Human immunodeficiency virus type-1 (HIV-1) DNA and messenger RNA sequences in both cell lines and blood obtained directly from HIV-1-infected patients were amplified by polymerase chain reaction and hybridized to fluorescein-labeled probes in situ, and the individually labeled cells were analyzed by flow cytometry. After flow cytometric analysis, heterogeneous cell populations were reproducibly resolved into HIV-1-positive and -negative distributions. Fluorescence microscopy showed that the cellular morphology was preserved and intracellular localization of amplified product DNA was maintained. Retention of nonspecific probe was not observed. Analysis of proviral DNA and viral messenger RNA in cells in the blood of HIV-1-infected patients showed that the HIV-1 genome persists in a large reservoir of latently infected cells. With the use of this technique it is now possible to detect single-copy DNA or low-abundance messenger RNA rapidly and reproducibly in a minor subpopulation of cells in suspension at single-cell resolution and to sort those cells for further characterization.

Latent and productive viral infections are at the extremes of the spectrum of virus-cell interactions that are thought to play a major role in the ability of such important human pathogens as human immunodeficiency virus (HIV) to elude host defenses and cause disease. The recent development of PCR-based methods to amplify target sequences in individual cells in routinely fixed tissues affords opportunities to directly examine the subtle and covert virus-ceDl relationships at the latent end ofthe spectrum that are inaccessible to analysis by conventional in situ hybridization techniques. We have now used PCR in situ with in situ hybridization to document latent and permissive HIV infection in routinely fixed and paraffinembedded tissue. In one of the first specimens we examined, a tumor biopsy from an HIV-infected individual, we found many of the lymphocytes and lymphocytes infiltrating the tumor had HIV DNA that was detectable only by PCR in situ. The fraction of positive cells varied regionally, but there were foci where most of the cells contained HIV DNA. Most of these lymphocytes and macrophages are latently infected, as we could detect HIV RNA in fewer than one in a thousand ofthese cells. We also detected HIV RNA, surprisingly, in 6% of the tumor cells, where the number of copies of viral RNA per cell was equivalent to productively infected cell lines. The alternative states of HmV-gene expression and high local concentration of latently infected lymphocytes and monocytes revealed by these studies conceptually supports models of lentiviral pathogenesis that attribute persistence to the reservoir of latently infected cells and disease to the consequences of viral-gene expression in this population. The magnitude of infection of lymphocytes documented in this report is also consistent with the emerging view that HIV infection per se could contribute substantially to depletion of immune cells in AIDS.Human immunodeficiency virus (HIV), the etiological agent of AIDS, can establish productive or latent infections in CD4+ lymphocytes and monocytes in culture, and these alternative states of viral-gene expression could readily account for the devastating consequences of infection -and the formidable difficulties in developing a protective vaccine (1-3). Latently infected cells could escape detection and destruction by host defenses and disseminate infection in and between individuals in the face of natural or vaccine-induced immunity. With activation of viral-gene expression, the immune system will eventually be destroyed, and the individual will succumb to a variety of opportunistic infections and unusual tumors. One critical prediction of this reconstruction of pathogenesis is that in vivo one can show that there is a population of infected cells that harbor the HIV provirus in a transcriptionally silent state with no detectable viral RNA and a second population in which viral transcripts are abundant. We have now used PCR-based technologies with singlecell resolution, originally developed for studies o...

In a retrospective analysis, patients with bacteremia due to Enterococcus faecalis with and without high-level gentamicin resistance (GRE; MIC greater than 2000 micrograms/ml) were compared. Bacteremic patients with GRE (n = 32) had significantly higher rates of nosocomial acquisition and bladder catheterization, longer hospitalizations, and more frequent prior treatment with cephalosporins than did bacteremic patients without high-level resistance (n = 19). Overall mortality was significantly associated with septic shock, high-risk source (intraabdominal, wound, respiratory tract, multiple, unknown), and polymicrobial bacteremia. Higher mortality was observed in GRE bacteremia (47%) than in bacteremia without high-level resistance (37%), but this difference was not statistically significant. For patients with monomicrobial bacteremia, low-risk source (genitourinary tract, intravascular), or treatment with antibiotics appropriate for the enterococcus, higher mortality with GRE bacteremia approached statistical significance. These results suggest that high-level resistance adversely affects survival with a pure E. faecalis bacteremia or low-risk bacteremic source. Also, response to antibiotic therapy may be diminished by high-level resistance.

Injection drug use is the strongest predictor of HCV infection, but sexual risk factors are also independently associated.

Using plasma preparation tubes for the collection and storage of plasma resulted in factitious, low-level human immunodeficiency virus type 1 (HIV-1) viremia among patients receiving highly active antiretroviral therapy who incurred unnecessary additional clinic visits, laboratory testing, and medication changes. We caution clinicians against the routine use of plasma preparation tubes for collection of blood samples for HIV-1 level 1 quantification.