Neuroinvasion of fluorescein-positive monocytes in acute simian immunodeficiency virus infection

Abstract

Monocytes and macrophages play a central role in the pathogenesis of human immunodeficiency virus (HIV)-associated dementia. They represent prominent targets for HIV infection and are thought to facilitate viral neuroinvasion and neuroinflammatory processes. However, many aspects regarding monocyte brain recruitment in HIV infection remain undefined. The nonhuman primate model of AIDS is uniquely suited for examination of the role of monocytes in the pathogenesis of AIDS-associated encephalitis. Nevertheless, an approach to monitor cell migration from peripheral blood into the central nervous system (CNS) in primates had been lacking. Here, upon autologous transfer of fluorescein dye-labeled leukocytes, we demonstrate the trafficking of dye-positive monocytes into the choroid plexus stromata and perivascular spaces in the cerebra of rhesus macaques acutely infected with simian immunodeficiency virus between days 12 and 14 postinfection (p.i.). Dye-positive cells that had migrated expressed the monocyte activation marker CD16 and the macrophage marker CD68. Monocyte neuroinvasion coincided with the presence of the virus in brain tissue and cerebrospinal fluid and with the induction of the proinflammatory mediators CXCL9/MIG and CCL2/MCP-1 in the CNS. Prior to neuroinfiltration, plasma viral load levels peaked on day 11 p.i. Furthermore, the numbers of peripheral blood monocytes rapidly increased between days 4 and 8 p.i., and circulating monocytes exhibited increased functional capacity to produce CCL2/MCP-1. Our findings demonstrate acute monocyte brain infiltration in an animal model of AIDS. Such studies facilitate future examinations of the migratory profile of CNS-homing monocytes, the role of monocytes in virus import into the brain, and the disruption of blood-cerebrospinal fluid and blood-brain barrier functions in primates.

FIG. 1.

Induction of monocytes, plasma viral load, and chemokines in acute SIV infection. (A) Plasma viral loads in SIV-infected animals (animals 29677, 30019, and 32222) were measured by SIV RNA TaqMan analysis between days 0 and 14 p.i. Virus loads peaked on day 11 p.i. in all three animals (shown here for representative animal 29677). Data are expressed as log₁₀ SIV RNA copies/ml. (B) Absolute peripheral blood monocyte numbers per microliter of blood were determined during the first 2 weeks p.i. (shown here for representative SIV-infected animal 29677). The production of the proinflammatory cytokine IFN-γ and the chemokine CXCL9 in plasma was measured by a protein ELISA in triplicate wells (blue, monocytes per microliter of blood; red, CXCL9; asterisks indicate the time point of maximal IFN-γ production and the concentration). (C) Peripheral blood monocytes derived from both animal groups (acutely SIV infected and uninfected macaques) were cultured for 24 h in the presence of rhesus IFN-γ and serum. Monensin was added during the last 6 h of stimulation prior to fixation, permeabilization, and cell staining. The gating strategy for flow cytometry analysis of peripheral blood monocytes is illustrated for representative acutely SIV infected animal 29677. Monocytes were gated based on forward- and side-scatter parameters, excluding smaller lymphocytes and lymphocyte blasts. Monocytes were subsequently gated based on CD14 expression to determine the frequency of CCL2⁺ monocytes (right panel). (D) Bar graph represents a summary of CCL2 flow cytometry experiments and CCL2⁺ cell frequencies measured for all animals (in both the SIV-infected and uninfected groups; n = 3 for each group). Frequencies of CCL2⁺ monocytes are shown for monocytes gated based on size, granularity, and CD14 expression (as shown in panel C). A one-tailed, nonparametric Student t test was used to calculate the P value for differences between animal groups. (open bars, uninfected animals; solid bars; SIV-infected animals).

FIG. 2.

Production of proinflammatory mediators in the CSF and cerebrum in acute SIV infection. (A) Proinflammatory chemokine CXCL9 levels were measured by an anti-human CXCL9 ELISA in the CSF of SIV-infected and uninfected animals at necropsy (day 14 p.i. for the SIV-infected group) in triplicate wells. CXCL9 levels were significantly higher in the CSF of acutely SIV infected animals (solid bars) than in the uninfected group (open bars) (P = 0.01 by unpaired repeated-measures analysis). (B) Relative transcript levels for proinflammatory CXCL9 were determined by RNA TaqMan PCR analysis in cerebral tissue sections of uninfected (animals 30892 and 30893) and acutely SIV infected (animals 29677, 30019, and 32222) macaques following RNA isolation from 30-μm-thick tissue sections and target gene-specific preamplification of cDNA. Transcript levels were normalized based on glyceraldehyde-3-phosphate dehydrogenase transcription, and values were calibrated against the lowest expressed target gene in the uninfected group. ND, not determined; neg, negative (undetectable). (C) Monocyte chemoattractant CCL2 protein levels in plasma (shaded bars) and CSF (solid bars) were determined by an anti-human CCL2 ELISA on day 14 p.i. for the SIV-infected group. Values are expressed in picograms per milliliter and represent data from triplicate wells. CCL2 protein levels were significantly higher in CSF than in plasma for acutely SIV infected animals (P < 0.0001 by paired repeated-measures analysis). (D) Monocyte chemoattractant CCL2 protein levels were also measured in the plasma (shaded bars) and CSF (open bars) of uninfected animals by an anti-human CCL2 ELISA at necropsy. No significant differences in CCL2 levels between plasma and CSF were identified for the uninfected animal group (P = 0.66 by paired analysis). *, the CCL2 level in plasma for animal 30892 was measured from a single well only.

FIG. 3.

Detection of SIV RNA⁺ cells and infiltrating fluorescein dye-labeled monocytes in the choroid plexus and cerebrum. (A and B) SIV RNA⁺ cells (arrows) in SIV-infected animals were detected on day 14 p.i. by ISH of choroid plexus tissue sections using ³⁵S-UTP-labeled riboprobes for SIV gag, pol, and env sequences (shown here for representative animals 29677 [A]and 30019 [B]). Tissue sections were counterstained with hematoxylin. (C) Infiltrating CFSE dye-labeled monocytes/macrophages in the choroid plexus were identified by double-immunofluorescence approaches with antibodies directed against CFSE (fluorescein epitope; indirect detection) and CD16. Secondary detection was performed with Alexa 488-conjugated (fluorescein; green) and Alexa 568-conjugated (CD16; red) antibodies, and sections were stained with the nuclear dye DAPI (blue). Enlarged views are shown on the right. A CD16 single-positive cell (enlarged view, panel 1, arrow) was observed along with several examples of CFSE dye-labeled CD16⁺ double-positive cells (enlarged view, panels 2 to 4, arrows) (shown here for animal 30019; top and center rows, single-color staining with DAPI; bottom row, triple-color overlay). (D and E) SIV RNA⁺ cells were identified in cerebral tissue sections by ISH utilizing methods identical to those for SIV detection in the choroid plexus (A and B). SIV RNA⁺ cells were observed near vessels in the gray matter of the cerebral cortex in representative SIV-infected animals 32222 (D) and 29677 (E). (F and G) Infiltrating CFSE dye-labeled monocytes/macrophages were identified in cerebral tissue sections from acutely SIV infected animals using double immunohistochemistry, shown here for representative SIV-infected animals 32222 (F) and 29677 (G). CFSE dye-labeled cells (dark purple) were visualized using BCIP-NBT substrate, and CD68 (red) was visualized with Fast Red substrate. To enhance the visibility of both substrates in double-positive cells, no counterstain was used on cerebral tissue sections. Arrowheads, CD68 single-positive cells; arrows, double-positive (CFSE dye-labeled, CD68⁺) cells.