Ceftriaxone-mediated upregulation of the glutamate transporter GLT-1 contrasts neurotoxicity evoked by kainate in rat organotypic spinal cord cultures

Accepted Manuscript Title: Ceftriaxone-mediated upregulation of the glutamate transporter GLT-1 contrasts neurotoxicity evoked by kainate in rat organotypic spinal cord cultures Authors: Dzejla Bajrektarevic, Andrea Nistri PII: DOI: Reference:

S0161-813X(17)30033-5 http://dx.doi.org/doi:10.1016/j.neuro.2017.02.013 NEUTOX 2147

To appear in:

NEUTOX

Received date: Revised date: Accepted date:

15-11-2016 1-2-2017 27-2-2017

Please cite this article as: Bajrektarevic Dzejla, Nistri Andrea.Ceftriaxonemediated upregulation of the glutamate transporter GLT-1 contrasts neurotoxicity evoked by kainate in rat organotypic spinal cord cultures.Neurotoxicology http://dx.doi.org/10.1016/j.neuro.2017.02.013 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Ceftriaxone-mediated upregulation of the glutamate transporter GLT-1

contrasts neurotoxicity evoked by kainate in rat organotypic spinal cord

cultures

Dzejla Bajrektarevic1 and Andrea Nistri2

Neuroscience Department, International School for Advanced Studies (SISSA), Trieste, Italy

Address correspondence to:

Prof. Andrea Nistri

Neuroscience Department, International School for Advanced Studies (SISSA)

Via Bonomea 265, Trieste, Italy

Phone: +39-40-3756518, Fax: +39-40-3756502

E-mail: [email protected]

2

Highlights 

The antibiotic ceftriaxone enhanced glutamate transporter expression



Ceftriaxone pretreatment strongly contrasted excitotoxicity in spinal cultures



Inhibiting glutamate uptake increased excitotoxic damage



Neither ceftriaxone nor uptake inhibition exerted neurotoxicity when applied alone

Abstract Excitotoxicity is a major pathological trigger of neurodegenerative diseases like amyotrophic lateral sclerosis. This process is caused by excessive release of the transmitter glutamate that overwhelms the capacity of astroglia transporters to maintain a low extracellular level of this aminoacid and strongly stimulates neurons. Using an in vitro model of rat organotypic spinal slice culture, we explored if the excitotoxicity caused by the potent glutamate analogue kainate, widely employedas a paradigm to evoke neurotoxicity in the central nervous system, was prevented by the antibiotic ceftriaxone known to enhance glutamate transporter expression. We also tested if excitotoxicity was made worse by inhibiting glutamate uptake with dl-threo-βbenzyloxyaspartate (TBOA). These experiments were aimed at clarifying the relative contribution to neurotoxicity by kainate-activation of glutamate receptors or kainate-mediated release of glutamate. Neither ceftriaxone nor TBOA had adverse effects. Ceftriaxone (10 µM; 3 days) significantly decreased delayed cell death induced by kainate (100 µM; 1 h) and limited neuronal damage especially to motoneurons. This effect was associated to stronger astrocytic immunostaining of the glutamate transporter GLT-1. Conversely, pharmacological inhibition of glutamate uptake with TBOA was per se unable to induce neurotoxicity, yet it intensified cell

3

death evoked by kainate. These data indicate that kainate-mediated glutamate release was critical to damage to neurons, an effect prevented by up regulating glutamate uptake. These data suggest that modulating glutamate uptake is an important strategy to preserve neuronal networks.

Keywords: excitotoxicity; astrocyte; neuroprotection; motoneuron

1

β-lactam

antibiotic;

glutamate

uptake

blocker;

Introduction

Glutamate is the predominant excitatory neurotransmitter in the mammalian central nervous system (CNS) and is also essential to support synaptic plasticity and development (Fontana, 2015; Yamada et al., 1998). On the other hand, excessively large release of glutamate caused by network hyperexcitability and/or cell damage induces excitotoxicity (Gardoni and Di Luca, 2006), namely a pathological process whereby overexcitation of central neurons leads to extensive neuronal loss and is believed to underlie the early phase of diseases like amyotrophic lateral sclerosis (ALS) (Rothstein et al., 1995; Van Den Bosch et al., 2006), Alzheimer´s disease (Li et al., 1997; Scott et al., 2011), stroke (Chao et al., 2010; Martin et al., 1997), traumatic brain injury (TBI) (Yi and Hazzel, 2006), Huntington´s disease (Estrada Sanchez and Rebec, 2012), neuropathic pain (Tao et al., 2005), Parkinson´s disease (Ambrosi et al., 2014) and spinal cord injury (SCI; Park et al., 2004). Kainate is a widely used neurotoxin to produce models of neurodegeneration ranging from epilepsy to SCI (Ben-Ari and Cossard, 2000; Taccola et al., 2008). Experimental excitotoxicity can be induced by applying kainate (a stable glutamate analogue) to the rat spinal cord in vitro (Taccola et al., 2008) and organotypic spinal cultures (Mazzone et al., 2010), thus, providing simple models for investigating the initial phase of SCI or neurodegeneration.

4

In order to regulate the extracellular glutamate concentration, glutamate levels are physiologically kept low by membrane transporters expressed by glia and neurons (Furuta et al., 1997; Kim et al., 2011; Yamada et al., 1998). Unsolved questions, however, remain such as the role played by uptake in limiting the extension of excitotoxicity and the relative contribution to excitotoxicity by direct activation of glutamate receptors (for instance by kainate) and by the release of endogenous glutamate (when uptake should presumably limit its extracellular concentration). Furthermore, any beneficial effect of enhanced glutamate uptake in contrasting excitotoxicity should be compared with the effects of increasing GABA-mediated inhibition (Bajrektarevic and Nistri, 2016) or depressing glutamate release (Mazzone and Nistri, 2011). Among the five subtypes of glutamate transporters, GLT-1 is responsible for more than 90% of total glutamate uptake (Furuta et al., 1997; Kim et al., 2011). Earlier studies have indicated that β-lactam antibiotics, such as ceftriaxone, can upregulate GLT-1 to provide neuroprotection in animal models of excitotoxic stress (Rothstein et al., 2005), such as stoke (Thone-Reineke et al., 2008), ALS (Rothstein et al., 2005), TBI (Goodrich et al, 2013), Parkinson´s disease (Leung et al., 2012) and Huntington´s disease (Rebec, 2013). In particular, ceftriaxone can decrease neuronal loss after glutamate uptake block in spinal cultures and in a genetic model of ALS (Rothstein et al., 2005). In the rat, repeated treatment with ceftriaxone upregulates expression of GLT-1 to ameliorate chronic pain (Ramos et al., 2010) and to facilitate recovery after traumatic SCI (Tajkey et al., 2014). The aims of the present study with organotypic spinal cultures were to clarify the effect of ceftriaxone pretreatment on glutamate uptake expression and kainate-mediated excitotoxicity, and to find out the impact of uptake inhibition on excitotoxicity. 2

Materials and methods

5

2.1 Experimental procedures 2.1.1 Preparation and maintenance of organotypic cultures Organotypic slice cultures of spinal cord were prepared form Wistar rat at 13 days of gestation as previously reported (Bajrektarevic and Nistri, 2016; Mazzone et al., 2010). The procedures were in accordance with the European Union guidelines and were approved by the Scuola Internazionale Superiore di Studi Avanzati (Trieste, Italy) Ethics Committee. All efforts were made to reduce the number of animals used and to minimize animal suffering. Slices were maintained in a medium containing 82% Dulbecco’s Modified Eagle’s medium, 8% sterile water for tissue culture, 10% fetal bovine serum, osmolarity 300 mOsm, pH 7.35. From each dissection, 30–40 slices were prepared from the thoracic as well as the lumbar segments, and kept in culture for 22 days in vitro (DIV) before use. Dulbecco’s Modified Eagle’s medium with high glucose (DME/HIGH) was purchased from Euroclone (Devon, UK), to which the following compounds were added; gentamicin (Sigma, Milan, Italy), fetal calf serum (FBS; Invitrogen, Carlsbad, CA, USA), nerve growth factor (D.B.A. Italia, Segrate, Italy), chicken plasma (Innovative, Novi, MI, USA), and thrombin (Merck, (Darmstadt, Germany). It is noteworthy that, without detectable consequence on culture viability, the antibiotic penicillin was omitted from the culture medium since it stimulates GLT-1 expression (Rothstein et al., 2005).

2.1.2 Drugs The β-lactam antibiotic ceftriaxone (Rocephin, Roche, Basel, Switzerland) was dissolved in sterile water at 0.29 g/mL and stored at - 20 °C because lower storage temperature improved the chemical stability of ceftriaxone for up to 76 days (Diego et al., 2010). Ceftriaxone solution (10

6

µM in complete medium) was freshly prepared immediately before treatment. This dose was chosen from previous studies which have reported that 10 µM ceftriaxone markedly enhances GLT-1 expression in organotypic spinal cord slices (Rothstein et al., 2005). The glutamate uptake blocker TBOA (Tocris, Bristol, UK) was used at 50 µM concentration suitable as an experimental protocol to damage brainstem motor network (Cifra et al., 2012a; Corsini et al., 2016).

2.1.3 Protocol to study excitotoxicity Figure 1 A summarizes the experimental protocol, used at 22 DIV, for the following four groups: control, ceftriaxone, kainate, and pre-treatment with ceftriaxone before kainate. Controls were in each experiment untreated sister cultures maintained in vitro for 4 days. In the ceftriaxone group, this antibiotic (10 µM) was applied once a day for 3 days and then washed out with complete medium for 23 h. In the kainate group, kainate (100 µM; Ascent Scientific, Cambridge, UK) was dissolved in complete medium, applied for 1 h and washed out with complete medium for 23 h. In the ceftriaxone pre-treatment group the antibiotic was applied once a day for 3 days (with daily medium change), washed out and replaced with the kainate solution for 1 h and washed out again with complete medium for 23 h. The 100 µM concentration of kainate was chosen because it leads to irreversible loss of neurons and functional network deficit in the in vitro spinal cord (Mazzone et al., 2010).

2.1.4 Protocol to test the effect of glutamate uptake blocker TBOA Figure 1 B summarizes the experimental protocol used for testing the effect of glutamate uptake blocker TBOA. The following experiments were run in parallel: control group, TBOA group,

7

kainate group, and TBOA co-applied with kainate group. Control was in each experiment untreated sister cultures maintained in vitro for 2 days. In the TBOA group, this drug (50 µM) was applied for 4 h and then washed out with complete medium for 44 h. In the kainate group, kainate was applied for 1 h and washed out with complete medium for 47 h. In the co-applied TBOA and kainate group, these drugs were applied for 1 h and washed out with complete medium for 47 h.

2.1.5 Immunofluorescence of organotypic cultures Organotypic slices were processed for immunohystochemistry as detailed recently (Bajrektarevic and Nistri, 2016, Cifra et al., 2012b, Mazzone et al., 2010). Cryoprotective slices were cryostatsectioned with cryostat OTF 5000 (Bright Instruments, Luton, UK) to obtain, on average, 6–8 tissue sections (16 µm thick) from each organotypic culture, which were collected sequentially on histology slides (Mazzone et al., 2013; Perez-Gomez and Tasker, 2012). After blocking with 3% FBS, 3% bovine serum albumin (BSA; Sigma, Dorset, UK) and 0.3% Triton (Sigma, Dorset, UK) in PBS (blocking solution) for 1 h at room temperature, primary antibodies (prepared in blocking solution) were applied overnight at 4 °C. Thus, we used mouse monoclonal NeuN (1:200; Millipore, Billerica, MA, USA) for neurons, SMI32 (mouse monoclonal, 1:1000 dilution; Covance, Berkeley, CA) for motoneurons, S100 (rabbit polyclonal, 1:1000 dilution; Dako, Glostrup, Denmark) for astrocytes, and GLT-1 (rabbit polyclonal, 1:1000 dilution; Abcam, Cambridge, UK) for this glutamate transporter. Previous studies have validated the use of this GLT-1 antibody with Western immunoblotting of organotypic spinal cultures (Rothstein et al., 2005). The primary antibodies were visualized using corresponding secondary fluorescent antibody (Alexa Fluor 488 or 546, 1:500, Invitrogen, Carlsbad, CA, USA). To visualize cell

8

nuclei, slices were incubated with 4´, 6-diamidino-2-phenylindole (DAPI; 1 µg/ml; Sigma, Dorset, UK) for 2 h and mounted using a fluorescence mounting medium (DAKO, Milan, Italy). The average number of NeuN positive cells and pyknotic cells (identified with DAPI nuclear staining as condensed chromatin in the nucleus) was obtained by counting 6–8 tissue sections in three regions of interest (ROIs), namely dorsal, central and ventral (Mazzone et al., 2010) with a Leica DM6000 (40x magnification) and quantified with ‘eCELLence’ software (Glance VisionTech, Trieste, Italy). SMI32 immunoreactive motoneurons were identified with the following criteria, namely cell body clustered around the ventral median fissure and somatic diameter ≥20 µm (Avossa et al., 2003, Mazzone et al., 2010). The number of SMI32 positive cells was obtained by counting 6-8 tissue sections and visualized with DM6000 (20x magnification). The GLT-1 signals were collected in three ROIs as mean fluorescence intensity with densitometry analysis using a Zeiss Axioskop2 microscope and MetaVue software (Molecular Devices, Sunnyvale, CA, USA). No difference in the basal expression of GLT-1 among ventral, central and dorsal ROIs was observed.

2.2 Statistics Data were analyzed using SigmaStat 3.11 (Systat Software, Chicago, IL, USA). Results are expressed as mean ± SD (standard deviation); N refers to the number of experiments, while n refers to the number of organotypic culture slices. As directed by the software, after distinguishing between parametric and non-parametric data, parametric values were analyzed with one-way ANOVA for multiple comparisons, with Tukey-Kramer post hoc test. Nonparametric values were analyzed with the Kruskal–Wallis test. When two groups were compared, the Student’s t-test for parametric data or the Mann–Whitney Rank Sum test for non-parametric

9

data was applied. The accepted level of significance was always <0.05.

3

Results

3.1 Ceftriaxone decreased cell damage after kainate application Ceftriaxone (10 µM for 3 days; see Fig. 1A) did not produce toxic effects as the number of pyknotic nuclei in the three test ROIs remained negligible as exemplified in Fig. 2B. Likewise, the number of NeuN-positive neurons was unaffected by this treatment (Fig. 2 D). We next explored if the 3 day pre-treatment with ceftriaxone might be neuroprotective against the strong excitotoxicity evoked by 100 µM kainate (1 h; see Fig. 2A for widespread pyknosis). On average, the number of pyknotic cells induced by 100 µM kainate in ceftriaxone-treated slices was significantly smaller in all ROIs (Fig. 2 A, B), an observation associated with a higher number of residual neurons vs. the protocol with kainate alone (Fig. 2 C, D). Complete neuroprotection was, however, not achieved as the number of surviving neurons remained below control. Because motoneurons are very vulnerable to excitotoxic damage (Mazzone et al., 2010), we also analyzed if pretreatment with ceftriaxone contrasted the kainate-induced fall in motoneuron number (examined with SMI32 immunostaining; Fig. 3A, B). In comparison to the 2/3rd loss of motoneurons, the antibiotic pretreatment enabled large preservation of the motoneuron number though not up to control value (Fig. 3 A, B). It may be noted that the SMI32 antibody did not produce any significant immunoreactivity in the central and dorsal region of the spinal slice (Fig. 3 C, D; see also Cifra et al., 2012b) confirming its selectivity for motoneurons (Avossa et al., 2003).

10

3.2 Effect of ceftriaxone on GLT-1 localization and expression We looked for the localization and expression of GLT-1 in S100-immunolabelled astrocytes that are the principal cells to express this membrane transporter (Kim et al., 2011). In accordance with this notion, Fig. 4 D, E shows lack of GLT-1 immunoreactivity in neurons or motoneurons (stained in pseudo green colour). Fig. 4 A, B, C indicates that ceftriaxone enhanced the GLT-1 signal averaged over the three ROIs when compared to control. Interestingly, 100 µM kainate per se significantly increased the GLT-1 immunoreactivity and, when this agent was applied after pretreatment with ceftriaxone for 3 days, a further rise in GLT-1 signal was apparent (Fig. 4 A, B, C). These results indicate that increased GLT-1 expression by ceftriaxone was an important phenomenon to confer neuroprotection against excitotoxic stress. 3.3 Effects of TBOA on cell loss in the presence of kainate Since pharmacological upregulation of GLT-1 could contrast excitotoxicity, we further looked for the proof of principle that blocking the glutamate uptake would be highly toxic. This test was performed by exposing the organotypic slice cultures to TBOA for 4 h at 50 µM concentration, a protocol previously shown in our laboratory to be efficient for suppressing glutamate uptake in rat brainstem slices (Cifra et al., 2012a; Corsini et al., 2016). Fig. 5 B, D shows that TBOA alone (4 h) induced no significant cell damage after 2 days (Fig. 5 B, D). Nevertheless, when TBOA was co-applied with 100 µM kainate, a significant increase in the percentage of pyknosis compared to kainate alone was detected in all ROIs (Fig. 5A, B). Neuronal numbers were also decreased in all three ROIs in comparison with the effect of kainate alone (Fig. 5 C, D), suggesting synergy of neurotoxicity.

4

Discussion

11

The principal finding of the present report is the demonstration that pretreatment with the antibiotic ceftriaxone strongly protected spinal networks from kainate-mediated excitotoxicity throughout the three main regions of the spinal slice and that this effect was associated to a significant rise in the astrocytic expression of GLT-1. Conversely, inhibiting glutamate uptake intensified kainate neurotoxicity. While former investigations have shown that upregulation of GLT-1 is neuroprotective against brain or spinal trauma in vivo (Cui et al., 2014; Goodrich et al., 2012; Leung et al., 2012; Tajkey et al., 2014; Thone-Reineke at al. 2008), Rothstein et al. (2005) have shown that, on organotypic spinal cord cultures, ceftriaxone and related antibiotics are neuroprotective when pre-applied for a few days before administration of an inhibitor of glutamate uptake. The present study has employed a different protocol that, by applying the potent glutamate analogue kainate, is designed to mimic in vitro the transient excitotoxic stimulation thought to occur at the time of acute SCI (Kuzhandaivel et al., 2011; Mazzone et al., 2010). Thus, the present model is suitable to explore potential methods to pharmacologically constrain neurotoxic damage and to limit secondary evolution of the injury that slowly develops even when the primary cause of lesion is removed (Kuzhandaivel et al., 2011; Mazzone and Nistri, 2013). In view of the common use of kainate as neurotoxin to lesion central neurons, the current results may be useful to interpret data with other preparations as well.

4.1 Glutamate uptake systems and excitotoxicity Glutamate is cleared from the extracellular space by Na+-dependent transporters located in neurons and astrocytes (Camacho and Massieu, 2006; Sheldon et al., 2007). The activity of glutamate transporters depends on electrochemical gradient generated by the membrane Na +/K+

12

ATPases. During energy failure due, for instance, to acute excitotoxicity, ionic gradients collapse with inability to transport glutamate intracellularly (Camacho and Massieu, 2006). In various CNS disorders the expression of the main glutamate transporter GLT-1 is downregulated with subsequent imbalance in glutamate homeostasis, a process which contributes to excitotoxicity (Cui et al., 2014; Goodrich et al., 2012; Leung et al., 2012; Thone-Reineke at al. 2008). Astrocytes largely express GLT-1 and, therefore, play a central role in regulating extracellular glutamate homeostasis (Putatunda et al., 2014). The early effects of excitotoxicity are multifarious as they include activation of glia, loss of GABAergic neurons, decreased uptake of glutamate (and its conversion to glutamine), and Ca2+ overload eventually contributing to delayed, widespread cell damage (Putatunda et al., 2014; Verma et al., 2010). Because upregulation of GLT-1 provides neuroprotection in different models (Chu et al., 2007; Hota et al., 2008: Lee et al., 2008), we explored its neuroprotective potential in spinal cord networks using ceftriaxone, an antibiotic previously shown to exert neuroprotection via upregulation of GLT-1 in organotypic spinal cultures (Rothstein et al., 2005). Former studies with ceftriaxone have employed a variety of protocols with longer pre- or posttreatment (5-7 days) and higher doses (200 mg/kg/day) (Cui et al., 2014; Goodrich et al., 2012; Leung et al., 2012; Tajkey et al., 2014; Thone-Reineke at al. 2008) than those used here. Our protocol involved 3 day pre-treatment with 10 µM ceftriaxone, since it was previously shown with Western immunoblotting that ceftriaxone increased the expression of GLT-1 after 48 h (Rothstein et al., 2005). In line with these results, the present study validated that immunohistochemical expression of GLT-1 was significantly larger after ceftriaxone. Furthermore, using an immunohistochemical approach, we confirmed that GLT-1 was not expressed by neurons, and therefore confined to astroglia. Hence, our results demonstrate that

13

beneficial actions by ceftriaxone do not necessarily require very high doses and/or very long treatments. The present report is based on the use of organotypic spinal cord cultures originally obtained from embryonic rats. This is an experimentally advantageous system that enables to investigate neuronal circuit formation and resilience for weeks as the basic cytoarchitecture and the dorsalventral orientation of a spinal segment are preserved (Avossa et al., 2003; Mazzone and Nistri, 2013, 2014). Notwithstanding the limitations intrinsic to the use of a tissue culture, our previous investigation has analyzed in detail the developmental maturation of rat organotypic spinal cord cultures and, when compared it with the characteristics of the in situ spinal cord, observed a similar developmental profile for neurons and motoneurons (Cifra et al., 2012b).

4.1 Neuroprotective effect of GLT-1 The present data show that ceftriaxone or the uptake inhibitor TBOA had no apparent toxic action when applied alone to organotypic cultures. While the first result accords with the safety of ceftriaxone in a 6-month rodent toxicology study (Ratti et al., 2015), the lack of TBOA toxicity implies that in our experimental control conditions basal release of glutamate was rather low (Mazzone and Nistri,2011) and that perhaps diffusion of this transmitter through the extracellular space was the main process regulating excitatory neurotransmission and allowing glutamate spillover to nearby neurons and glia (Kullmann et al., 1998; Kullmann et al., 1999; Sykova, 2004). It was, however, clear that organotypic cultures expressed GLT-1, that its level was readily augmented 24 h after the excitotoxic stress by kainate, and that was further increased by ceftriaxone. Conversely, TBOA enhanced kainate evoked pyknosis. Thus, GLT-1 was endowed with strong adaptation to intense changes in the microenvironment like for example the rapid surge in glutamate release elicited by the excitotoxic stimulation (Mazzone et al., 2014).

14

Interestingly, although TBOA did not further decrease neuronal numbers, it enhanced kainateelicited pyknosis and neuronal loss throughout the three ROIs. One interpretation of the present observation may be that the neurotoxic effect of kainate was partly contrasted by GLT-1 activity: once the latter was pharmacologically inhibited by TBOA, the effect due to kainate direct depolarization plus kainate-mediated glutamate release became manifested and more intensively toxic. Spinal motoneurons are very vulnerable to excitotoxicity (Taccola et al., 2008; Mazzone et al., 2010) and their loss is a major determinant of deficit in locomotor network function (Kuzhandaivel et al., 2011). Indeed, previous studies have reported that, in a number of ALS patients, there is decreased glutamate transport (Rothstein et al., 1992) due to ineffective operation of GLT-1 (Rothstein et al., 1995) in the spinal cord. In support of this notion, other studies have demonstrated that long-lasting administration of ceftriaxone prevents motoneurons loss due to chronic block of GLT-1 with the pharmacological inhibitor THA (Rothstein et al., 2005), improves motor function and axonal regeneration after rat experimental SCI (Tajkey et al., 2014), increases GLT-1 expression and suppresses glial activation accompanying neuropathic pain (Ramos et al., 2014). Thus, upregulation of GLT-1 restores the normal concentration of extracellular glutamate and prevents the development of excitotoxicity in the rat (Nicholson et al., 2014).

4.2 Conclusions A multicenter clinical trial is currently in progress to test the effect of ceftriaxone on ALS patients: initial results indicate good tolerability when the antibiotic was administered daily to exceed the cerebrospinal fluid concentration of 1 µM (Berry et al., 2013). The study is now

15

investigating if this antibiotic prolongs survival and slows decline in function due to ALS (https://clinicaltrials.gov/ct2/show/NCT00349622). Our own observations provide further support to the principle of boosting glutamate uptake with ceftriaxone to antagonize neuronal loss.

Acknowledgements We thank Drs. Beatrice Pastore for helpful support with culture preparation, and Prof. Miranda Mladinic for her help with immunohistochemistry. This work was supported by an intramural grant from SISSA

16

References Ambrosi G, Cerri S, Blandini F. A further update on the role of excitotoxicity in the pathogenesis of Parkinson's disease. J Neur Transm 2014; 121:849-859. Avossa D, Rosato-Siri MD, Mazzarol F, Ballerini L. Spinal circuits formation: a study of developmentally regulated markers in organotypic cultures of embryonic mouse spinal cord. Neuroscience 2003; 122: 391- 405. Bajrektarevic D, Nistri A. Delayed application of the anesthetic propofol contrasts the neurotoxic effects of kainate on rat organotypic spinal slice cultures. Neurotoxicology 2016;54:1-10. Ben-Ari Y, Cossart R. Kainate, a double agent that generates seizures: two decades of progress. Trends Neurosci 2000; 23:580-587. Berry JD, Shefner JM, Conwit R, Schoenfeld D, Keroack M, Felsenstein D, Krivickas L, David WS, Vriesendorp F, Pestronk A, Caress JB, Katz J, Simpson E, Rosenfeld J, Pascuzzi R, Glass J, Rezania K, Rothstein JD, Greenblatt DJ, Cudkowicz ME. Design and initial results of a multi phase randomized trial of ceftriaxone in amyotrophic lateralsclerosis. PLoS One 2013;8:e61177. Camacho A, Massieu L. Role of glutamate transporters in the clearance and release of glutamate during ischemia and its relation to neuronal death. Arch Med Res 2006; 37:11-18. Chao XD, Fei F, Fei Z. The role of excitatory amino acid transporters in cerebral ischemia. Neurochem Res 2010; 35:1224-1230. Cifra A, Nani F, Nistri A. Riluzole is a potent drug to protect neonatal rat hypoglossal motoneurons in vitro from excitotoxicity due to glutamate uptake block. Eur J Neurosci 2012a; 33:899-913. Cifra A, Mazzone GL, Nani F, Nistri A, Mladinic M. Postnatal developmental profile of neurons and glia in motor nuclei of the brainstem and spinal cord, and its comparison with organotypic slice cultures. Dev Neurobiol 2012b; 72: 1140-1160. Corsini S. Tortora M. Nistri A. Nicotinic receptor activation contrasts pathophysiological bursting and neurodegeneration evoked by glutamate uptake block on rat hypoglossal motoneurons. J Physiol 2016; 594:6777-6798. Cui C, Cui Y, Gao J, Sun L, Wang Y, Wang K, Li R, Tian Y, Song S, Cui J. Neuroprotective effect of ceftriaxone in a rat model of traumatic brain injury. Neurol Science 2014;35:695-700. Diego M, Godoy G, Mennickent S. Chemical stability of ceftriaxone by a validated stabilityindicating liquid chromatographic method. J Chil Chem Soc2010; 55: 335–337. Estrada-Sanchez AM, Rebec GV. Corticostriatal dysfunction and glutamate transporter 1 (GLT1) in Huntington's disease: interactions between neurons and astrocytes. Basal Ganglia 2012; 2:5766.

17

Fontana AC. Current approaches to enhance glutamate transporter function and expression. J Neurochem 2015; 134:982-1007. Furuta A, Rothstein JD, Martin LJ. Glutamate transporter protein subtypes are expressed differentially during rat CNS development. J Neurosci 1997; 17:8363–8375. Gardoni F, Di Luca M. New targets for pharmacological intervention in the glutamatergic synapse. Eur J Pharmacol 2006; 545:2-10. Goodrich GS, Kabakov AY, Hameed MQ, Dhamne SC, Rosenberg PA, Rotenberg A. Ceftriaxone treatment after traumatic brain injury restores expression of the glutamate transporter, GLT-1, reduces regional gliosis, and reduces post-traumatic seizures in the rat. J Neurotrauma 2013; 30:1434-1441. Hota SK, Barhwal K, Ray K, Singh SB, Ilavazhagan G. Ceftriaxone rescues hippocampal neurons from excitotoxicity and enhances memory retrieval in chronichypobaric hypoxia. Neurobiol Learn Mem 2008; 89:522-532. Kim K, Lee SG, Kegelman TP, Su ZZ, Das SD, Dash R, Dasgupta S, Barral PM, Hedvat M, Diaz P, Reed JC, Stebbins JL, Pellecchia M, Sarkar D, Fisheri PB. Role of excitatory amino acid transporter-2 (EAAT2) and glutamate in neurodegeneration: opportunities for developing novel therapeutics. J Cell Physiol 2011; 226: 2484–2493. Kullmann DM, Asztely F. Extrasynaptic glutamate spillover in the hippocampus: evidence and implications. Trends Neurosci 1998; 21:8-14. Kullmann DM, Min MY, Asztely F, Rusakov DA. Extracellular glutamate diffusion determines the occupancy of glutamate receptors at CA1 synapses in the hippocampus. Phil Trans Roy Soc London B 1999; 354:395-402. Kuzhandaivel A, Nistri A, Mazzone GL, Mladinic M. Molecular mechanisms underlying cell death in spinal networks in relation to locomotor activity after acute injury in vitro. Front Cell Neurosci 2011; 5:9. Lee SG, Su ZZ, Emdad L, Gupta P, Sarkar D, Borjabad A, Volsky DJ, Fisher PB. Mechanism of ceftriaxone induction of excitatory amino acid transporter-2 expression and glutamate uptake in primary human astrocytes. J Biol Chem 2008; 283:13116-13123. Leung TC, Lui CN, Chen LW, Yung WH, Chan YS, Yung KK. Ceftriaxone ameliorates motor deficits and protects dopaminergic neurons in 6 hydroxydopamine-lesioned rats. ACS Chem Neuroscience 2012; 3: 22-30. Li S, Mallory M, Alford M, Tanaka S, Masliah E. Glutamate transporter alterations in Alzheimer disease are possibly associated with abnormal APP expression. J Neuropathol Exp Neurol 1997; 56:901–911.

18

Martin LJ, Brambrink AM, Lehmann C, Portera-Cailliau C, Koehler R, Rothstein J, Traystman RJ. Hypoxia-ischemia causes abnormalities in glutamate transporters and death of astroglia and neurons in newborn striatum. Ann Neurol 1997; 42:335–348. Mazzone GL, Margaryan G, Kuzhandaivel A, Nasrabady SE, Mladinic M, Nistri A. Kainateinduced delayed onset of excitotoxicity with functional loss unrelated to the extent of neuronal damage in the in vitro spinal cord. Neuroscience 2010; 168: 451–462. Mazzone GL, Nistri A. Electrochemical detection of endogenous glutamate release from rat spinal cord organotypic slices as a real-time method to monitor excitotoxicity. J Neurosci Meth 2011; 197: 128–132. Mazzone GL, Nistri A. Excitotoxic cell death induces delayed proliferation of endogenous neuroprogenitor cells in organotypic slice cultures of the rat spinal cord. Cell Death Dis 2013; 4:e902 Mazzone GL, Nistri A. S100β as an early biomarker of excitotoxic damage in the spinal cord organotypic cultures. J Neurochem 2014; 130: 598-604. Nicholson KJ, Gilliland TM, Winkelstein BA. Upregulation of GLT-1 by treatment with ceftriaxone alleviates radicular pain by reducing spinal astrocyte activation and neuronal hyper excitability. J Neurosci Res 2014; 92:116-29. Park E, Velumian AA, Fehlings MG. The role of excitotoxicity in secondary mechanisms of spinal cord injury: a review with an emphasis on the implications for white matter degeneration. J Neurotrauma 2004; 21:754-774. Pérez-Gómez A, Tasker RA. Enhanced neurogenesis in organotypic cultures of rat hippocampus after transient subfield-selective excitotoxic insult induced by domoic acid. Neuroscience 2012; 208: 97–108. Putatunda R, Hala TJ, Chin J, LeporeAC. Chronic at-level thermal hyperalgesia following rat cervical contusion spinal cord injury is accompanied by neuronal and astrocyte activation and loss of the astrocyte glutamate transporter, GLT1, in superficial dorsal horn. Brain Res 2014;1581:64-79. Ramos KM, Lewis MT, Morgan KN, Crysdale NY, Kroll JL, Taylor FR, Harrison JA, Sloane EM, Mairand SF, Watkins LR. Spinal upregulation of glutamate transporter GLT-1 by ceftriaxone: therapeutic efficacy in a range of experimental nervous system disorders. Neuroscience 2010; 169: 1888–1900. Ratti E, Berry JD, Greenblatt DJ, Loci L, Ellrodt AS, Shefner JM, Cudkowicz ME. Preclinical rodent toxicity studies for long term use of ceftriaxone. Toxicol Rep 2015; 2:1396-1403. Rebec GV. Dysregulation of corticostriatal ascorbate release and glutamate uptake in transgenic models of Huntington's disease. Antioxidants Redox Sign 2013; 19:2115-2128.

19

Rothstein JD, Martin LJ, KunclR W. Decreased glutamate transport by the brain and spinal cord in amyotrophic lateral sclerosis. New Eng J Med 1992; 326:1464-1468. Rothstein JD, Van Kammen M, Levey AI, Martin LJ, Kuncl RW. Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann Neurol 1995; 38:73–84. Rothstein JD, Patel S, Regan MR, Haenggeli C, Huang YH, Bergles DE, Jin L, Hoberg MD, Vidensky S, Chung DS, Toan SV, Bruijn LI, Su ZZ, Gupta P, Fisher PB. β-lactam antibiotics offer neuroprotection by increasing glutamate transporter expression. Nature 2005; 433:73-77. Scott HA, Gebhardt FM, Mitrovic AD, Vandenberg RJ, Dodd PR. Glutamate transporter variants reduce glutamate uptake in Alzheimer's disease. Neurobiol Aging 2011; 32: 553.e1-11. Sheldon AL, Robinson MB. The role of glutamate transporters in neurodegenerative diseases and potential opportunities for intervention. Neurochem Int 2007; 51:333-355. Syková E. Extrasynaptic volume transmission and diffusion parameters of the extracellular space. Neuroscience 2004;129:861-876. Taccola G, MargaryanG, Mladinic M, Nistri A. Kainate and metabolic perturbation mimicking spinal injury differentially contribute to early damage of locomotor networks in the in vitro neonatal rat spinal cord. Neuroscience 2008; 155: 538–555. Tajkey J, Ramazani A, Biglari A, Mazlomzadeh S, Asl BH. Ceftriaxone improves neuron protection and functional recovery in rat model of spinal cord injury. Ann Res Rev Biology 2014; 4: 1958-1967. Tao YX, Gu J, Stephens RL Jr. Role of spinal cord glutamate transporter during normal sensory transmission and pathological pain states. Mol Pain 2005; 21:1-30. Thone-Reineke C, Neumann C, Namsolleck P. The beta-lactam antibiotic, ceftriaxone, dramatically improves survival, increases glutamate uptake and induces neurotrophins in stroke. J Hypertension 2008; 26:2426-2435. Van Den Bosch L, Van Damme P, Bogaert E, Robberecht W. The role of excitotoxicity in the pathogenesis of amyotrophic lateral sclerosis. Biochimica Biophys Acta 2006; 1762:1068-1082. Verma R, Mishra V, Sasmal D, RaghubirR. Pharmacological evaluation of glutamate transporter 1 (GLT-1) mediated neuroprotection following cerebral ischemia/reperfusion injury. Eur J Pharmacol 2010;638:65-71. Yamada K, Watanabe M, Shibata T, Nagashima M, Tanaka K, Inoue Y. Glutamate transporter GLT-1 is transiently localized on growing axons of the mouse spinal cord before establishing astrocytic expression. J Neurosci1998; 18:5706-5713.

20

Yi JH, Hazell AS. Excitotoxic mechanisms and the role of astrocytic glutamate transporters in traumatic brain injury. Neurochem Int 2006; 48: 394-403.

21

Figure legends Figure 1. Experimental Protocol (A) Schematic diagram to illustrate experimental protocols for excitotoxic insult (KA = kainate) and the potential neuroprotective effect by ceftriaxone (CEF). (B) Schematic diagram to illustrate experimental protocols for effects of glutamate uptake blocker TBOA.

22 Figure 2.Pyknosis and neuronal loss could be counteracted by ceftriaxone pretreatment (A, C) Examples of DAPI-stained cells showing pyknosis (condensed chromatin) (A) and NeuN-immunoreactive positive cells (C) in the ventral ROI in control condition, or after kainate (KA; 100 µM) application for 1h, or after 3 days pre-treatment with ceftriaxone (CEF; 10 µM) followed by kainate 100 µM application for 1 h. (B, D) Histograms showing average percent of pyknosis (B) and the absolute number of NeuN positive cells (D) in the three ROIs. Average data are from 4-6 experiments in which 12-18 slices were used. ***P < 0.001 when comparing ceftriaxone + kainate vs. kainate (Kruskal – Wallis One Way analysis of Variance on Ranks) for pyknosis and **P < 0.041, ***P < 0.001(One Way Analysis of Variance with Tukey Test) for NeuN.

23 Figure 3. Quantification of motoneuronal staining after ceftriaxone pretreatment (A) Images showing the SMI32 positive neurons in the ventral ROI in control, 24 h after 100 µM kainate (KA), or after 3 days pre-treatment with ceftriaxone (CEF; 10 µM) followed by kainate 100 µM application for 1 h. (C, D) Lack of SMI32 staining in the central and dorsal region. (B) Histograms showing number of motoneurons. Average data are from 3-4 experiments in which 9-16 slices were used. *P < 0.002 when comparing ceftriaxone + kainate vs kainate (one way analysis of Variance with Tukey test).

24 Figure 4: Characterization of ceftriaxone effect on the expression of GLT-1 with or without application of kainate. (A) Representative S100 immunopositive astrocytes (green) and GLT-1 immupositive glutamate transporter (red) in the ventral ROI after kainate (KA; 100 µM), or 3 days pre-treatment with ceftriaxone (CEF;10 µM) followed by kainate application for 1 h. (B) The same images as in Fig. 4A with addition of DAPI staining. (D, E) Lack of GLT1 staining in motoneurones (D) and neurons (E). (C) Histogram shows the GLT-1 mean fluorescence intensity expressed as percent of control after ceftriaxone treatment in the three ROIs. GLT-1 immunostaining was quantified (with densitometry analysis of 670 x 500 µm area) to provide mean data for at least 7-10 experiments in which 1430 slices were used. ***P < 0.001 when comparing ceftriaxone vs kainate, ceftriaxone + kainate vs kainate (Kruskal-Wallis one way analysis of variance on ranks).

25 Figure 5: Characterization of the cell loss induced by TBOA (A, C) Example of ventral ROI showing DAPI (A) and NeuN (C) positive cells in control, or after 1 h application of 100 µM kainate (KA), or when TBOA was co-applied with kainate for 1 h. Pyknosis and neuronal loss evoked by kainate was increased by TBOA co-application. (B, D) Histograms showing quantification of cell loss in the three ROIs. Average data are from 3-7 experiments in which 9-21 slices were used. **P < 0.017, ***P < 0.001 when comparing TBOA + kainate vs kainate (Kruskal-Wallis one way analysis of variance on ranks) for pyknosis and ***P < 0.001 (one way analysis of Variance with Tukey test) for NeuN.