Genome sequencing accuracy by RCA-seq versus long PCR template cloning and sequencing in identification of human papillomavirus type 58

Abstract

Background: Genome variations in human papillomaviruses (HPVs) are common and have been widely investigated in the past two decades. HPV genotyping depends on the finding of the viral genome variations in the L1 ORF. Other parts of the viral genome variations have also been implicated as a possible genetic factor in viral pathogenesis and/or oncogenicity.

Results: In this study, the HPV58 genome in cervical lesions was completely sequenced both by rolling-circle amplification of total cell DNA and deep sequencing (RCA-seq) and by long PCR template cloning and sequencing. By comparison of three HPV58 genome sequences decoded from three clinical samples to reference HPV-58, we demonstrated that RCA-seq is much more accurate than long-PCR template cloning and sequencing in decoding HPV58 genome. Three HPV58 genomes decoded by RCA-seq displayed a total of 52 nucleotide substitutions from reference HPV58, which could be verified by long PCR template cloning and sequencing. However, the long PCR template cloning and sequencing led to additional nucleotide substitutions, insertions, and deletions from an authentic HPV58 genome in a clinical sample, which vary from one cloned sequence to another. Because the inherited error-prone nature of Tgo DNA polymerase used in preparation of the long PCR templates of HPV58 genome from the clinical samples, the measurable error rate in incorporation of nucleotide into an elongating DNA template was about 0.149% ±0.038% in our studies.

Conclusions: Since PCR template cloning and sequencing is widely used in identification of single nucleotide polymorphism (SNP), our data indicate that a serious caution should be taken in finding of true SNPs in various genetic studies.

Figure 1

Enrichment of HPV genomic DNA by RCA from cervical samples. (A) HPV58 genome copy numbers before and after RCA enrichment. Real-time PCR (qPCR) was performed with an HPV58-specific primer pair on ~100 pg of sample DNA (sample 10 and sample 13) either before or after RCA enrichment. A 10-fold serial dilution, starting from 100 pg (~1.3 x 10⁷ copies) of the plasmid pXW59-1 which contains an HPV58 DNA fragment from nt 6906 to 3695 was amplified using the same primer set by qPCR to create a standard curve. The threshold cycle (Ct) values of qPCR data from 2 repeats were calculated for copy number analysis. GAPDH was used as an internal control. (B) HPV58 DNA in the sample 10 was under detection level before RCA, but became detectable by agarose gel electrophoresis after enrichment by RCA to 20481 copies as quantified by qPCR.

Figure 2

Nucleotide substitutions identified in HPV58 isolates to the reference HPV58 [32] by RCA-seq and by cloning sequencing. RCA-seq results were compared to the sequencing results of two individual bacterial clones of each plasmid for samples 9 (A), 10 (B), and 13 (C). S1 and S2 denotes Sanger sequence #1 (clone #1) and #2 (clone #2), respectively. Common nucleotide substitutions at positions in the reference HPV58 genome seen from RCA-seq to cloning sequencing were colored in red. The nucleotide substitutions identified only by cloning sequencing were shown in black.

Figure 3

Restriction enzyme digestion distinguishes HPV58 amplicons from RCA products to their corresponding plasmids. (A) Diagrams of HPV58 amplicons from RCA samples and their corresponding plasmid clones with a new restriction enzyme cutting site. (B) Restriction enzyme digestion of RCA products and their corresponding plasmid clones from each clinical sample. RCA products and pXHW57-1 from the sample 9 were amplified using a primer pair of Pr743 and Pr1424 and digested with AfeI. RCA products and pXHW55-1 from the sample 10 were amplified using a primer pair of Pr4974 and Pr5181 and digested with EcoRV. RCA products and pXHW59-1 from the sample 13 were amplified using a primer pair of Pr2417 and Pr2960 and digested with FspI. The digested products were resolved in 1.5% agarose gel. D: digested; ND: not digested.