Evidence of blood-brain barrier alteration and activation in HIV-1 gp120 transgenic mice

Objective: To verify whether HIV envelope protein gp120 changes the blood-brain barrier in vivo, as a fundamental mechanism of early central nervous system damage by HIV-1.

Design: Analysis of the functional integrity and immune activation of the blood-brain barrier in brains of HIV-1 gp120 transgenic mice secreting circulating gp120 at levels similar to those detected in AIDS patients.

Methods: Number of vessels/mm2 section area with perivascular albumin and percentage of vessels expressing adhesion molecules (ICAM-1 and VCAM-1) were determined by immunohistochemistry in frozen brains from autopsied transgenic and non-transgenic mice. The percentage of vessels showing substance P immunoreactivity was also calculated, as this neuropeptide is known to mediate the increase in permeability of the rat brain endothelium in vitro caused by HIV-1 gp120.

Results: The number of vessels with albumin extravasation was significantly higher in transgenic than non-transgenic mice brains (P = 0.0003). A greater percentage of ICAM-1- and VCAM-1-positive brain vessels in transgenic than non-transgenic mice was shown (P = 0.0017 and P = 0.0008 respectively). Significant immunoreactivity for substance P was detected in brain vessels in transgenic mice and a significant correlation was found between the percentage of substance P-positive and ICAM-1-positive brain vessels (P < 0.0001) in transgenic mice.

Conclusions: These findings demonstrate that HIV-1 gp120 is capable of changing and activating in vivo the vascular component of the blood-brain barrier.

Introduction

The central nervous system is an early target of HIV-1 infection. Intrathecal synthesis of anti-HIV-1 antibodies has been demonstrated in neurologically asymptomatic seropositive patients [^1,2] and in infected chimpanzees with no neurological signs[³]. However, the mechanism underlying the early invasion of nervous system by HIV-1 remains obscure. In the last few years, several pathology studies on brain tissues of HIV-positive autopsy cases have provided evidence of possible impairment of the blood-brain barrier in the early stages of infection without neurological signs[^4,5]. We recently demonstrated that recombinant gp120, in amounts similar to those detected in HIV-1-positive patients, increased the permeability of rat brain endothelium cultures, a representative model of the vascular component of the blood-brain barrier. This functional impairment is caused by dose-dependent secretion of substance P by the endothelium and its subsequent binding to the brain endothelial cell surface. The resulting changes in cell morphology allow the passage of macromolecules and possibly even cells[⁶]. These findings suggest that an HIV-1 envelope protein shed by the cell surface as a result of high virus turnover [⁷] impairs the blood-brain barrier before viral invasion of the central nervous system. To test whether gp120 also has this role in vivo, we performed an immunohistochemical study of brain tissue of transgenic mice for gp120, secreting circulating levels of gp120 similar to those detected in AIDS patients[⁸]. We demonstrated alteration and further activation of the blood-brain barrier in this model, confirming that gp120 impairs the blood-brain barrier in vivo.

Materials and methods

Transgenic mice

Transgenic mice were obtained by microinjection of a DNA construct derived by inserting a 1557 bp SalI fragment encoding HIV gp120 (nucleotides 1633-3190) into the XhoI site of the Thy-1Eμ gene cassette as described previously [⁸] into pronuclei of fertilized eggs derived from C57BL/6 3 CB6F1 matings. All lines generated expressed transgene mRNA in thymus and brain. Secretion of gp120 protein was checked by analysing mice serum samples by Western blot and capture ELISA as described previously[⁹].

Northern blot analysis

Northern blot analysis of gp120 mRNA was performed on 10 μg of total RNA extracted from thymus and brain of transgenic animals from three different lines; non-transgenic littermates were used as a control. RNA was transferred to a nylon membrane (Biorad, Hercules, California, USA) and hybridized as recommended by the manufacturer using the 32P-labelled 1.5 kb SalI fragment of the gp120 coding region as probe. Exposure was to Kodak XOmat film, overnight using an intensifying screen.

Immunohistochemical studies

Ten brains from 3-month-old transgenic mice and 10 from non-transgenic mice were analysed. The mice were killed under anaesthesia with chloral hydrate (40 mg/100 g body weight) and the brains immediately removed and frozen in liquid nitrogen and isopentane; 8-10 μm sections were cut and mounted on polylysine-coated slides. At least three sections for each brain were studied blindly.

For perivascular albumin analysis - as a reliable parameter of blood-brain barrier function - sections were fixed with frozen acetone for 10 min at room temperature, washed with phosphate buffered saline (PBS) pH 7.4 and overlaid with methanol/0.3% H2O2 (to block endogenous peroxidase) for 20 min. After washing, sections were incubated with 10% normal goat serum (NGS) in PBS pH 7.4 to block non-specific antibody binding, then with polyclonal rabbit anti-albumin antiserum (Dakopatts, Copenhagen, Denmark) diluted 1 : 800 in PBS/2% NGS for 2 h at room temperature. The sections were washed and overlaid with biotin-conjugated goat anti-rabbit IgG (Boehringer Mannheim, Mannheim, Germany; 1 : 250 in PBS/2% NGS) for 1 h at room temperature, washed again and overlaid with peroxidase-conjugated streptavidin (Boehringer Mannheim; 500 mU/ml in PBS/2% NGS) for 30 min at room temperature. Colour was developed with the substrate diaminobenzidine (Sigma, St. Louis, Missouri, USA; 1 mg/ml 0.05 M Tris-HCl pH 7.6) containing 0.03% H2O2 and stopped with PBS pH 7.4. Sections were counterstained with Mayer‚s hematoxylin, dehydrated with ethanol and xylol and finally mounted under a coverslip. The number of vessels showing albumin extravasation was determined by visual scanning of albumin immunoreactivity with a square grid of area 1 mm2 (Zeiss, Oberkochen, Germany) and expressed as number per mm2 section area. For substance P analysis, sections were treated as for albumin except that polyclonal rabbit anti-substance P antiserum (UCB-Bioproducts, Braine-l‚Alleud, Belgium) was used as primary antibody. Substance P-positive vessels were expressed as a percentage of the total number of vessels in each section.

For ICAM-1 analysis biotin conjugated hamster anti-mouse CD54 (ICAM-1) monoclonal antibody (Pharmingen, San Diego, California, USA; 4 μg/ml in PBS pH 7.4/2% NGS) was used as primary antibody followed by peroxidase-conjugated streptavidin. Vessels expressing moderate or strong ICAM-1 immunoreactivity were considered positive and ICAM-1-positive vessels were expressed as a percentage of total vessels in the section. For VCAM-1 analysis biotin-conjugated-rat anti-mouse CD106 (VCAM-1) (Pharmingen; 4 μg/ml PBS pH 7.4/2% NGS) was used as primary antibody and the percentage of VCAM-1-positive vessels was calculated as above.

For immunohistochemical detection of gp120 the sections were processed as for albumin, then biotin-conjugated goat anti-gp120 polyclonal antibody (Biogenesis, Poole, UK; 5 μg/ml in 2% NGS/1% BSA) was added for 2 h at room temperature followed by peroxidase-conjugated streptavidin and diaminobenzidine as substrate. For all markers studied, a negative control was made by omitting primary antibody with NGS and processing sections as for positive specimens.

Statistical analysis

Analysis of the mean values was performed by the non-parametric Mann-Whitney test. Correlations were analysed by calculating Pearson‚s r coefficient. P values < 0.05 were considered significant.

Results

Gp120 levels and mRNA expression

Serum circulating gp120 levels of transgenic mice are given in Table 1. HIV-1 gp120 mRNA expression in brain and thymus of transgenic mice is shown in Fig. 1. High levels of RNA were expressed in thymus and brain of transgenic mice, but not in non-transgenic brain. However, no significant gp120 immunoreactivity was found on transgenic brain vessels or in other cells (data not shown).

Table 1

Fig. 1

Albumin permeability

Transgenic mice brains showed wide albumin extravasation around large and small vessels, indicating extensive impairment of the blood-brain barrier function (Fig. 2a and b). The number of vessels showing albumin leakage was significantly higher in transgenic than in non-transgenic mouse brains (P =0.0003) (Fig. 3a). To verify whether perivascular albumin could be secondary to non-specific post-mortem extravasation, we performed immunohistochemical analysis of albumin in brains of 6 control normal BALB/C mice (Morini, S. Polo Denza, Italy) killed and processed in the same way as for transgenic mice. No immunoreactivity was found around either large or small vessels (data not shown).

Fig. 2

Fig. 3

Substance P immunoreactivity

Strong immunostaining for substance P was found in many brain vessels of transgenic mice (Fig. 2c and d). The percentage of substance P-positive vessels was higher in transgenic than in non-transgenic mice (P = 0.002) (Fig. 3b).

Adhesion molecules

Surface expression of adhesion molecules involved in the recruitment and transendothelial migration of peripheral blood lymphocytes and monocytes were analysed to verify whether blood-brain barrier impairment was associated with immunological activation of brain endothelium. Strong immunostaining for ICAM-1 (Fig. 2e and f) and moderate immunoreactivity for VCAM-1 (Fig. 2g and h) was observed on endothelial cells in transgenic mice brains. However, the percentage of ICAM-1-positive and VCAM-1-positive brain vessels was significantly higher in transgenic than in non-transgenic brains (P = 0.0017 and P = 0.0008 respectively; Fig. 3c and d).

Correlations

There was no correlation between circulating gp120 levels and number of vessels showing perivascular albumin (r = 0.32; P = 0.36) and percentage of substance P- positive brain vessels (r = 0.16; P = 0.65) or ICAM-1-positive (r = 0.05; P = 0.89) or VCAM-1-positive brain vessels (r = 0.58; P = 0.08) in transgenic mice. In contrast, in transgenic mice a very significant correlation was found between percentage of substance P-positive and ICAM-1-positive (r = 0.97; P < 0.0001) (Fig. 4) but not VCAM-1-positive brain vessels (r = -0.03; P = 0.93).

Fig. 4

Discussion

The results of this study show that mice transgenic for HIV-1 gp120 have a substantially impaired and activated blood-brain barrier. The significance of this transgenic model is highlighted by the finding that mice expressing gp120 mRNA in the brain and thymus, secreted circulating levels of gp120 similar to those detected in HIV-positive patients at different stages of infection[¹⁰]. The lack of correlation between circulating levels of gp120 and blood-brain barrier changes in our transgenic model may be explained by the fact that circulating levels do not constitute all of the gp120 produced by transgenic mice; intracellular molecules have to be considered. In addition, the absence of gp120 immunoreactivity in transgenic brains is consistent with a similar finding obtained previously in another neuronal transgenic model for gp120[¹¹]. However, it has been found that in brain tissue, neurotoxic fragments of gp120 may be generated making undetectable intact soluble molecule[¹²]. The increased permeability to albumin was found to be associated with substance P immunoreactivity at the endothelial cell surface, confirming in vivo our recent demonstration that spantide, a potent substance P antagonist, and anti-substance P antibody block increased albumin permeability and related morphological changes in rat brain endothelium cultures stimulated with gp120[⁶]. This implies that substance P plays a role in these alterations. Substance P bound to endothelium, is secreted by the endothelium itself as demonstrated by our recent observation that proinflammatory cytokines stimulated brain endothelium cultures to produce and secrete substance P, that in turn bound to endothelial cell surface by triggering an autocrine circuit[¹³]. It is therefore conceivable that gp120 acts as a proinflammatory cytokine on the blood-brain barrier, stimulating substance P production and changing endothelium function. The strong correlation between substance P and ICAM-1 expression on transgenic brain vessels is consistent with the capacity of substance P to increase lymphocyte-cell adhesion in vitro, to recruit T cells through the interaction of ICAM-1 with its ligand LFA-1 and to upregulate ICAM-1 expression on endothelial cells[¹⁴]. The lack of correlation between substance P and VCAM-1 expression in transgenic brain vessels, however, suggests that substance P does not induce up-regulation of this adhesion molecule. Increased ICAM-1 and VCAM-1 expression in brain endothelial cells co-cultured with HIV-infected monocytes and in the brain vessels of AIDS autopsy cases [¹⁵] recently suggested that adhesion molecules play a part in the recruitment of HIV-1-infected monocytes. In the early stages of HIV-1 infection, the viral load in the central nervous system remains low, as demonstrated by recent studies performed with sensitive molecular biology techniques. Indeed, significantly more HIV-1 DNA copies and HIV-1 RNA have been found in brains of AIDS patients with or without dementia than in asymptomatic HIV-1-positive subjects[^16,17]. This raises the possibility that early impairment of the blood-brain barrier facilitates the passage of macromolecules and neurotoxic substances produced by HIV-1-infected monocytes prior to direct entry of the virus into the central nervous system. This would explain the cognitive and neurological abnormalities that precede the clinical manifestations of AIDS in 20% of cases[¹⁸]. Our findings provide further evidence that soluble gp120 plays a part in inducing alterations of the blood-brain barrier in vivo with no direct participation of virus. These transgenic mice could be a reliable experimental model for examining the role of substance P in gp120-induced blood-brain barrier impairment and for formulating new therapeutic strategies to prevent AIDS-related neurological disease.