Optimisation for subtype A amplification

Having selected the Life Technologies PCR system, the reaction conditions needed optimisation such that subtypes A, B, C and D were amplified with comparable efficiency. The multiple subtype optimisation was begun using a subtype A clinical specimen, using the same reaction conditions used during the Roche/Life Technologies comparison. Although these conditions amplified subtype A, the highest amplification achieved was 31% using the 1:2 cDNA dilution (Figure 4_7a). As this dilution equated to a total input of 100 template molecules per reaction, the reaction efficiency was clearly suboptimal. The reaction conditions were therefore optimised to increase the efficiency of subtype A amplification.

Firstly, the primer concentration was increased to 0.30µM per reaction, and 8 replicates of a clinical specimen extracted at 20,000 and 40,000 copies were run (Figure 4_7b). 7/8 (75%) of the 20,000 copies reactions showed amplification at the correct size, although 1 of these also showed evidence of mispriming. The 40,000 copies reactions showed 5/8 (63%) positive reactions, of which 2/5 (40%) also showed mispriming. 3/8 (38%) of reactions were negative. These conditions clearly improved the subtype A reaction efficiency, however could potentially negatively impact the reactions for other subtypes, so at this stage reactions using subtype A, B, C and D specimens were run at primer concentrations of 0.30 and 0.35µM per reaction.

Figure 4_7c shows the results from the primer titration reactions. Overall, 16 replicates of each of subtype A, B, C and D were performed using a primer concentration of 0.35µM per reaction, and eight replicates of each subtype were performed using a primer concentration of 0.30µM per reaction. At 0.35µM, there were 4/16 (25%) positive reactions for subtype A, 6/16 (38%) positive reactions for subtype B, 12/16 (75%) positive reactions for subtype C and 9/16 (56%) positive reactions for subtype D. The reactions for subtypes A and D showed widespread mispriming (11/16, 69% and 7/16, 44%, respectively). The best results were for subtype C. The reactions using a primer concentration of 0.30µM showed 3/8 (38%) positive reactions for subtypes A and B, 7/8 (88%) positive reactions for subtype C and 5/8 (63%) positive reactions for subtype D. Mispriming was still evident in the subtype A and D reactions (5/8, 62% and 2/8, 25%, respectively), however, this was reduced.

115

Figure 4_7. Optimisation of PCR reactions to work successfully with subtypes A, B, C and D. a) Initial amplification of a subtype A clinical specimen. 16 reactions each using undiluted cDNA and cDNA dilutions of 1:2, 1:4 and 1:10 were performed. Lane 1 of each row contained Hyperladder I. Lanes 2-17 of the top row contained neat cDNA (3/16 positive, 19%); lanes 18-34 contained the1:2 cDNA dilution (5/16, 31% positive). In the middle row, lanes 2-17 contained the 1:4 cDNA dilution (2/16 positive,13%) and lanes 18-34 contained the 1:10 cDNA dilution (1/16, 6%). The highest number of positive reactions (5/16, 31%) was gained from using a 1:2 cDNA dilution. Lane 35 contained the negative control. This low rate of successful reactions indicated suboptimal reaction conditions for the amplification of subtype A specimens. b) Optimisation of subtype A reaction conditions. 8 replicates of a subtype A clinical specimen extracted at both 20,000 and 40,000 copies were amplified using an increased primer concentration of 0.35µM per reaction. Lane 1 in each row contained Hyperladder I. Lanes 2-9 in the top row contained the amplified products from the 20,000 copies extraction. Lanes 2-9 in the bottom row contained the products from the 40,000 copies extraction, and lane 10 contained the negative control. 7/8 of the 20,000 copies reactions showed amplification at the correct size; 5/8 of the 40,000 copies extraction showed amplification at the correct size. c) Optimisation of primer concentrations across subtypes A, B, C and D. 16 reactions using a primer concentration of 0.35µM each primer per reaction and 8 reactions using a primer concentration of 0.3µM each primer per reaction were performed using extracts from clinical specimens of subtypes A, B, C and D. Lane 1in each row contained Hyperladder I; The top and second rows contained the subtype A and B reactions using 0.35µM (top row) and 0.30µM (second row). The third and bottom rows contained the subtype C and D reactions using 0.35µM (third row) and 0.30µM (bottom row). Top and second rows: lanes 2-17 contained subtype A reactions; lanes 18-34 contained subtype B reactions. Third and bottom rows: lanes 2-17 contained subtype C reactions and

116

lanes 18-34 contained subtype D reactions. Although the reactions across both concentrations were generally robust when using the subtype C extract, the higher primer concentration caused multiple mispriming events across all four subtypes, particularly subtypes A and D. The lower primer concentration of 0.3µM produced fewer mispriming events.

Further testing showed that a primer concentration at 0.25µM per reaction reduced the degree of mispriming, but also reduced the overall reaction efficiencies for subtypes A and D. The number of PCR cycles in the first round was increased from 35 to 40 to help alleviate this.

The annealing temperature was optimised across subtypes A, B, C, and D. A preliminary gradient PCR was performed using annealing temperatures of 50°C, 53°C, 56°C, 59°C, 62°C, 65°C and four replicates each of clinical specimens with 20,000 RNA copies extracted . The reactions for subtypes B and C viruses were robust over a range of temperatures, but the reactions for subtypes A and D were not. On this basis, the gradient PCR was repeated using eight replicates of subtypes A and D only, in order to perform a more rigorous assessment for these subtypes. The results showed that, whilst mispriming was still occurring with subtype A, the best temperature was between 59°C and 62°C for both subtypes. Based on the results from both gradient PCRs, an annealing temperature of 60°C was chosen.

Two final checks were performed to ensure that the selected reaction conditions were optimal. Firstly, 16 replicates each of a subtype A specimen with 20,000 copies extracted were amplified as neat cDNA, and dilutions of 1:2 and 1:4 at primer concentrations per reaction of both 0.25 and 0.30µM. The results showed comparable amplification success at neat cDNA (10/16, 63%) for both primer concentrations, but better success at the 1:4 cDNA dilution when using the 0.25µM concentration (7/15, 47% vs. 3/14, 21%). In terms of mispriming, the 0.25µM reactions showed generally less mispriming (7/16, 44%, 8/16, 50%, 3/16, 19% for neat cDNA, 1:2 and 1:4, respectively) than the 0.3µM reactions (12/16, 75%, 7/16, 44%, 4/16, 25% for neat cDNA, 1:2 and 1:4, respectively). These results suggested that mispriming was related as much to the template concentration as the primer concentration.

117

One of the potential recombinant specimens that had a high viral load and a large sample volume available (500,000 copies/ml and 1500µl, respectively) was amplified using eight replicates each at the following cDNA dilutions: neat, 1:4, 1:10, 1:50 and 1:100. The degree of mispriming found in these reactions was 7/8 (88%) using neat cDNA, 3/8 (38%) using the 1:4 dilution, 2/8 (25%) using the 1:10 dilution, 0/8 using 1:50, and 0/8 using 1:100, respectively. Given that the aim of the optimisation was to optimise for low numbers of template molecules, these results suggested that the assay was optimised sufficiently.

The final reaction conditions chosen represented the best available combination that achieved consistent amplification across all four validation subtypes. These conditions differed from the original protocol, in that the optimised assay used a different PCR system, different primer concentrations, a different annealing temperature and a different number of PCR cycles. Table 4_1 summarises the main features of the finalised protocol compared with the Nadai et al. and unadapted CHAVI protocols.

Nadai et al. (2008) CHAVI (2009) Final protocol

Template Plasma RNA Plasma RNA Plasma RNA

No. of fragments to amplify HIV-1 genome

2 - 3 1 1

Genomic coverage (HXB2)

623 - 9636 552 - 9636 552 - 9636

Optimum HIV-1 RNA copies (input)

50,000 - 375,000 20,000 - 40,000 10,000 - 20,000

Optimised HIV-1 subtypes

B B, C A, B, C, D

Method 2-step, nested RT-PCR 2-step, nested RT-PCR 2-step, nested RT-PCR

cDNA synthesis method Oligo dT or gene-specific priming (UNINEF 7')

Gene-specific priming (1.R3.B3R)

Gene-specific priming (1.R3.B3R)

PCR Reagents Expand Long Template

PCR kit (Roche Diagnostics)

Expand Long Template PCR kit (Roche Diagnostics)

Platinum PCR Supermix High Fidelity (Life Technologies) Forward and Reverse

Primer concentrations (µM) 0.4 0.3 0.25 Number of PCR cycles first round 30 35 40 Number of PCR cycles second round 30 35 45 Annealing Temperature (⁰C) 60 55 60

Table 4_1. Comparison of the final optimised RT-PCR protocol for near full-length HIV-1 amplification with the unadapted Nadai et al. and CHAVI protocols.

118