Increasing the CpG dinucleotide abundance in the HIV-1 genomic RNA inhibits viral replication

Abstract

Background: The human immunodeficiency virus type 1 (HIV-1) structural protein Gag is necessary and sufficient to form viral particles. In addition to encoding the amino acid sequence for Gag, the underlying RNA sequence could encode cis-acting elements or nucleotide biases that are necessary for viral replication. Furthermore, RNA sequences that inhibit viral replication could be suppressed in gag. However, the functional relevance of RNA elements and nucleotide biases that promote or repress HIV-1 replication remain poorly understood.

Results: To characterize if the RNA sequence in gag controls HIV-1 replication, the matrix (MA) region was codon modified, allowing the RNA sequence to be altered without affecting the protein sequence. Codon modification of nucleotides (nt) 22-261 or 22-378 in gag inhibited viral replication by decreasing genomic RNA (gRNA) abundance, gRNA stability, Gag expression, virion production and infectivity. Comparing the effect of these point mutations to deletions of the same region revealed that the mutations inhibited infectious virus production while the deletions did not. This demonstrated that codon modification introduced inhibitory sequences. There is a much lower than expected frequency of CpG dinucleotides in HIV-1 and codon modification introduced a substantial increase in CpG abundance. To determine if they are necessary for inhibition of HIV-1 replication, codons introducing CpG dinucleotides were mutated back to the wild type codon, which restored efficient Gag expression and infectious virion production. To determine if they are sufficient to inhibit viral replication, CpG dinucleotides were inserted into gag in the absence of other changes. The increased CpG dinucleotide content decreased HIV-1 infectivity and viral replication.

Conclusions: The HIV-1 RNA sequence contains low abundance of CpG dinucleotides. Increasing the abundance of CpG dinucleotides inhibits multiple steps of the viral life cycle, providing a functional explanation for why CpG dinucleotides are suppressed in HIV-1.

Keywords: CpG dinucleotide; Genomic RNA; HIV-1; Viral replication.

Fig. 1

Codon modification of nucleotides 22-261 in gag inhibits viral replication. a Schematic representation of p55^Gag in HIV-1_NL4-3, HIV-1 CM22-261, HIV-1 CM22-165 and HIV-1 CM166-261. b The amount of HIV-1 CA (p24^Gag) in supernatants from 293T cells transfected with pHIV-1_NL4-3, pHIV-1 CM22-261, pHIV-1 CM22-165 or pHIV-1 CM166-261 were quantified by p24^Gag ELISA. The bar chart is the average of three independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation. c Jurkat cells were infected with 25 ng of p24^Gag for each indicated virus. The amount of infectious virus present at each time point was measured in TZM-bl cells. This is representative of three independent experiments

Fig. 2

Codon modification of nucleotides 22-261 in gag inhibits infectious virus production. HeLa cells were transfected with pHIV-1_NL4-3, pHIV-1 CM22-261, pHIV-1 CM22-165 or pHIV-1 CM166-261. a Culture supernatants were used to infect TZM-bl reporter cells to measure viral infectivity. The bar charts show the average values of six independent experiments normalized to the value obtained for HIV-1_NL4-3. The average relative light units (RLU) for HIV-1_NL4-3 is 3,680,747. The dashed line represents 1000 RLU, which is approximately the limit of the assay for reproducible differential results. Error bars represent standard deviation. b, c Gag expression in the media (b) and cell lysate (c) was detected using quantitative immunoblotting. The bar charts show the average of four independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation

Fig. 3

Codon modification of nucleotides 22-261 in gag decreases gRNA abundance. HeLa cells were transfected with pHIV-1_NL4-3, pHIV-1 CM22-261, pHIV-1 CM22-165 or pHIV-1 CM166-261. The bar charts show the average values of three independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation. a RNA was extracted from cell lysates and gRNA abundance was quantified by qRT-PCR. b RNA was extracted from the media and gRNA abundance was quantified by qRT-PCR. c Equivalent amounts of HIV-1 genomes were used to infect TZM-bl reporter cells to measure infectivity. The bar charts show the average values of three independent experiments normalized to the value obtained for HIV-1_NL4-3. The average RLU for HIV-1_NL4-3 is 1,146,196. The dashed line represents 1000 RLU, which is approximately the limit of the assay for reproducible differential results

Fig. 4

Codon modification but not deletion of nucleotides 22-261 or 22-378 in gag inhibits infectious virus production. a Schematic representation of p55^Gag in HIV-1_NL4-3, HIV-1 CM22-261, HIV-1 Δ22-261, HIV-1 CM22-378 or HIV-1 Δ22-378. b, d HeLa cells were transfected with pHIV-1_NL4-3, pHIV-1 CM22-261, pHIV-1 Δ22-261, pHIV-1 CM22-378 or pHIV-1 Δ22-378 and pGFP (b) or pVSV-G (d). The amount of infectious virus in the culture supernatants was measured in TZM-bl cells. The average RLU for HIV-1_NL4-3 + GFP and HIV-1_NL4-3 + VSV-G is 13,701,427 and 16,981,387, respectively. The dashed line represents 1000 RLU, which is approximately the limit of the assay for reproducible differential results. c, e Gag expression in the media and cell lysate were measured by quantitative western blotting. b–e The bar charts show the average of three independent experiments relative to HIV-1_NL4-3. Error bars represent standard deviation

Fig. 5

Codon modification of nucleotides 22-261 and 22-378 in gag decreases gRNA abundance and stability. HeLa cells were transfected with pHIV-1_NL4-3, pHIV-1 CM22-261 or pHIV-1 CM22-378. a RNA was extracted from cell lysates at the 0 h timepoint and gRNA abundance was quantified by qRT-PCR. The bar charts show the average values of three independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation. b, c Actinomycin D was added to inhibit RNA polymerase II transcription and the abundance of MYC mRNA (b) or gRNA (c) was measured at 0, 1, 2, 4 and 6 h post-addition. Each value is relative to the 0 h timepoint and is an average of three independent experiments. Error bars represent standard deviation

Fig. 6

Schematic representation of nucleotides 1-420 in gag for HIV-1_NL4-3, HIV-1 CM22-261, HIV-1 CM22-165, HIV-1 CM166-261, HIV-1CM22-261_lowCpG, HIV-1 CM22-378 and HIV-1 CM22-378_lowCpG. Mutations that are the result of codon modification and do not introduce CpG dinucleotides are highlighted in turquoise. CpG dinucleotides present in HIV-1_NL4-3 are highlighted in yellow and CpG dinucleotides introduced by codon modification are highlighted in red

Fig. 7

Decreasing the CpG abundance within HIV-1 CM22-261 and HIV-1 CM22-378 restores infectious virus production. HeLa cells were transfected with pHIV-1_NL4-3, pHIV-1 CM22-261, pHIV-1 CM22-261_lowCpG, pHIV-1 CM22-378 and pHIV-1 CM22-378_lowCpG. The bar charts show the average of three independent experiments relative to HIV-1_NL4-3. Error bars represent standard deviation. a, c The amount of infectious virus present in the media was measured in TZM-bl cells. b, d Virion production and intracellular Gag expression was measured using quantitative western blotting

Fig. 8

Introduction of CpG dinucleotides into gag inhibits HIV-1 replication in Jurkat cells. a The amount of HIV-1 CA (p24^Gag) in supernatants from 293T cells transfected with pHIV-1_NL4-3, pHIV-1 CpG22-165, pHIV-1 CpG22-261 or pHIV-1 CpG22-378 were quantified by p24^Gag ELISA. The bar chart is the average of three independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation. b Jurkat cells were infected with 25 ng of p24^Gag for each indicated virus. The amount of infectious virus present at each time point was measured in TZM-bl cells. This is representative of three independent experiments

Fig. 9

Introduction of CpG dinucleotides into gag inhibits infectious virus production in HeLa cells. HeLa cells were transfected with pHIV-1_NL4-3, pHIV-1 CpG22-165, pHIV-1 CpG22-261 or pHIV-1 CpG22-378. a Culture supernatants were used to infect TZM-bl reporter cells to measure viral infectivity. The bar charts show the average values of four independent experiments normalized to the value obtained for HIV-1_NL4-3. b Gag expression in the media and cell lysate was determined by quantitative immunoblotting. The bar charts show the average of three independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation. RNA was extracted from cell lysates and media (c, d) and gRNA abundance was quantified by qRT-PCR. b–d The bar charts show the average of three independent experiments normalized to HIV-1_NL4-3. Error bars represent standard deviation