This will be an uncontrolled, open-labelled pilot study exploring whether minocycline has a measurable and selective effect on HIV infection of the central nervous system.

This study is founded on a sequence of related hypotheses: 1. inflammatory responses related to activation of macrophages importantly contribute to the magnitude of CNS HIV infection by increasing the local production of viral progeny; 2. the tetracycline, minocycline, has anti-inflammatory properties which likely underlie studies showing that this drug can inhibit HIV-1 infection in macrophages and microglia in vitro and reduce simian immunodeficiency virus (SIV) encephalitis in macaques; 3. by reducing CNS monocyte/macrophage/microglial activation, minocycline will therefore reduce CNS HIV infection; 4. CSF will reflect or parallel (and thus serve as a 'model' of) brain infection and inflammation in this setting; 5. therefore, longitudinal CSF monitoring can assess the effect of minocycline on both CNS HIV infection and inflammation; 6. because the brain injury underlying AIDS dementia complex (ADC) and its pathological substrate, HIV encephalitis, critically involve inflammatory processes and, in the broad sense, immunopathology, minocycline might eventually prove useful as an adjunct to antiviral therapy in accelerating recovery from this condition (though importantly, this pilot study will not include ADC patients).

This will be an uncontrolled, open-labelled pilot study exploring whether minocycline has a measurable and selective effect on CSF HIV RNA concentration. There are no previous studies examining this effect in humans. We define a priori a 'biologically meaningful' effect to be an increase in the Δplasma-CSF HIV concentration of >0.5 log10 copies/mL of HIV RNA (i.e. an increase in the difference between plasma and CSF of >0.5 log10 copies/mL of HIV RNA compared to the baseline difference) in the face of unchanged or reduced plasma HIV RNA. Reductions in the absolute levels of CSF and plasma HIV as well as reductions in CSF inflammatory markers and T cell activation will also be of interest.

This study will serve as an initial exploration of the possible therapeutic effect of minocycline on CNS HIV infection. Our overall strategy is to begin with this pilot study, and if the results look promising (biological effect and lack of toxicity), to use these results to design a controlled trial, either as a single or multi-institutional study.

Additionally, this study shares an almost identical design with another proposed study examining the effects of atorvastatin on CSF HIV infection. While neither of these studies is controlled, they will yield pilot comparative results.

Inclusion Criteria:

1. HIV infection with plasma and CSF HIV RNA concentrations (using Roche Amplicor assay) > 1,000 copies/ mL (available after baseline LP).
2. Off antiretroviral therapy (ART) for > 6 weeks before the study and no plans to begin treatment for the study duration. (The decision of whether or not a subject takes antiretroviral therapy will be made by the subject in consultation with his/her primary care provider prior to screening for this study.)
3. Predicted adherence to the medication.
4. Capable of providing informed consent.
5. > 18 years old
6. CD4 cell counts >150 cells/μL (though likely most, if not all, will be >250 cells/μL).
7. When available, subjects will be screened for stability of blood CD4 and HIV RNA levels.

Exclusion Criteria:

1. Taking a tetracycline within 6 months or history of adverse reaction to minocycline or another tetracycline.
2. Enhanced risk from lumbar puncture, including documented or suspected cerebral mass lesion predisposing to brain herniation or bleeding diathesis.
3. Pregnancy or expectation of pregnancy during the study.
4. Active opportunistic infection or active neurological disease that might confound evaluation.
5. ADC Stage > 1.
6. Hemoglobin < 10 Gms/dL.
7. BUN or creatine above the normal limits.
8. Taking other drugs known to reduce the metabolism of minocycline and thus increase the probability of toxicity.

HIV-infected subjects >18 years old not on antiretroviral therapy who are willing to volunteer to take minocycline and undergo the four study LPs and other evaluations recruited from the San Francisco Bay area.