Template-dependent multiple displacement amplification for profiling human circulating RNA

Abstract

Multiple displacement amplification (MDA) is widely used in whole-genome/transcriptome amplification. However, template-independent amplification (TIA) in MDA is a commonly observed phenomenon, particularly when using high concentrations of random hexamer primers and extended incubation times. Here, we demonstrate that the use of random pentamer primers with 5´ ends blocked by a C18 spacer results in MDA solely in a template-dependent manner, a technique we have named tdMDA. Together with an optimized procedure for the removal of residual genomic DNA during RNA extraction, tdMDA was used to profile circulating RNA from 0.2 mL of patient sera. In comparison to regular MDA, tdMDA demonstrated a lack of quantifiable DNA amplification in the negative control, a remarkable reduction of unmapped reads from Illumina sequencing (7 ± 10.9% versus 58.6 ± 39%, P = 0.006), and increased mapping rates of the serum transcriptome (26.9 ± 7.9% versus 5.8 ± 8.2%, P = 3.8 × 10-4). Transcriptome profiles could be used to separate patients with chronic hepatitis C virus (HCV) infection from those with HCV-associated hepatocellular carcinoma (HCC). We conclude that tdMDA should facilitate RNA-based liquid biopsy, as well as other genome studies with biological specimens having ultralow amounts of genetic material.

Keywords: C18 spacer; liquid biopsy; multiple displacement amplification.

Figure 1. List of primers tested in multiple displacement amplification (MDA)

Group 1 primers have phosphorothioate bonds placed at different positions to test the minimum lengths of primers capable of initiating DNA polymerization. Group 2 primers are modified internally with a C3 spacer to minimize self-priming. Group 3 primers are the group 2 primers modified with the inclusion of locked nucleic acid (LNA) to stabilize the binding between primers and templates. Groups 4 and 5 are constrained primers and RNA primers (with or without a C3 spacer), respectively. Random primers blocked at the 5′ end are designated as group 6 (hexamer) or group 7 (pentamer) primers. In these two groups of primers, the blockage at the 5′ end was tested using different modifications, including inverted, 2′,3′dideoxy-dT base (5′ inverted ddT), dSpacer, C3 spacer, C9 spacer, C18 spacer, amino modifier C6 (AmMC6) or alkyne. Degenerate bases are indicated using standard International Union of Pure and Applied Chemistry (IUPAC) codes. Each primer was tested in triplicate at a concentration of 50 μM in the standard MDA protocol with (blue) or without (red) template [1 ng of the hepatitis C virus (HCV) plasmid]. The MDA reaction with the group 7 primers was incubated at 28°C, while the other groups of primers were amplified at 30°C. The error bars indicate SD.

Figure 2. Real-time reverse transcription–template dependent multiple displacement amplification (RT-tdMDA) using three hepatitis C virus (HCV) patient serum samples and a negative control (H₂O)

These samples covered the range of the RNA yield extracted from 200 μL serum (7.5–19.2 ng), as quantitated in the final 14 μL of elution by the QIAGEN miRNA kit prior to HL-DNase digestion. An aliquot of 10.6 μL RNA was used for RT in a reaction containing 200 U SuperScript III, 80 μM 5′-end-blocked random pentamer primer (5′-/iSpC3/NNN*N*N-3′; asterisks denote phosphorothioate bonds), and 2 mM dNTPs in a 20-μL volume. An aliquot of 4 μL of the RT reaction was used in a 40-μL tdMDA reaction containing 300 U phi29 DNA polymerase (Epicentre), 80 μM primer, and 0.1× SYBR Green I (Thermo Fisher Scientific). The reaction was incubated at 28°C for 24 h on the ABI TaqMan 7500, in which fluorescent intensities were monitored through the SYBR Green channel. Estimated amount of cDNA input in tdMDA = [10.6 / (14 + 0.5 (HL-DNase) + 1.4 (buffer)] × [total RNA amount] × 0.2. Template-independent amplification was completely inhibited, as indicated by the negative control.

Figure 3. Transcriptome sequencing of circulating RNA

(A) RNA extracted from seven patient serum samples was amplified using real-time reverse transcription–template dependent multiple displacement amplification (RT-tdMDA). Lanes 1–2: negative control (water); Lanes 3–9: seven hepatitis C virus (HCV) and hepatocellular carcinoma (HCC) patient serum samples; M: lambda DNA/HindIII marker. (B) Comparisons of the percentages of unmapped reads and TopHat mapping rates between tdMDA and regular MDA. (C) Differential gene expression (DGE) analysis in edgeR identified 457 genes showing ≥2 fold changes either up- (n = 424) or down-regulation (n = 23) in the HCC group. (D) Unsupervised clustering analysis among HCV and HCC patients based on the leading log-fold-change of detected gene transcripts.