HIV-1 Infection Induces Differentiation of Immature Neural Cells through Autocrine Tumor Necrosis Factor and Nitric Oxide Production

Virology 261, 193–204 (1999) Article ID viro.1999.9848, available online at http://www.idealibrary.com on

HIV-1 Infection Induces Differentiation of Immature Neural Cells through Autocrine Tumor Necrosis Factor and Nitric Oxide Production Eva Obrego´n,* Carmen Punzo´n,† Eduardo Ferna´ndez-Cruz,* Manuel Fresno,† and M. Angeles Mun˜oz-Ferna´ndez* ,1 *Division of Immunology, Hospital Universitario Gregorio Maran˜o´n; and †Centro de Biologı´a Molecular, Universidad Auto´noma de Madrid, Madrid, Spain Received January 13, 1999; accepted June 9, 1999 Immature neural cell lines could be productively infected by HIV-1. Interestingly, this infection was associated with a differentiation to a mature neuronal phenotype, characterized by the expression of mature neurofilaments and cell adhesion molecules, intercellular cell adhesion molecule-1, and vascular cell adhesion molecule-1. Infection also induced TNF-a and IL-1b mRNA expression, as well as the synthesis of inducible nitric oxide synthase by neuroblastoma cells. Exogenous addition of TNF-a, but not of IL-1b or many other cytokines, including nerve growth factor, mimicked those effects induced by infection. Moreover, blocking endogenous TNF-a or NO production in cultures of infected cells with a neutralizing anti-TNF-a antibody or inducible nitric oxide synthase inhibitors prevented the expression of the mature cell phenotype as well as expression of intercellular cell adhesion molecule-1 and vascular cell adhesion molecule-1. Addition of NO generators and TNF-a activated NF-kB- and intercellular cell adhesion molecule-1-dependent promoter transcription, whereas inducible nitric oxide synthase inhibitors prevented the transcriptional activation of intercellular cell adhesion molecule-1 promoter that was induced by TNF-a. Those results suggest that HIV can infect immature neural cells and this infection induces their neural development via a TNF-a- and NO-mediated mechanism. © 1999 Academic Press

debate. In particular, it is unclear whether the functional impairment is due to a direct (virus infection) or indirect (immune activation) mechanism or both. Recent results have indicated that viral load of the CNS correlates with neurological symptoms (Cinque et al., 1998; Stefano et al., 1997). In addition, there is a good correlation in the most severe cases of disease with markers of immune activation, such as cytokines, especially TNF-a (DuboisDalcq et al., 1995). TNF-a but not IL-1 or IL-6 mRNA levels are augmented in the brains of ADC patients (Glass and Johnson, 1996; Wesselingh et al., 1994). Moreover, TNF-a levels are elevated in the CSF and serum of AIDS patients and those levels are closely correlated with the severity of dementia (Dubois-Dalcq et al., 1995; Wesselingh et al., 1994). Expression of cell adhesion molecules (CAM) is also up-regulated coincidentally with TNF-a (Obrego´n et al., 1996). According to several authors, clinical manifestations such as neuronal cell loss correlate with macrophage infiltration (Brew et al., 1995) and initial studies suggested that infiltrating mononuclear phagocytes were the major HIV-1-infected cells in the brain (Koenig et al., 1986). However, this hypothesis does not explain the presence of the virus in the brain before any sign of inflammation (An and Scaravilli, 1997). Recent studies have demonstrated that direct infection of the neuronal, astroglial, and microglial elements of the brain takes place in HIV-1-associated dementia. Viral particles have been observed in astrocytes from pediatric AIDS patients

INTRODUCTION HIV-1 infection is often associated with a remarkable array of neurological symptoms, such as demyelinating neuropathies, meningitis, and vacuolar myelopathy peripheral nervous system vasculitis, even in the early asymptomatic phases (reviewed in Glass and Johnson, 1996; Kolson et al., 1998). Among those, the AIDS dementia complex (ADC) is the most abundant. Its pediatric counterpart, commonly called HIV encephalopathy (Belman, 1994), can often be the only clinical manifestation of AIDS in children. Around 30% of adults and 50% of children have neurological impairment due to HIV-1 infection of the central nervous system (CNS). At least 90% of AIDS patients demonstrate neuropathological abnormalities at autopsy (Janssen et al., 1989). Those neuropathological hallmarks have been well described, consisting of encephalitis, reactive microglial cells, infiltration of macrophages, and widespread astrocytosis (Glass and Johnson, 1996). Neuropathological changes also include morphological alteration and even destruction of cortical and basal ganglia neurons (Masliah et al., 1996). However, the pathogenic mechanisms of neurological alteration in HIV-infected individuals are yet a matter of

1 To whom correspondence and reprint requests should be addressed at Hospital General Universitario Gregorio Maran˜on, Servicio de Inmunologia, c/Dr. Esquerdo 47, 28007 Madrid, Spain. Fax: 34-915868018. E-mail [email protected].

193

0042-6822/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

OBREGO´N ET AL.

194

(Epstein et al., 1985). Besides, HIV-1-infected neurons have also been described in the brain from AIDS patients (Budka, 1989; Nuovo et al., 1994; Saito et al., 1994) and primary human neuroblasts (Ensoli et al., 1995) as well as neuronal cell lines can be infected by HIV-1 in vitro (Ensoli et al., 1994; Harouse et al., 1989; Li et al., 1990; Mizrachi et al., 1994). On the other hand, certain viral products can be neurotoxic. Among those, the external gp120 glycoprotein is the most likely candidate, although regulatory proteins, such as Tat, have also been involved (Price, 1996). They may exert neurotoxicity to neurons either directly or indirectly through neurotoxic mediators, such as nitric oxide (NO) and TNF-a (An and Scaravilli, 1997; Chen et al., 1997; Gelbard et al., 1994; Power et al., 1995). NO influences many aspects of CNS physiology, including synaptic plasticity, neuronal development, and behavioral responses (reviewed in Bredt and Snyder, 1994; Mun˜oz-Ferna´ndez and Fresno, 1998). Moreover, a pathological role of NO in the CNS is also evident (Bredt and Snyder, 1994; Mun˜oz-Ferna´ndez and Fresno, 1998). NO is synthesized from L-arginine by the NADPH-dependent enzyme NO synthase (NOS), of which there are three types (reviewed in Knowles and Mocada, 1994; Nathan and Xie, 1994). Isoform I, or nNOS, is Ca 21dependent and it is constitutively present in neuronal and certain epithelial cells. Isoform III, or eNOS, expressed in endothelial cells, is also Ca 21-dependent. In contrast, isoform II, or iNOS, is Ca 21-independent and is inducible by lipopolysaccharide and cytokines, requiring activation of the gene and de novo synthesis of iNOS protein. It is now clear that NOS isoforms are not confined to separate cell lineages, but they may coexist in the same cell type. Besides, recent results have shown an increase in the expression of iNOS mRNA and protein in the brains of severe ADC patients and pediatric patients with advanced HIV encephalitis, which correlated with the severity of the disease (Adamson et al., 1996; Bukrinsky et al., 1995). We have shown here that HIV-1 infection promotes differentiation of human neuroblastoma (NB) cells, as detected by the induced expression of intercellular cell adhesion molecule (ICAM)-1 and vascular cell adhesion molecule (VCAM)-1 as well as mature neurofilament protein (NFP) expression. Moreover, those cells chronically infected with HIV-1 synthesize IL-1 and TNF-a as well as iNOS. Interestingly, this HIV-1-induced NB differentiation is controlled by the autocrine production of TNF-a and NO. RESULTS Infection of neuroblastoma cells by HIV-1 Although neuronal cell lines can be infected in vitro by HIV-1 and infected neurons are detected in vivo, very little is known about the effect of viral infection on the host cells. To study this, tumor cell lines of neuronal

FIG. 1. Infection of NB cell lines by different HIV-1 isolates. SK-N-MC or SK-N-SH NB cells were infected with (0.03 m.o.i.) of the primary isolates HIV LRH, HIV CGP, HIV FGA, as well as with the established strain HIV pNL4.3 or mock infected with heat-inactivated pNL434 (10 m.o.i.). (A) Agp24 was quantified in the supernatants of cultures 0, 3, 6, and 12 days after infection. (B) At 12 days after infection HIV-1 viral DNA was detected in the cell cultures by a nested PCR of the pol region of HIV. Lanes 1–4, NB infected with HIV LRM, HIV CGP, HIV FCD, and HIV pNL43, respectively; lane 5, mock-infected cells; and lane 6, positive control of an infected T cell line.

origin, SK-N-SH and SK-N-MC, representing immature differentiation stages (Gross et al., 1992), were infected with primary and established HIV-1 isolates at a low m.o.i. (0.03) and viral replication was assessed after infection by monitoring p24 viral core antigen production in the cell supernatants at serial time points. A representative experiment is shown in Fig. 1. Both cell lines were productively infected by HIV-1, as shown by increasing Agp24 levels in the culture supernatants with time (Fig. 1A). Moreover, HIV-1 DNA was detected by nested PCR in HIV-1-infected SK-N-SH and SK-N-MC cells (Fig. 1B). Cell viability was not altered after infection, did not change with time, and was always greater than 95% for both cell lines (data not shown). Infection of NB with a higher m.o.i. resulted in a dose response increase in the amount of Agp24 released to the supernatant (data not shown). Between 35 and 50% (depending on the experiment) of the NB cells were infected, as detected by infection with recombinant HIV-1-GFP (expressing green fluorescence protein) (data not shown), in agreement with previously published data on the same NB (Vesanen et al., 1992). Although only performed with four strains, it seemed that viruses with a SI/RH phenotype were much more infectious than those of NSI/SL. To further confirm that in HIV-infected NB cell lines the Agp24 detected represent infectious virus, the highly sensitive MT-2 cell line was cocultured with supernatant from NB-infected cells, harvested 3 days after infection,

HIV INDUCES NEURONAL DIFFERENTIATION VIA TNF AND iNOS

195

FIG. 2. Infection of MT-2 cell line with culture supernatants from HIV-1-infected NB cell lines. MT-2 cell line was incubated with supernatants (200 pg Agp24/ml/10 5 cells) from SK-N-SH or SK-N-MC infected with pNL4.3 strain harvested 12 days after infection. (A) After 24 h, MT-2 cells were washed thoroughly and 0, 3, 6, and 12 days later, virus production was quantified by measuring Agp24 in the supernatant. (B) Syncytia formation 10 days after infection. Panel (1), culture supernatants from SK-N-SH; (2), supernatant from SK-N-MC; and (3), with infectious HIV pNL43 virus.

and tested again for productive infection. As seen in Fig. 2A, high levels of Agp24, which increased over time, were again detected in those cultures, suggesting that infectious virus is produced by NB cells. Moreover, syncytia formation was also detected in MT-2 cultures (Fig. 2B). Taken together the above findings suggest that productive infection can take place in these neuronal cell lines and that infection is not limited to one particular strain of HIV-1. Effect of HIV-1 infection on neuronal phenotype To study the effect of HIV infection on cell differentiation, NB cells were infected with HIV and differentiation was monitored 72 h postinfection through expression on the cell surface of adhesion molecules (ICAM-1 and VCAM-1) by flow cytometry. The higher expression of those CAM is directly related to a more differentiated phenotype (Gross et al., 1992). The results of a representative example are shown in Fig. 3. Infection of SK-N-SH and SK-N-MC human NB lines strongly up-regulated the expression of both cell adhesion molecules, ICAM-1 and VCAM-1, at the cell surface. Thus, 3 days after infection almost all NB cells expressed ICAM-1 compared to

around 40% in uninfected cultures. VCAM-1 expression increased from 15–20% to 40–70%. Moreover, the mean fluorescence intensity of positive cells was also strongly increased. The induction of a differentiation phenotype of NB cells by HIV infection was also evaluated by indirect immunofluorescence assay using monoclonal antibodies specific to mature NFPs of neurons and neuroblasts, 72 h after HIV-1 infection. Changes in the type and amount of intermediate NFP have been shown to associate with neuronal differentiation (Biedler et al., 1988; Liem et al., 1978; Shaw and Weber, 1982). No significant changes were observed after infection in the fluorescence with the 68-kDa NFP, which is expressed in immature and mature neural cells (Fig. 4). By contrast, the expression of 200-kDa NFP, expressed mainly by mature cells, was significantly more pronounced in NB after infection with HIV-1 (compare Figs. 4e and 4f). It is noteworthy that a shift in fluorescence localization from a diffuse somatic to an intense perinuclear pattern was observed in HIV-1-infected NB with both antibodies. In addition, much of the fluorescence was also localized in neuritic processes in differentiated cells (Fig. 4).

196

OBREGO´N ET AL.

FIG. 3. Induction of ICAM-1 and VCAM-1 expression by HIV-1 infection in NB cells. SK-N-SH and SK-N-MC were infected or mock infected with 0.1 m.o.i. of HIV pNL4.3 as described under Materials and Methods. Three days after infection, cells were collected and stained with ICAM-1- or VCAM-1-specific monoclonal antibodies or with an isotype-matched irrelevant antibody. Cell surface expression was detected by flow cytometry. Shown inside the panels are the percentage of positive cells with respect to the irrelevant isotype matched (control) monoclonal antibody.

Role of TNF-a and NO in HIV-induced neuroblastoma differentiation It has previously been shown that some cytokines are excellent inducers of ICAM-1 and VCAM-1 (Springer, 1995). Thus we tested the effect of exogenously added cytokines to uninfected SK-N-SH cells in order to see if they could mimic induction of cell adhesion molecules induced by HIV-1 infection. Of all cytokines tested only TNF-a was able to upregulate ICAM-1 and VCAM-1 (Table 1). Surprisingly, IL-1b, which was effective on inducing CAM in other cell types (Springer, 1995), had no effect. Similar results were obtained with SK-N-MC (data not shown). Since only TNF-a induced the expression of ICAM-1 and VCAM-1 in NB cells, we investigated the possibility that HIV-1 infection in NB caused TNF-a synthesis, which in turn mediated differentiation of SK-N-MC

and SK-N-SH by an autocrine/paracrine mechanism. Results in Fig. 5 indicated that HIV-1 infection augmented TNF-a mRNA in NB cells. IL-1b mRNA was, however, induced by HIV-1 infection (Fig. 5). Previously we have shown that NO was involved in TNF-a-induced NB differentiation (Obrego´n et al., 1997). Therefore, we examined if this mechanism was also taking place during NB differentiation induced by HIV-1 infection. For this, we assayed the NOS enzymatic activity in human NB 72 h postinfection, as determined by measuring the conversion of L-[ 3H]arginine to L-[ 3H]citrulline by a dialyzed cytosolic fraction. NOS activity was not detectable in the cytosol of uninfected NB cells even though Ca 21 was added to the reaction mixture (Fig. 6). However, Ca 21-independent NOS activity was detected in infected NB. Thus, HIV-1 infection induces in NB cells

HIV INDUCES NEURONAL DIFFERENTIATION VIA TNF AND iNOS

197

fication product of 259 bp (Fig. 7). Thus, the differentiation of infected SK-N-MC and SK-N-SH was accompanied by expression of iNOS mRNA. Interestingly, this effect was completely blocked by addition of a neutralizing anti-TNF-a antibody (Fig. 7), suggesting that iNOS was induced by a TNF-a autocrine/paracrine mechanism in these infected NB cells. Furthermore, the adhesion/differentiation process induced by HIV-1 infection in human NB, detected by cell surface expression of ICAM-1 in the infected NB cells, was greatly prevented by the specific iNOS inhibitor, as well as by the addition of a neutralizing anti-TNF-a antibody (Fig. 8). Considering the percentage of positive cells and the mean fluorescence intensity together, both treatments reduced HIV-1-induced CAM expression by more than 90%. The low basal levels of CAM expression in NB were not affected by L-NIL or anti-TNF-a treatment (data not shown). Similar effects with anti-TNF-a antibody and L-NIL were observed in the HIV-1-induced expression of the 200-kDa NFP (see Fig. 4). Most of the effects of HIV-1 on NB cells described above required a productive viral infection. Thus, treatment of NB cultures with azidothymidine (AZT), which blocks viral replication, prevented in a dose response manner ICAM-1 and VCAM-1 upregulation, iNOS induction, as well as recovery of syncytium-inducing infectious virus from NB supernatants in both SK-N-SH (Table 2) and SK-N-MC (data not shown). Furthermore, there was a good correlation between the ability of AZT to block viral replication and the rest of the effects mentioned. Those effects were not due to contaminating mycoplasma or lipopolysaccharide in the viral preparation since control heat-inactivated virus, even at a 100-fold higher m.o.i., induced no up-regulation of ICAM-1 and VCAM-1 (Table 2). FIG. 4. Induction of mature neurofilament expression by HIV-1 infection in NB cells. SK-N-SM and SK-N-MC were mock infected (a, e) or infected with HIV pNL4.3 strain (0.1 m.o.i) in the absence (b, f) or in the presence of (c, g) anti-TNF-a or (d, h) L-NIL. Three days after infection cells were stained with monoclonal antibody against 68-kDa NFP (a–d) or with a polyclonal antibody against 200-kDa NFP (e–h). A representative optical field is shown.

a Ca 21-independent NOS comparable to the inducible form of activated macrophages or hepatocytes (Kowles and Mocada, 1994). iNOS expression in HIV-1-infected NB was confirmed by using RT–PCR with specific primers to detect iNOS mRNA. Uninfected SK-N-SH or SK-N-MC cells did not show any specific product amplified. When those uninfected cells were treated with exogenous TNF-a, a known inducer of iNOS (Nathan and Xie, 1994), an iNOSspecific amplification product of 259 bp was detected. When the cells were infected with HIV, the RT–PCR of isolated mRNA also resulted in an iNOS-specific ampli-

Effect of NO on ICAM-1 transcription To study the above actions at the molecular level, we analyzed the effect of NO and TNF-a in the transcription activity of the ICAM-1 promoter. As shown in Fig. 9, NB cells had a high level of ICAM-1 basal transcription. However, TNF-a induced a dose response increase in the transcriptional activity of a luciferase plasmid under control of the ICAM-1 promoter. Interestingly, addition of the inhibitors of the iNOS, L-NMMA or L-NIL, largely prevented this induction. Moreover, addition of NO generators, such as SNAP or SNP, significantly increased ICAM-1 transcription, although at a lower level than TNF-a (Fig. 9A). ICAM-1 transcription is strongly dependent on the activation of the NF-kB transcription factor (Baeuerle and Henkel, 1994; Stratowa and Audette, 1995). Therefore, we tested the effect of TNF-a and NO in the transcriptional activity of a luciferase plasmid under control of the kB binding sites. Similarly to the results found with ICAM-1 promoter, NB cells had a high level of

OBREGO´N ET AL.

198

TABLE 1 Effect of Cytokines on CAM Expression by Neuroblastoma Cells 4h

12 h

24 h

Cytokine added

ICAM-1

VCAM-1

ICAM-1

VCAM-1

ICAM-1

VCAM-1

None TNF-a IFN-g IL-6 IL-4 IL-10 IL-1b

20 36 18 21 19 21 21

7 20 7 8 10 8 9

21 38 22 22 23 20 19

6 23 6 8 10 10 9

23 46 20 22 20 20 22

9 27 10 10 11 8 9

Note. SK-N-MC cells were incubated with 100 U/ml of the indicated cytokines for 24 h. After 4, 12, or 24 h of incubation, cells were collected and stained with ICAM-1- or VCAM-1-specific monoclonal antibodies or with an isotype-matched irrelevant antibody. Cell surface expression was detected by flow cytometry.

basal transcription. Both TNF-a and NO generators activated NF-kB-dependent transcription in NB cells, having additive effects. Moreover, the presence of iNOS inhibitors completely or partially abrogated the induction of NF-kB activity induced by TNF-a in NB cells (Fig. 9B). DISCUSSION The pathogenic mechanisms of neurological alteration in HIV-infected individuals are yet a matter of debate. In particular, it is unclear whether the functional impairment is due to a direct (virus infection) or indirect (immune activation) mechanism or both. HIV-1-infected neurons have been described in brain from AIDS patients (Budka, 1989; Nuovo et al., 1994; Saito et al., 1994) and primary human neuroblasts (Ensoli et al., 1995) can be infected in vitro. Moreover, recent studies have indicated that direct infection of the neurons may contribute to HIV-1-associated dementia. Thus, neurological damage in AIDS and

FIG. 5. Induction of TNF-a and IL-1b by HIV-1 infection in NB cells. SK-N-MC (lanes 1 and 4), SN-N-SH (lanes 2 and 5), as well as the glioma U-87 (lanes 3 and 6), used as a control, were mock infected with HIV (lanes 1–3) or infected with HIV pNL4.3 strain (0.1 m.o.i) (lanes 4–6). (A) IL-1b mRNA and (B) TNF-a mRNA were analyzed by RT–PCR. Lanes 7, negative controls, and lanes 8, positive controls with mRNA of IL-1b or TNF-a, respectively.

clinical severity correlated with the percentage of neurons infected (Bukrinsky et al., 1995). In favor of a direct infection of neurons is the fact that LTR-driven HIV-1 transcription may take place in neurons but not in astrocytes of LTR transgenic mice, suggesting that developing neurons are fully competent to support HIV-1 replication (Buzy et al., 1995). Furthermore, transgenic mice expressing a full HIV-1 genome in neurons underwent axonal degeneration even in the absence of detectable viral proteins (Thomas et al., 1994), suggesting that restricted expression of the HIV-1 genome may cause degeneration. Thus, direct neurological infection may also contribute to the neurological dysfunction. All these results have raised important questions as to the potential role of restricted neuronal infection in the induction of neuronal cell death and/or cellular dysfunction in ADC. We have tried to address some of those questions using cell lines established from several NBs,

FIG. 6. Induction of iNOS enzymatic activity by HIV-1 infection in NB cells. SK-N-SH and SK-N-MC were mock infected or infected with HIV pNL4.3 strain (0.1 m.o.i.). Infected cells were also treated with TNF-a (300 U/ml). After 3 days, cells were lysed and NOS activity was determined by conversion of L-[ 3H]arginine to L-[ 3H]citrulline in the cytosolic fraction in the presence or the absence of exogenous added Ca 21. Result shown are the mean 6 SEM of two different experiments.

HIV INDUCES NEURONAL DIFFERENTIATION VIA TNF AND iNOS

FIG. 7. Induction of iNOS mRNA by HIV-1 infection in NB cells. SK-N-SH (A) and SK-N-MC (B) were mock infected (lane 1), treated with TNF-a (300 U/ml) (lane 2), infected with HIV pNL4.3 strain (0.1 m.o.i.) (lane 3), or HIV-1 infected in the presence of a neutralizing anti-TNF-a antibody (lane 4). Three days later, iNOS mRNA was detected by RT–PCR with specific primers. Lane 5, positive control.

199

likely representing the in vivo differentiated neuronal stages, providing a useful model for analysis of NB phenotypic differentiation (Gross et al., 1992). SK-N-SH and SK-N-MC human NB lines are relatively undifferentiated cells that can be induced to differentiate toward neuroblasts or Schwannlike mesoectodermal cells by a variety of stimuli (Biedler et al., 1988). Our results indicate that those neuronal cell lines can be productively infected by HIV-1, in agreement with previous results in other NB cells (Ensoli et al., 1994; Harouse et al., 1989; Li et al., 1990; Mizrachi et al., 1994). This was independent of the viral phenotype of the infective virus. However, the level of viral replication was low and did not kill NB cells, suggestive of a chronic restricted viral replication. HIV-1 infectivity was associated with an immature phenotype of the NB cells. This is in agreement with previous

FIG. 8. Prevention of HIV-1-induced ICAM-1 and VCAM-1 expression by anti-TNF-a and iNOS inhibitors. SK-N-MC and SK-N-SH were infected with HIV-1 (HIV pNL4.3 strain, 0.1 m.o.i.) in the absence (control) or in the presence of neutralizing anti-TNF-a antibody or the iNOS inhibitor L-NIL. Three days later cell surface expression was evaluated by flow cytometry as in Fig. 3. The percentages of positive cells with respect to an irrelevant isotype-matched monoclonal antibody are shown inside the panel. Uninfected SK-N-SH cell cultures had 45% positive cells for ICAM-1 and 21% for VCAM-1, whereas SK-N-MC had 41% for ICAM-1 and 22% for VCAM-1.

OBREGO´N ET AL.

200

TABLE 2 The Effects of HIV-1 on NB Cells Are Dependent on Viral Replication

Virus addition

Agp24 formation

None pNL pNL (AZT 10) pNL (AZT 25) pNL (hi)

0 123 6 16 78 6 8 76 3 0

ICAM-1 expression 21 6 4 83 6 12 59 6 8 32 6 8 24 6 3

(p , .005) (p , .01) (ns) (ns)

VCAM-1 expression 17 6 3 39 6 4 30 6 2 21 6 3 23 6 2

(p , .001) (p , .01) (ns) (ns)

Syncytia formation

iNOS induction

2 111 11 2 2

2 1 1 2 2

Note. SK-N-MC or SK-N-SH NB cells were infected with the indicated amount of the primary isolate HIV LRH or with the established strain HIV pNL4.3 or mock infected with heat-inactivated virus (hi). AZT at 25 or 10 mg/ml was added to the cultures where indicated. Three days after infection, cells were collected and stained by flow cytometry. with ICAM-1 and VCAM-1 monoclonal antibodies. The percentages of the mean positive cells 6 SEM with respect to the irrelevant isotype-matched monoclonal antibody of two different experiments are shown. Agp24 was measured in the supernatants of NB at day 3 postinfection. Syncytia formation was measured in MT-2 cells using supernatants of NB cultures 3 days after infection (their relative amount is indicated by a plus or minus sign). iNOS induction was measured by RT–PCR as described under Materials and Methods. The presence or the absence of an amplified specific band in the gel is indicated by a plus or minus sign.

results showing that immature neurons as well as fetal astrocytes are more susceptible to HIV-1 infection (Ensoli et al., 1995; Tornatore et al., 1994; Truckenmiller et al., 1993; Vesanen et al., 1994), suggesting that the immature elements of the developing nervous systems are more susceptible to HIV-1 infection. Interestingly, our results indicate that HIV-1 infection induces differentiation of human NB cells, measured by CAM and mature NFP expression. HIV infection also induced iNOS mRNA, although this was mediated by an autocrine/paracrine TNF-a mechanism. Those effects require productive HIV-1 replication in NB cells. Moreover, our results suggest that HIV-1 infection increases TNF-a, although this cannot be accurately quantified by the RT–PCR technique and induces IL-1b synthesis. The production of TNF-a by HIV-1 infection has been previously described in several hematopoietic cell lines (Fauci, 1996). More importantly, NB differentiation was also controlled by the autocrine production of TNF-a via the production of NO production. Similar to what we found in NB, in human macrophages HIV-1 infection induced iNOS expression (Bukrinsky et al., 1995). Moreover, a direct correlation between endogenous NO production and viral replication was found (Bukrinsky et al., 1995). Recent results have shown an increase in the expression of iNOS mRNA and protein in the brains of severe ADC patients and pediatric patients with advanced HIV encephalitis (Adamson et al., 1996; Bukrinsky et al., 1995). These results indicate that HIV-1 can induce the expression of iNOS in vivo. This enhanced iNOS can be due to direct infection by HIV-1 or mediated by HIV-1 envelope proteins (gp160 or gp41), either directly or indirectly through TNF-a induction. Neurons may be induced to up-regulate NO production by HIV gp120 (Power et al., 1995). Furthermore, gp41 induces iNOS in primary cultures of mixed rat neuronal and glial cells (Adamson et al., 1996). However, we think that this mech-

FIG. 9. Activation of ICAM-1- and NF-kB-dependent transcription by TNF-a and NO. (A) ICAM-1 promoter. SK-N-SH cells were transfected with a reporter plasmid having luciferase under the control of the 1.3-kb upstream region of the human ICAM-1 promoter. (B) NF-kB-dependent transcription. SK-N-SH cells were transfected with a reporter plasmid having luciferase under the control of three kB sites. After transfection, cells were incubated with TNF-a (300 U/ml), SNAP (25 mM), SNP (30 mM) L-NIL (200 mM), and L-NMMA (200 mM) alone or in combination where indicated. Four hours later luciferase activity was determined. Results shown are the mean 6 SEM of two different experiments.

HIV INDUCES NEURONAL DIFFERENTIATION VIA TNF AND iNOS

anism, involving viral proteins such as gp 120 and/or gp41, is unlikely to take place in our system since heatinactivated virus even at high concentrations did not up-regulate iNOS or ICAM-1. Until very recently NO production by neuronal cells was thought to be constitutive in response to neurotransmitters (Knowles and Mocada, 1994; Nathan and Xie, 1994). However, several recent results clearly show that neurones can express iNOS in vitro and in vivo in pathophysiological circumstances (Kifle et al., 1996; Obrego´n et al., 1997; Ogura and Esumi, 1996; Peunova et al., 1996). Our present results confirm previous reports that suggested a physiologic role for iNOS-derived NO in neural differentiation (Obrego´n et al., 1997; Peunova and Enikolopov, 1995; Peunova et al., 1996). Besides, an increase in iNOS immunoreactivity accompanies regenerative processes in peripheral ganglia, further suggesting that NO, the product of iNOS activity, could be involved in nerve regeneration (Magnusson et al., 1996). Taken together those data suggest that iNOS-derived NO may be an important mediator in the induction of neuronal cell differentiation by certain cytokines. On the other hand, we have shown here for the first time that NO induces ICAM-1 transcription. This takes place most likely by activating NF-kB, since it is known that ICAM-1 transcription is dependent on NF-kB activation (Stratowa and Audette, 1995) and basically the same results were obtained with NO generators in NF-kBdependent transcription. Very interestingly, TNF-a-induced ICAM-1- and NF-kB-dependent transcription were almost completely abrogated by blocking iNOS enzymatic activity. This suggests that iNOS, besides being a NF-kB-dependent gene (Baeuerle and Henkel, 1994), is also a downstream regulator of NF-kB, at least in NB cells. In agreement with that role, NO has been shown to modulate NF-kB activity, either positively or negatively (Lander et al., 1993; Peng et al., 1995). In summary, HIV-1 infection can enhance the production of several proinflammatory cytokines, such as TNF-a, which in turn may cause increased iNOS production in infected as well as in neighboring uninfected macrophages, astrocytes, or neurons, altering their functional properties. This NO and TNF-a thus released may cause the astrocytosis and gliosis often observed in HIV neurological disease (Glass and Johnson, 1996; Mun˜ozFerna´ndez and Fresno, 1998) and may contribute not only to degeneration but also to other neuronal dysfunctions, since TNF-a also affects electrophysiological currents on neurons (Ko¨ller et al., 1997). Furthermore, the secretion of TNF-a by HIV-1-infected neural cells may be involved the induction of virus replication in infected cells as well as in neighboring cells in an autocrine/paracrine manner, similar to what takes place in lymphocytes and monocytes (Fauci, 1996; Mun˜oz-Ferna´ndez et al., 1997). Various mechanisms may thus converge on target neurons, where differentiation, as well as direct cytopathic effect,

201

may lead to cellular dysfunction and cell death, which could be responsible for several of the symptoms associated with AIDS dementia. Moreover, our results open new therapeutic possibilities in HIV-1 patients with clinical symptoms of neurologic disease, by using drugs able to neutralize iNOS, as has been used for treatment of autoimmune encephalomyelitis, where TNF-a and iNOS play a central role in pathogenesis. MATERIALS AND METHODS Materials Recombinant human TNF-a was a generous gift of Antibioticos-Pharma (Madrid, Spain). L-N G-monomethylarginine (L-NMMA), D-N G-monomethylarginine (D-NMMA), 6 L-N -(1-imidoethyl)lysine (L-NIL), and S-nitroso-Nacetylpenicillamine (SNAP) were purchased from Alexis Corporation (San Diego, CA). Sodium nitroprusside (SNP) was purchased from Sigma Chemical Co (St. Louis, MO). Monoclonal antibodies against the 68-kDa neurofilament protein (NFP) complex (Mouse-Ascite Fluid, Clone NR4) and against the 200-kDa NFP (developed in rabbit, IgG fraction of antiserum) were purchased from Sigma Chemical Co. Anti-human ICAM-1 and anti-human VCAM-1 monoclonal antibodies were purchased from Ancell, Immunology Research Products (Bayport, U.S.A.). The neutralizing anti-TNF-a monoclonal antibody B13.2 was developed in our laboratory. It does not cross-react with lymphotoxin (TNF-b) and it is able to neutralize the cytotoxic and T cell costimulatory activity of TNF-a (Pimentel-Muin˜os et al., 1994). Cell lines and virus stocks The human NB lines SK-N-SH (a generous gift of Dr. F. Sa´nchez-Madrid. Hospital de la Princesa, Madrid, Spain) and SK-N-MC (from ATCC HTB10) were routinely grown in RPMI 1640 (Biochrom KG Seromed, Berlin, Germany), supplemented with 10% heat-inactivated fetal calf serum (FCS), 1% penicillin/streptomycin, and 2 mM L-glutamine (ICN Pharmaceuticals, Costa Mesa, CA) at 37°C in a humidified atmosphere of 5% CO 2. Three primary HIV-1 isolates (HIV-1 LRH, HIV-1 CGP, HIV1 FGA), isolated by us from HIV-1-infected patients, and a T lymphocyte-adapted strain (HIV-1 pNL4.3, kindly provided by Dr. Alcamı´, Hospital 12 de Octubre, Madrid, Spain) were used. The phenotypes of those HIV-1 isolates are different. They were classified according to: (a) their syncytium inducing (SI) capacity by cocultivating the patient’s PBMC with MT-2 cells and (b) the replicative capacity in PBMC operatively defined as rapid/high (R/H) or slow/ low (S/L) (Mun˜oz-Ferna´ndez et al., 1996). Based on that, the HIV-1 isolates used were HIV-1 LRH (SI, R/H), HIV-1 CGP (SI, S/L), HIV-1 FGA (NSI, S/L), and HIV-1 pNL43, (SI, R/H). Virus stocks were prepared by expanding viral isolates in PBMC and were titrated using the end point dilution

202

OBREGO´N ET AL.

method (Ka¨rber, 1931). Briefly, 10 6 MT-2 cells were infected with the corresponding virus at a m.o.i of 0.1, in a final volume of 5 ml of complete medium. After 3–4 days, when the cells presented 75% syncytia, 30 3 10 6 MT-2 cells in 30 ml of complete medium were added to the culture and incubated again until 75% of syncytia were observed in the culture. The cells were then pelleted at 1200 rpm for 10 min. The virus was further purified by ultracentrifugation (Aoki-Sei et al., 1992) and titrated. The remaining supernatant was used as a negative control in the same experiment. Mock infection was also carried out with heat-inactivated viral preparations. HIV infection of neuroblastoma cells SK-N-SH or SK-N-MC cells were exposed to HIV-1, at different m.o.i. (0.01 to 0.1), depending on the experiment, for 2 h at 37°C. At the end of this period, the culture medium was removed, cells were extensively washed with PBS, and 3 ml of complete medium was added to each well. Virus titers were evaluated in the last washing buffer (time 0). Cell supernatants were harvested every 3 days postinfection to monitor p24 viral core antigen production using an antigen capture (INNOTEST HIV antigen mAb, Innogentic N.V., Belgium). The sensitivity of the assay was at least 4 pg/ml. Viral infection in NB was also detected by polymerase chain reaction (PCR) with a set of nested primers specific for a region of the pol gene (JA17 to JA20) described previously (Mun˜oz-Ferna´ndez et al., 1996). Briefly, the samples were first amplified for 24 cycles with the outer primers, and then 0.1 (5 ml) of the product from the first PCR reaction was amplified for 30 cycles with the inner primers. Only samples that gave positive signals with primers specific for the human housekeeping GAPDH gene were therefore suitable and used for amplification. The PCR product was analyzed by electrophoresis on 1.5% agarose gels stained with ethidium bromide. In parallel cultures, and at the same time after infection, NB cells were removed from the plates by trypsinization, washed, and counted by trypan blue dye exclusion. In all experiments, cell viability was above 95% for both infected cell lines. ICAM-1 and VCAM-1 expression The expression of ICAM-1 and VCAM-1 in NB was evaluated by direct flow cytometry as previously described (Obrego´n et al., 1997). Briefly, cells (1–2 3 10 5/ 100 ml) were incubated in RPMI supplemented with 10% FCS, in the absence or the presence of different stimuli. The cells were recovered by treatment of the monolayers with 0.02% EDTA in phosphate-buffered saline and washed twice in EDTA-containing RPMI medium. They were subsequently incubated with fluorescein isothiocyanate-labeled anti-ICAM-1 and phycoeritrin antiVCAM-1 monoclonal antibodies or with irrelevant fluorescein isothiocyanate or phycoerithrin-labeled mono-

clonal antibodies as negative controls, for 30 min at 4°C. Then the cells were washed in the above buffer and surface fluorescence was determined in a FACS-Scan spectrofluorimeter (Becton–Dickinson). A minimum of 10,000 cells per point were analyzed. Immunofluorescence analysis Cell monolayers, infected or mock infected with HIV-1, were washed four times with PBS and cultured for 48 h with 200 mM L-NIL and 10 mg/ml anti-TNF-a as indicated. After the incubations, cells were fixed with methanol for 15 min at 4°C. The coverslips were extensively washed with PBS and blocked for 15 min at room temperature with 2% bovine serum albumin in PBS. Monoclonal anti68-kDa NFP or rabbit IgG against 200-kDa NFP was added to the coverslips and incubated 45 min at room temperature. After washing once with PBS and twice with 2% bovine serum albumin in PBS, a second immunoglobulin-fluoresceinated rabbit anti-mouse or goat anti-rabbit-rhodaminated immunoglobulin antibody, respectively, was added and incubated for additional 45 min at room temperature. After being washed four times with PBS and once with ethanol, coverslips were mounted in Moviol (Sigma). The preparations were examined in a Zeiss Axioskop microscope and photographed on Kontax. Determination of NOS isoforms Cells were infected or mock infected with HIV-1 and in 3 days the NO synthase activity was measured by monitoring the conversion of L-[ 3H]citrulline as previously described (Obrego´n et al., 1997). Briefly, enzyme assays contained 25 ml of soluble tissue, 50 ml of 100 nM [ 3H]arginine and 1 mM NADPH, and 25 ml of 3 mM CaCl 2. After a 15-min incubation at room temperature, the reaction was stopped by addition of 3 ml of 20 mM HEPES (pH 5.5) containing 2 mM EDTA. The reaction mixtures were then applied to 0.5 ml Dowex AG50WX8 (Na 1 form) columns. [ 3H]Citrulline was quantified by liquid scintillation spectroscopy of the 3-ml flow-through. In addition, determination of NOS isoforms was carried out by RT–PCR. Cells infected or not (10 5/ml) were incubated in RPMI 1640 supplemented with 5% (v/v) FCS in Eppendorf tubes in the presence or the absence of TNF-a. At the indicated times cells were washed in PBS and the pellet was frozen at 270°C until further analysis. The mRNA from 10 5 cells was isolated with oligo(dT)coated magnetic beads and subsequent reverse transcription (PolyAtract Series 9600, Promega Corp., Madison, WI), according to the manufacturer’s instructions. For amplification of the desired cDNA, the following gene-specific primers were used: iNOS sense (59-CCAGAAGCAGAATGTGACCA-39) and iNOS antisense (59-TACATGCTGGAGCCAAGGCCAAA-39) (Geller et al., 1993). The reaction mixture contained 5 ml cDNA (16 of the

HIV INDUCES NEURONAL DIFFERENTIATION VIA TNF AND iNOS

isolated cDNA), 1 mM sense and antisense primers, 200 mM deoxynucleotide triphosphates, and 2.5 U Taq DNA polymerase (Perkin–Elmer) in a final volume of 50 ml. The cycle program was set to denature at 94°C for 45 s, to anneal at 58°C for 45 s, and to extend at 72°C for 2 min for a total of 40 cycles. Electrophoresis of the PCR products was performed on 1.5% agarose gels (FMC Bioproducts, Rockland U.S.A.) containing 1 mg/ml ethidium bromide. Plasmids and transient transfection assays pBHluc 1.3, a luciferase plasmid containing 1.3 kbp of the ICAM-1 59 flanking region (Voraberger et al., 1991), was kindly provided by Dr. C. Stratowa. The reporter pNF-kB-luc expression vector contains three tandem copies of the NF-kB site of the conalbumin promoter driving the luciferase reporter gene and was provided by Dr. J. Alcami (Alcami et al., 1995). SK-N-SH and SK-N-MC cells (5 3 10 5) were plated at near confluence on 60-mm culture dishes and 24 h later were transfixed in OPTIMEM medium (Life Technologies) containing 3 mg of lipofectin (Life Technologies) and 1 mg of plasmid DNA pBHluc 1.3 for 24 h. After removal of the lipofectin-containing transfection mixture, cells were resuspended in completed medium and incubated at 37°C for 24 h. Then transfected cells were exposed to different stimuli for 4 h, and luciferase activity was measured according to the instructions of a luciferase system kit (Promega Corp.). The light emission was measured for 30 s in the luminometer (Monolight 2010, Analitical Luminescence Laboratory). ACKNOWLEDGMENTS This work was supported by grants from Direccio´n General de Investigacio´n Cientı´fica y Te´cnica and Comunidad de Madrid to MAMF Direccio´n General de Investigacı´o´n Cientı´fica y Te´cnica, Fondo de Investigaciones Sanitarias, Comunidad de Madrid, and Fundacio´n Ramo´n Areces to M.F. We thank Dolores Garcia-Alonso, Josefa Gonza´lez-Nicola´s, and Maria Chorro for their excellent technical assistance and Dr. Christan Stratowa of Bender Wien for the ICAM-1 promoter construct.

REFERENCES Adamson, D. C., Wildemann, B., Sasaki, M., Glass, J. D., McArthur, J. C., Christov, V. I., Dawson, T. M., and Dawson, V. L. (1996). Immunologic NO synthase: Elevation in severe AIDS dementia and induction by HIV-1 gp41. Science 274, 1917–1921. Alcami, J., Lain de Lera, T., Folgueira, L., Pedraza, M. A., Jacque, J.-M., Bachelerie, F., Noriega, A. R., Hay, R. T., Harrich, D., Gaynor, R. B., Virelizier, J.-L., and Arenzana-Seisdedos, F. (1995). Absolute dependence on kB responsive elements for initiation and Tat-mediated amplification of HIV transcription in blood CD4 T lymphocytes. EMBO J. 14, 1552–1560. An, S. F., and Scaravilli, F. (1997). Early HIV-1 infection of the central nervous system. Arch. Anat. Cytol. Pathol. 45, 94–105. Aoki-Sei, S., Yarchoan, R., Kageyama, S., Hoekzema, D. T., Pluda, J. M., Wyvill, K. M., Broder, S., and Mitsuya, H. (1992). Plasma HIV-1 viremia

203

in HIV-1 infected individuals assessed bypolymerase chain reaction. AIDS Res. Hum. Retroviruses 8, 1263–1270. Baeuerle, P. A., and Henkel, T. (1994). Function and activation of NF-kB in the immune system. Annu. Rev. Immunol. 12, 141–179. Belman, A. L. (1994). HIV-1-associated CNS disease in infants and children. Res. Publ. Assoc. Res. Nerv. Ment. Dis. 72, 289–310. Biedler, J. L., Spengler, B. A., Chang, T. D., and Ross, R. A. (1988). Differential expression of intermediate filaments and fibronectin in human NBs. Adv. Neuroblastoma Res. 2, 277–289. Bredt, D. S., and Snyder, S. H. (1994). Nitric oxide: A physiologic messenger molecule. Annu. Rev. Biochem. 63, 175–195. Brew, B., Rosenblum, M., Cronin, K., and Price, R. (1995). The AIDS dementia complex and HIV-1 brain infection: Clinical-virological correlations. Ann. Neurol. 38, 563–570. Budka, H. (1989). Human immunodeficiency virus (HIV)-induced disease of the central nervous system: Pathology and implications for pathogenesis. Acta Neuropathol. 77, 225–236. Bukrinsky, M. I., Nottet, H. S., Schmidtmayerova, H., Dubrovsky, L., Flanagan, C. R., Mullins, M. E., Lipton, S. A., and Genderman, H. E. (1995). Regulation of nitric oxide synthase activity in human immunodeficiency virus type 1 (HIV-1)-infected monocytes: Implications for HIV-associated neurological disease. J. Exp. Med. 181, 735–745. Buzy, J. M., Lindstrom, L. M., Zink, M. C., and Clements, J. E. (1995). HIV-1 in the developing CNS: developmental differences in gene expression. Virology 210, 361–371. Chen, P., Mayne, M., Power, C., and Nath, A. (1997). The Tat protein of HIV-1 induces tumor necrosis factor-a production. J. Biol. Chem. 272, 22385–22388. Cinque, P. L., Vago Ceresa, D., Mainini, F., Terreni, M. R., Vagani, A., Torri, W., Bossolasco, S., and Lazzarin, A. (1998). Cerebrospinal fluid HIV-1 RNA levels: Correlation with HIV encephalitis. AIDS 12, 389– 394. Dubois-Dalcq, M., Altmeyer, R., Chiron, M., and Wilt, S. (1995). HIV interactions with cells of the immune system. Curr. Opin. Neurobiol. 5, 647–655. Ensoli, F., Cafaro, A., Fiorelli, V., Vannelli, B., Ensoli, B., and Thiele, C. J. (1995). HIV-1 infection of primary human neuroblasts. Virology 210, 221–225. Ensoli, F., Ensoli, B., and Thiele, C. J. (1994). HIV-1 gene expression and replication in neuronal and glial cell lines with immature phenotype: effects of nerve growth factor. Virology 200, 668–676. Epstein, L. G., Sharer, L. R., Joshi, V. V., Fojas, M. M., Koenigsberger, M. R., and Oleske, M. (1985). Progressive encephalopathy in children with acquired immunodeficiency syndrome. Ann. Neurol. 17, 488– 496. Fauci, A. S. (1996). Host factors and the pathogenesis of HIV-induced disease: Implications for therapy. Science 384, 529–534. Gelbard, H. A., Nottet, H. S., Swindells, S., Jett, M., Dzenko, K. A., Genis, P., White, R., Wang, L., Choi, Y. B., Zhang, D., Lipton, S. A., Tourtellotte, W. W., Epstein, L. G., and Gendelman, H. E. (1994). Platelet-activating factor: A candidate human immunodeficiency virus type 1-induced neurotoxin. J. Virol. 68, 4628–4235. Geller, D. A., Nussler, A. K., and Di Silvio, M. (1993). Cytokines, endotoxin, and glucocorticoids regulate the expression of inducible nitric oxide synthase in hepatocytes. Proc. Natl. Acad. Sci. USA 90, 522– 526. Glass, J. D., and Johnson, R. T. (1996). Human immunodeficiency virus and the brain. Annu. Rev. Neurosci. 19, 1–26. Gross, J., Favre, B., and Meyer, M. (1992). Differentiation-related expression of adhesion molecules and receptors on human neuroblastoma tissues, cell lines and variants. Int. J. Cancer 52, 85–91. Harouse, J. M., Kunsch, C., Hartle, H. T., Laughlin, M. A., Hoxie, J. A., Wigdahl, B., and Gonzalez-Scararano, F. (1989). CD4-independent infection of human neural cells by human immunodeficiency virus type 1. J. Virol. 63, 2527–2533. Janssen, R. S., Cornblath, D. R., Epstein, L. G., McArthur, J., and Price, R. W. (1989). Human immunodeficiency virus (HIV) infection and the

204

OBREGO´N ET AL.

nervous system: Report from the American Academy of Neurology, AIDS task force. Neurology 39, 119–122. Ka¨rber, G. (1931). Beitrage¨ zur kollektiven behandlung pharmalologischer reihenversuche. Arch. Exp. Pathol. Pharmakol. 162, 480–483. Kifle, Y., Monnier, J., Chesrown, S. E., Raizada, M. K., and Nick, H. S. (1996). Regulation of the manganese superoxide dismutase and inducible nitric oxide synthase gene in rat neuronal and glial cells. J. Neurochem. 6, 2128–2135. Knowles, R. G., and Moncada, S. (1994). Nitric oxide synthases in mammals. Biochem. J. 298, 249–259. Koenig, S., Gendelmam, H. E., Orenstein, J. M., Canto, M. C. D., Pezeshkpour, G. H., Yungbluth, M., Janotta, F., Aksamit, A., Martin, M. A., and Fauci, A. S. (1986). Detection of AIDS virus in macrophages in brain tissue from AIDS patients with encephalopathy. Science 233, 1089– 1093. Ko¨ller, H., Siebler, M., and Hartung, H. P. (1997). Immunologically induced electrophysiological dysfunction: Implications for inflammatory diseases of the CNS and PNS. Prog. Neurobiol. 52, 1–26. Kolson, D. L., Lavi, E., and Gonzalez-Scarano, F. (1998). The effects of human immunodeficiency virus in the central nervous system. Adv. Virus. Res. 50, 1–47. Lander, H. M., Sehajpal, P., Levine, D. M., and Novogrodsky, A. (1993). Activation of human peripheral blood mononuclear cells by nitric oxide-generating compounds. J. Immunol. 150, 1509–1516. Li, X. L., Moudgil, T., Vinters, H. V., and Ho, D. D. (1990). CD4-independent, productive infection of a neuronal cell line by human immunodeficiency virus type 1. J. Virol. 64, 1383–1387. Liem, R. K. H., Yen, S. H., Salomon, G. D., and Shelanski, M. L. (1978). Intermediate filaments in nervous tissues. J. Cell Biol. 79, 637–645. Magnusson, S., Almand, P., and Kanje, M. (1996). Inducible nitric oxide synthase increases in regenerating rat ganglia. Neuroreport 7, 2046– 2050. Masliah, E., Ge, N., and Mucke, L. (1996). Pathogenesis of HIV-1 associated neurodegeneration. Crit. Rev. Neurobiol. 10, 57–67. Mizrachi, Y., Rodriguez, I., Sweetnam, P. M., Rubinstein, A., and Volsky, D. J. (1994). HIV type 1 infection of human cortical neuronal cells: Enhancement by select neuronal growth factors. AIDS Res. Hum. Retroviruses 10, 1593–1596. Mun˜oz-Ferna´ndez, M. A., and Fresno, M. (1998). The role of tumor necrosis factor, interleukin 6, interferon-g, and inducible nitric oxide synthase in the development and pathology of the nervous system. Prog. Neurobiol. 56, 307–340. Mun˜oz-Ferna´ndez, M. A., Obrego´n, E., Navarro, J., Bo¨rner, C., Gurbindo, M. D., Sampelayo, T. H., and Fernandez-Cruz, E. (1996). Relationship of virologic, immunologic, and clinical parameters in infants with vertically acquired human immunodeficiency virus type 1 infection. Pediatr. Res. 40, 597–602. Mun˜oz-Ferna´ndez, M. A., Navarro, J., Garcia, A., Punzo´n, C., Ferna´ndezCruz, E., and Fresno, M. (1997). Replication of human immunodeficiency virus-1 in primary human T cells is dependent on the autocrine secretion of tumor necrosis factor through the control of nuclear factor-NF-kB activation. J. Allergy Clin. Immunol. 100, 838–845. Nathan, C., and Xie, Q.-W. (1994). Nitric oxide synthases: Roles, tolls, and controls. Cell 78, 915–918. Nuovo, G. J., Gallery, F., MacConnell, P., and Braun, A. (1994). In situ detection of polymerase chain reaction-amplified HIV-1 nucleic acids and tumor necrosis factor-a RNA in the central nervous system. Am. J. Pathol. 144, 659–666. Obrego´n, E., Bo¨rner, C., Navarro, J., Gurbindo, M. D., Fe´rnandez-Cruz, E., and Mun˜oz-Ferna´ndez, M. A. (1996). Elevated levels of circulating intercellular adhesion Molecule 1 and tumor necrosis factor a in the serum of children vertically infected with HIV-1. Pediatr. AIDS HIV Infect. 7, 413–417. Obrego´n, E., Punzo´n, C., Gonzalez-Nicola´s, J., Ferna´ndez-Cruz, E., Fresno, M., and Mun˜oz-Ferna´ndez, M. A. (1997). Induction of adhesion/differentiation of human neuroblastoma cells by tumour necro-

sis factor-alpha requires the expression of an inducible nitric oxide synthase. Eur. J. Neurosci. 9, 1184–1193. Ogura, T., and Esumi, H. (1996). Nitric oxide synthase expression in human neuroblastoma cell line induced by cytokines. Neuroreport 7, 853–856. Peng, H. B., Libby, P., and Liao, J. K. (1995). Induction and stabilization of I kappa B alpha by nitric oxide mediates inhibition of NF-kappa B. J. Biol. Chem. 270, 14214–14219. Peunova, N., and Enikolopov, G. (1995). Nitric oxide triggers a switch to growth arrest during differentiation of neuronal cells. Nature 375, 68–73. Peunova, N., Kuzin, B., Roberts, I., O’Kane, C., and Enikolopov, G. (1996). Nitric oxide, cell multiplication and cell survival. Cold Spring Harbor Symp. Quant. Biol. 56, 417–426. Pimentel-Muin˜os, F. X., Mun˜oz-Ferna´ndez, M. A., and Fresno, M. (1994). Control of T lymphocyte activation and interleukin 2 receptor expression by endogenously secreted lymphokines. J. Immunol. 152, 5714– 5722. Power, C., McArthur, J. C., Johnson, R. T., Griffin, D. E., Glass, J. D., Dewey, R., and Chesebro, B. (1995). Distinct HIV-1 env sequences are associated with neurotropism and neurovirulence. Curr. Top. Microbiol. Immunol. 202, 89–104. Price, R. W. (1996). AIDS dementia complex: a complex, slow virus “model” of acquired genetic neurodegenerative disease. Cold Spring Harbor Symp. Quant. Biol. 61, 759–770. Saito, Y., Sharer, L. R., Epstein, L. G., Michaels, J., Mintz, M., Louder, M., Golding, K., Cvetkovich, T. A., and Blumberg, B. M. (1994). Overexpression of nef as a marker for restricted HIV-1 infection of astrocytes in postmortem pediatric central nervous tissues. Neurology 44, 474–481. Shaw, G., and Weber, K. (1982). Differential expression of neurofilament triplet proteins in brain development. Nature 298, 277–279. Springer, T. A. (1995). Traffic signals on endothelium for lymphocyte recirculation and leukocyte emigration. Annu. Rev. Physiol. 57, 827– 872. Stefano, M., Monno, L., Ramon Fiore, J., and Angarano, G. (1997). In vivo evidence of HIV-1 productive infection in cerebrospinal fluid of patients with HIV-1 encephalopathy. AIDS 11, 698–699. Stratowa, C., and Audette, M. (1995). Transcriptional regulation of the human intercellular adhesion molecule-1 gene: a short overview. Immunobiology 193, 293–304. Thomas, F. P., Chalk, C., Lalonde, R., Robitaille, Y., and Jolicoeur, P. (1994). Expression of human immunodeficiency virus type 1 in the nervous system of transgenic mice leads to neurological disease. J. Virol. 68, 7099–7107. Tornatore, C., Chandra, R., Berger, J. R., and Major, E. O. (1994). HIVinfection of subcortical astrocytes in the pediatric central nervous system. Neurology 44, 481–487. Truckenmiller, M. E., Kulaga, H., Coggiano, M., Wyatt, R., Snyder, S. H., and Sweetnam, P. (1993). Human cortical neuronal cell line: a model for HIV-1 infection in an immature neuronal system. AIDS Res. Hum. Retroviruses 9, 445–453. Vesanen, M., Salminen, M., Wessman, M., Lankinen, H., Sistonen, P., and Vaheri, A. (1994). Morphological differentiation of human SHSY5Y neuroblastoma cells inhibits human immunodeficiency virus type 1 infection. J. Gen. Virol. 75, 201–206. Vesanen, M., Wessman, M., Salminen, M., and Vaheri, A. (1992). Activation of integrated immunodeficiency virus type 1 in human neuroblastoma by the cytokines tumor necrosis factor alpha and interleukin-6. J. Gen. Virol. 73, 173–1760. Voraberger, G., Schafer, R., and Stratowa, C. (1991). Cloning of the human gene for intercellular adhesion molecule 1 and analysis of its 59-regulatory region: Induction by cytokines and phorbol ester. J. Immunol. 147, 2777–2786. Wesselingh, S. L., Glass, J., McArthur, J. C., Griffin, J. W., and Griffin, D. E. (1994). Cytokine dysregulation in HIV-associated neurological disease. Adv. Neuroimmunol. 4, 199–206.