In order to determine the rate of primer-independent cDNA synthesis at increased reverse transcription reaction temperatures, we replaced M-MuLV Reverse Transcriptase with RevertAid Premium Reverse Transcriptase, which has increased thermostability and is active up to 60°C. Samples and studied genes remained the same as in the previous experiment. We carried out reverse transcription without RT primers in parallel with separate reactions where gene-specific sequence-modifying primers were included (Figure 1). PCR priming sites were chosen so that the short 55–65 bp amplicons spanned an AT- or GC-rich segment of typically 5–7 consecutive A/T or C/G nucleotides. The sequence-modifying reverse transcription primers were designed to alter this sequence by substituting A or T for C or G nucleotides, or vice versa. This resulted in a 3°–5°C increase or decrease in the melting temperature of the PCR amplicons derived from specifically primed cDNA transcripts. Sequence-modifying reverse transcription primers were designed specifically avoiding nonspecific priming upstream to the intended priming site, since this would cause reverse transcription of un-modified target cDNA. OLIGO Primer Analysis Software, version 6.8 (Molecular Biology Insights, Cascade, CO, USA) was used for designing all the reverse transcription and PCR primers. The priming efficiency (PE value) of the PCR primers was kept significantly higher than that of the sequence-modifying reverse transcription primers in order to prevent any undigested reverse transcription primers from annealing during PCR amplification.
Reverse transcription was performed at 50°, 55°, and 60°C in eight replicates for each sample. The cDNA synthesis products were used as templates for PCR amplification. Unmodified amplification products derived from nonspecifically primed reverse transcription were distinguished from sequence-modified PCR products by postamplification melting curve analysis. Real-time PCR revealed that the rate of primer-independent reverse transcription was highly dependent on the reverse transcription temperature (Figure 2). The amplification of primer-independent cDNA transcripts was maximally attenuated when reverse transcription was performed at 60°C. At this temperature, however, the efficiency of specific primer-initiated reverse transcription also decreased for eight out of 15 tested genes (Supplementary Figure S1a). Several of these eight genes were relatively AT-rich at the reverse transcription primer binding site and we speculate that decreased priming efficiency was, rather than enzyme inactivation, the limiting factor for improving the specificity of the reverse transcription reaction. At 55°C, there was only minimal impairment of the reverse transcription efficiency of the gene-specific RT primers, whereas PCR amplification of nonspecifically primed cDNA templates was attenuated by 6 to 12 cycles (Supplementary Figure S1b).
By modifying the sequence of specifically primed cDNA transcripts during reverse transcription, the proportion of primer-independent cDNA synthesis can be quantitatively determined separately for each reaction. The sequence modification of the specifically primed cDNA transcripts results in an alteration of the melting properties of the subsequent PCR amplicons. This modification allows PCR amplicons derived from primer-initiated cDNA and nonspecific primer-independent cDNA transcripts to be separated by postamplification melting curve analysis, and their proportional amounts can be determined from the relative melting curve peak areas. To compare the proportion of primer-independent cDNA synthesis at increased reverse transcription temperatures with that occurring at standard reverse transcription conditions, we carried out multiplex reverse transcription at 42°, 50°, 55° and 60°C using sequence-modifying reverse transcription primers for the same 15 genes as in previous experiments.
At 42°C, the proportion of primer-independent cDNA synthesis was in the range of 11%–57% (Figure 3A), indicating that primer-independent cDNA synthesis composed a significant portion of the total cDNA products. Contamination with gDNA or previous amplification products was excluded as no amplification product was seen in any of the reverse transcriptase-negative controls. The percentage of primer-independent cDNA synthesis decreases significantly as the temperature of reverse transcription increases (Figure 3 B-D). At 55° and 60°C the proportion of nonspecifically primed cDNA accounted less than 10% of the total amount of cDNA template for 12 out of 15 tested genes. For two genes, MUC2 and PLAU, however, the impact of raising the reverse-transcription temperature was significantly smaller and the proportion of nonspecifically primed cDNA at 55°C was 48% and 27%, respectively. The expression level of these two genes was close to the average expression level of the 15 genes tested and reason for the higher proportion of nonspecific priming was not fully elucidated.