receptor protein immunocomplex in human and mouse cortical nerve endings amplifies complement-induced glutamate release

Neuroscience Letters 600 (2015) 50–55

Contents lists available at ScienceDirect

Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet

Research article

Antibody/receptor protein immunocomplex in human and mouse cortical nerve endings amplifies complement-induced glutamate release Elisa Merega a,1 , Silvia Di Prisco a,1 , Paolo Severi b , Fotios Kalfas b , Anna Pittaluga a,c,∗ a

Department of Pharmacy, Pharmacology and Toxicology Section, University of Genoa, Genoa, Italy Division of Neurosurgery, Galliera Hospital, Genoa, Italy c Center of Excellence for Biomedical Research, University of Genoa, Genoa, Italy b

h i g h l i g h t s • • • • •

The classic pathway of activation is not involved in the complement-evoked glutamate release. Complement glutamate release is significantly increased in synaptosomes incubated with anti-NH2- CCR5 antibody. C1q deprived serum-induced glutamate release is unaltered in terminals incubated with anti-NH2- CCR5 antibody. Complement glutamate release is unmodified in anti-COOH-CCR5 entrapped synaptosomes. Antibodies recognizing the outer sequence of receptors triggers the activation of complement trough the classic pathway.

a r t i c l e

i n f o

Article history: Received 9 April 2015 Received in revised form 31 May 2015 Accepted 1 June 2015 Available online 3 June 2015 Abbreviations: BBB, blood-brain barrier RANTES, regulated upon activation normal T cells expressed and secreted CCRs, C–C chemokine receptors Keywords: CCR5 Complement Anti-ccr5 antibody Glutamate Human cortex Mouse cortex

a b s t r a c t Previous studies have demonstrated that complement alone releases glutamate from human and mouse cortical terminals in an antibody-independent manner. In order to expand our knowledge on complement-mediated effects, we investigated whether the presence of an antigen-antibody complex in synaptosomal plasmamembranes could also trigger complement-induced functional responses that might affect neurotransmitter release. To this aim, we focused on the chemokine 5 receptor (CCR5) expressed in human and mouse cortical glutamate terminals, whose activation by CCL5 elicits [3 H]Daspartate ([3 H]D-ASP) release. Preincubating synaptosomes with an antibody recognizing the NH2 terminus of the CCR5 protein (anti-NH2 -CCR5 antibody) abolished the CCL5-induced [3 H]D-ASP release. Similarly, enriching synaptosomes with an antibody recognizing the COOH terminus of CCR5 (antiCOOH-CCR5 antibody) prevented the CCL5-induced [3 H]D-ASP release. The antagonist-like activity of the anti-NH2 -CCR5 antibody turned to facilitation when anti-NH2 -CCR5-treated synaptosomes were exposed to complement. In these terminals, the releasing effect was significantly higher than that elicited by complement in untreated synaptosomes. On the contrary, the complement-induced [3 H]D-ASP release from anti-COOH-CCR5 antibody-entrapped synaptosomes did not differ from that from untreated synaptosomes. Preincubating synaptosomes with anti-beta tubulin III antibody, used as negative control, neither prevented the CCL5-induced releasing effect nor it amplified the complement-induced [3 H]D-ASP release. Finally, serum lacking the C1q protein, i.e. the protein essential to promote the antibody-mediated activation of complement, elicited a comparable [3 H]D-ASP release from both untreated and anti-NH2-CCR5 antibody-treated synaptosomes. Thus, we propose that antibodies raised against the outer sequence of a receptor protein can trigger the activation of the complement through the classic, C1q-mediated antibody-dependent pathway, which results in an abnormal release of glutamate that could be deleterious to central nervous system. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction ∗ Corresponding author at: Department of Pharmacy, Pharmacology and Toxicology Section, University of Genoa, Viale Cembrano 4–16148 Genoa, Italy. Tel.: +390103532049; fax: +390103993360. E-mail address: [email protected] (A. Pittaluga). 1 These authors equally contributed to the work. http://dx.doi.org/10.1016/j.neulet.2015.06.001 0304-3940/© 2015 Elsevier Ireland Ltd. All rights reserved.

Complement is a complex network of soluble and membrane proteins that drives the humoral and cellular immune responses. Complement activation is proposed to occur through three differ-

E. Merega et al. / Neuroscience Letters 600 (2015) 50–55

ent pathways. The first is the classic, antibody-dependent pathway, which represents the humoral response of complement to the presence of immunocomplexes with IgG1 and/or IgM antibodies. The second one is the lectin-dependent pathway, that is initiated by lectin recognizing sugar structures, while the third is the alternative pathway, which is triggered by the degradation of the C3 component of the complement system [1,2]. Very recently, complement was shown to release glutamate from glutamatergic nerve endings isolated from different central nervous system (CNS) regions, without affecting the release of other neurotransmitters (i.e. noradrenaline, acetylcholine or GABA as well, [3]). The pathway of complement activation involved in the releasing effect has not so far been investigated. However, since the presence of antibody/receptor protein complexes in synaptosomal plasmamembranes could not in principle be excluded, we have investigated the possibility that the classic pathway could be involved. Two different approaches were used to address this question. First, synaptosomes were incubated with an antibody recognizing the outer sequence of receptor proteins known to be present in the plasmamembranes of glutamatergic nerve endings. This allowed us to determine whether the presence of an antibody/receptor protein complex in synaptosomal plasmamembranes could affect the complement-induced releasing efficiency in these terminals. To this aim, we focused on the chemokine receptor subtype 5 (CCR5). Beside their well-known expression in macrophages and T-lymphocytes, CCR5s are also widely distributed in non-immunocompetent cells. In particular, as far as the CNS is concerned, CCR5 exists in glutamate neurones as well as in astrocytes where its activation controls glutamate release [4–6]. Taking into consideration that the natural ligand of CCR5, the chemokine CCL5 (RANTES, regulated upon activation normal T cells expressed and secreted), is actively produced and secreted by glial cells, the CCL5-induced glutamate release could represent a mechanism of glial to neuron communication, which is crucial to neuronal plasticity and developmental processes, and potentially relevant in neuroinflammatory diseases [7,8]. In recent years we showed that antibodies recognizing the outer sequence of the CCR5 receptor protein (anti-NH2 CCR5-antibody) can abolish the chemokine-mediated glutamate release possibly by impeding the binding of the natural ligand CCL5 at CCR5 then [6,9]. Synaptosomes bearing anti-NH2 -CCR5antibody/CCR5 protein complexes were exposed in superfusion to complement, in order to assess whether the presence of the immunocomplex could elicit unexpected functional responses that could emerge as changes in glutamate outflow. The complementinduced releasing effect in anti-NH2 -CCR5-antibody pretreated synaptosomes was significantly higher than that from untreated synaptosomes, suggesting that other releasing mechanism(s) took place because of the presence of the antibody/receptor protein complexes. As a second approach, we directly verified the involvement of the classic pathway of activation of complement by analysing the impact of C1q deprived serum in human and mouse cortical glutamatergic synaptosomes. The C1q component is essential to trigger the complement activation through the classic pathway, so that the use of the deprived serum could allow us to investigate the involvement of this pathway in the complement-induced releasing effects. Interestingly, the anti-NH2 -CCR5-antibody-mediated complement-induced releasing effect was not observed in synaptosomes exposed to the C1q-deprived serum. This observation is compatible with the idea that the classic pathway of activation of complement had a role in eliciting the antibody-mediated releasing effect.

51

2. Materials and methods 2.1. Tissue samples and synaptosomes preparation Samples of human cerebral cortex (parts of frontal and temporal lobes) were obtained from informed and consenting patients (7 women, 4 men, aged 27-58 years, for premedication and anaesthesia [6,10]) undergoing neurosurgery to reach deeply seated tumours. Adult male mice (C57BL/6, 20–25 g, Charles River, Calco, Italy) were killed by decapitation and the cortices were rapidly removed. The animal experimental procedures were in accordance with the European legislation (European Communities Council Directive of September, 22, 2010, n◦ 2010/63/EU) and the Italian legislation (L.D. no. 116 /1992 and 26/2014), and they were approved by the Italian Ministry of Health (protocol number n◦ 29,823-10). Tissues were placed in a physiological salt solution at 2–4 ◦ C and synaptosomes were prepared within 30 min as previously described [7]. 2.2. Release experiments Purified synaptosomes were resuspended in a physiological solution having the following composition (mM): NaCl, 140; KCl, 3; MgSO4 , 1.2; CaCl2 , 1.2; NaH2 PO4 , 1.2; NaHCO3 , 5; HEPES, 10; glucose, 10; pH 7.2–7.4 and incubated at 37 ◦ C, for 15 min in the presence of [2,3-3 H]D-aspartate ([3 H]D-ASP, specific activity 11.3Ci/mmol, f.c. 50 nM, Perkin Elmer, Boston, MA), an unmetabolizable analogue of glutamate often used in release studies to monitor glutamate release [11]. Synaptosomes were pre-incubated with mouse anti-CCR5 antibody raised against the NH2 terminus of the CCR5 receptor protein (CCR5-03,IgM, recognising the 13–28 amino acid sequence of the human CCR5 receptor, NIBSC, Centralized Facility for AIDS reagents, Hertfordshire, UK, the anti-NH2 -CCR5 antibody in the text, 1:2000) or with mouse anti␤-tubulin III (Sigma–Aldrich, Milan, Italy, 1:400), used as negative control. In control experiments, synaptosomes were incubated with a rabbit anti-CCR1 antibody recognizing the NH2 terminus of the CCR1 receptor protein (rabbit anti-CCr1, IgG, Novus Biological, Cambridge, UK, the anti-NH2 -CCR1 antibody in the text 1:2000). When indicated, rabbit anti-CCR5 antibody recognizing the COOH terminus of the receptor (CKR-5, IgG, ANASPEC, S. Jose, CA, the antiCOOH-CCR5 antibody throughout the text, 1:400) were entrapped within synaptosomes by homogenizing human tissue specimens with sucrose containing the antibody [6,10]. Based on estimates made by entrapping [3 H]sucrose, the intrasynaptosomal concentration of the antibody in the cytosol should amount to about the 5% of the original concentration in the homogenization medium. Identical portions of the synaptosomal suspensions were layered at the bottom of superfusion chambers in a Superfusion Apparatus (Ugo Basile, Comerio, Varese, Italy, for technical details see Ref. [12]) at 37 ◦ C and were superfused for 36 min to equilibrate the system. Three consecutive 3-min fractions (termed b1 to b3) were collected and counted for radioactivity. Human and mouse (lyophilized, Low Tox) complement (Cederlane, Ontario, Canada), complement C1qdepleted human serum (Calbiochem, EMD Millipore Inc., Billerica, MA, USA) or CCL5 (Sigma–Aldrich, Milan, Italy) were introduced at the end of the first fraction collected and then replaced with standard medium after 3 min of superfusion. Superfusate fractions and synaptosomes were then counted for radioactive content with liquid scintillation. 2.3. Calculations and statistics Fractions collected and superfused synaptosomes were counted for radioactivity that was expressed as a percentage of the total

52

E. Merega et al. / Neuroscience Letters 600 (2015) 50–55

synaptosomal tritium content at the start of the fraction collected. Drug effects were expressed as increase over basal release and were evaluated from the ratio between the percentage of release in the b3 fraction (corresponding to the maximal complement-induced effect) and that in the b1 fraction (basal release). This ratio was compared with the corresponding ratio obtained when no drug was added. Analysis of variance was performed by ANOVA followed by multiple-comparisons test (Dunnett’s test or Tukey’s test, as appropriate); data were considered significant for P < 0.05 at least.

3. Results Human cortical nerve endings preloaded with [3 H]D-ASP were exposed in superfusion to the CCR5 natural ligand, the chemokine CCL5. When added up to 1 nM, the chemokine elicited a significant release of the radioactive tracer [6]. Incubating human cortical synaptosomes with an anti CCR5-antibody raised against the NH2 terminus of the receptor protein (the anti-NH2 -CCR5 antibody) prevented the CCL5-induced releasing effect (Fig. 1A). Similarly, entrapping human cortical synaptosomes with an anti CCR5-antibody raised against the COOH terminus (the anti-COOHCCR5 antibody) antagonized the CCL5-mediated [3 H]D-ASP. On the other hand, preincubation of synaptosomes with an unrelated antibody (the anti ␤-tubulin III antibody in this case) was devoid of activity (Fig. 1A). . The lack of activity of the anti-␤-tubulin III antibody pre-treatment ruled out the possibility that indirect effects could be involved in the antibody-mediated antagonism of the CCL5-mediated releasing effect. Fig. 1B shows that exposure of superfused human cortical synaptosomes to selected complement (1:100 and 1:10) dilutions increases the spontaneous release of [3 H]D-ASP (Fig. 1B). Of note, in human cortical synaptosomes pre-incubated with anti-NH2 -CCR5 antibody, the complementinduced releasing effect was significantly higher than that observed in control, antibody-untreated, synaptosomes (Fig. 1B). However, the complement-induced [3 H]D-ASP release from human cortical synaptosomes pretreated with an unrelated antibody (anti-␤ tubulin III, Fig. 1B), as well as entrapped with the anti-COOHCCR5 antibody (Fig. 1B) did not significantly differ from the complement-induced [3 H]D-ASP release from untreated synaptosomes. The classical pathway of activation of complement is triggered by IgG or IgM antibodies that bind to antigens. The binding of the antibodies to the antigens exposes a site on the antibody which is the preferential binding site for the first complement component, C1q, the activation of which allows the cleavage of the other components of the complement complex. The serum deprived of the C1q component of the complement represents therefore a useful tool to assess whether the C1q-dependent classic pathway of activation could have a role in the complement-induced releasing effect observed in control and in anti-NH2 -CCR5 antibody pretreated human cortical synaptosomes. Accordingly, Fig. 1C shows that the C1q-depleted human serum can still evoke the release of preloaded [3 H]D-ASP release from control, untreated, human cortical synaptosomes. However, the releasing effect elicited by the C1q-deprived serum in human cortical synaptosomes pre-exposed to anti-NH2 CCR5 antibody was lower than that evoked by complement in this synaptosomal preparation. Moreover, this effect did not significantly differ from that elicited by C1q-depleted serum from untreated synaptosomes. Altogether, these observations suggest that the complement-induced releasing effect in anti-NH2 -CCR5 antibody pretreated synaptosomes consists of two components, one dependent on a molecular cascade of events occurring through the antibody-dependent classic pathway, and the second relying on other pathway(s).

Similar results were obtained with mouse cortical synaptosomes. Fig. 2A shows that mouse complement (1:1000 to 1:100) releases [3 H]D-ASP from untreated mouse cortical synaptosomes. The complement-induced release was significantly increased in anti-NH2 -CCR5 antibody pre-treated mouse cortical synaptosomes when compared to the control (i.e., untreated cortical synaptosomes), but was unmodified in anti ␤-tubulin III pre-treated nerve endings (Fig. 2A). Furthermore, the release of [3 H]D-ASP evoked by complement (1:300) from anti-COOH-CCR5 antibody entrapped nerve endings did not significantly differ from that observed from untreated synaptosomes (1:300 complement in control synaptosomes: 189 ± 35; 1:300 complement in anti-COOH2 -CCR5 antibody pretreated synaptosomes: 217 ± 42, P > 0.05, results are expressed as percent increase over basal release, data are means ± SEM from three experiments run in triplicate). The facilitation of the complement-evoked releasing effect observed in the anti-NH2 CCR5 antibody pre-incubated mouse cortical synaptosomes could not be observed when anti-NH2 -CCR5 antibody pre-treated nerve terminals were exposed to the C1-deprived serum (Fig. 2B). Finally, experiments were carried out in order to assess whether also antibodies recognizing the outer sequence of receptor proteins other than CCR5 could unveil complement-induced releasing activity. To this aim, mouse cortical synaptosomes were preincubated with an antibody recognizing the NH2 sequence of the CCR1 protein. This antibody was shown to null the CCL5-induced releasing effect in mouse cortical nerve endings [9]. Again, complement-induced releasing activity was significantly increased in anti-NH2 -CCR1 antibody preincubated synaptosomes when compared to control (1:300 complement in control synaptosomes: 201 ± 28; 1:300 complement in anti-COOH2 -CCR5 antibody pretreated synaptosomes: 354 ± 22, P < 0.05, results are expressed as percent increase over basal release, data are means ± SEM from three experiments run in triplicate).

4. Discussion The first observation resulting from this work is that the presence of an antibody/receptor complex in synaptosomal plasma membrane triggers a complement-mediated cascade of events leading to glutamate release that is additive to that elicited by complement alone. This is indicated by the fact that complement-mediated releasing effect was significantly increased in synaptosomes bearing anti-NH2- CCR5 antibody/CCR5 protein complex at the outer side of nerve ending membranes, when compared to that observed in untreated control nerve endings. The observation is of particular relevance when considering that the rationale for the use in therapy of antibodies recognizing a selected receptor protein is to silence the signalling elicited by that receptor [13]. Indeed, our previous studies demonstrated that antibodies raised against the NH2 sequence of the CCR5 receptor protein prevents efficiently the CCL5-induced CCR5-mediated releasing effects in both human and mouse cortical glutamatergic nerve endings [6,9]. However, the antagonist-like activity exerted by anti-NH2− CCR5 antibodies on the CCL5-induced releasing activity in both human and mouse cortical nerve endings turns to an overt facilitation of glutamate outflow when anti-NH2 −CCR5 treated synaptosomes are concomitantly exposed to complement. A comparable antagonism-like activity was retained by the anti-COOH-CCR5 antibody, when these antibodies were entrapped into synaptosomes by means of the so-called ‘entrapping technique,’ (i.e. an experimental approach that favours the entry of impermeant large size molecules within the terminals, [6]). Unlike the anti-NH2- CCR5 antibody pretreated synaptosomes, however, the complement-mediated releasing activity was not significantly modified in anti-COOH-CCR5 entrapped synaptosomes, nor was

E. Merega et al. / Neuroscience Letters 600 (2015) 50–55

53

Fig. 1. Effects of CCL5 and complement on the [3 H]D-aspartate release from human cortical synaptosomes pretreated with anti CCR5 or anti-␤ tubulin III antibodies. A: effect of CCL5 (1 nM) on [3 H]D-aspartate ([3 H]D-ASP) release from untreated (control, white bar) synaptosomes, from anti-NH2 -CCR5 antibody pretreated synaptosomes (black bar), from anti-COOH-CCR5 antibody entrapped synaptosomes (dark grey bar) and from in anti-␤ tubulin III antibody (light grey bar) pretreated synaptosomes. Results are expressed as percent of increase of [3 H]D-ASP release over basal and represent the mean ± SEM of four experiments run in triplicate (three superfusion chambers for each experimental condition). # P < 0.05 versus untreated synaptosomes B: effect of anti-CCR5 antibody raised against the NH2 (black bar) and the COOH (dark grey bar) terminus of the CCR5 receptor protein and of an anti-␤ tubulin III (light grey bar) on the complement-evoked [3 H]D-aspartate ([3 H]D-ASP) (white bar) from human cortical synaptosomes in superfusion. Results are expressed as percent of increase of [3 H]D-aspartate release over basal and represent the mean ± SEM of three to five experiments run in triplicate (three superfusion chambers for each experimental condition). # P < 0.05 versus 1:100 complement in control synaptosomes; * P < 0.05 versus 1:10 complement in control synaptosomes; ◦ P < 0.05 versus 1:100 complement in anti-NH2 -CCR5 antibody pretreated synaptosomes; + P < 0.05 versus 1:100 complement in anti-COOH-CCR5 antibody pretreated synaptosomes; ˆ P < 0.05 versus 1:100 complement in anti-␤ tubulin III antibody pretreated synaptosomes. C Effects of the complement and of the C1q deprived human serum on the release of [3 H]D-ASP from control (white bar) and from anti-NH2 -CCR5 antibody pretreated (black bar) human cortical synaptosomes. Results are expressed as percent of increase of [3 H]D-aspartate release over basal and represent the mean ± SEM of three to five experiments run in triplicate (three superfusion chambers for each experimental condition). # P < 0.05 versus 1:100 complement in control synaptosomes; * P < 0.05 versus 1:100 complement in anti-NH2 -CCR5 antibody pretreated synaptosomes.

it changed by pre-exposure to an unrelated antibody (anti-beta tubulin III, in the present case). This is consistent with the idea that the anti-COOH-CCR5 antibody/CCR5 protein complex cannot elicit the additive component of the complement-induced release described above. Whether the lack of efficacy of anti-COOH-CCR5 might depend on the inability of complement to either reach the inner side of the synaptosomal plasmamembrane, or to efficiently activate the complement itself, remains matter of discussion. The latter hypothesis however seems unlikely. Actually, although IgGs are weak activator of complement, our results indicate that IgG antibodies (i.e. the anti-NH2− CCR1 antibody) can reinforce the complement–mediated releasing effect. The second finding of the present work is that the complementinduced releasing activity recently described to occur in human and mouse cortical glutamatergic nerve endings still could be detected in synaptosomes exposed in superfusion to a serum lacking the C1q

protein. C1q is the complement component that, upon its binding to an antibody, sets in motion a domino effect leading to the of complement activation through the classic pathway. This finding sheds new light on the molecular events accounting for the complement-mediated releasing activity recently described [3], strongly supporting the involvement of an antibody-independent pathway in that event. Notably, the increased releasing efficiency could not be observed in synaptosomes pretreated with the anti-NH2- CCR5 antibody that were subsequently exposed to the C1q deprived serum, consistent with the idea that the classic antibody-dependent pathway of complement was involved in the antibody–mediated releasing effect. Of note, the complement component C1q has been proposed to cooperate with pentraxins at glutamatergic synapses. The C1q-pentraxin interaction is thought to have a role in the mechanisms dedicated to remodelling of synapses in the adult CNS, as well as in the structural changes

54

E. Merega et al. / Neuroscience Letters 600 (2015) 50–55

circulating anti-CCR5 molecules could be expected to gain access to the CNS and to bind CCR5 presynaptically located on glutamate nerve endings, then impeding the CCL5-mediated releasing efficacy. In these patients, therefore, the expected final outcome of the antibody administration could be the impairment of one of the mechanisms that control glutamate bioavailability in CNS [6]. Taking this into consideration, besides its role as neurotransmitter, glutamate also has chemotactic activity, the antagonist-like activity exerted by anti-CCR5 antibodies in CNS could be neuroprotective, delaying the progression of neuroinflammation. In addition to antibodies, other immune-competent molecules such as the complement, can cross the BBB and enter the CNS, where they could alter nerve terminals activity through different mechanisms. Based on the present results, it is possible that in anti-CCR5 antibody administered patients, complement could be activated because of the presence of the anti-NH2-CCR5 antibody/CCR5 receptor protein complex in nerve terminal plasmamembranes triggering an anomalous release of glutamate. This event might have detrimental effects in CNS in patients treated with neutralizing antibodies, i.e. favouring the onset of neuropsychiatric disorders and atypical seizures [17]. 5. Conclusion

Fig. 2. Effects of complement on the release of [3 H]D-aspartate from mouse cortical synaptosomes pretreated with anti-CCR5 or anti-␤ tubulin III antibodies. A: Effect of complement (dilution as indicated) on the release of [3 H]D-aspartate ([3 H]D-ASP) from untreated (white bar) from anti-CCR5-treated (black bar, antibody recognizing the NH2 terminus) and from anti-␤ tubulin III-treated (light grey bar) mouse cortical synaptosomes. Results are expressed as percent of increase of [3 H]D-aspartate release over basal and represent the mean ± SEM of three to five experiments run in triplicate (three superfusion chambers for each experimental condition). # P < 0.05 versus 1:1000 complement in control synaptosomes; + P < 0.05 versus ◦ 1:1000 complement in anti-NH2 -CCR5 antibody pretreated synaptosomes; P < 0.05 versus 1:1000 complement in anti-␤ tubulin III antibody pretreated synaptosomes; ˆ P < 0.05 versus 1:300 complement in control synaptosomes; $ P < 0.05 versus 1:300 complement in anti-NH2 -CCR5 antibody pretreated synaptosome; & P < 0.05 versus 1:300 complement in in anti-␤ tubulin III antibody pretreated synaptosomes. B: Effects of the C1q deprived human serum on the release of [3 H]D-ASP from control (white bar) and from anti-NH2 -CCR5 antibody pretreated (black bar) mouse cortical synaptosomes. Results are expressed as percent of increase of [3 H]D-aspartate release over basal and represent the mean ± SEM of three to five experiments run in triplicate (three superfusion chambers for each experimental condition). # P < 0.05 versus 1:100 complement in control synaptosomes; * P < 0.05 versus 1:100 complement in anti-NH2 -CCR5 antibody pretreated synaptosomes.

in synaptic connections that occur in neurodegenerative diseases [14]. The C1q-dependent glutamate release evoked by complement described here could have a role in these events. Neutralizing antibodies recognizing the outer sequence of the chemokine receptor subtype 5 (CCR5) have been proposed as therapeutics to prevent HIV-1 infection, based on the observation that the binding of the HIV-1 viral protein gp120 to this receptor is a prerequisite for HIV-1 cell infection [15]. Anti-CCR5 antibodies have successfully completed the proof of concept studies in HIV-infected individuals for their use in AIDS therapy [16]. Although antibodies have a limited access to the brain, inflammation increases the blood-brain barrier (BBB) permeability, allowing their entry into the CNS [17]. Here, antibodies target the related proteins, then altering their physiological functions. If this is the case, in anti-CCR5 antibody administered HIV-1+ patients,

To conclude, the events described here highlight a mechanism through which complement could have a deleterious impact when circulating antibodies reach CNS. We propose that the antibodymediated activation of complement described in this study might represent a common event in mammals, by which neutralizing antibodies (including those used in therapy) could exert, at least under specific pathological conditions (i.e. neuroinflammation), central unexpected deleterious effects at nerve terminals, leading to impaired central neurotransmission and deleterious excitotoxic processes that could worsen the progression of central diseases. Contribution of all the authors Elisa Merega carried out the release experiments and the analysis of results. Silvia Di Prisco was involved in the release experiments and analysis of data. Paolo Severi and Fotios Kalfas were involved in the neurosurgery. Anna Pittaluga planned the experiments, analysed by a statistical point of view the results and edited the manuscript. Conflict of interest The authors state no conflict of interest. Acknowledgements We are grateful to the Centre for AIDS Reagents, NIBSC HPA UK for antibodies to CCR5 receptor. We wish to thank Silvia E. Smith, Ph.D. (University of Utah, School of Medicine) for reviewing the manuscript and Maura Agate for editorial assistance. This work was supported by University of Genoa, Ricerca di Ateneo, Principal investigator Anna Pittaluga. References [1] P. Gasque, Complement: a unique innate immune sensor for danger signals, Mol. Immunol. 41 (2004) 1089–1098. [2] C. Ehrnthaller, A. Ignatius, F. Gebhard, et al., New insights of an old defense system: structure, function, and clinical relevance of the complement system, Mol. Med. 17 (2011) 317–329.

E. Merega et al. / Neuroscience Letters 600 (2015) 50–55 [3] E. Merega, S. Di Prisco, M. Lanfranco, et al., Complement selectively elicits glutamate release from nerve endings from different regions of mammal central nervous system, J. Neurochem. 29 (2014) 473–483. [4] O. Meucci, A. Fatatis, A.A. Simen, et al., Chemokines regulate hippocampal neuronal signaling and gp120 neurotoxicity, Proc. Natl. Acad. Sci. USA 95 (1998) 14500–14505. [5] R.S. Klein, K.C. Williams, X. Alvarez-Hernandez, et al., Chemokine receptor expression and signaling in macaque and human fetal neurons and astrocytes: implications for the neuropathogenesis of aids, J. Immunol. 163 (1999) 1636–1646. [6] V. Musante, F. Longordo, E. Neri, et al., Rantes modulates the release of glutamate in human neocortex, J. Neurosci. 28 (2008) 12231–12240. [7] S. Di Prisco, E. Merega, M. Milanese, et al., Ccl5-glutamate interaction in central nervous system: early and acute presynaptic defects in eae mice, Neuropharmacology 75 (2013) 337–346. [8] S. Di Prisco, E. Merega, M. Lanfranco, et al., Acute desipramine restores presynaptic cortical defects in murine experimental autoimmune encephalomyelitis by suppressing central ccl5 overproduction, Br. J. Pharmacol. 171 (2014) 2457–2467. [9] S. Di Prisco, M. Summa, V. Chellakudam, et al., Rantes-mediated control of excitatory amino acid release in mouse spinal cord, J. Neurochem. 121 (2012) 428–437.

55

[10] V. Musante, M. Summa, E. Neri, et al., The HIV-1 viral protein tat increases glutamate and decreases gaba exocytosis from human and mouse neocortical nerve endings, Cereb. Cortex 20 (2010) 1974–1984. [11] E. Luccini, V. Musante, E. Neri, M. Raiteri, A. Pittaluga, N-methyl-d-aspartate autoreceptors respond to low and high agonist concentrations by facilitating, respectively, exocytosis and carrier-mediated release of glutamate in rat hippocampus, J. Neurosci. Res. 85 (2007) 3657–3665. [12] M. Summa, S. Di Prisco, M. Grilli, et al., Presynaptic mGlu7 receptors control GABA release in mouse hippocampus, Neuropharmacology 66 (2013) 215–224. [13] D. Bellet, V. Dangles-Marie, Humanized antibodies as therapeutics, Med. Sci. 21 (2005) 1054–1062. [14] V.H. Perry, V. O’Connor, C1q: the perfect complement for a synaptic feast? Nat. Rev. Neurosci. 9 (2008) 807–811. [15] W.C. Olson, J.M. Jacobson, Ccr5 monoclonal antibodies for hiv-1 therapy, Curr. Opin. HIV Aids 4 (2009) 104–111. [16] J. Lalezari, G.K. Yadavalli, G. Richmond, et al., Safety, pharmacokinetics and antiviral activity of HGS004, a novel fully human IgG4 monoclonal antibody against CCR5 in HIV-1-infected patients, J. Infect. Dis. 197 (2008) 721–727. [17] P. Martinez-Martinez, P.C. Molenaar, M. Losen, et al., Autoantibodies to neurotransmitter receptors and ion channels: from neuromuscular to neuropsychiatric disorders, Front. Genet. 4 (2013) 181.