A template-dependent dislocation mechanism potentiates K65R reverse transcriptase mutation development in subtype C variants of HIV-1

Abstract

Numerous studies have suggested that the K65R reverse transcriptase (RT) mutation develops more readily in subtype C than subtype B HIV-1. We recently showed that this discrepancy lies partly in the subtype C template coding sequence that predisposes RT to pause at the site of K65R mutagenesis. However, the mechanism underlying this observation and the elevated rates of K65R development remained unknown. Here, we report that DNA synthesis performed with subtype C templates consistently produced more K65R-containing transcripts than subtype B templates, regardless of the subtype-origin of the RT enzymes employed. These findings confirm that the mechanism involved is template-specific and RT-independent. In addition, a pattern of DNA synthesis characteristic of site-specific primer/template slippage and dislocation was only observed with the subtype C sequence. Analysis of RNA secondary structure suggested that the latter was unlikely to impact on K65R development between subtypes and that Streisinger strand slippage during DNA synthesis at the homopolymeric nucleotide stretch of the subtype C K65 region might occur, resulting in misalignment of the primer and template. Consequently, slippage would lead to a deletion of the middle adenine of codon K65 and the production of a -1 frameshift mutation, which upon dislocation and realignment of the primer and template, would lead to development of the K65R mutation. These findings provide additional mechanistic evidence for the facilitated development of the K65R mutation in subtype C HIV-1.

Figure 1. Sequence comparison of the RT region of the pol gene derived from subtype B and C HIV-1 spanning codons 62 to 68.

(A) Outline of the reverse-transcription process in subtype B HIV-1 leading to either wild-type or K65R-containing transcripts. The K65R mutation results from an AAA-to-AGA mutation in subtype B HIV-1. (B) Outline of the reverse-transcription process in subtype C HIV-1 leading to either wild-type or K65R-containing transcripts. The K65R mutation results from an AAG-to-AGG mutation in subtype C HIV-1. The genomic sequences of the subtype B and C RT segments of pol spanning codons 62 to 68 are indicated. (−)ssDNA synthesis from the viral (+)ssRNA genome and (+)dsDNA synthesis from the (−)ssDNA intermediate are also depicted. Underlined are the homopolymeric nt stretches of both templates at their specific locations. In bold are the nt polymorphisms that exist between both subtypes in this region. Highlighted in a square is the base that, once mutated, gives rise to the K65R drug resistance mutation.

Figure 2. K65R production rates with subtype B RT on subtype B and C templates with primers containing the mutagenic nt.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (-)ssDNA intermediate with subtype B RT on the subtype B template. The full-length product is observed as a single band at the FL position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype C template. The full-length product is observed as two distinct bands at the FL and FL-1nt positions, which is indicative of dislocation on the subtype C template. (B) Graphical representation of the amount of transcripts containing the mutagenic nt produced with subtype B RT on both subtype B and C templates. The values indicated with an asterisk have a p-value <0.05 when the amount of K65R-production between both subtypes at the given time-point is compared. More transcripts with the mutagenic G at position 65 are produced with the subtype C template than with the subtype B template. (C) Depiction of the primer and template systems used. The primers contain a G base on their 3′-end that becomes mismatched on the T on the template strand thus yielding transcripts with the mutagenic nt. The homopolymeric regions of both templates are underlined and the base responsible for the K65R mutation is indicated in bold.

Figure 3. K65R production rates with subtype C RT on subtype B and C templates with primers containing the mutagenic nt.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype C RT on the subtype B template. The full-length product is observed as a single band at the FL position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype C RT on the subtype C template. The full-length product is observed as two distinct bands at the FL and FL-1nt positions, which is indicative of dislocation on the subtype C template. The results are similar to those obtained when subtype B RT was used on the same templates, thus highlighting the template-specific and enzyme-independent nature of the mechanism. (B) Graphical representation of the amount of transcript production with subtype C RT on both subtype B and C templates. The values indicated with an asterisk have a p-value <0.05 when the amount of K65R-production between both subtypes at the given time-point is compared. More transcripts containing the mutagenic nt are produced with the subtype C template than with the subtype B template. The same trend is present when subtype B RT was used on the same templates. (C) Depiction of the primer and template systems used. The primers contain a G base at their 3′-end that becomes mismatched on the T on the template strand thus yielding K65R-containing transcripts. The homopolymeric regions of both templates are underlined and the base responsible for the K65R mutation is indicated in bold.

Figure 4. Single incorrect nt incorporation and K65R production rates with subtype B RT on subtype B and C templates.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype B template. The full-length product is observed as a single dGTP incorporation occurred at the P+1nt position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype C template. The full-length product is observed as containing two distinct dGTP incorporations at the P+1nt and P+2nt positions, indicative of dislocation with the subtype C template. (B) Graphical representation of transcript production with subtype B RT on both the subtype B and C templates. The values indicated with an asterisk have a p-value <0.05 when the amount of transcripts containing the mutagenic nt produced between both subtypes at the given time-point is compared. More transcripts containing the mutagenic G nt at codon 65 are produced with the subtype C template than with the subtype B template. (C) Depiction of the primer and template systems used. dGTP was the only nt employed in the reaction to allow for the production of the mutagenic transcripts. The homopolymeric regions of both templates are underlined and the base responsible for the K65R mutation is indicated in bold.

Figure 5. Single incorrect nt incorporation and K65R production rates with subtype C RT on subtype B and C templates.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype C RT on the subtype B template. The full-length product contains a single dGTP incorporation at the P+1nt position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype C RT on the subtype C template. The full-length product contains two distinct dGTP incorporations at the P+1nt and P+2nt positions, indicative of dislocation with the subtype C template. (B) Graphical representation of transcript production with subtype C RT on both the subtype B and C templates. The same trend as when the subtype B-derived RT enzyme was employed was observed, further reinforcing the template-specific and enzyme independent notion of the mechanism. (C) Depiction of the primer and template systems used. The homopolymeric regions of both templates are underlined and the base responsible for the K65R mutation is indicated in bold.

Figure 6. Multiple incorrect nt incorporations at the K65 position with subtype B RT on subtype B and C templates.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype B template. The full-length product contains a single dGTP, dCTP or dTTP incorporation opposite the T in the template strand at the P+1nt position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype C template. Dislocation is observed with the subtype C template as two distinct nt incorporations at the P+1nt and P+2nt positions. (B) Graphical representation of DNA synthesis with subtype B RT on the subtype B and C templates. The values indicated with an asterisk have a p-value <0.05 when the amount of transcript-production between both subtypes at the given time-point is compared. (C) Depiction of the primer, template and nt employed in the system. The homopolymeric regions of both templates are underlined and the base responsible for the K65R generation is bolded.

Figure 7. Multiple incorrect nt incorporations at the K65 position with subtype C RT on subtype B and C templates.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype C RT on the subtype B template. The full-length product is observed as a single dGTP, dCTP or dTTP incorporation occurred opposite the T in the template strand at the P+1nt position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype C RT on the subtype C template. Dislocation is only present with the subtype C template. (B) Graphical representation of DNA synthesis with subtype C RT on the subtype B and C templates. The same trend is observed regardless of origin of the RT enzyme used. (C) Depiction of the primer, template and nt used in the reaction. The homopolymeric regions are underlined and the base responsible for the K65R mutation is indicated in bold.

Figure 8. Evaluation of the dislocation products in the various subtype C primer/template conditions.

(A) Graphical representation of K65R-containing transcript production with primers already harboring the mutagenic nt. Left: K65R production with subtype B RT on the subtype C template. Right: K65R production with subtype C RT on the subtype C template. The product of dislocation without realignment (FL-1nt product) is depicted as empty circles and the products of dislocation with realignment or misincorporation without dislocation (FL product) are depicted as empty triangles. When the mutagenic nt is already incorporated into the primer strand, the amount of FL and FL-1nt products is identical. (B) Schematic of the dislocation without realignment (FL-1nt) and the dislocation with realignment or misincorporation without dislocation (FL) products with the primer strand already containing the mutagenic nt. The sequence of the unextended primer is depicted in bold. (C) Graphical representation of K65R-containing transcript production with dGTP incorporation at the site of K65R development. Left: K65R production with subtype B RT on the subtype C template. Right: K65R production with subtype C RT on the subtype C template. The product of dislocation without realignment (P+1nt product) is depicted as empty circles and the products of dislocation with realignment or misincorporation without dislocation (P+2nt product) are depicted as empty triangles. During the process of incorporation of the mutagenic nt, the amount of P+1nt product reaches a maximum at 60 min whereas the amount of P+2nt product continues to rise steadily beyond 60 min. (D) Schematic of the dislocation without realignment (P+1nt) and the dislocation with realignment or misincorporation without dislocation (P+2nt) products during dGTP incorporation at the site of K65R development. The sequence of the unextended primer is depicted in bold.

Figure 9. K65K, K65T and K65I production as a result of DNA synthesis with primers containing the three nt alternatives and their impact on dislocation.

(A) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype B template. The full-length product is observed as a single band at the FL position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype C template. The full-length product is observed again as a distinct band at the FL and positions without any dislocation products. When A is present at the 3′-end of the primer strand, it correctly binds with the T of the template strand to yield wild-type K65K-containing transcripts. (B) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype B template. The full-length product is observed as a single band at the FL position. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype C template. No dislocation products were observed. When C is present at the 3′-end of the primer strand, it incorrectly binds with the T of the template strand to yield K65T-containing transcripts. (C) Lanes 1 through 10 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype B template. The full-length product is observed as a single band at the FL position without dislocation. Lanes 11 through 20 depict (+)dsDNA synthesis from the (−)ssDNA intermediate with subtype B RT on the subtype C template. When T is present at the 3′-end of the primer strand, it incorrectly binds with the T of the template strand to yield K65I-containing transcripts. Dislocation products are only observed when G is included on the primer strand yielding K65R-containing transcripts.

Figure 10. Schematic of dislocation mutagenesis and the development of the K65R mutation in subtype C HIV-1.

Step 1: DNA synthesis approaches the end of a homopolymeric nt stretch that ends precisely at the location of K65R development. Step 2: At the end of the sequence, the RT enzyme exhibits characteristic pausing of DNA synthesis. The template-strand folds onto itself and exposes a C in the folded-over template strand. Step 3: A dGTP nt correctly binds opposite the C base of the misaligned template strand as DNA synthesis continues. Step 4: The primer and template strands realign and the same C base becomes re-exposed on the now correctly aligned template strand. Step 5: A second dGTP becomes incorporated opposite the re-exposed C on the correctly aligned primer/template strands and DNA synthesis continues normally. This series of events yields the AAG-to-AGG change that is responsible for the more facilitated appearance of the K65R mutation in subtype C HIV-1.