HIV-1 Nef Is Associated with Complex Pulmonary Vascular Lesions in SHIV-nef–infected Macaques

Animal Selection

From the NENPRC, we obtained eight animals: two infected with a chimeric SHIV-nef virion (23, 24) and six infected with SIV_mac251 or SIV_E660 containing the native SIV nef allele. From the California National Primate Research Center (CNPRC; University of California at Davis, Davis, CA) we obtained five macaque lungs: one infected with SIV_mac239 and four infected with SHIV-nef chimeric viruses (25). From the Retrovirus Laboratory, Department of Comparative Medicine, Johns Hopkins University (JHU; Baltimore, MD) we obtained 12 lungs from macaques infected with a mixture of lymphocytotropic and monocytotropic SIV strains (26). Table 1 describes the different animals and their known characteristics. No hemodynamic parameters were available from any of the macaque cohorts. All animal procedures conformed to the requirements of the Animal Welfare Act and protocols were approved before implementation by the Institutional Animal Care and Use Committees at the University of California at Davis, JHU, and Harvard University.

Human Tissues

Archived, paraffin-embedded lung tissue sections were obtained from two patients with IPAH, two patients with HRPH, and one patient with no clinical or histologic evidence of IPAH. These tissues were used for immunohistochemistry of the HIV-1 Nef protein. Patient characteristics have already been described by Cool and coworkers (6, 27).

Tissue and Histologic Examination

All tissues were obtained at necropsy, and were fixed and embedded in paraffin. Serial sections were prepared and shipped to the University of Colorado at Denver and Health Sciences Center (Denver, CO) for hematoxylin and eosin (H&E) staining and immunohistochemistry as indicated. The sections were masked concerning the identity of the experimental groups and examined independently by two pulmonary pathologists (C.D.C. and R.M.T.) for the presence of pulmonary vascular lesions as recommended by Pietra and coworkers (28). A total of 10 and 12 serial sections from a single lung block were obtained from the NENPRC and CNPRC, respectively. From JHU, we received one H&E-stained section for each animal (n = 12). Representative sections were performed on the basis of random selection by the respective institutions. Therefore, we had to limit our analyses to the archived material already collected for other studies. Nevertheless, we are confident that the random nature of the archival process allowed us an analysis of representative samples from many different areas of the lung.

Immunohistochemistry

Statistics

A Fisher's exact test of a 4 × 4 matrix was used to analyze the observations, with p < 0.05 considered significant.

Pulmonary Pathologies in SIV- and SHIV-nef–infected Macaques

We examined a total of 24 different monkey lungs from two separate primate centers based in New England (NENPRC), California (CNPRC), and Johns Hopkins University (JHU), with a total of 18 monkeys infected with SIV and six infected with SHIV-nef (Table 1). The SIV strains contained the simian nef allele whereas the SHIV-nef strains were chimeric viruses with HIV nef from various HIV isolates cloned in place of the native SIV nef sequence. Figure E1 of the online supplement shows a diagrammatic depiction of the genomic structures of parental and chimeric SIV strains.

From the NENPRC, we obtained eight animals: two infected with a chimeric SHIV-nef virion that contained the SIV_mac251 backbone and the HIV nef gene, and six infected with SIV_mac251 or SIV_E660 containing the native SIV nef allele. We found that the two SHIV-nef animals (designated NENPRC7 and NENPRC8 in Table 1) showed pulmonary vessels with medial smooth muscle cell hypertrophy and lumenal obstruction with endothelial cells (Figures 1B and 1C). The lesions were histologically similar to the plexiform lesions seen in patients with HRPH. One animal infected with SIV_mac251 (Figure 1A and Table 1; designated as NENPRC1), showed significant perivascular cellular infiltration, but little disruption of the intimal or medial layers of the pulmonary vasculature. Furthermore, none of the animals infected with parental SIV strains (SIV_mac251 or SIV_E660, NENPRC1–NENPRC6 in Table 1) had any evidence of endothelial layer disruption or smooth muscle cell proliferation. Several of the prominent histopathologic features in these lungs included pulmonary edema, organizing thromboemboli, nonnecrotizing perivascular lymphocytic infiltrates, neutrophilic infiltration, and hyperplasia of bronchial-associated lymphoid tissue (Table 1). In four animals, alveolar proteinosis from Pneumocystis carinii was also evident, suggesting severe immunodeficiency with opportunistic infections.

No measurements of pulmonary artery pressure or cardiac output were taken in the NENPRC early study. On the other hand, examination of echocardiographic data from monkeys infected with the parental SIV strains showed no evidence of right ventricular changes in these monkeys (i.e., no echocardiographic evidence of right ventricular hypertrophy in approximately 30 SIV-infected animals examined [R.P.S., unpublished data]), despite evidence of widespread myocarditis and left ventricular changes.

We then compared the findings in the studies from the NENPRC with those in five additional monkey lungs obtained from the CNPRC (animals designated as CNPRC1–CNPRC5 in Table 1). Animal CNPRC1 had been infected with SIV_mac239 and animal CNPRC2 had been infected with a chimeric SHIV-nef containing the HIV-1 nef_SF33 allele in the SIV_mac239 backbone. CNPRC1 had no involvement of the pulmonary vasculature but showed neutrophilic and lymphocytic infiltration of the perivascular space (Figure 1D). One SHIV-nef–infected monkey (CNPRC2) showed mild vascular remodeling (see Figure E2). Another animal infected with SHIV-nef_SF33 (tissues unavailable) died of immunodeficiency, but its SHIV-nef virus was isolated, recovered, and used to infect three additional macaques (CNPRC3, CNPRC4, and CNPRC5 in Table 1). The nef allele, which had acquired four amino acid mutations during passage in vivo, was designated as Nef _SF33A (A, for adapted) (25). Two animals (CNPRC3 and CNPRC4; Figures 1E and 1F, respectively) showed abnormal vascular lesions with medial hypertrophy and a disrupted, cuboidal endothelial cell layer. Pentachrome staining corroborated both medial and intimal hypertrophy of the pulmonary arteries (see Figure E3). CNPRC5 died of severe wasting and opportunistic infections 3 mo postinfection, with lung samples showing extensive inflammation, but no vascular remodeling (see Figure E2). As with animals in the NENPRC cohort, no evidence of cardiac abnormalities was found in young animals infected with parental SIV_mac251 at the CNPRC (A.F.T., unpublished data).

To further confirm that SIV alone is not associated with complex pulmonary vascular lesions, we examined an additional 12 monkeys infected with an immunosuppressive swarm SIV-ΔB670 and a macrophage-tropic molecular clone SIV/17E-Fr mixture (JHU1–JHU12; Table 1) from the Retrovirus Laboratory, Department of Comparative Medicine, Johns Hopkins University. These macaques developed a rapid and lethal immunodeficiency within 3 mo of infection (26). Although significant inflammatory lung involvement with inflammation and alveolar simplification was observed, no complex vascular lesions were detected (see Figure E4).

Combining the results from the three geographically separate primate centers, four of the six SHIV-nef–infected animals were found to have complex vascular lesions, and one additional animal had evidence of less prominent pulmonary vascular remodeling. Thus, five of the six SHIV-nef–infected animals (83%) with remodeling were chronically infected (more than 12 mo), whereas none of the animals infected with the parental SIV strains containing the SIV nef allele (n = 18) developed complex pulmonary vascular lesions. Taken together, there is a statistically significant association between SHIV-nef infection in macaques and the presence of pulmonary vascular lesions (Fisher's exact test, p < 0.0001).

Characterization of Pulmonary Vascular Lesions in SHIV-nef–infected Animals

To further characterize the lesions, we immunostained the sections with the endothelial cell marker factor VIII, and the smooth muscle markers muscle-specific actin (MSA) and α-smooth muscle actin (Figures 2 and 3). Animals NENPRC7 and NENPRC8 showed marked medial hypertrophy along with a mass of endothelial cells at a site of bifurcation (Figures 2D–2F). The pattern of factor VIII staining (Figures 2B, 2E, 2H, and 2K) highlighted the endothelial cell layer and intimal thickening, and the muscle-specific actin staining pattern (Figures 2C, 2F, 2I, and 2L) indicated medial hypertrophy. In the SHIV-nef animals CNPRC3 and CNPRC4 (Figure 2, rows 3 and 4), the pattern of factor VIII staining highlighted the disrupted, cuboidal endothelial cells, and the muscle-specific actin staining indicated medial hypertrophy. In animal NENPRC7, in a vessel at a bifurcation point, the lesion showed a core of factor VIII–positive endothelial cells obliterating the lumen (Figures 3A and 3B). Furthermore, at higher magnification, the smooth muscle actin–positive cells (Figures 3C and 3D) can be seen as interdigitating strings surrounded by an endothelial cell core.

Presence of Nef in the Lungs of HIV-1–infected Patients and SHIV-nef–infected Macaques

The presence of plexiform lesions in the pulmonary vasculature of SHIV-nef but not in the SIV-infected animals suggested a role for HIV-1 Nef in the pathogenesis of HRPH. As the presence of Nef in the lungs of patients with HRPH has not been reported previously, we examined paraffin-embedded lung sections from a normal human lung, two patients with IPAH, two patients with HRPH, and two SHIV-nef–infected macaques (CNPRC3 and CNPRC4) for the presence of the HIV-1 Nef protein. Figure 4 (first column) shows some representative lung sections stained with a monoclonal HIV-1 Nef antibody, whereas a separate section and region of the same subject in the remaining panels show dual-immunofluorescence stains for factor VIII and Nef. We detected no Nef staining in an uninfected control (Figure 4A; and see Figure E5A) or in patients with IPAH (Figure 4E; and see Figure E5B), illustrating the specificity of the antibody and the absence of any molecular mimicry (29). On the other hand, the lungs from patients with HRPH showed distinct positive staining in alveolar mononuclear cells (most likely lymphocytes and macrophages) as well as in endothelial cells lining the vessels (see Figures E5C and E6). The occluding cells and some lumenal cells found in the complex lesion in the patient with HRPH also stained positive for Nef (Figure 4I), and Nef-positive cells were observed clustering at sites adjacent to the vessel walls. In addition, Nef-positive cells were observed in lymphoid-associated tissues (see Figure E6), suggesting a potential source for the viral protein.

Figures 4 and 5 show immunofluorescence staining for factor VIII and Nef in human and macaque lungs, respectively. The pattern of staining in the merged images clearly indicates that Nef colocalizes with endothelial cells. This pattern in the SHIV-nef macaque lungs recapitulated that found in the human lungs, with staining in several endothelial cells within the vasculature as well as in perivascular sites (Figure 5I). We also found factor VIII-negative, Nef-positive medial and parenchymal cells. Interestingly, some adventitial cells in HRPH stained positive for both factor VIII and Nef (see Figure E6). As expected, lymphoid tissue showed staining for Nef, but no factor VIII positivity (Figure 5). These findings provide additional evidence that HIV-1 Nef may be a critical player in the development of complex pulmonary vascular lesions.