Human immunodeficiency virus-1 evolutionary patterns associated with pathogenic processes in the brain

Abstract

The interplay between pathology and human immunodeficiency virus (HIV) expansion in brain tissues has not been thoroughly assessed in the highly active antiretroviral therapy (HAART) era. HIV-associated dementia (HAD) is marked by progressive brain infection due to recruitment and migration of macrophages in brain tissues; however, the cellular and viral events occurring prior to HAD development and death are under debate. In this study, 66 brain tissues from 11 autopsies were analyzed to assess HIV-1 DNA concentration in brain tissues. In most patients without HAD, it was impossible to amplify HIV-1 from brain tissues. Amplifiable DNA was obtained from three cases of patients on HAART who died due to primary pathology other than HAD: (1) cardiovascular disease, a disease associated with HAART therapy; (2) bacterial infections, including Mycobacterium avium complex, rapid occurrence of extreme dementia; and (3) acquired immunodeficiency syndrome (AIDS)-related lymphoma with meningeal involvement. HIV-1 DNA was also amplified from multiple tissues of two HAD patients. Analysis of HIV-1 nef, gp120, and gp41 sequences showed reduced viral evolution within brain tissues for the non-HAD cases relative to patients with clinical and histological HAD. The present study is the first to show a potential correlation between HIV-1 evolutionary patterns in the brain and different neuropathologies.

Figure 1. V3 loop consensus sequences

The 35 consensus amino acids that span the V3 loop domain along with the overall charge of the domain are shown for each patient. Conserved positions, those that occur at a level of 95% or more, are shown in black and variable positions are shown in red. Lower case letters indicate a position that is conserved at a level of 51–94%, whereas a question mark indicates a position where no majority is observed.

Figure 2. Nef, gp120 and gp41 genetic variation in sequences generated from brain tissues of different patients

Each two-letter code on the x-axis indicates the subject ID. Subject pathology is shown above the graph (CVD = cardiovascular disease, ARL = AIDS-related lymphoma, HAD = HIV-1 associated dementia, MAC = Mycobacterium avium complex). Colored bars represent percent nucleotide diversity within brain sequences in specific gene regions according to the color legend to the right of the graph. Standard error bars are also given. The number along each colored bar indicates the percentage of recombinant sequences detected in that gene region. The number of sequences used in each calculation is shown in parentheses below the subject ID.

Figure 3. Bayesian phylogenetic analysis of nef proteins from sequences derived from lymphoid and non-lymphoid tissues

The Bayesian MAP trees were inferred as detailed in the methods section. The subject name appears above each analysis (Panels A-E). Tropism for patient DY was not determined experimentally and was implied by computational tests and viral behavior. Branches containing sequences generated from non-lymphoid (brain) tissues are shown in red whereas those generated from lymphoid tissues are shown in black. Branch lengths are drawn in scale proportional to nucleotide substitutions per site. The two lines at the bottom of each tree represents the length of the entire tree (top line) and the length of the tree up to the earliest brain clade (bottom line) in units of nucleotide substitutions per site (bold numbers). Each number along an internal branch represents the posterior probability for that branch. Only posterior probabilities >75% are indicated.

Figure 3. Bayesian phylogenetic analysis of nef proteins from sequences derived from lymphoid and non-lymphoid tissues

The Bayesian MAP trees were inferred as detailed in the methods section. The subject name appears above each analysis (Panels A-E). Tropism for patient DY was not determined experimentally and was implied by computational tests and viral behavior. Branches containing sequences generated from non-lymphoid (brain) tissues are shown in red whereas those generated from lymphoid tissues are shown in black. Branch lengths are drawn in scale proportional to nucleotide substitutions per site. The two lines at the bottom of each tree represents the length of the entire tree (top line) and the length of the tree up to the earliest brain clade (bottom line) in units of nucleotide substitutions per site (bold numbers). Each number along an internal branch represents the posterior probability for that branch. Only posterior probabilities >75% are indicated.

Figure 3. Bayesian phylogenetic analysis of nef proteins from sequences derived from lymphoid and non-lymphoid tissues

The Bayesian MAP trees were inferred as detailed in the methods section. The subject name appears above each analysis (Panels A-E). Tropism for patient DY was not determined experimentally and was implied by computational tests and viral behavior. Branches containing sequences generated from non-lymphoid (brain) tissues are shown in red whereas those generated from lymphoid tissues are shown in black. Branch lengths are drawn in scale proportional to nucleotide substitutions per site. The two lines at the bottom of each tree represents the length of the entire tree (top line) and the length of the tree up to the earliest brain clade (bottom line) in units of nucleotide substitutions per site (bold numbers). Each number along an internal branch represents the posterior probability for that branch. Only posterior probabilities >75% are indicated.