The extent of sequence complementarity correlates with the potency of cellular miRNA-mediated restriction of HIV-1

Abstract

MicroRNAs (miRNAs) are 22-nt non-coding RNAs involved in the regulation of cellular gene expression and potential cellular defense against viral infection. Using in silico analyses, we predicted target sites for 22 human miRNAs in the HIV genome. Transfection experiments using synthetic miRNAs showed that five of these miRNAs capably decreased HIV replication. Using one of these five miRNAs, human miR-326 as an example, we demonstrated that the degree of complementarity between the predicted viral sequence and cellular miR-326 correlates, in a Dicer-dependent manner, with the potency of miRNA-mediated restriction of viral replication. Antagomirs to miR-326 that knocked down this cell endogenous miRNA increased HIV-1 replication in cells, suggesting that miR-326 is physiologically functional in moderating HIV-1 replication in human cells.

Figure 1.

In silico screening of human miRNA target sites in the HIV genome. (A) Positions of 39 predicted target sites in the HIV NL4-3 genome for the indicated 22 human miRNAs are shown. The miRNAs with multiple putative target sites are shown at the top, and those miRNAs with single putative target sites are shown at the bottom. (B) Examples of predicted target duplexes for four miRNAs, miR-326, miR-133b, miR-138 and miR-149, are shown. The miRNA sequence (top) is shown with its seed sequence underlined, and its complementary HIV sequence indicated below. The numbering is based on the pNL4.3 sequence (left, nucleotide position numbers). (C) Three examples of published miRNA–mRNA target site duplex with mismatches or G:U pairings in the seed sequence are shown.

Figure 2.

Expression of human miRNAs suppress HIV-1 replication in cell culture. (A) Effects of transfection of the indicated individual miRNAs on HIV-1 replication in cultured cells. The 42CD4 cells were transfected with the indicated miRNAs, and cells were infected with HIV-1 NL4-3 24 h later. Viral replication was monitored by measuring RT production in culture supernatants, 3 days after infection. Values are expressed as percent of a control irrelevant RNA oligonucleotide (oligo fluo). Arrows indicate miRNA chosen for follow-up experiments. (B) Time course replication curves spanning 2, 3, 4 and 6 days post-infection in cells transfected with miR-138, miR-149, miR-326, miR-195 and miR-29b (indicated by an arrow in panel A) and irrelevant control RNA oligonucleotide (oligo fluo); miR-29b is a positive control; oligo fluo is a negative control. Results are representative of three independent experiments. Values are the mean of three separate experiments with standard deviations.

Figure 3.

Restriction of an HIV-1 genome containing a 3’ LTR site engineered to have perfect complementarity with miR-326. (A) A schematic representation of engineered HIV genomes containing a pNL4-3 Wt miR-326 target site or a target site created by changing the indicated nucleotides to create a sequence fully complementary to human miR-326 (Tgt). (B) and (C) Comparisons of the replication of Tgt and Wt viruses in 42CD4 (B) and Jurkat (C) cells. Cells were infected with the same amount of Wt or Tgt virus (normalized using reverse transcriptase measurements), and the infections were monitored by measuring supernatant RT 2, 3 and 4 days (42CD4) or 4, 6 and 9 days (Jurkat) post-infection. (D) TZMbl indicator cells were infected with RT equivalents of Wt or Tgt viruses. On completing 24 h post-infection, cells were fixed and infectivity was measured by β-galactosidase assay. P-value was determined by Student’s t-test comparing Wt with Tgt. Values are the mean of three separate experiments with standard deviations.

Figure 4.

Deficit in Tgt replication is miR-326-specific. (A) SiRNA against Dicer (siDicer) equalizes the replication of Wt and Tgt viruses in TZMbl cells. HeLa-derived TZMbl indicator cells were transfected with siRNA against EGFP (control) or Dicer. After 24 h post-transfection, the cells were infected with equal amounts of Wt or Tgt virus normalized by RT. The cells were fixed and the amount of infectious virus was determined by β-galactosidase staining at 24 h post-infection. Values of Wt and Tgt viruses are plotted relative to each other. (B, C) Anti-miR326 antagomir increases relative Tgt-virus replication to a greater extent than Wt-virus replication. Anti-miR326 antagomir or control ‘scrambled’ antagomir was transfected into 42CD4 (B) or Jurkat (C) cells. After 24 h post-transfection, the cells were infected with equal amounts of Wt or Tgt virus. Infections were monitored by measuring supernatant RT at 2 days (42CD4) or at 6 days (Jurkat). P-values were determined using Student’s t-test to compare Wt with Tgt. Values of Wt and Tgt viruses are plotted independently of each other. Note that in each case (Wt or Tgt) anti-326 (+) virus sample was graph-normalized to the respective control scrambled antagomir (−) sample, which was arbitrary set as ‘100’. In actual value, the Tgt scrambled antagomir (−) sample arbitrarily set at ‘100’ is ∼60% (Figure 3D) of the value relative to the Wt scrambled antagomir (−) sample, which is also graphed as arbitrarily set at ‘100’. Because the values of Wt and Tgt viruses are plotted independently of each other, the graphs in (B) and (C) as drawn are intended to compare the Tgt (−) anti-326 relative to Tgt (+) anti-326 and the Wt (−) anti-326 relative to Wt (+) anti-326, but not intended to cross-compare the Wt samples with Tgt samples (in which case, the latter values should be corrected by a factor of ∼0.6).

Figure 5.

Increased miRNA–mRNA complementarity correlates with greater reduction of expression in reporter assays. (A) Schematic representation of CMV promoter-driven reporter constructs engineered to contain three tandem copies of either the Wt or Tgt target sites positioned downstream of the luciferase gene. (B) Effect of miR-326 on the Wt and Tgt target sites. The 293T cells were transfected with Luc-3X Wt or Luc-3X Tgt together with an internal control pMIR-βgal plasmid. After 24 h post -transfection, cells were lysed and luciferase and βgal activities were measured. The luciferase activity from Luc-3X Tgt, after normalization to βgal, was expressed relative to that from the Luc-3X Wt construct. (C) Effect of transfected miR-326 on Luc-3X Wt and Luc-3X Tgt. miR-326 (+) or negative control (−) siRNA (negative control siRNA #1, Ambion) was transfected into 293 T cells with either the Luc-3X Wt or Luc-3X Tgt constructs and the internal control pMIR-βGal. Luciferase and βgal activities were monitored 24 h post-transfection. Values (+) are given as percent of the value from the negative control siRNA (−) transfected cells which was set to 100%. Values are the mean of three separate experiments presented with standard deviations.

Figure 6.

Virus replication competition assays in cells co-infected with Wt and Tgt viruses. (A) Quantification of Wt and Tgt viral RNAs in co-infected cells. Jurkat cells were co-infected with 1:1 or 1:4 (Wt:Tgt) ratios (normalized by RT) of input viruses. On completing 9 days after infection, cells were harvested and total cellular RNA was extracted. Viral RNAs for Wt and Tgt virus were then quantified using real time RTQPCR with primers specific for Wt and Tgt viruses. The corresponding RNA copy numbers are graphed. (B) Assessment of the Tgt RNA:Wt RNA ratios at the different rounds of infection. Supernatants from 9 days Wt+Tgt (1:1 and 1:4) co-infected cells (infectious round 1) were used to infect new Jurkat cells (round 2). After 9 days of infection, supernatants from round 2 of infection were used to initiate a third round of infection (round 3). The Wt and Tgt viral RNAs were quantified by QRT–PCR in infected cells at the end of each infection to calculate their relative amount at the end of each infection cycle (in percent of total virus genomic RNA). Data points represent geometric mean values from duplicate experiments, with error bars indicating standard deviation.

Figure 7.

Creation of multiple miR326 target site mutants. Sequences are shown of various target sites with the indicated changed nucleotides designed to increase complementarity with miR-326. Top two rows show the miR-326 sequence and the Wt target sequence. Three mismatched positions between miR-326 and Wt (G, T, A) are shown in gray. The next three rows (+1 match) show single nucleotide changes to the Wt sequence that increase complementarity with miR-326. In 1.1, G is changed to T; in 1.2, T is changed to G; in 1.3, A is changed to G. The following three rows (+2 matches) show two nucleotide changes to the Wt sequence that increase complementarity with miR-326. In 2.1, the G, T are changed to T, G; in 2.2, the T, A are changed to G, G; in 2.3, the G, A are changed to T, G. In the last two rows (+3 matches), the G, T, A bases in the Wt sequence are changed to T, G, G in 3.1; and in Tgt, three other bases T, G, G are changed to C, A, A. Light and bold green represents either G-U or classical A-U/G-C base pairing, respectively. Red box indicates changed nucleotide.

Figure 8.

Increasing the complementarity of HIV sequence to miR-326 leads to greater restriction of the virus. Co-infection experiments in Jurkat cells with mutant and Wt viruses. Each of the seven changed viruses (1.1, 1.2, 1.3, 2.1, 2.2, 2.3 or 3.1) in A was used in head-to-head co-infection experiments with the Wt virus for three successive rounds of infection. The relative amounts of Wt and mutant RNA (in percent of total genomic RNA) at rounds 1, 2 and 3 are given for co-infection of Wt with 1.1, 1.2 or 1.3 virus (left graph) or Wt with 2.1, 2.2, 2.3 or 3.1 virus (right graph).