Skip to main content
PMC home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2006 Dec 5;103(50):19182–19187. doi: 10.1073/pnas.0604678103

Proteolytic processing of SDF-1α reveals a change in receptor specificity mediating HIV-associated neurodegeneration

David Vergote *,†, Georgina S Butler ‡, Martine Ooms *, Jennifer H Cox §, Claudia Silva †, Morley D Hollenberg , Jack H Jhamandas *, Christopher M Overall ‡,§, Christopher Power *,†,‖
PMCID: PMC1748196  PMID: 17148615

Abstract

Proteolytic cleavage of constitutively expressed proteins can generate peptides with novel bioactive properties. Matrix metalloproteinase (MMP)-2 cleaves the 4 amino-terminal residues of the chemokine, stromal cell-derived factor (SDF)-1α, yielding a highly neurotoxic molecule, SDF(5-67), which fails to bind to its cognate receptor, CXCR4. Herein, we detected SDF(5-67) in brain monocytoid cells of HIV-infected persons, particularly in those with HIV-associated dementia. SDF(5-67) activated cell type-specific expression of proinflammatory genes including IL-1β, TNFα, indoleamine 2′,3′-dioxygenase (IDO), and IL-10 in both astrocytic and monocytoid cells (P < 0.05). Unlike SDF-1α, SDF(5-67) caused neuronal membrane perturbations with ensuing neurotoxicity and apoptosis (P < 0.05) through engagement of an inducible receptor. CXCR3 antagonists and siRNA-mediated knockdown of CXCR3 inhibited SDF(5-67)-stimulated neurophysiological changes, neuronal death, and neuroimmune activation (P < 0.05). Moreover SDF(5-67) bound directly to CXCR3 in a competitive manner, mediated by its amino terminus. In vivo neuroinflammation, neuronal loss, and neurobehavioral abnormalities caused by SDF(5-67) (P < 0.05) were prevented by a CXCR3 antagonist. These studies reveal additive neuropathogenic properties exerted by a proteolytically cleaved chemokine as consequences of a change in receptor specificity, culminating in neurodegeneration.

Keywords: apoptosis, chemokine, human immunodeficiency virus, neuron, stromal cell-derived factor-1α

In the brain, chemokines play pivotal roles in the formation and maintenance of the neuronal network architecture during development, in trophic support of neurons and modulation of synaptic transmission, together with controlling cellular navigation (see (1) for review). For instance, stromal cell-derived factor (SDF)-1/CXCL12, a constitutively expressed chemokine in the brain, is vital for normal growth and development within the nervous system (2). Chemokines are also implicated in the pathogenesis of numerous neurodegenerative diseases with both intrinsic causes, such as multiple sclerosis or Alzheimer's disease (3, 4), or extrinsic/infectious causes including prion diseases or HIV-associated dementia (HAD) (5, 6). Moreover, it is widely recognized that chemokines released by activated glial cells act as neuroimmunomodulatory factors (6, 7). Several proteases have been reported to process specific chemokines, cleaving either their amino- or carboxyl-terminal ends with ensuing altered structure and function, leading to the highly topical area of chemokine degradomics (8–10). Proteolytic cleavage of the amino-terminal end preceding the two first conserved cysteinyl residues of a chemokine can specifically change a chemokine's actions, leading either to the activation or inactivation of the chemokine (11, 12), a switch from a receptor agonist to an antagonist (13, 14) or an increase in chemokine affinity for its respective receptor (15, 16).

Earlier studies reported that the chemokine stromal cell-derived factor (SDF)-1/CXCL12 undergoes precise proteolytic cleavage by matrix metalloproteinase (MMP)-2, removing the four amino-terminal residues from the full-length molecule (17, 18). Of importance, MMP-2 is induced in macrophages after HIV-1 infection and was associated with the development of HIV-induced neurodegeneration (19). In fact, activation and proliferation of monocytoid cells in the brain is assumed to be a key determinant of HIV neuropathogenesis (20). The proteolytic product, SDF(5-67), shows no chemotactic properties but acquired selective neurotoxic effects (18). Although the cognate receptor for SDF(5-67) was unknown, we showed that its effects were mediated by a G protein-coupled receptor, differing from SDF-1α's well characterized cognate receptor, CXCR4 (17, 18). Herein, we demonstrate that SDF(5-67) is highly immunogenic contributing to neurodegeneration through engagement of the chemokine receptor, CXCR3.

Results

SDF(5-67) Is Increased in HIV-Infected Brains and Is Immunogenic.

Although an interaction initiated by HIV-1 infection gave rise to proteolytic cleavage of SDF-1α by MMP-2 resulting in the production of a cleaved form of SDF-1α, SDF(5-67), which was highly neurotoxic to human and mouse neurons (18), SDF(5-67)'s expression in the nervous system was unknown. Hence, we examined SDF(5-67) abundance in the brains of HIV-1 infected or uninfected patients using an antiserum showing specific recognition of SDF(5-67) [supporting information (SI) Fig. 5]. SDF(5-67) immunoreactivity in brain (52 kDa) was associated with HIV-1 infection and particularly with the diagnosis of HAD (Fig. 1a). In parallel, an increase of the immunoreactivity detected with an antibody recognizing full length SDF-1α (47 kDa) was observed in the same samples (Fig. 1a). The higher molecular weight of SDF(5-67) and SDF-1α immunoreactivities compared with the theoretical weights of SDF(5-67) and SDF-1α (7.4 and 7.8 kDa, respectively) might be due to either an association of the chemokine with other molecules or an aggregation of the chemokines because SDF(5-67) and SDF-1α are able to form oligomers spontaneously (SI Fig. 6). The higher SDF-1α abundance was associated with a higher transcript abundance of SDF-1α in HIV-1 positive brains, which is in accordance with other studies (21), with no significant differences between nondemented and demented patients suggesting that regulation of SDF(5-67) and SDF-1α abundance involved posttranscriptional events (SI Fig. 6). Because the cellular sources of the SDF(5-67) within the brain were unclear, we used the SDF(5-67)-specific neoepitope antibody to define the cell type(s) producing SDF(5-67). SDF(5-67) immunoreactivity was observed in CD45-positive macrophages/microglia in brain sections from patients with HIV-1 infection (SI Fig. 7). In fact, the SDF(5-67) immunopositive cells also expressed MMP-2 (SI Fig. 7), underscoring the role of parenchymal monocytoid cells in HIV neuropathogenesis. Treatment of monocytoid cells with 1 nM SDF(5-67) resulted in activation of the inflammation genes, IL-1β and indoleamine 2′,3′-dioxygenase (IDO) (Fig. 1b), but in astrocytic cells, IDO, TNFα and IL-10 were induced (Fig. 1c) compared with untreated cells. At matched concentrations, SDF-1α exhibited no modulatory effects on the same genes in both monocytoid (Fig. 1b) and astrocytic cells (Fig. 1c). Primary cells (monocyte-derived macrophages and human fetal astrocytes) also exhibited specific inflammatory gene responses after SDF(5-67) treatment (SI Fig. 8). In contrast to SDF-1α, SDF(5-67) failed to induce calcium mobilization in monocytoid cells. Moreover, SDF(5-67) was not able to repress the SDF-1α-induced calcium mobilization suggesting that these two molecules act through independent signaling pathways (SI Fig. 9). Thus, these results disclosed that SDF(5-67) was detectable in the brains of HIV-infected persons and possessed immunogenic properties, which distinguished it from those induced by SDF-1α in both monocytoid and astrocytic cells.

Fig. 1.

Fig. 1.

Open in a new tab

SDF(5-67) is increased in HIV-infected brains and is immunogenic. (a) Immunoreactivity of both SDF-1 isoforms on Western blot analysis was higher in HIV-1-infected brains [HIV(+), four right lanes], compared with uninfected human brains [HIV(−), two left lanes]. Both SDF-1α and SDF(5-67) immunoreactivity were highest in samples from patients diagnosed with HAD (two right lanes) compared with nondemented (ND) HIV-1-infected samples. (b and c) Differential immunogenic effects of SDF(5-67) on monocytoid (U-937) (b) and astrocytic (U373) (c) cells compared with SDF-1α. Expression of IL-1β, IDO, TNFα, and IL-10 were measured by real time-PCR and expressed as relative fold change (RFC) ± SEM. ANOVA: ∗, P < 0.05; ∗∗, P < 0.01; compared with the control.

Neurotoxic Properties of SDF(5-67).

Because previous studies reported that SDF-1α modulates the spontaneous excitability of rat hippocampal and cerebellar neurons through a calcium-dependent mechanism (22), we investigated SDF(5-67)'s neurophysiological properties in primary rat neurons. Patch-clamp recordings from neurons revealed that SDF(5-67) depressed whole-cell currents whereas SDF-1α had no effect on such currents at these concentrations (Fig. 2a). To characterize the conductances affected by SDF(5-67), we tested the effects of cells that responded to the chemokine in the presence of iberiotoxin (IBTX), a blocker of calcium-activated potassium conductance (BK) channels. The SDF(5-67)-induced reduction in whole-cell currents was almost completely abolished in the presence of IBTX (Fig. 2b). To further define SDF(5-67)'s neurotoxic properties, we applied SDF(5-67) to neurons, which resulted in a decrease in both neurite length and soma size of neurons (SI Fig. 10). Moreover, a synthetic peptide corresponding to the amino-terminal end of SDF(5-67), residues 5–19 of the SDF-1α, mimicked SDF(5-67) effects in terms of neurite retraction and soma atrophy. These findings suggested that the amino-terminal end of the SDF(5-67) played a major role in its pathogenic effects, although it failed to recapitulate the SDF(5-67)-mediated neuronal death (Fig. 2c). Of interest, other MMP-2-cleaved and noncleaved chemokines did not cause neuronal death (SI Fig. 10). Immunolabeling experiments showed that SDF(5-67)-induced neurotoxicity involved activation of caspase-3 (Fig. 2d) and p53 (SI Fig. 10) indicating activation of programmed cell death in neuronal cells. Thus, these data emphasized the membrane-mediated effects of SDF(5-67) compared with those of SDF-1α in neurons, resulting in neuronal apoptosis.

Fig. 2.

Fig. 2.

Open in a new tab

SDF(5-67) induces electrophysiological changes and apoptosis in neurons. (a) Whole-cell currents recorded from primary rat neurons are reduced after exposure to SDF(5-67) (10 nM), whereas SDF-1α did not exhibit similar effects. (Inset) The current reduction induced by SDF-1α (n = 4) and SDF(5-67) (n = 5). (b) In the presence of iberiotoxin (50 nM; IBTX), a blocker of calcium-activated potassium conductance, the SDF(5-67)-induced reduction of currents is abolished (n = 5). (c) Neuronal death, expressed as relative fold change, was induced by 100 nM SDF(5-67) but not by SDF-1α, SDF(5-19), or SDF(20-33) peptides. (d) Caspase-3 activation is involved in cell death induced by 100 nM SDF(5-67). Data in bar graphs are presented ± SEM. Unpaired Student's t test (a and b) and ANOVA (c and d): ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.

CXCR3 Mediates Neurotoxic and Immunogenic Effects of SDF(5-67).

We and others have reported that amino-terminally cleaved SDF-1α does not bind to SDF-1α's cognate receptor CXCR4 (17, 23). Previous reports indicated that SDF-1α exhibited affinity for CXCR7/RDC1, and CXCR3 whose ligands are known to be neurotoxic (24–26). Human neuronal (LAN-2), monocytoid (U-937) and astrocytic (U373) cells expressed CXCR4 and CXCR3 (inset, Fig. 3a). Although CXCR7 gene is expressed in human brain and in cultured cells (SI Fig. 11), earlier studies showed that the SDF-1α amino terminus was required for its binding to CXCR7 (24). Moreover, the CXCR3 ligand, IP-10, mimicked SDF(5-67)-induced membrane conductance perturbations on primary rat neurons (SI Fig. 12). Indeed, a CXCR3 blocking antibody and the CXCR3-specific antagonist, I-TAC(5-73), a truncated product of the chemokine I-TAC (14), were efficient competitors of SDF(5-67)-mediated neurotoxicity, whereas an anti-CXCR4 antibody had no effects on SDF(5-67)'s neurotoxic features (Fig. 3a). I-TAC(5-73) pretreatment of neurons reduced the electrophysiological depressive effects induced by both SDF(5-67) and IP-10 (Fig. 3b and SI Fig. 12) together with its immunogenicity in monocytoid and astrocytic cells (SI Fig. 13). Of interest, neuronal differentiation was associated with an increase in CXCR3 abundance (SI Fig. 14), which was consistent with the use of a differentiation-induced receptor by SDF(5-67) (SI Fig. 15). We extended these studies by showing that siRNA-mediated knockdown of CXCR3 in glia and neurons resulted in reduced CXCR3 expression at the protein level, together with diminished SDF(5-67)-induced immunogenicity and neurotoxicity in astrocytic cells (SI Fig. 16) and, in primary human neurons (Fig. 3c), respectively. The above experiments did not assess whether SDF(5-67) was directly binding to CXCR3 or whether this receptor was involved secondarily as a receptor for a SDF(5-67)-induced ligand. To investigate this latter question, the expression of CXCR3 ligand, IP-10, which possesses neurotoxic properties (26, 27), was evaluated in neuronal and astrocytic cells. Whereas neuronal cells did not show any IP-10 expression, astrocytic cells constitutively expressed IP-10 (SI Fig. 17). However, IP-10's transcript abundance is markedly reduced in presence of SDF(5-67), indicating that IP-10 was unlikely to mediate SDF(5-67)'s neurotoxicity.

Fig. 3.

Fig. 3.

Open in a new tab

SDF(5-67) effects are mediated through CXCR3. (a) SDF(5-67) (100 nM) neuronal (LAN-2) cytotoxicity (expressed as relative fold change) was inhibited with an anti-CXCR3 neutralizing antibody (15 μg/ml) and the specific CXCR3 antagonist, I-TAC(5-73) (100 nM), but not with an anti-CXCR4 antibody (15 μg/ml). These antagonists had no effect on staurosporine-mediated neurotoxicity. (Inset) The three cell lines used in this study expressed both CXCR3 and CXCR4 (lanes 1, 2, and 3 are LAN-2, U-937, and U373, respectively). (b) I-TAC(5-73) blocked the reduction of whole-cell currents in rat primary neurons induced by SDF(5-67) (at +30 mV) (n = 5). (Inset) CXCR3 was detected on human neuronal (LAN-2) cells (lane 1) and primary rat neurons by Western blot analysis (lane 2). (c) siRNA-mediated knockdown of CXCR3 prevented 1 μM SDF(5-67)-induced neurotoxicity in primary human fetal neurons, whereas a control siRNA targeting the HPRT gene did not prevent SDF(5-67)'s toxicity. (d and e) Biot-SDF(5-67) bound specifically to monocytoid (U-937) cells, and binding occurred via the CXCR3 receptor, as shown by the ability of I-TAC(5-73) to compete with the biot-SDF(5-67) binding detected by FACS. (d) Representative results obtained over five experiments. (e) Quantitative representation expressed in percentage of control. Data in bar graphs are presented ± SEM. ANOVA (a, c, and e) and paired Student's t test (b): ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.

To examine SDF(5-67)'s binding properties, we developed a binding assay, using biotinylated SDF(5-67). Biotinylation of SDF(5-67) did not affect its neurotoxicity (SI Fig. 18). The binding specificity was confirmed by the competing efficiency of unlabeled SDF(5-67), whereas unlabeled SDF-1α did not compete with SDF(5-67) binding, analyzed by FACS (Fig. 3d). Quantitative analysis of SDF(5-67) binding disclosed that unlabeled SDF(5-67) competitively reduced biotinylated SDF(5-67) binding in a concentration dependent manner (Fig. 3e). Competition with biotinylated SDF(5-67) binding was observed with I-TAC(5-73) in the same order as that observed with the unlabeled SDF(5-67), thus demonstrating that SDF(5-67) was a ligand for CXCR3. Moreover, as with the above neurotoxicity studies, the SDF (5–19) peptide, unlike SDF (20–33), seemed to be responsible for the binding to CXCR3 as the 5–19 peptide also displayed competition with SDF(5-67) binding (Fig. 3 d and e). In summary, these studies revealed that metalloproteinase cleavage of the CXC chemokine SDF-1α gave rise to a peptide with a change in receptor preference.

SDF(5-67) in Vivo Neuropathogenesis Is Prevented by a CXCR3 Antagonist.

Given that we detected SDF(5-67) in vivo in human brains with HIV infection, we examined its in vivo effects in an established mouse model of neuroAIDS in which SDF(5-67) was stereotactically implanted into the striatum, an area particularly vulnerable to HIV-induced immune activation and neuronal injury (28). Compared with PBS-implanted controls (Fig. 4a–c), 14 days after implantation intrastriatal astrogliosis, microgliosis and loss of neurons were observed in the brains of SDF(5-67)-implanted mice (Fig. 4 d–f and j and SI Fig. 19). Conversely, animals treated with I-TAC(5-73) or SDF-1α did not display similar neuropathological changes (data not shown). However, concomitant implantation of I-TAC(5-73) with SDF(5-67) prevented the astrogliosis, microgliosis and neuronal loss (Fig. 4 g–j and SI Fig. 19). SDF(5-67)-induced neuroinflammation, characterized by increased TNFα and IL-10 transcript expression was also inhibited by I-TAC(5-73) (SI Fig. 19). Implanted SDF(5-67) induced neurobehavioral abnormalities in mice, reported as the mean disability score (MDS), whereas the implantation of SDF-1α did not cause similar neurobehavioral effects (Fig. 4k). However, we observed that SDF(5-67)'s neuropathogenic effects were significantly reduced with coimplantation of I-TAC(5-73) (Fig. 4k), implying that CXCR3-mediated SDF(5-67) effects within the brain caused a neurological disease phenotype.

Fig. 4.

Fig. 4.

Open in a new tab

SDF(5-67)-induced in vivo neuronal injury, neuroinflammation, and behavioral impairment are prevented by CXCR3 blockade. (a–i) Neuropathological analysis of the striatum from PBS- (a–c), SDF(5-67)- (d–f), or SDF(5-67) plus I-TAC(5-73)- (g–i) implanted mice after GFAP (a, d, and g), Iba-1 (b, e, and h), and NeuN (c, f, and i) immunostaining. The striatum of SDF(5-67)-implanted animals displayed greater GFAP and Iba-1 immunoreactivities (d and e, respectively) and neuronal loss (f) compared with PBS-implanted mice (a–c). These neuropathological effects were reversed in animals implanted simultaneously with SDF(5-67) and I-TAC(5-73) (g–i). (j) Neuronal counts (nuclei per area ± SEM) were performed in animals 14 days after implantation of PBS, ITAC(5-73), SDF(5-67), SDF(5-67) plus ITAC(5-73), or SDF-1α. Mice receiving SDF(5-67) showed reduced neuronal counts near the implantation site, which was attenuated by ITAC(5-73) coimplantation. (k) SDF(5-67)-implanted mice exhibited significant neurobehavioral deficits, expressed in mean disability score (MDS), that were eliminated by coimplantation of I-TAC(5-73), whereas SDF-1α-implanted mice did not exhibit any significant neurobehavioral deficits (n = 6 for each group). All data in bar graphs are presented ± SEM. (Magnification, ×400.) ANOVA: ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001 relative to the PBS-implanted group.

Discussion

Protein degradation or processing modulate stability and permit recycling of extra- or misfolded proteins but are also crucial evolutionary strategies for generating bioactive molecules. Indeed, the maturation of numerous neuropeptides involves sequential proteolytic cleavages of a precursor protein by different proteases leading to peptide products with pleiotropic effects (see (29) for review). A requisite maturation by proteolysis has also been reported for molecules involved in immune response including inflammatory cytokines, receptors or components of the complement cascade (30). Several proteins acquire neuropathogenic properties after a proteolytic processing; for example, in Alzheimer's disease, the pathogenicity of amyloid peptides depends on proteases involved in amyloid precursor protein maturation (31).

Herein, we assessed the association between the expression of a cleavage product of the chemokine, SDF-1α, and a neurodegenerative disease, HAD. Chemokine involvement in HIV neuropathogenesis is well established because of (i) chemokines' ability to attract HIV-infected immune cells into the brain, (ii) their neuroinflammatory properties, and (iii) chemokine receptors involvement as HIV-1 coreceptors, in particular the ability of SDF-1α to compete with the HIV-1 gp120 for CXCR4 (32). We also show that the neurotoxic molecule, SDF(5-67) acts on both neurons and glial cells, not through SDF-1α's cognate receptor CXCR4, but by engagement of another CXC receptor, CXCR3. In the context of HIV neuropathogenesis, SDF-1α cleavage is remarkable for the change in receptor preference, because the maturation (i) decreases the abundance of a recognized competitor (SDF-1α) for HIV-1 binding to CXCR4 and (ii) enhances the virus's pathogenic effects without interfering with its replication in the brain. Chemokines can be proteolytically cleaved at either amino or carboxyl termini by numerous enzymes with substantial effects on the chemokines' structural and biological properties (33, 34). Of importance, a change in receptor preference change after a proteolytic cleavage of a chemokine has been described for RANTES, in which its cleavage by dipeptidylpeptidase IV alters its receptor preference from CCR1 to CCR5 thus increasing its antiviral activity (35); conversely whereas MMP-2 cleavage of SDF-1α abolished its antiviral properties (17). CXCR3 is an inducible chemokine receptor reported to bind Mig/CXCL9, IP-10/CXCL10 and I-TAC/CXCL11. Interestingly, Booth et al. (36) have shown that SDF-1α shared some critical structural features with IP-10 in domains involved in receptor binding. In fact, the CXCR3 ligand, IP-10, has recently been reported to have neurotoxic properties (26, 27). However, our results showed that SDF(5-67) diminished astrocytic IP-10 expression, which was not detected in neurons. Nonetheless, SDF(5-67) and IP-10 shared a common feature in their inability to induce calcium release in monocytoid cells (data not shown) and to depress whole-cell currents in neurons.

Our results suggest that SDF(5-67) also exerts both direct and indirect effects contributing to neuronal death in vitro and in vivo. We showed that SDF(5-67) directly induced neuronal apoptosis by activating caspase-3 and p53. Moreover, SDF(5-67) evoked a reduction of an IBTX-sensitive calcium-activated potassium conductance, which we have previously shown to result in an increase in neuronal excitability (37). These observations, indicated that SDF(5-67) could impair basal neuronal function at subtoxic concentrations immediately upon exposure to the neurotoxin. SDF(5-67)-induced electrophysiological effects were accompanied by morphological changes, also induced by a synthetic peptide encoding the SDF(5-67) amino terminus. These findings indicate that the N-loop of the SDF(5-67) peptide was sufficient for both binding and activation of the CXCR3 receptor but nevertheless insufficient to mimic SDF(5-67)-mediated neurotoxicity implying that the SDF(5-67) amino terminus initiates the neurotoxic mechanism, but the full neurotoxic effects of SDF(5-67) might require other domains of the molecule.

In addition to the direct neurotoxic effect, our results also suggest that SDF(5-67) could amplify the HIV-induced “bystander” neurotoxicity. Indeed, several proinflammatory molecules induced by SDF(5-67) in both astrocytic and monocytoid cells have neurotoxic actions (38). This indirect effect occurred at a concentration range 100 times lower than the concentration required to induce direct neuronal death. The present observations imply that SDF(5-67)-induced neuroinflammation may be an early event in the HIV-induced neuropathogenesis cascade, whereas the direct SDF(5-67) neurotoxicity is a later, or a more localized, mechanism. Although both CXCR4 and CXCR7 engage full-length SDF-1α, we identified the receptor responsible for SDF(5-67)'s effects as the CXC chemokine receptor, CXCR3, thus identifying an undescribed chemokine/receptor association. Our in vitro findings were confirmed in vivo by demonstrating that intrastriatal implantation of SDF(5-67) induced neurological impairment in mice which was blocked by coimplantation of the CXCR3 antagonist, ITAC(5-73). This mechanism is highly plausible in vivo, especially as SDF(5-67) abundance was increased in brains of HAD patients. We propose that the induction of these mechanisms including the activation of MMP-2 with ensuing SDF-1α cleavage, together with direct and indirect neurotoxic properties of the cleaved product is a feature acquired during the rapid adaptation of HIV-1 to host cells and particularly to the CNS environment, in keeping with the virus' widely recognized neurotropic properties (39).

Methods

Cell Cultures.

Primary rat and human neurons were isolated as described (18, 37). Human neuroblastoma (LAN-2), monocytoid (U-937), or astrocytic (U373) cell lines were grown in MEM, RPMI medium, and DMEM, respectively, supplemented with 10% FBS and 1% penicillin–streptomycin. LAN-2 cells were differentiated as described (40).

Human Brain Samples.

Human CNS tissue (frontal lobe) was collected at autopsy from HIV-1 sero-negative or -positive patients with consent and stored at −80°C. HIV-1 sero-positive patients were diagnosed premortem with or without HAD, as described (41).

Electrophysiological Studies.

Whole-cell patch-clamp recordings of neurons were performed under voltage-clamp conditions as described (42). A more detailed description of the methods is provided in SI Methods.

Neurotoxicity Assay.

Neurotoxicity was evaluated in differentiated neuronal (LAN-2) cells by trypan blue exclusion after 48 h of treatment (41). Cellular injury was quantified in neuronal cultures immunolabeled with MAP-2 antibodies after 48 h of treatment by an examiner unaware of the treatment, as described (43). Activated caspase-3 was assessed by infrared immunofluorescence (LI-COR Biosciences, Lincoln, NE) with an activated caspase-3-specific antibody (1:100 dilution; Cell Signaling, Danvers, MA) (40) and expressed relative to β-tubulin abundance detected with a monoclonal anti-β-tubulin III antibody (1:800 dilution; Sigma, Oakville, ON, Canada).

Western Blot Analysis.

Protein extracts were prepared from human frontal cortex or cells by homogenization within cell lysis buffer as described (40). Western blots were probed with polyclonal anti-SDF(5-67) neoepitope antibody [1:100 dilution, preadsorbed for 30 min with SDF-1α to optimize its specificity against SDF(5-67); see SI Methods] anti-SDF-1α (1:250 dilution; R&D Systems, Minneapolis, MN) monoclonal antibody or anti-p53 (1:250 dilution; Santa Cruz Biotechnology), anti-CXCR3 and anti-CXCR4 polyclonal antibodies (1:250 dilution; R&D Systems), or HRP-conjugated anti-β-actin antibody (1:200 dilution; Chemicon, Temecula, CA).

Real-Time RT-PCR.

Total cellular RNA was isolated from SDF-1α- or SDF(5-67)-treated cells or from mice brains by using TRIzol reagent (Invitrogen, Burlington, ON, Canada), and cDNAs were subsequently synthesized as reported (40). Real-time quantitative PCR was performed by using the iCycler IQ system (Bio-Rad, Mississauga, ON, Canada), as described (40). The RNA levels of the gene of interest were expressed relative to GAPDH RNA and expressed as relative fold change over RNA levels in nontreated cells. All real-time RT-PCR quantifications were performed in duplicate and repeated with cDNAs from two or more different independent experiments. Primers used are described in SI Table 1.

Calcium Mobilization Assay.

Changes in intracellular calcium concentration induced in U-937 by SDF-1α or SDF(5-67) were measured with the intracellular calcium indicator fluo-3 acetoxymethyl ester (Molecular Probes, Eugene, OR) as described (44).

Binding Assay.

The biotinylation of SDF(5-67) was performed by using the Sulfo-NHS-Biotin reagent (Pierce, Rockford, IL) according to the manufacturer's guidelines. Biotinylation effectiveness was assessed by MALDI-TOF mass spectrometry. Cells were incubated for 1 h on ice with ≈1 μg of biotinylated SDF(5-67) [biot-SDF(5-67)] in binding buffer (0.3% BSA in PBS). For competition assays, cells were incubated with potential competitors for 1 h on ice before adding the biot-SDF(5-67). Cells were then washed with ice-cold binding buffer, and biot-SDF(5-67) binding was revealed with a streptavidin-PE conjugate (0.25 μg/ml in binding buffer for 30 min on ice; AnaSpec, San Jose, CA). After another washing step, cells were fixed in 2% paraformaldehyde, and the fluorescence associated with the biot-SDF(5-67) binding was evaluated with a FACScan (Becton Dickinson, Oakville, ON, Canada).

Immunohistochemistry/Immunofluorescence.

Immunohistochemical labeling was performed by using 10-μm paraffin-embedded serial brain sections prepared as described in ref. 40 and SI Methods.

Implantation of Mice and Assessment of Neurological Impairment.

CD-1 male mice 4 weeks of age were housed according to the guidelines of the Canadian Animal Care Committee. Animals (n = 6 for all groups) were placed in a stereotaxic frame under ketamine/xylazine anesthesia. PBS containing 100 nM SDF-1α, SDF(5-67), I-TAC(5-73), or both SDF(5-67) and I-TAC(5-73) was delivered into the striatum through a 27-gauge needle inserted 2 mm lateral and 3 mm posterior relative to bregma at a depth of 2.5 mm over 6 min in a 3-μl volume. Behavioral tests were performed as described in SI Methods.

Statistical Analysis.

Statistical analyses were performed by using GraphPad InStat version 3.0 (GraphPad Software) for both parametric and nonparametric comparisons. P values of <0.05 were considered significant.

Supplementary Material

Supporting Information
pnas_0604678103_index.html (19.2KB, html)

Acknowledgments

We thank Kim H. Harris, Kristofor K. Ellestad, Shuhong Liu, and Andrea Sullivan for technical assistance. This work was supported by the Canadian Institutes for Health Research (CIHR), the Canadian Foundation for AIDS Research (CANFAR), and National Institutes of Health Grant 1R01MN07568301A1 (to C.P., C.M.O., and M.D.H.). C.P., J.H.J., and C.M.O. hold Canada Research Chairs (T1) in Neurological Infection and Immunity, Alzheimer Research, and Metalloproteinase Proteomics and Systems Biology, respectively.

Abbreviations

HAD

HIV-associated dementia

I-TAC

interferon-inducible T cell chemoattractant

IDO

indoleamine 2′,3′-dioxygenase

IP-10

interferon-inducible protein-10

MMP

matrix metalloproteinase

SDF

stromal cell-derived factor.

Footnotes

The authors declare no conflict of interest.

This article is a PNAS direct submission.

This article contains supporting information online at www.pnas.org/cgi/content/full/0604678103/DC1.

References

  • 1.Cartier L, Hartley O, Dubois-Dauphin M, Krause KH. Brain Res Brain Res Rev. 2005;48:16–42. doi: 10.1016/j.brainresrev.2004.07.021. [DOI] [PubMed] [Google Scholar]
  • 2.Lapidot T, Kollet O. Leukemia. 2002;16:1992–2003. doi: 10.1038/sj.leu.2402684. [DOI] [PubMed] [Google Scholar]
  • 3.Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, Cooper NR, Eikelenboom P, Emmerling M, Fiebich BL, et al. Neurobiol Aging. 2000;21:383–421. doi: 10.1016/s0197-4580(00)00124-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Rottman JB, Slavin AJ, Silva R, Weiner HL, Gerard CG, Hancock WW. Eur J Immunol. 2000;30:2372–2377. doi: 10.1002/1521-4141(2000)30:8<2372::AID-IMMU2372>3.0.CO;2-D. [DOI] [PubMed] [Google Scholar]
  • 5.Burwinkel M, Riemer C, Schwarz A, Schultz J, Neidhold S, Bamme T, Baier M. Int J Dev Neurosci. 2004;22:497–505. doi: 10.1016/j.ijdevneu.2004.07.017. [DOI] [PubMed] [Google Scholar]
  • 6.Gonzalez-Scarano F, Martin-Garcia J. Nat Rev Immunol. 2005;5:69–81. doi: 10.1038/nri1527. [DOI] [PubMed] [Google Scholar]
  • 7.Biber K, Zuurman MW, Dijkstra IM, Boddeke HW. Curr Opin Pharmacol. 2002;2:63–68. doi: 10.1016/s1471-4892(01)00122-9. [DOI] [PubMed] [Google Scholar]
  • 8.Overall CM, McQuibban GA, Clark-Lewis I. Biol Chem. 2002;383:1059–1066. doi: 10.1515/BC.2002.114. [DOI] [PubMed] [Google Scholar]
  • 9.McQuibban GA, Gong JH, Wong JP, Wallace JL, Clark-Lewis I, Overall CM. Blood. 2002;100:1160–1167. [PubMed] [Google Scholar]
  • 10.Davis DA, Singer KE, De La Luz Sierra M, Narazaki M, Yang F, Fales HM, Yarchoan R, Tosato G. Blood. 2005;105:4561–4568. doi: 10.1182/blood-2004-12-4618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Struyf S, De Meester I, Scharpe S, Lenaerts JP, Menten P, Wang JM, Proost P, Van Damme J. Eur J Immunol. 1998;28:1262–1271. doi: 10.1002/(SICI)1521-4141(199804)28:04<1262::AID-IMMU1262>3.0.CO;2-G. [DOI] [PubMed] [Google Scholar]
  • 12.Ludwig A, Schiemann F, Mentlein R, Lindner B, Brandt E. J Leukocyte Biol. 2002;72:183–191. [PubMed] [Google Scholar]
  • 13.Proudfoot AE, Buser R, Borlat F, Alouani S, Soler D, Offord RE, Schroder JM, Power CA, Wells TN. J Biol Chem. 1999;274:32478–32485. doi: 10.1074/jbc.274.45.32478. [DOI] [PubMed] [Google Scholar]
  • 14.Clark-Lewis I, Mattioli I, Gong JH, Loetscher P. J Biol Chem. 2003;278:289–295. doi: 10.1074/jbc.M209470200. [DOI] [PubMed] [Google Scholar]
  • 15.Nufer O, Corbett M, Walz A. Biochemistry. 1999;38:636–642. doi: 10.1021/bi981294s. [DOI] [PubMed] [Google Scholar]
  • 16.Macphee CH, Appelbaum ER, Johanson K, Moores KE, Imburgia CS, Fornwald J, Berkhout T, Brawner M, Groot PH, O'Donnell K, et al. J Immunol. 1998;161:6273–6279. [PubMed] [Google Scholar]
  • 17.McQuibban GA, Butler GS, Gong JH, Bendall L, Power C, Clark-Lewis I, Overall CM. J Biol Chem. 2001;276:43503–43508. doi: 10.1074/jbc.M107736200. [DOI] [PubMed] [Google Scholar]
  • 18.Zhang K, McQuibban GA, Silva C, Butler GS, Johnston JB, Holden J, Clark-Lewis I, Overall CM, Power C. Nat Neurosci. 2003;6:1064–1071. doi: 10.1038/nn1127. [DOI] [PubMed] [Google Scholar]
  • 19.Conant K, McArthur JC, Griffin DE, Sjulson L, Wahl LM, Irani DN. Ann Neurol. 1999;46:391–398. doi: 10.1002/1531-8249(199909)46:3<391::aid-ana15>3.0.co;2-0. [DOI] [PubMed] [Google Scholar]
  • 20.Patrick MK, Johnston JB, Power C. J Virol. 2002;76:7923–7931. doi: 10.1128/JVI.76.16.7923-7931.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Peng H, Erdmann N, Whitney N, Dou H, Gorantla S, Gendelman HE, Ghorpade A, Zheng J. Glia. 2006;54:619–629. doi: 10.1002/glia.20409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Limatola C, Giovannelli A, Maggi L, Ragozzino D, Castellani L, Ciotti MT, Vacca F, Mercanti D, Santoni A, Eusebi F. Eur J Neurosci. 2000;12:2497–2504. doi: 10.1046/j.1460-9568.2000.00139.x. [DOI] [PubMed] [Google Scholar]
  • 23.Valenzuela-Fernandez A, Planchenault T, Baleux F, Staropoli I, Le-Barillec K, Leduc D, Delaunay T, Lazarini F, Virelizier JL, Chignard M, et al. J Biol Chem. 2002;277:15677–15689. doi: 10.1074/jbc.M111388200. [DOI] [PubMed] [Google Scholar]
  • 24.Balabanian K, Lagane B, Infantino S, Chow KY, Harriague J, Moepps B, Arenzana-Seisdedos F, Thelen M, Bachelerie F. J Biol Chem. 2005;280:35760–35766. doi: 10.1074/jbc.M508234200. [DOI] [PubMed] [Google Scholar]
  • 25.Weng Y, Siciliano SJ, Waldburger KE, Sirotina-Meisher A, Staruch MJ, Daugherty BL, Gould SL, Springer MS, DeMartino JA. J Biol Chem. 1998;273:18288–18291. doi: 10.1074/jbc.273.29.18288. [DOI] [PubMed] [Google Scholar]
  • 26.Sui Y, Stehno-Bittel L, Li S, Loganathan R, Dhillon NK, Pinson D, Nath A, Kolson D, Narayan O, Buch S. Eur J Neurosci. 2006;23:957–964. doi: 10.1111/j.1460-9568.2006.04631.x. [DOI] [PubMed] [Google Scholar]
  • 27.van Marle G, Henry S, Todoruk T, Sullivan A, Silva C, Rourke SB, Holden J, McArthur JC, Gill MJ, Power C. Virology. 2004;329:302–318. doi: 10.1016/j.virol.2004.08.024. [DOI] [PubMed] [Google Scholar]
  • 28.Glass JD, Fedor H, Wesselingh SL, McArthur JC. Ann Neurol. 1995;38:755–762. doi: 10.1002/ana.410380510. [DOI] [PubMed] [Google Scholar]
  • 29.Hallberg M, Nyberg F. Curr Protein Pept Sci. 2003;4:31–44. doi: 10.2174/1389203033380313. [DOI] [PubMed] [Google Scholar]
  • 30.Gasque P. Mol Immunol. 2004;41:1089–1098. doi: 10.1016/j.molimm.2004.06.011. [DOI] [PubMed] [Google Scholar]
  • 31.Hooper NM. Biochem Soc Trans. 2005;33:335–338. doi: 10.1042/BST0330335. [DOI] [PubMed] [Google Scholar]
  • 32.Hesselgesser J, Liang M, Hoxie J, Greenberg M, Brass LF, Orsini MJ, Taub D, Horuk R. J Immunol. 1998;160:877–883. [PubMed] [Google Scholar]
  • 33.McQuibban GA, Gong JH, Tam EM, McCulloch CA, Clark-Lewis I, Overall CM. Science. 2000;289:1202–1206. doi: 10.1126/science.289.5482.1202. [DOI] [PubMed] [Google Scholar]
  • 34.Van Damme J, Struyf S, Wuyts A, Van Coillie E, Menten P, Schols D, Sozzani S, De Meester I, Proost P. Chem Immunol. 1999;72:42–56. [PubMed] [Google Scholar]
  • 35.Schols D, Proost P, Struyf S, Wuyts A, De Meester I, Scharpe S, Van Damme J, De Clercq E. Antiviral Res. 1998;39:175–187. doi: 10.1016/s0166-3542(98)00039-4. [DOI] [PubMed] [Google Scholar]
  • 36.Booth V, Keizer DW, Kamphuis MB, Clark-Lewis I, Sykes BD. Biochemistry. 2002;41:10418–10425. doi: 10.1021/bi026020q. [DOI] [PubMed] [Google Scholar]
  • 37.Jassar BS, Harris KH, Ostashewski PM, Jhamandas JH. J Neurophysiol. 1999;81:234–246. doi: 10.1152/jn.1999.81.1.234. [DOI] [PubMed] [Google Scholar]
  • 38.Deshpande M, Zheng J, Borgmann K, Persidsky R, Wu L, Schellpeper C, Ghorpade A. Neurotox Res. 2005;7:183–192. doi: 10.1007/BF03036448. [DOI] [PubMed] [Google Scholar]
  • 39.Power C. Trends Neurosci. 2001;24:162–169. doi: 10.1016/s0166-2236(00)01737-9. [DOI] [PubMed] [Google Scholar]
  • 40.Noorbakhsh F, Vergnolle N, McArthur JC, Silva C, Vodjgani M, Andrade-Gordon P, Hollenberg MD, Power C. J Immunol. 2005;174:7320–7329. doi: 10.4049/jimmunol.174.11.7320. [DOI] [PubMed] [Google Scholar]
  • 41.Johnston JB, Silva C, Holden J, Warren KG, Clark AW, Power C. Ann Neurol. 2001;50:434–442. doi: 10.1002/ana.1131. [DOI] [PubMed] [Google Scholar]
  • 42.Jhamandas JH, Harris KH, Cho C, Fu W, MacTavish D. J Neurophysiol. 2003;89:2923–2930. doi: 10.1152/jn.01138.2002. [DOI] [PubMed] [Google Scholar]
  • 43.Zhu Y, Jones G, Tsutsui S, Opii W, Liu S, Silva C, Butterfield DA, Power C. J Immunol. 2005;175:1118–1126. doi: 10.4049/jimmunol.175.2.1118. [DOI] [PubMed] [Google Scholar]
  • 44.Kawabata A, Saifeddine M, Al-Ani B, Leblond L, Hollenberg MD. J Pharmacol Exp Ther. 1999;288:358–370. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
pnas_0604678103_index.html (19.2KB, html)
pnas_0604678103_16.pdf (21.6KB, pdf)
pnas_0604678103_17.pdf (12.4KB, pdf)
pnas_0604678103_1.pdf (37.6KB, pdf)
pnas_0604678103_2.pdf (54.3KB, pdf)
pnas_0604678103_3.pdf (39KB, pdf)
pnas_0604678103_4.pdf (28.7KB, pdf)
pnas_0604678103_5.pdf (46.8KB, pdf)
pnas_0604678103_6.pdf (30.3KB, pdf)
pnas_0604678103_7.pdf (38KB, pdf)
pnas_0604678103_8.pdf (21.4KB, pdf)
pnas_0604678103_9.pdf (30.4KB, pdf)
pnas_0604678103_10.pdf (31.3KB, pdf)
pnas_0604678103_11.pdf (25.5KB, pdf)
pnas_0604678103_12.pdf (23.8KB, pdf)
pnas_0604678103_13.pdf (27.6KB, pdf)
pnas_0604678103_14.pdf (14.1KB, pdf)
pnas_0604678103_15.pdf (24.2KB, pdf)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES