at both ends of the assembly DNA will be digested. When the DNA fragment concentration before assembly is high, the assembly time can be appropriately prolonged. In practice, the primer matching sites from the first PCR should be properly shifted outward from both target sequence ends if the DNA fragments were amplified from the genome for OE-PCR fusion. In another case, the universal primers matching the vector sequence can be selected for the first PCR round while cloning site-directed mutagenesis based on the recombinant plasmid.

Conclusion

This protocol, insert a Gibson assembly step between the two rounds of DNA amplification of OE-PCR, integrate two DNA fusion methods exquisitely, significantly improving gene amplification efficiency, particularly in the fusion of long genes and multi-fragments.

Future perspective

Various innovative gene site-directed mutagenesis approaches, such as rolling cycle amplification, mega-primer PCR and Gibson assembly [17,22,23], have been developed in recent years, but each has its sets of drawbacks [18,24,25]. Despite this, the OE-PCR procedure remains a popular technique. In addition, OE-PCR can also be used for DNA fragment fusion. Many times, the amplification efficiency of OE-PCR is influenced by so many factors, including the total number of fragments, GC%, length and even the type of polymerase. Repeatedly adjusting reaction conditions is time-consuming and inefficient. New efficient approaches are still required to overcome bottlenecks regarding difficult OE-PCR, such as in the fusion of long genes or multiple DNA fragments. Gibson assembly is mainly used for ultralong DNA splicing, and this method is simple to operate but with strict sequence requirements, high cost and low efficiency [12,13]. Combining the two methods could compensate for their respective drawbacks. The protocol proposed in this article has the potential to be helpful for scientists facing difficult OE-PCR or when direct Gibson assembly is inconvenient, such as TA cloning after splicing multiple exons; when it is challenging to prepare overlapping sequences in DNA fragments and linear vector, making it impossible for Gibson assembly; and when the tremendously low Gibson assembly efficiency cannot screen out positive clones.