Closing Bacterial Genomic Sequence Gaps with Adaptor-PCR

Abstract

Shotgun sequencing has become the standard strategy for sequencing genomes. When the shotgun phase of a sequencing project has been completed, the assembly typically consists of a number of contigs, which are separated by gaps. The purpose of finishing is to close all gaps and to generate a continuous sequence for a natural DNA molecule such as a chromosome or a plasmid. Most of the captured gaps can be closed with primer walks on subclones. The uncaptured gaps can be closed by attempting PCR with every possible pair-wise combination of primers picked close to uncaptured gaps. This can be done in a relatively fast and efficient way when the number of uncaptured gaps is <5. However, the number of PCRs required grows rapidly when the number of uncaptured gaps increases; for example, 180 PCRs are needed when there are 10 uncaptured gaps, and 740 PCRs are required in the case of 20 uncaptured gaps. It is costly, time-consuming, and usually labor-intensive to set up so many PCRs in smallto medium-sized finishing centers. At this stage, other PCR-based technology, such as multiplex PCR (1), inverse PCR (2,3), restriction site PCR (4), and capture PCR (5), can be used to close uncaptured gaps as well. However, these methods have their own limits. For example, the success rate of multiplex PCR is low without substantial fine-tuning of primer combinations. A limit of inverse PCR is that a pair of primers must be picked within a restriction fragment that extends into the gap. If there is minimal extension, other enzymes will have to be tried, likely requiring the design of new primers. To ease the process of closing gaps, we describe here a modification of the anchored PCR method (6–8) that involves the use of single strand adaptor PCR. The scheme of the procedure is shown in Figure 1. This adaptor PCR includes three main steps: (i) digestion of genomic DNA; (ii) ligation of adaptor with DNA fragments; and (iii) PCR amplification using an adaptor primer (AP) and a specific primer (SP) for the target. The DNA has to be digested with a restriction enzyme that leaves 5′ end overhangs that are four bases or longer. Preferably, the digested DNA fragments will be mainly 5–10 kb in length. An enzyme that works with one genome might not be suitable for another. Since two restriction digests are needed for the procedure, several restriction enzymes may need to be tried to find two suitable enzymes. The adaptors (Table 1) consist of a chimeric sequence that contains, at its 5′ end, a sequence complementary to the overhang of the enzyme used and a unique sequence at its 3′ end, which is 5′-CCCTATAGTGAGTCGTATTAAC3′. The adaptor is phosphorylated at its 5′ end. The AP contains a sequence complementary to the unique region in the adaptor. Taq DNA polymerase should be the main enzyme for the PCR, because it lacks strand replacement properties. We have successfully employed this method in two of our finishing projects, Exiguobac ter ium and Shewanella baltica OS1155, which have genome sizes of 3 and 5 Mb, respectively. Both contain numerous gaps with unaligned scaffolds and were deemed good candidates for using this technology in gap closure. Genomic DNA used in this study was supplied by J.M. Tiedje (Michigan State University) and J. Fredrickson (Pacific Northwest National Laboratory). Specific primers were picked with Autofinish, a tool that is part of the consed package (9). Restriction enzymes tested included XhoI, PaeR7I, NgoMIV, NheI, XbaI, BglII, SalI, BclI, EcoRI, BamHI, and HindIII (New England Biolabs, Beverly, MA, USA). Digestion was performed on 100–150 ng total genomic DNA. Five units of enzyme were combined with 1 μL of the appropriate buffer and water, bringing the final reaction volume to 10 μL. Samples were incubated at 37°C for 1.5 h in an MJ Research PTC-225 Thermal Cycler (Bio-Rad Laboratories, Hercules, CA, USA). One microliter of each sample was analyzed on an agarose gel to check for completeness of digestion. DNA should appear as a smear with the majority of fragments in the range of 3–10 kb. The ligation reactions were performed using 9 μL uncleaned digest, 1 μL T4 ligase (New England Biolabs), 2 μL 10× ligase buffer with 10 mM ATP (New England Biolabs), 1 μL phosphorylated adaptor (final concentration: 2.5 μM), and 7 μL water for a total volume of 20 μL. Reactions were processed in a PTC-225 Thermal Cycler under the