Overview of experimental and analytical procedures. (A) Experimental pipeline. (B) Bioinformatics pipeline.

Analysis of cellular transcripts in mock- and HIV-infected cells. Average tag density analysis in mock-infected (black lines) and HIV-1-infected (red lines) samples for total cellular transcripts (lines), repetitive elements (dashed lines), and HERV elements (dotted lines) is shown. Most cellular transcripts were detected with 1 to 1,000 tags (log 0 to 3; average, 2.9 log) across the complete transcriptome. Only a few transcripts were identified in larger numbers (e.g., above 10,000; log 4). A smaller number of tags was observed for the expressed repetitive elements (average, 2.3 log), including HERVs (average, 1.9 log). Statistical differences between mock infection and HIV-1 infection distributions were assessed by Wilcoxon test.

Distribution of HIV-specific tags along the viral genome. (A) The HIV vector genome is depicted at the top, and nucleotide positioning, NlaIII restriction sites (vertical black bars), the common sense transcription start site (TSS; red arrow), and the polyadenylation signal [poly(A); red vertical bar] are shown below. Multiple antisense transcription start sites (distributed between positions 9175 and 8714; only two are drawn here for figure simplicity; green arrows) and the poly(A) signal (vertical green bar), as described in Landry et al. (26), are indicated. HIV-1 open reading frames also are indicated, including that of the putative antisense protein (ASP). The lower portion of this panel indicates the density of tags (vertical axis) distributed along the HIV-1 vector genome (x axis) for both HIV-1 samples (HIV rep1 and HIV rep2) and mock infection samples (Mock rep1 and Mock rep2). The five major peaks are numbered at the bottom, displaying the number of tags at each peak as well as the proportion of these tags representing sense or antisense transcription. (B) Total RNA from mock-infected or HIV-based vector-infected cells was extracted using miRvana or Trizol at 22 and 24 h postexposure and subjected to Northern blot hybridization. Left panel, total RNA gel electrophoresis; right two panels, Northern blots using strand-specific oligonucleotide probes aligning with the corresponding peak of tags detected by SAGE-Seq. Probes were designed to anneal with sense transcripts (red probes; top) or antisense transcripts (green probes; bottom) with the nucleic acid sequence complementary to the identified peak tag sequence. The plasmid encoding the HIV vector genome digested with BglII and AflII was used as a positive control for hybridization (+).