LCN2R) in the normal rat hippocampus and after kainate-induced excitotoxicity

Accepted Manuscript Title: Expression and localisation of brain-type organic cation transporter (BOCT) in the hippocampus after kainate-induced excitotoxicity Author: Wan-Jie Chia, Francis Chee Kuan Tan, Wei-Yi Ong, Gavin S. Dawe PII: DOI: Reference:

S0197-0186(15)00082-0 http://dx.doi.org/doi:10.1016/j.neuint.2015.04.009 NCI 3706

To appear in:

Neurochemistry International

Received date: Revised date: Accepted date:

18-3-2014 6-4-2015 14-4-2015

Please cite this article as: Wan-Jie Chia, Francis Chee Kuan Tan, Wei-Yi Ong, Gavin S. Dawe, Expression and localisation of brain-type organic cation transporter (BOCT) in the hippocampus after kainate-induced excitotoxicity, Neurochemistry International (2015), http://dx.doi.org/doi:10.1016/j.neuint.2015.04.009. 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

Expression and Localisation of Brain-Type Organic Cation Transporter (BOCT) in the Hippocampus after Kainate-Induced Excitotoxicity

Wan-Jie Chiaa,b,c, Francis Chee Kuan Tan a,c,d, Wei-Yi Ongc,e*, Gavin S. Dawea,b,c,d*

a

Department of Pharmacology, Yong Loo Lin School of Medicine, National University

Health System, National University of Singapore, 10 Medical Drive, Singapore, 117597, b

National University of Singapore Graduate School for Integrative Sciences and

Engineering, National University of Singapore, Centre for Life Sciences, 28 Medical Drive, Singapore 117456 cNeurobiology and Ageing Programme, Life Sciences Institute, National University of Singapore, Centre for Life Sciences, 28 Medical Drive, Singapore 117456, d

Singapore Institute for Neurotechnology (SINAPSE), Centre for Life Sciences, 28 Medical

Drive, Singapore 117456. eDepartment of Anatomy, Yong Loo Lin School of Medicine, National University Health System, National University of Singapore, 4 Medical Drive, Singapore, 117597,

*Correspondence to Dr. Gavin S. Dawe ([email protected]); Department of Pharmacology, Yong Loo Lin School of Medicine, National University Health System, National University of Singapore, Centre for Life Sciences, Level 5, 28 Medical Drive, Singapore 117456. Tel: (+65) 6516 8864; Fax: (+65) 6777 3271; OR Dr. Wei-Yi Ong ([email protected]); Department of Anatomy, National University of Singapore, Singapore 119260. Tel: (+65) 6516 3662; Fax: (+65) 6778 7643

Page 1 of 38

2

Abbreviations apo-LCN2, iron-lacking LCN2; BOCT, brain-type organic cation transporter; CNS, central nervous system; DIV, days in vitro; Fe:Ent, iron:enterochelin; icv, intracerebroventricular; ip, intraperitoneal; holo-LCN2; iron-loaded LCN2; KA, kainate; LCN2, lipocalin 2; LPS, lipopolysaccharide; NGAL, neutrophil gelatinase-associated lipocalin; OCT, organic cation transporter; rLCN2, recombinant LCN2; rLCN:Fe:Ent, rLCN2:iron:enterochelin; Slc22A17, solute carrier family 22 (organic cation transporter), member 17.

Page 2 of 38

3

Highlights - Lipocalin 2 (LCN2) and BOCT expression were studied after kainate excitotoxicity. - BOCT was expressed in pyramidal neurons, but induced in reactive microglia. - Hippocampal neurons express BOCT in culture. - LCN2 treatment resulted in increased Bim expression and reduced neuronal survival. Abstract The iron siderophore binding protein lipocalin 2 (LCN2, also known as 24p3, NGAL and siderocalin) may be involved in iron homeostasis, but till date, little is known about expression of its putative receptor, brain-type organic cation transporter (BOCT, also known as BOCT1, 24p3R, NGALR and LCN2R), in the brain during neurodegeneration. The present study was carried out to elucidate the expression of LCN2 and BOCT in hippocampus after excitotoxicity induced by the glutamate analog, kainate (KA) and a possible role of LCN2 in neuronal injury. As reported previously, a rapid and sustained induction in expression of LCN2 was found in the hippocampus after intracerebroventicular injection of KA. BOCT was expressed in neurons of the saline-injected control hippocampus, and immunolabel for BOCT protein was preserved in pyramidal neurons of CA1 at 1 day post-KA injection, likely due to the delayed onset of neurodegeneration after KA injection. At 3 days and 2 weeks after KA injections, loss of immunolabel was observed due to degenerated neurons, although remaining neurons continued to express BOCT, and induction of BOCT was found in OX-42 positive microglia. This resulted in an overall decrease in BOCT mRNA and protein expression after KA treatment. Increased expression of the pro-apoptotic marker, Bim, was found in both neurons and microglia after KA injection, but TUNEL staining indicating apoptosis was found primarily in Bim- expressing neurons, but not microglia. Apo-LCN2 caused no significant differences in neuronal Bim expression or cell survival, whereas holo-LCN2 increased Bim mRNA expression and

Page 3 of 38

4

decreased cell survival. Together, the results suggest that LCN2 and BOCT may have a role in neuronal injury.

Keywords Lipocalins; Lipocalin 2; megalin; BOCT; Bim; apoptosis; iron; neurodegeneration

Page 4 of 38

5

Introduction Lipocalin-2 (LCN2) also known as 24p3 in rodents and siderocalin or neutrophil gelatinase-associated lipocalin (NGAL) in humans is a member of the lipocalin protein family. It is an acute phase protein (Liu and Nilsen-Hamilton, 1995, Marques et al., 2008) and is often upregulated in inflammatory conditions (Alpizar-Alpizar et al., 2009, Dittrich et al., 2010, Borkham-Kamphorst et al., 2011). LCN2 is an important component of innate immunity against bacterial infection (Flo et al., 2004, Berger et al., 2006); immune cells increase the production and secretion of LCN2 in response to invading bacteria, and LCN2 then limits bacterial growth by sequestering bacterial catecholate-type ferric siderophore. LCN2 is also an adipokine (Yan et al., 2007, Zhang et al., 2008) and plays a role in cell death/survival (Kehrer, 2010), obesity, insulin-resistance (Wang et al., 2007, Yan et al., 2007) and hyperglycemia (Wang et al., 2007). LCN2 plays a role in kidney development and differentiation (Yang et al., 2002a). It mediates iron transport in kidney cells (Yang et al., 2002b, Schmidt-Ott et al., 2006, Schmidt-Ott et al., 2007), and traffics iron to endosomes via a different pathway from transferrin (Goetz et al., 2002). In some cell types, iron-loaded LCN2 (holo-LCN2) increases intracellular iron levels and inhibited apoptosis, while ironlacking LCN2 (apo-LCN2) decreases intracellular iron levels, and induces expression of the pro-apoptotic protein, Bim, leading to apoptosis (Devireddy et al., 2005, Richardson, 2005). Pro-apoptotic effects of LCN2 have also been found in primary astrocytes, C6 glioma and B35 neuroblastoma cells that are mediated by Bim and iron (Lee et al., 2009, Lee et al., 2011). Two cell-surface receptors have been identified for LCN2: megalin, a multi-ligand scavenger receptor that is highly expressed in kidney epithelial cells which binds LCN2 with high affinity (Hvidberg et al., 2005), and brain-type organic cation transporter (BOCT), also known as BOCT1, 24p3 receptor (24p3R) in rodents, NGAL receptor (NGALR) in humans,

Page 5 of 38

6

LCN2 receptor (LCN2R ) or solute carrier family 22 (organic cation transporter), member 17 (Slc22a17) (Bennett et al., 2011), which has been reported to be a specific cell-surface receptor of LCN2 (Devireddy et al., 2005). Our previous studies have shown increase in the level of iron in brain tissues undergoing degeneration after excitotoxic injury induced by the glutamate analog, kainate (KA) (Ong et al., 1999). Together with the increase in iron, we observed an upregulation in the level of LCN2 expression in reactive astrocytes in the degenerating CA fields (Chia et al., 2011), but thus far, the expression of lipocalin 2 receptors in these fields is unknown. In view of the importance of iron in many neurodegenerative diseases (Thomas and Jankovic, 2004, Zecca et al., 2004), the present study was carried out to elucidate the expression and distribution of BOCT in the normal hippocampus and after KA lesions. The interaction between LCN2 and BOCT and their relation to iron loading and neuronal injury was also studied, as this could yield insights into the role of iron in neurodegenerative diseases.

Materials and methods Animals and intracerebroventricular injections of kainate or lipopolysaccharide All procedures involving animals were approved by the Institutional Animal Care and Use Committee of the National University of Singapore and were in accordance with the guidelines of the National Advisory Committee for Laboratory Animal Research (NACLAR), Singapore, and the Guide for the Care and Use of Laboratory Animals, National Research Council of the National Academies, USA. Male Wistar rats weighing approximately 200 g each were used in this study. They were injected intracerebroventricularly with KA to induce excitotoxic neuronal injury and gliosis in the hippocampus, and the expression and cellular localisation of LCN2 and BOCT, analysed. A small number of animals were injected with

Page 6 of 38

7

lipopolysaccharide to mimic the effect of inflammation on BOCT. Animals were anaesthetised by intraperitoneal (ip) injection of ketamine (75 mg/kg) and xylazine (10 mg/kg), and intracerebroventricular (icv) injection of KA (1.2 μl of 1 mg/ml) or LPS (1 µl of 25 mg/ml) made by stereotaxic injection into the right lateral ventricle (coordinates: 1.0 mm caudal to bregma, 1.5 mm lateral to the midline, 4.5 mm from the surface of the cortex) using a microliter syringe. Experimental control rats were either injected with 1.2 μl or 1 μl of normal saline instead of KA or LPS, respectively.

Primary hippocampal neuronal cultures Primary neuronal cultures were obtained from the hippocampus of newborn Wistar pups. Brains were removed and placed in Dissection Medium (Hank’s balanced salts solution (HBSS, 0.01 M HEPES/NaOH). The hippocampi were dissected, cut into halves, and digested with papain in HBSS (1.5 mM CaCl2, 0.2 ug/ul of L-cysteine, 0.5 mM EDTA, 20 units/ml DNase I, 15 units/ml papain) at 37°C for 30 min followed by mechanical trituration using a 1 ml pipette tip. Dissociated cells were harvested by centrifugation and resuspended in Neurobasal-A medium (supplemented with B27, 2 mM GlutaMAX-1 and 1% penicillin-streptomycin) for plating on 0.01% poly-L-lysine solution (Sigma, St. Louis, MO, USA)-coated plates. Cells were seeded at 20,000 cells/coverslip in 24-well plates for Duolink assays; 160,000 cells/well in 12-well plates for RNA extraction for real-time RTPCR analysis; 28,750 cells/well in 96-well plates for cell survival assays and were cultured at 37°C in a humidified 5% CO2 incubator. HBSS, HEPES/NaOH, penicillin/streptomycin, GlutaMAX-1, sodium pyruvate, B27 supplement, and Neurobasal-A were from Invitrogen (Carlsbad, CA, USA); DNase I was from Roche (Roche Diagnostic GmbH, Mannheim, Germany); papain and Glucose D-(+) were from Sigma-Aldrich (MO, USA); and CaCl2 was from Merck (NJ, USA). Primary hippocampal neurons were analysed on DIV 10 by Duolink,

Page 7 of 38

8

real-time RT-PCR and cell survival assays. Recombinant rat LCN2 (rLCN2) was reconstituted in MES buffer (25 mM MES and 150 mM NaCl, pH 6.5) and used for treatment (Minneapolis, MN, USA). To obtain the tricomplex, “rLCN2:Fe:Ent”, ferric enterochelin (0.7 kDa) (EMC Microcollections, Tϋbingen, Germany) was pre-incubated with rLCN2 protein for 1 hr at 4°C. Vehicle controls consisted of treatment with MES buffer alone.

Real-time RT-PCR analyses Animals were anaesthetized and killed by decapitation. The brains were removed and the right hippocampus dissected out, immersed in RNAlater (Ambicon, CA, USA), flash frozen and kept at -80°C until analysis. Total RNA was isolated using TRIzol reagent (Invitrogen, CA, USA) according to the manufacturer’s protocol, and RNeasy Mini Kit (Qiagen Inc., CA, USA) was used to purify the RNA. Samples were reverse transcribed using High-Capacity cDNA Reverse Transcription Kits (Applied Biosystems, CA, USA). The reaction conditions were 25°C for 10 min, 37°C for 120 min and 85°C for 5 min. Real-time RT-PCR amplification was carried out using the 7500 Real-time RT-PCR system (Applied Biosystems, CA, USA) with Taq-Man Universal PCR Master Mix (Applied Biosystems, CA, USA) and gene-specific primers and probes [Assay ID: LCN2 (Rn00590612_m1); Slc22a17 (Rn00598583_m1); Bim (Rn00674175_m1)] according to manufacturer’s protocols. β-actin (Part no: 4352340E) was used as an internal control. All primers and probes were synthesized by Applied Biosystems. The PCR conditions were: an initial incubation of 50°C for 2 min and 95°C for 10 min followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. All reactions were carried out in triplicates. The threshold cycle, CT, which correlates inversely with the levels of target mRNA, was measured as the number of cycles at which the reporter fluorescence emission exceeds the preset threshold level. Gene expression analyses was also carried out on primary hippocampal neurons plated on 12-well plates

Page 8 of 38

9

and treated with rLCN2, rLCN2:Fe:Ent or vehicle control (MES buffer) at 10 days in vitro (DIV) for 24, 48 or 72 hrs.

Western blot analyses The right hippocampus was quickly dissected out and homogenised in cold lysis buffer containing Tris–HCl, pH 7.4, 10 mM EDTA, 150 mM NaCl, 0.5% Triton X-100 with protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN, USA) and placed on a shaker for 1 hr at 4°C. After centrifugation at 12,000 g for 30 min, the supernatant was collected, and protein concentration determined, using the BCA protein assay kit (Pierce Biotechnology, Rockford, IL, USA). Total proteins (60 μg) were denatured by heating in the presence of SDS and DTT for 10 min at 95-100°C, resolved in 12% SDS-polyacrylamide gel under reducing conditions, and electro-transferred to a nitrocellulose membrane (BioRad, CA, USA). Non-specific binding sites on the membrane were blocked by incubation with 5% non-fat milk for 1 hr. The membrane was then incubated overnight with polyclonal goat anti-LCN2 antibody at 1:200 (AF3508, R&D systems, Minneapolis, MN, USA), polyclonal rabbit anti-BOCT (Slc22a17) at 1:2000 (Prosci Incorporated, Poway, CA, USA) and monoclonal rabbit goat anti-Bim antibody at 1:1000 (Cell Signaling Technology, MA, USA). Specificity of the BOCT antibody was determined by incubation with 5 ng/µl (10x) of BOCT blocking peptide (Prosci Incorporated, Poway, CA, USA) and 0.5 ng/µl of BOCT antibody. Specificity of the Bim antibody was confirmed by blocking with 230ng/ml (10x) and 460ng/ml (20x) of recombinant Bim peptide (Cell Signaling Technology, MA, USA) and 23ng/ml Bim antibody. After washing with 0.1% Tween-20 in TBS, the membrane was incubated with horseradish peroxidase-conjugated mouse anti-goat or goat anti-rabbit immunoglobulin lgG (Thermo Fisher Scientific, Rockford, IL, USA) for 1 hr at room temperature. Proteins were visualised with an enhanced chemiluminescence kit (Thermo

Page 9 of 38

10

Fisher Scientific, Rockford, IL, USA) according to the manufacturer’s instructions. Membranes were treated with stripping buffer (Restore Western Blot Stripping Buffer, Thermo Fisher Scientific, Rockford, IL, USA) before reprobing with anti-mouse β-actin (Sigma, St. Louis, MO, USA) as loading controls. Exposed films were scanned, and the optical density of LCN2 bands analysed by densitometry using ImageJ software (NIH, Maryland, USA).

Immunohistochemistry The animals were anaesthetised and perfused with Ringer solution, followed by 4% paraformaldehyde in 0.1M phosphate buffer (pH 7.4). The brains were dissected out, and blocks containing the hippocampus sectioned coronally at 20 µm using a freezing microtome. Sections were washed to remove traces of sucrose and permeabilised with PBS-0.3% Triton X-100 for 10 min. They were then blocked with 5% goat serum in PBS-0.1% Triton X-100 for 1 hr, followed by overnight incubation with goat polyclonal to LCN2 (1:200, AF3508, R&D systems, Minneapolis, MN, USA) and rabbit polyclonal to GFAP (1:1000, DakoCytomation, Glostrup, Denmark); rabbit polyclonal to BOCT (1:500, Prosci Incorporated, Poway, CA, USA) and mouse monoclonal to MAP2a (1:200, Sigma, St. Louis, MO, USA) or mouse monoclonal to OX-42 (1:200, Chemicon, Temecula, CA, USA). Sections were also incubated overnight with rabbit polyclonal antibody to Bim, and mouse monoclonal antibody to OX-42 (1:200, Chemicon, Temecula, CA, USA) or NeuN (1:200, Chemicon, Temecula, CA, USA). Specificity of the BOCT antibody was determined by incubation with 10 ng/μl (5x) and 20 ng/μl (10x) of blocking peptide (Prosci Incorporated, Poway, CA, USA) and 2 ng/µl of BOCT antibody. Specificity of the Bim antibody was confirmed by blocking with 230 ng/ml (10x) recombinant Bim peptide (Cell Signaling Technology, MA, USA) and 23 ng/ml Bim antibody. Sections were then washed in PBS,

Page 10 of 38

11

and incubated for 1 hr at room temperature in 1:200 dilution with Alexa Fluor lgG (H+L) secondary antibodies. They were washed and mounted with ProLong® Gold antifade reagent with DAPI (Invitrogen, Carlsbad, CA, USA) and examined using a laser scanning confocal microscope (LSM 510, Carl Zeiss, Göttingen, Germany). Immunocytochemistry to localise BOCT in primary hippocampal neurons was carried out on DIV 10. Cells on coverslips were fixed with 4% paraformaldehyde for 10 min, followed by permeabilisation with PBS-01% Triton X-100 for 5 min. They were then blocked with 3% BSA in PBS for 1 hr at room temperature, followed by incubation with rabbit polyclonal antibody to Slc22A17/BOCT at 1:2000 (Cell Signaling Technology, MA, USA) and mouse monoclonal MAP2a (1:200, Sigma-Aldrich, MO, USA) overnight at 4°C. Cells were washed in PBS, incubated for 1 hr at room temperature in 1:200 dilution of Alexa Fluor IgG (H+L) secondary antibodies (Invitrogen, CA, USA), washed and coverslipped. Possible double labelling was determined by analysing the overlap between different labels by orthogonal reconstruction throughout the entire z-stack (LSM 510, Carl Zeiss Göttingen, Germany).

In situ cell death detection (TUNEL assay) TUNEL assay was performed in accordance with the manufacturer’s instructions for the in situ cell death detection kit, Fluorescein (Roche, Mannheim, Germany). Immunofluorescence of Bim was carried out, followed by incubation with Proteinase K (10 µg/ml in 10mM Tris/HCI, pH 7.4) at 37°C for 30 min, and incubation with TUNEL reaction mixture with 40 μl/section, at 37°C for 1 hr. For positive controls, DNAase I recombinant was added at (30 U/ml) at 37°C for 30 min to induce DNA strand breaks. Negative controls were incubated with label solution but not enzyme solution. Sections were then washed, coverslipped with ProLong® Gold antifade reagent with DAPI (Invitrogen, Carlsbad, CA,

Page 11 of 38

12

USA) and viewed using a fluorescence microscope (Olympus BX51; Olympus Corporation, Tokyo, Japan).

Native polyacrylamide gel electrophoresis 500 ng of recombinant rat LCN2 (rLCN2) (Minneapolis, MN, USA) or rLCN2 preincubated with five-fold molar excess of ferric enterochelin (0.7kDa) (EMC Microcollections, Tϋbingen, Germany) were loaded for native gel electrophoresis. Tris-glycine buffer without SDS, at pH 7.4 and pH > 8.0 were used as the running and transfer buffer respectively. Post-transfer denaturation of the electro-transferred proteins was done with 10% SDS in TBS buffer at 70°C for 10 min. Reducing agents were avoided in the procedures prior to the post-transfer denaturation. Membranes were then incubated overnight with polyclonal goat anti-LCN2 antibody at 1:200 (AF3508, R&D systems, Minneapolis, MN, USA). After washing with 0.1% Tween-20 in TBS, membranes were incubated with horseradish peroxidase-conjugated mouse anti-goat immunoglobulin lgG (1:10,000; Thermo Fisher Scientific, Rockford, IL, USA) for 1 hr and proteins visualised with SuperSignal West Pico chemiluminescent substrate (Thermo Fisher Scientific, Rockford, IL, USA) according to the manufacturer’s instructions.

Duolink in situ proximity ligation assay (PLA) Coverslips with primary hippocampal neurons were fixed, permeabilised, blocked with 3% BSA in PBS and incubated with primary antibodies, goat anti-LCN2 and rabbit antiBOCT overnight. Interactions were detected by the proximity ligation assay Duolink kit (PLA probe anti-rabbit minus, anti-goat plus and detection Kit 563). The PLA probe anti-rabbit MINUS binds to the rabbit BOCT antibody, while the PLA probe anti-goat PLUS binds to the goat LCN2 antibody. When both proteins are closer than 40nm, the PLA probes generate a

Page 12 of 38

13

signal, indicating an interaction. In situ PLA was performed according to the manufacturer’s instructions (OLINK Bioscience, Uppsala, Sweden). After performing the Duolink assay, cells were incubated with mouse monoclonal MAP2a (1:200, Sigma-Aldrich, MO, USA), followed by goat anti-mouse IgG (H+L) Alexa Fluor 488 (Invitrogen, CA, USA). They were then washed and mounted with Duolink mounting medium (OLINK Bioscience, Uppsala, Sweden). Duolink and MAP2a colocalisation were examined using a laser scanning confocal microscope and analysing the overlap between the different labels by orthogonal reconstruction throughout the entire z-stack (LSM 510, Carl Zeiss, Göttingen, Germany).

Cell survival assay: MTS assay Primary hippocampal neurons plated (28,750 cells/well) in 96-well plates were pretreated with rLCN2, rLCN2:Fe:Ent or MES buffer (vehicle) on DIV 10 for 48 hrs. 20 and 50 µg/ml of rLCN2 were used in both rLCN2 and rLCN2:Fe:Ent treatments. Cells were then post-treated with 2 µg/ml tunicamycin (Sigma-Aldrich MO, USA) or PBS (vehicle) for 24 hrs. Cell survival assay was performed on DIV 13 with CellTiter 96® Aqueous Non-Radioactive Cell Proliferation / Survival Assay (MTS assay) (Promega, WI, USA). CellTiter 96® AQueous One Solution reagent was added according to the manufacturer’s instructions. The cells were incubated at 37ºC and spectrophotometer readings obtained at 1, 2, 3 and 4 hours. Absorbance values were normalised to vehicle controls of each set of hippocampal neurons and the mean was calculated from 3 sets of neurons. Possible significant differences were analysed by one-way ANOVA with Tukey’s HSD post-hoc comparisons. p < 0.05 was considered significant.

Page 13 of 38

14

Data analysis For real-time RT-PCR analyses in ex vivo samples, four to five KA-injected animals and four to five saline-injected control animals were used at each of three time points, 1 day, 3 days and 2 weeks after injection. In addition, four LPS injected rats at 24 hours after injection, and three untreated rats were used to study the cellular distribution of BOCT in the normal brain. All reactions were carried out in triplicate. Amplified transcripts were quantified using the comparative CT method (Livak and Schmittgen, 2001), with the formula for relative fold change = 2–ΔΔCT. The mean was calculated, and possible significant differences were analysed between two groups using Student’s t-test and for the time course data by two-way ANOVA for time and treatment effects followed-up by Tukey’s HSD post-hoc comparisons using SPSS version 22 (SPSS Inc., Chicago, IL, USA). p < 0.05 was considered significant. For cell culture experiments at least three independent replicates were performed for each treatment. For real-time RT-PCR analyses in cell culture, mRNA levels were determined, and normalised to untreated cells (media only added) from each set of hippocampal neurons. All reactions were carried out in triplicate. The mean was calculated and possible significant differences were analysed between two groups using Student’s ttest and between multiple groups by ANOVA followed-up by Tukey’s HSD post-hoc comparisons using SPSS version 22 (SPSS Inc., Chicago, IL, USA). p < 0.05 was considered significant. For western blot studies in ex vivo samples, a total of 18 animals were used, three KA-injected animals versus three saline-injected control animals in each group were used at 1 day, 3 days and 2 weeks after injection. Three normal rats were also used to analyse BOCT and LCN2 protein levels in the rat hippocampus. The densitometry readings were normalised to that of β-actin, and possible significant differences were analysed between

Page 14 of 38

15

two groups using Student’s t-test and for the time course data by two-way ANOVA for time and treatment effects followed-up by Tukey’s HSD post-hoc comparisons using SPSS version 22 (SPSS Inc., Chicago, IL, USA). p < 0.05 was considered significant.

For immunohistochemistry on ex vivo samples, three KA-injected rats and three saline-injected rats, at each time point, were used at 1 day, 3 days and 2 weeks post icv injection. All images shown are representative examples of at least three independent animals. For immunohistochemistry on cell culture, at least three independent replicates were performed for each treatment. The TUNEL assay was performed on tissues from three KA-treated animals at each of three time points 1 day, 3 days and 2 weeks postinjection. The native polyacrylamide gel electrophoresis was performed in triplicate. The Duolink in situ PLA was performed in triplicate in independent cultures for each treatment condition. The MTS cell survival assay was performed in triplicate in independent cultures for each treatment condition. Absorbance values were normalised to vehicle controls of each set of hippocampal neurons and the mean was calculated from 3 sets of neurons. Possible significant differences were analysed by one-way ANOVA with Tukey’s HSD posthoc comparisons. p < 0.05 was considered significant.

Results LCN2 and BOCT expression in the normal hippocampus RT-PCR revealed bands for LCN2 and BOCT in normal rat hippocampus (Fig. 1A), while western blots showed discrete bands for LCN2 and BOCT in normal hippocampal lysates (Fig. 1B). The specificity of the LCN2 antibody has previously been demonstrated by blocking experiments, in our previous study (Chia et al., 2011). Blots incubated with 10X

Page 15 of 38

16

excess of antigen-absorbed BOCT antibody showed absence of bands in saline-injected hippocampal lysates, indicating specificity of BOCT antibody (Fig. 1C).

Localisation of BOCT in the normal hippocampus BOCT immunofluorescence labelling was observed in neurons of the normal rat hippocampal CA fields (Fig. 2A, D, G, J and M) and the dentate gyrus (Fig. 2M). Double labelling of BOCT with MAP2a confirmed its localisation in neurons (Fig. 2C and F). Control experiments performed by incubating the BOCT antibody together with its blocking peptide in 5x and 10x excess showed absence of staining in fields CA1 (Fig. 2H and I), CA3 (Fig. 2K and L) and the dentate gyrus (Fig. 2N and O), indicating specificity of the BOCT antibody.

LCN2, megalin and BOCT mRNA expression in the hippocampus after LPS injection or KA injury Increase in LCN2 mRNA expression was detected by real-time RT-PCR in the hippocampus after LPS injection (side of the icv injection) to 20.8-fold (p < 0.01) that of saline-injected controls (Fig. 3A). In addition, western blots showed denser LCN2 bands in lysates from the LPS-treated hippocampus compared to controls (Fig. 3B, top panel), and a 19.8- (p < 0.01) fold increase in LCN2 protein levels was found by densitometric analysis (Fig. 3B, bottom panel). In contrast to LCN2, decrease in mRNA expression of LCN2 receptors, BOCT and megalin was detected in the hippocampus after LPS injection, to 0.68- (p < 0.01) and 0.74(p < 0.01) folds that of saline-injected controls (Fig. 3C). Decrease in mRNA levels of BOCT and megalin was also found in the hippocampus two weeks after KA injection, to 0.32- (p < 0.01) and 0.50- (p < 0.001) folds compared to controls (Fig. 3D). Protein expression of

Page 16 of 38

17

BOCT was selected for further analyses in the remainder of this study, due to availability of a specific commercial antibody.

LCN2 and BOCT mRNA and protein expression in the hippocampus after KA injury Over 1 day, 3 days and 2 weeks post-KA injection, LCN2 mRNA expression was significantly increased [F(5,20)=5.0340, p<0.005]. There was a significant effect of the KA treatment on LCN2 mRNA expression [F(1,20)=19.0479, p<0.0005] but the treatment x days interaction was not significant [F(2,20)=2.0363, n.s.]. At 1 day, 3 days and 2 weeks post-KA injection, LCN2 mRNA expression was 5.6- (n.s.), 8.6- (p < 0.05) and 14.2- (p < 0.05) fold relative to their saline-injected controls, respectively (Fig. 4A). Upregulation of LCN2 mRNA expression was further analysed by western blots (Fig. 4C top panel). Over 1 day, 3 days and 2 weeks post-KA injection, LCN2 protein expression was significantly increased [F(5,12)=41.8274, p<0.0001]. There was a significant effect of the KA treatment on LCN2 protein expression [F(1,12)=91.7956, p<0.0001] and a significant treatment x days interaction [F(2,12)=29.3319, p<0.0001]. Densitometric analysis of 1 day, 3 days and 2 weeks post-KA injection samples showed 2.7 (n.s.), 16.3 (p < 0.01) and 31.2 (p < 0.0001) times increase in LCN2 protein respectively, compared to their respective saline-injected controls (Fig. 4D top panel). In contrast to LCN2, over 1 day, 3 days and 2 weeks post-KA injection, BOCT levels were reduced [F(5,20)=17.5144, p<0.0001]. There was a significant effect of the KA treatment on BOCT mRNA expression [F(1,20)=85.9437, p<0.0001] but the treatment x days interaction was not significant [F(2,20)=0.0774, n.s.]. BOCT mRNA levels were decreased, to 0.28- (p < 0.005), 0.22- (p < 0.0005 and 0.30- (p < 0.005) fold at 1 day, 3 days and 2 weeks post-KA injection compared to saline-injected controls (Fig. 4B). Over 1 day, 3 days and 2 weeks post-KA injection, BOCT protein expression was likewise

Page 17 of 38

18

significantly reduced [F(5,12)=15.5837, p<0.0001]. There was a significant effect of the KA treatment on BOCT protein expression [F(1,12)=69.2328, p<0.0001] but the treatment x days interaction was not significant [F(2,12)=2.1714, n.s.]. Reduced BOCT protein expression was also found by western blots after KA treatment (Fig. 4C, middle panel) and BOCT protein expression was 0.53 times (n.s.) that of saline-injected controls at 1 day postKA injection, and further decreased to 0.23 (p < 0.001) and 0.18 (p < 0.005) times that of controls, at 3 days and 2 weeks post-KA injection (Fig. 4D lower panel).

Cellular localisation of LCN2 in the hippocampus after kainate injury To determine the cellular localisation of LCN2 after KA injury, double immunolabelling of LCN2 was performed with astrocyte (GFAP), oligodendrocyte (CNPase), microglia (OX-42) and neuronal markers (NeuN and MAP2a). Little LCN2 expression was observed in the normal hippocampus, while upregulation of LCN2 was found in GFAPlabelled astrocytes (Fig. 5A-B) but not microglia, oligodendrocytes, or neurons at 2 weeks post-KA injection (Fig. 5C-J). Control sections showed little LCN2 staining in CA1 (Fig. 5K) and CA3 (Fig. 5L), whereas LCN2 was colocalised with GFAP-positive astrocytes in the lesioned CA fields at 1 day (Fig. 5M-N), 3 days (Fig. 5O-P) and 2 weeks (Fig. 5Q-R) postinjection. Z-series reconstruction of cells showed colocalisation of LCN2 with GFAP-positive astrocytes. LCN2 staining was observed in intense dots, which could be secretory vesicles (Fig. 5S).

Cellular localisation of BOCT in the hippocampus after kainate injury BOCT immunofluorescence was detected in neurons in saline-injected controls (Fig. 6A-B). At 1 day post-KA injection, BOCT was localised in pyramidal neurons in the unlesioned field CA1 (Fig. 6C), but was present in microglia in lesioned CA3 (Fig. 6D). At 3

Page 18 of 38

19

days (Fig. 6E-F) and two weeks (Fig. 6G-H) post-KA injection, BOCT continued to be expressed in pyramidal neurons in unlesioned CA fields, whereas microglia in lesioned fields CA1 and CA3 were double labelled for OX-42 and BOCT. Orthogonal projections at higher magnifications confirmed BOCT expression in OX42 positive microglia in lesioned CA fields at 3 days (Fig. 6I) and 2 weeks (Fig. 6J, 7A) postKA injection but was expressed in neurons, in unlesioned areas (Fig. 7B).

Bim mRNA and protein expression and cellular localisation in the hippocampus after KA injury Over 1 day, 3 days and 2 weeks post-KA injection, Bim mRNA expression was significantly increased [F(5,20)=7.6166, p<0.0005]. There was a significant effect of the KA treatment on Bim mRNA expression [F(1,20)=11.3742, p<0.005] and a significant treatment x days interaction [F(2,20)=7.2101, p<0.005]. RT-PCR showed no significant difference between KA- and saline-injected hippocampus in Bim mRNA expression at 1 day and 3 days post-KA injection, but Bim expression increased to 2.1- (p < 0.005) fold in the hippocampus at 2 weeks post-KA injury, relative to saline-injected controls (Fig. 8A). Over 1 day, 3 days and 2 weeks post-KA injection, Bim protein expression was likewise significantly increased [F(5,12)=33.4037, p<0.0001]. There was a significant effect of the KA treatment on Bim protein expression [F(1,12)=58.2370, p<0.0001] and a significant treatment x days interaction [F(2,12)=27.1954, p<0.0001]. Western blots showed no significant differences between 1 day post-KA and saline injection samples, but denser bands at 3 days and 2 weeks post-KA injection, with 3.0- (p < 0.05) and 5.0- (p < 0.0001) fold, in KA-treated samples relative to the saline controls, respectively (Fig. 8B-C). Control experiments performed by incubating blots with 10x and 20x antigen-absorbed Bim antibody showed absence of bands in the 2 weeks post-KA injection hippocampal lysates

Page 19 of 38

20

(Fig. 8D). Although Bim has three major isoforms generated by alternative splicing, namely BimEL (extra-long), BimL (long) and BimS (short) (O'Connor et al., 1998), only BimEL was detected as a single band at 23kDa in the rat hippocampus, consistent with a previous study (Shinoda et al., 2004). Immunohistochemistry showed upregulation of Bim in the subiculum (Fig. 8E), CA1 (Fig. 8G) and CA3 (Fig. 8I) lesioned areas of the hippocampus. Sections incubated with Bim antibody pre-absorbed with 10x excess of blocking peptide showed absence of staining in the subiculum (Fig. 8F), CA1 (Fig. 8H) and CA3 (Fig. 8J) regions, indicating specificity of the Bim antibody. No Bim staining was observed in all saline-injected controls, as represented by 2 weeks post-saline-injected hippocampus (Fig. 9A), while Bim labelling was observed in neurons at 1 day (Fig. 9B) and also in microglia, at 3 days (Fig. 9C) and 2 weeks (Fig. 9D) post-KA injection. Orthogonal projections through the lesioned field CA1 at higher magnification showed upregulated Bim in OX-42-positive microglia (Fig. 9E). Staining was also observed in neurons after KA injection. No Bim was detected in saline-injected controls (Fig. 10A-C), but upregulation of Bim was found in NeuN-positive neurons of fields CA1 (Fig. 10D-F) and CA3 (Fig. 10G-I) at 1 day post-KA injection. Orthogonal projections through the lesioned CA1 showed that upregulated Bim was localised in NeuN-positive pyramidal neurons (Fig. 10J).

TUNEL staining in the hippocampus after KA injury TUNEL staining was observed in pyramidal neurons, in the CA1 and CA3 regions in the 1 day and 3 days post-KA-injected hippocampus (Fig. 11A-D). Few TUNEL-positive cells were observed at 2 weeks post-KA injection (Fig. 11E-F) at a time when almost all neurons had degenerated and intense glial reaction was present in the CA fields.

Page 20 of 38

21

Orthogonal projections through the CA1 lesioned region in 1 day post-KA-injected samples showed that TUNEL-positive neurons were Bim positive (Fig. 11G-I).

Putative interaction of LCN2 with BOCT receptor complex in vitro BOCT staining was observed in primary cultured hippocampal neurons and colocalised with MAP2a, indicating that BOCT was localised in neurons (Fig. 12A-C), similar to observations made in the hippocampus of normal animals. At higher magnification, orthogonal projections demonstrate that the punctate staining of BOCT was mostly present on the surface of the MAP2a-positive neurons (Fig. 12D). The Duolink assay demonstrates the possible presence of interaction between exogenous rLCN2 and endogenous BOCT receptor on neurons, when either rLCN2 or rLCN2:Fe:Ent is applied. Duolink signals were not obvious in vehicle (MES)-treated cells (Fig. 12E), but these signals were present and colocalised with MAP2a staining, in rLCN2treated (Fig. 12F) and rLCN2:iron:enterochelin (rLCN2:Fe:Ent)-treated (Fig. 12G) primary hippocampal neurons. Native gel electrophoresis was carried out to examine the ability of the exogenous iron:enterochelin (Fe:Ent) to bind recombinant LCN2 (rLCN2). Lanes loaded with rLCN2:Fe:Ent showed an increase in molecular weight, compared to that of rLCN2 (Fig. 12H). This is due to formation of a complex of rLCN2 and iron enterochelin (719 Da), and confirms that the commercial rLCN2 is capable of binding iron:bacterial siderophore.

Effect of rLCN2 and rLCN2:Fe:Ent treatment on Bim mRNA expression and cell survival Primary hippocampal neurons at DIV 10 were treated with recombinant LCN2 (rLCN2), to study the effect of astrocytic LCN2 on hippocampal neurons. No significant

Page 21 of 38

22

differences in Bim expression were observed between rLCN2-treated neurons and vehicle controls at all time points, whereas Bim mRNA levels were significantly increased at 24 and 48 hrs of rLCN2:Fe:Ent treatment (Fig. 13A). Post-hoc Tukey’s HSD tests confirmed significant differences between rLCN2 and rLCN2:Fe:Ent (p < 0.01) and between rLCN2:Fe:Ent and vehicle control at 24 hrs of treatment (p < 0.001), and Bim mRNA expression was 1.3-fold in rLCN2:Fe:Ent-treated neurons relative to vehicle controls. At 48 hrs of treatment Bim mRNA levels was significantly increased to 1.2-fold in rLCN2:Fe:Enttreated neurons, compared to controls. (p < 0.05) Treatment with tunicamycin, a cytotoxic agent that triggers Bim-induced apoptosis (Morishima et al., 2004, Puthalakath et al., 2007), and cell survival assay were performed to determine whether rLCN2:Fe:Ent (holo-LCN2) treatment augments or diminishes susceptibility of neurons to apoptosis. Tunicamycin treatment of primary hippocampal neurons for 24 hours resulted in increase in Bim mRNA levels of 2.7-fold (p < 0.05) relative to untreated cells (Fig. 13B left panel). MES-PBS-treated neurons had 93.2% of cell survival while MES-tunicamycin-treated neurons had 63.8% relative to the untreated cells. Hence, cell survival decreased about 30% after tunicamycin treatment (Fig. 13B right panel). Neurons were then pre-treated with rLCN2 or rLCN2:Fe:Ent (48hrs) and post-treated with tunicamycin (24hrs) to determine any effects of LCN2 on tunicamycin-induced injury. One-way ANOVA indicated that the pre-treatment (20 μg/ml of rLCN2 or tricomplex) resulted in significant effects on cell survival with PBS post-treatment [pre-treatment: F(2,6)=14.20, p<0.01], but not tunicamycin post-treatment [pre-treatment: F(2,6)=0.67, n.s.]. (Fig. 13C). In cells treated with 50 μg/ml of rLCN2 and rLCN2:Fe:Ent, one-way ANOVA confirmed significant effects on cell survival with both PBS post-treatment [pre-treatment: F(2,6)=7.34, p<0.05], as well as tunicamycin post-treatment [pre-treatment: F(2,6)=6.663, p<0.05].

Page 22 of 38

23

Discussion The present study was carried out to elucidate the distribution of BOCT in the normal hippocampus, and possible changes after neuronal injury by KA or LPS, as well as possible roles of LCN2 and BOCT in neuronal apoptosis. As in our previous study, a basal level of LCN2 expression in the normal hippocampus, but upregulation of LCN2 was found in astrocytes in the degenerating hippocampus after KA injury (Chia et al., 2011). Similar observations were made after LPS injection. LCN2 staining of round intense dots in astrocytes is consistent with its localisation in acidic vesicles/endosomes (Abergel et al., 2008). Other studies have shown that LCN2 is secreted by primary astrocytes (Lee et al., 2009) and upregulated after LPS administration in mice (Marques et al., 2008, Bonow et al., 2009, Ip et al., 2011). In addition, LCN2 expression has been reported in pyramidal and granule cell layers of the hippocampus following restraint stress (Mucha et al., 2011), and endothelial and microglial cells after intraperitoneal LPS administration in mice (Ip et al., 2011). The increase in LCN2 expression after KA lesions or LPS injections was accompanied by changes in lipocallin 2 receptors. BOCT was found in neurons of the normal hippocampus, and immunolabel for BOCT protein was preserved in pyramidal neurons of CA1 at 1 day post-KA injection likely due to the delayed onset of neurodegeneration after KA injection (Sperk et al., 1983, Tokuhara et al., 2007). At 3 days and 2 weeks after KA injections, loss of immunolabel was observed due to degenerated neurons, although remaining neurons continued to express BOCT, and induction of BOCT was found in OX-42 positive microglia. This resulted in an overall decrease in BOCT mRNA and protein expression after KA injections. A similar loss of LCN2 receptors was also

Page 23 of 38

24

observed after intracerebroventricular injection of LPS. The localisation of BOCT in normal neurons is consistent with previous reports of 24p3R (mouse ortholog of BOCT) mRNA in cortical neurons, hippocampal dentate gyrus and granule neurons and Purkinje neurons of the cerebellum in mice (Ip et al., 2011). Further studies were carried out to examine the expression of the pro-apoptotic protein, Bim, and apoptosis in the degenerating hippocampus after KA lesions, and a possible role of LCN2 and BOCT in affecting Bim expression in vitro. Increased expression of the pro-apoptotic marker, Bim, was found in both neurons and microglia after KA injection, but TUNEL staining indicating apoptosis was found primarily in neurons but not microglia. Importantly, double labelling showed that Bim positive neurons were also TUNEL-positive, consistent with the notion that Bim is a pro-apoptotic protein in hippocampal neurons undergoing excitotoxic injury. Previous studies have shown an increase in number of Fluoro-Jade positive and TUNEL-positive degenerating pyramidal neurons 1 day after KA administration (Korhonen et al., 2003, Shinoda et al., 2004), and presence of activated caspase 3 in the hippocampus, 1 day after excitotoxic injury in mice (Theofilas et al., 2009). The results suggest greater vulnerability of neurons than microglia to apoptosis. Our previous study on the KA-lesioned hippocampus showed increased expression of ferritin in microglia and oligodendrocytes (Huang and Ong, 2005), and one possibility is that iron delivered to activated microglia by holo-LCN2 could be effectively sequestered by the upregulated ferritin to prevent it from participating in catalysis of free radical formation. We further elucidated possible roles of LCN2 and BOCT in primary cultured hippocampal neurons, and showed that LCN2 treatment for 24 or 48 hours increased Bim expression in the presence of iron, but not its absence. The Duolink assay is an in situ PLA technology which enables the detection and visualisation of protein interactions in cell samples, and we demonstrated the possible ability of BOCT to bind and

Page 24 of 38

25

internalise LCN2, applied both as apo-LCN2 (rLCN2) and as holo-LCN2 (rLCN2:Fe:Ent) to neurons, consistent with receptor-mediated endocytosis (Yang et al., 2002b, Devireddy et al., 2005, Hanai et al., 2005). To mimic the effect of LCN2 release in the lesioned hippocampus, apo- or holo-LCN2 were added to primary hippocampal cultures to examine the effect on Bim expression and neuronal injury. Apo-LCN2 had no significant differences in Bim expression and cell survival, whereas holo-LCN2 increased Bim expression and decreased cell survival. This indicates that holo-LCN2 is pro-apoptotic in primary hippocampal neurons. Possible effects of apo- versus holo-LCN2 in affecting tunicamycin-induced apoptosis were also elucidated. Incubation of primary hippocampal neurons with tunicamycin resulted in upregulation of Bim mRNA and decrease in cell survival, consistent with the results of previous studies (Morishima et al., 2004, Puthalakath et al., 2007). No significant differences were observed between apo-LCN2 and vehicle treated-primary neurons. In contrast, the higher dose of holo-LCN2 (50 µg/ml) caused a further decrease in cell survival compared to apo-LCN2 (50 µg/ml) after tunicamycin treatment, indicating that holo-LCN2 potentiates tunicamycin-induced neuronal injury. There are many possible interpretations of the current results. One possible explanation invokes the possibility of close interaction between LCN2 and BOCT. Interaction between LCN2 and BOCT has been proposed (Devireddy et al., 2005, Fang et al., 2007). Recent reports characterised BOCT as a promiscuous high-affinity, multi-ligand receptor for internalisation of proteins and peptides in the distal nephron and intestinal epithelium (Langelueddecke et al., 2012, Langelueddecke et al., 2013). However, the possibility of interaction between LCN2 and BOCT is not without controversy. Other reports have been unable to confirm the region of BOCT interacting with LCN2 (Bennett et al., 2011, Correnti et al., 2012) and could not confirm effects on iron efflux or apoptosis in hematopoietic cells (Correnti et al., 2012).

Page 25 of 38

26

Nevertheless, assumption of an interaction between LCN2 and BOCT offers one of the most parsimonious explanations of the current findings. In this interpretation, the following putative mechanism is proposed (Fig. 14): Holo-LCN2 interacts with BOCT expressed on primary hippocampal neurons and either the entire complex or the holo-LCN2 is internalised to import iron into the cell. Iron can catalyse the formation of free radicals, and increase oxidative stress in neurons, resulting in neuronal injury. This is consistent with the notion that excess iron is detrimental to the CNS (Halliwell, 1992, Thomas and Jankovic, 2004). The results are different from that of previous studies involving non-neuronal cells. In one study, apo-LCN2 was proposed to be pro-apoptotic and holo-LCN2 to be anti-apoptotic in 24p3R-overexpressing HeLa cells (Devireddy et al., 2005); and in two other studies, LCN2-induced cell death sensitisation was abolished by addition of iron:siderophore complex in C6 glioma and BV-2 microglial cells (Lee et al., 2007, 2009). Therefore the proapoptotic effect of holo-LCN2 found in our study may be specific to neurons, and LCN2 may, indeed, have different effects depending on cell type (Hanai et al., 2005, Lee et al., 2009). A recent study has also reported detrimental effects of LCN2, after spinal cord injury (Rathore et al., 2011). We postulate that LCN2 and BOCT, whether or not they interact in either a direct or indirect manner, may play an important role in neuronal death in the hippocampus after KA induced excitotoxicity, and possibly other neurodegenerative conditions. Increase in iron is observed in the rat hippocampus after KA injury (Ong et al., 1999) and in many neurodegenerative diseases (Thomas and Jankovic, 2004, Zecca et al., 2004). In this context, accumulation of free iron and the rapid upregulation of LCN2 in the hippocampus could promote the formation of holo-LCN2 to bind BOCT expressed on neurons that are at the edge of the glial scar. Our in vitro data on interaction between LCN2 and BOCT suggest that these neurons may be particularly susceptible to injury, since they express BOCT and

Page 26 of 38

27

are close to areas where there is significant upregulation of LCN2 in reactive astrocytes. This proposed mechanism of neuronal vulnerability presupposes the presence of extracellular mammalian siderophore(s) in the hippocampus to bind free iron and LCN2 after KA injury to form holo-LCN2. However, despite recent studies reporting catechols as possible mammalian siderophores (Bao et al., 2010); and BDH2 as an enzyme which synthesizes the iron moiety (2,5-dihdroxybenzoic acid, 2,5-DHBA) of the mammalian siderophore (Devireddy et al., 2010), the exact identity of the endogenous siderophore(s) still remains to be determined. Further work is necessary to elucidate the factors regulating LCN2 and BOCT expression and their possible interactions, as this may yield insights into neurodegenerative and neuroinflammatory disease conditions.

Acknowledgements The authors declare no competing financial interests. This work was supported by a grant from the National Medical Research Council of Singapore (NMRC/EDG/1021/2010).

Page 27 of 38

28

References Abergel RJ, Clifton MC, Pizarro JC, Warner JA, Shuh DK, Strong RK, Raymond KN (2008) The siderocalin/enterobactin interaction: a link between mammalian immunity and bacterial iron transport. Journal of the American Chemical Society 130:11524-11534. Alpizar-Alpizar W, Laerum OD, Illemann M, Ramirez JA, Arias A, Malespin-Bendana W, Ramirez V, Lund LR, Borregaard N, Nielsen BS (2009) Neutrophil gelatinaseassociated lipocalin (NGAL/Lcn2) is upregulated in gastric mucosa infected with Helicobacter pylori. Virchows Arch 455:225-233. Bao G, Clifton M, Hoette TM, Mori K, Deng SX, Qiu A, Viltard M, Williams D, Paragas N, Leete T, Kulkarni R, Li X, Lee B, Kalandadze A, Ratner AJ, Pizarro JC, Schmidt-Ott KM, Landry DW, Raymond KN, Strong RK, Barasch J (2010) Iron traffics in circulation bound to a siderocalin (Ngal)-catechol complex. Nat Chem Biol 6:602-609. Bennett KM, Liu J, Hoelting C, Stoll J (2011) Expression and analysis of two novel rat organic cation transporter homologs, SLC22A17 and SLC22A23. Molecular and Cellular Biochemistry 352:143-154. Berger T, Togawa A, Duncan GS, Elia AJ, You-Ten A, Wakeham A, Fong HE, Cheung CC, Mak TW (2006) Lipocalin 2-deficient mice exhibit increased sensitivity to Escherichia coli infection but not to ischemia-reperfusion injury. Proc Natl Acad Sci U S A 103:1834-1839. Bonow RH, Aid S, Zhang Y, Becker KG, Bosetti F (2009) The brain expression of genes involved in inflammatory response, the ribosome, and learning and memory is altered by centrally injected lipopolysaccharide in mice. Pharmacogenomics J 9:116126. Borkham-Kamphorst E, Drews F, Weiskirchen R (2011) Induction of lipocalin-2 expression in acute and chronic experimental liver injury moderated by pro-inflammatory cytokines interleukin-1beta through nuclear factor-kappaB activation. Liver Int 31:656-665. Chia WJ, Dawe GS, Ong WY (2011) Expression and localization of the iron-siderophore binding protein lipocalin 2 in the normal rat brain and after kainate-induced excitotoxicity. Neurochemistry International. Correnti C, Richardson V, Sia AK, Bandaranayake AD, Ruiz M, Suryo Rahmanto Y, Kovacevic Z, Clifton MC, Holmes MA, Kaiser BK, Barasch J, Raymond KN, Richardson DR, Strong RK (2012) Siderocalin/Lcn2/NGAL/24p3 does not drive apoptosis through gentisic acid mediated iron withdrawal in hematopoietic cell lines. PLoS One 7:e43696. Devireddy LR, Gazin C, Zhu X, Green MR (2005) A cell-surface receptor for lipocalin 24p3 selectively mediates apoptosis and iron uptake. Cell 123:1293-1305. Devireddy LR, Hart DO, Goetz DH, Green MR (2010) A mammalian siderophore synthesized by an enzyme with a bacterial homolog involved in enterobactin production. Cell 141:1006-1017.

Page 28 of 38

29

Dittrich AM, Krokowski M, Meyer HA, Quarcoo D, Avagyan A, Ahrens B, Kube SM, Witzenrath M, Loddenkemper C, Cowland JB, Hamelmann E (2010) Lipocalin2 protects against airway inflammation and hyperresponsiveness in a murine model of allergic airway disease. Clinical and Experimental Allergy 40:1689-1700. Fang WK, Xu LY, Lu XF, Liao LD, Cai WJ, Shen ZY, Li EM (2007) A novel alternative spliced variant of neutrophil gelatinase-associated lipocalin receptor in oesophageal carcinoma cells. Biochem J 403:297-303. Flo TH, Smith KD, Sato S, Rodriguez DJ, Holmes MA, Strong RK, Akira S, Aderem A (2004) Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron. Nature 432:917-921. Goetz DH, Holmes MA, Borregaard N, Bluhm ME, Raymond KN, Strong RK (2002) The neutrophil lipocalin NGAL is a bacteriostatic agent that interferes with siderophoremediated iron acquisition. Mol Cell 10:1033-1043. Halliwell B (1992) Reactive oxygen species and the central nervous system. J Neurochem 59:1609-1623. Hanai J, Mammoto T, Seth P, Mori K, Karumanchi SA, Barasch J, Sukhatme VP (2005) Lipocalin 2 diminishes invasiveness and metastasis of Ras-transformed cells. Journal of Biological Chemistry 280:13641-13647. Huang E, Ong WY (2005) Distribution of ferritin in the rat hippocampus after kainateinduced neuronal injury. Exp Brain Res 161:502-511. Hvidberg V, Jacobsen C, Strong RK, Cowland JB, Moestrup SK, Borregaard N (2005) The endocytic receptor megalin binds the iron transporting neutrophil-gelatinaseassociated lipocalin with high affinity and mediates its cellular uptake. FEBS Lett 579:773-777. Ip JP, Nocon AL, Hofer MJ, Lim SL, Muller M, Campbell IL (2011) Lipocalin 2 in the central nervous system host response to systemic lipopolysaccharide administration. J Neuroinflammation 8:124. Kehrer JP (2010) Lipocalin-2: pro- or anti-apoptotic? Cell Biology and Toxicology 26:83-89. Korhonen L, Belluardo N, Mudo G, Lindholm D (2003) Increase in Bcl-2 phosphorylation and reduced levels of BH3-only Bcl-2 family proteins in kainic acid-mediated neuronal death in the rat brain. European Journal of Neuroscience 18:1121-1134. Langelueddecke C, Roussa E, Fenton RA, Thevenod F (2013) Expression and function of the lipocalin-2 (24p3/NGAL) receptor in rodent and human intestinal epithelia. PLoS One 8:e71586. Langelueddecke C, Roussa E, Fenton RA, Wolff NA, Lee WK, Thevenod F (2012) Lipocalin-2 (24p3/neutrophil gelatinase-associated lipocalin (NGAL)) receptor is expressed in distal nephron and mediates protein endocytosis. J Biol Chem 287:159169.

Page 29 of 38

30

Lee S, Lee J, Kim S, Park JY, Lee WH, Mori K, Kim SH, Kim IK, Suk K (2007) A dual role of lipocalin 2 in the apoptosis and deramification of activated microglia. J Immunol 179:3231-3241. Lee S, Lee WH, Lee MS, Mori K, Suk K (2011) Regulation by lipocalin-2 of neuronal cell death, migration, and morphology. Journal of Neuroscience Research. Lee S, Park JY, Lee WH, Kim H, Park HC, Mori K, Suk K (2009) Lipocalin-2 is an autocrine mediator of reactive astrocytosis. J Neurosci 29:234-249. Liu Q, Nilsen-Hamilton M (1995) Identification of a new acute phase protein. Journal of Biological Chemistry 270:22565-22570. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25:402-408. Marques F, Rodrigues AJ, Sousa JC, Coppola G, Geschwind DH, Sousa N, Correia-Neves M, Palha JA (2008) Lipocalin 2 is a choroid plexus acute-phase protein. J Cereb Blood Flow Metab 28:450-455. Morishima N, Nakanishi K, Tsuchiya K, Shibata T, Seiwa E (2004) Translocation of Bim to the endoplasmic reticulum (ER) mediates ER stress signaling for activation of caspase-12 during ER stress-induced apoptosis. Journal of Biological Chemistry 279:50375-50381. Mucha M, Skrzypiec AE, Schiavon E, Attwood BK, Kucerova E, Pawlak R (2011) Lipocalin2 controls neuronal excitability and anxiety by regulating dendritic spine formation and maturation. Proceedings of the National Academy of Sciences of the United States of America. O'Connor L, Strasser A, O'Reilly LA, Hausmann G, Adams JM, Cory S, Huang DC (1998) Bim: a novel member of the Bcl-2 family that promotes apoptosis. EMBO Journal 17:384-395. Ong WY, Ren MQ, Makjanic J, Lim TM, Watt F (1999) A nuclear microscopic study of elemental changes in the rat hippocampus after kainate-induced neuronal injury. J Neurochem 72:1574-1579. Puthalakath H, O'Reilly LA, Gunn P, Lee L, Kelly PN, Huntington ND, Hughes PD, Michalak EM, McKimm-Breschkin J, Motoyama N, Gotoh T, Akira S, Bouillet P, Strasser A (2007) ER stress triggers apoptosis by activating BH3-only protein Bim. Cell 129:1337-1349. Rathore KI, Berard JL, Redensek A, Chierzi S, Lopez-Vales R, Santos M, Akira S, David S (2011) Lipocalin 2 plays an immunomodulatory role and has detrimental effects after spinal cord injury. J Neurosci 31:13412-13419. Richardson DR (2005) 24p3 and its receptor: dawn of a new iron age? Cell 123:1175-1177.

Page 30 of 38

31

Schmidt-Ott KM, Mori K, Kalandadze A, Li JY, Paragas N, Nicholas T, Devarajan P, Barasch J (2006) Neutrophil gelatinase-associated lipocalin-mediated iron traffic in kidney epithelia. Curr Opin Nephrol Hypertens 15:442-449. Schmidt-Ott KM, Mori K, Li JY, Kalandadze A, Cohen DJ, Devarajan P, Barasch J (2007) Dual action of neutrophil gelatinase-associated lipocalin. J Am Soc Nephrol 18:407413. Shinoda S, Schindler CK, Meller R, So NK, Araki T, Yamamoto A, Lan JQ, Taki W, Simon RP, Henshall DC (2004) Bim regulation may determine hippocampal vulnerability after injurious seizures and in temporal lobe epilepsy. Journal of Clinical Investigation 113:1059-1068. Sperk G, Lassmann H, Baran H, Kish SJ, Seitelberger F, Hornykiewicz O (1983) Kainic acid induced seizures: neurochemical and histopathological changes. Neuroscience 10:1301-1315. Theofilas P, Bedner P, Huttmann K, Theis M, Steinhauser C, Frank S (2009) The proapoptotic BCL-2 homology domain 3-only protein Bim is not critical for acute excitotoxic cell death. Journal of Neuropathology and Experimental Neurology 68:102-110. Thomas M, Jankovic J (2004) Neurodegenerative disease and iron storage in the brain. Current Opinion in Neurology 17:437-442. Tokuhara D, Sakuma S, Hattori H, Matsuoka O, Yamano T (2007) Kainic acid dose affects delayed cell death mechanism after status epilepticus. Brain and Development 29:28. Wang Y, Lam KS, Kraegen EW, Sweeney G, Zhang J, Tso AW, Chow WS, Wat NM, Xu JY, Hoo RL, Xu A (2007) Lipocalin-2 is an inflammatory marker closely associated with obesity, insulin resistance, and hyperglycemia in humans. Clin Chem 53:34-41. Yan QW, Yang Q, Mody N, Graham TE, Hsu CH, Xu Z, Houstis NE, Kahn BB, Rosen ED (2007) The adipokine lipocalin 2 is regulated by obesity and promotes insulin resistance. Diabetes 56:2533-2540. Yang J, Blum A, Novak T, Levinson R, Lai E, Barasch J (2002a) An epithelial precursor is regulated by the ureteric bud and by the renal stroma. Developmental Biology 246:296-310. Yang J, Goetz D, Li JY, Wang W, Mori K, Setlik D, Du T, Erdjument-Bromage H, Tempst P, Strong R, Barasch J (2002b) An iron delivery pathway mediated by a lipocalin. Mol Cell 10:1045-1056. Zecca L, Youdim MB, Riederer P, Connor JR, Crichton RR (2004) Iron, brain ageing and neurodegenerative disorders. Nat Rev Neurosci 5:863-873. Zhang J, Wu Y, Zhang Y, Leroith D, Bernlohr DA, Chen X (2008) The role of lipocalin 2 in the regulation of inflammation in adipocytes and macrophages. Mol Endocrinol 22:1416-1426.

Page 31 of 38

32

Figure legends Fig. 1 LCN2 and BOCT mRNA and protein expression in the hippocampus. A RT-PCR was conducted on the cDNA of the hippocampus from a normal rat with specific primers to LCN2 and BOCT transcripts. B 60μg of hippocampus lysate from three individual normal animals were loaded for western blot analysis and probed with anti-LCN2 and anti-BOCT. n = 3 per group. C Blots of 2 weeks post-saline-injected hippocampal lysates, R7 and R8, were incubated with BOCT antibody (lane 1 and 2 from the left) and BOCT antibody preabsorbed with 10x excess of BOCT blocking peptide (lane 3 and 4) showed absence of bands.

Fig. 2 Cellular expression of BOCT in normal hippocampus. Hippocampal sections were double labelled with (A, D) BOCT (red) and (B, E) MAP2a (green), showing colocalisation at both (C) CA1 and (F) CA3 regions of the hippocampus. Bar = 20 μm. BOCT positive staining at (G) CA1, (J) CA3 hippocampal regions and (M) the dentate gyrus. Sections preincubated with BOCT antibody with (H, K and N) 5x and (I, L, and O) 10x excess of blocking peptide showed absence of staining, indicating the specificity of the antibody. G – L: bar = 200 μm; M – O: bar = 500 μm

Fig. 3 LCN2 and BOCT expression changes after LPS or KA treatment. A Real-time RTPCR analysis indicates a significant increase in LCN2 mRNA levels in the hippocampus after 24 hours LPS treatment relative to saline administration. Data are expressed as mean ± SEM, n = 4 per treatment group. B Western blot analysis of LCN2 expression in LPS vs. saline-injected animals (top panel) and densitometric analysis of western blot (bottom panel). LCN2 protein increased after LPS treatment. Data are expressed as mean ± SEM, n = 3 per treatment group. C Real-time RT-PCR analysis on mRNA changes of BOCT and

Page 32 of 38

33

megalin at 24 hrs post-LPS and -saline injection. D Real-time RT-PCR analysis of mRNA changes of BOCT or megalin, at 2 weeks post-KA and saline injection. Data are expressed as mean ± SEM, n = 4 per treatment group. Analysed by Student’s t-test, asterisks indicate significant difference (**p < 0.01, ****p < 0.001)

Fig. 4 LCN2 and BOCT mRNA and protein expression changes after KA injury. Real-time RT-PCR analysis of (A) LCN2 and (B) BOCT mRNA expression changes in the hippocampus at 1 day, 3 days and 2 weeks post-KA injection. Data are expressed as mean ± SEM, n = 4 to 5 per treatment group. C Western blot analysis of LCN2 and BOCT in the hippocampus and (D) densitometric analysis of LCN2 and BOCT protein expression in KAtreated vs. saline-injected controls (n = 3 per treatment group). Values are obtained after normalising to β-actin, then the saline-injected controls at the respective time points. Data are expressed as mean ± SEM, n = 3 per treatment group. Analysed by Tukey’s HSD post hoc test following two-way ANOVA, asterisks indicate significant difference (*p < 0.05, **p < 0.01, *** p < 0.005, ****p < 0.001, *****p < 0.0005, ******p < 0.0001).

Fig. 5 Cellular localisation of LCN2 after KA injury. A – J Double immunolabelling of LCN2 with astrocyte (GFAP), oligodendrocyte (CNPase), microglial (OX-42) and neuronal (NeuN and MAP2a) markers in 2 weeks post-saline and post-KA-injected sections. Bar = 50 μm. Negligible LCN2 staining was observed in (K) CA1 and (L) CA3 of 2 weeks post-saline injection hippocampus. Immunofluorescence staining for LCN2 (red) was upregulated in CA1 and CA3 lesioned regions and colocalised with GFAP (green) at (M, N) 1 day (O, P) 3 days and (Q, R) 2 weeks post-KA injection. Bar = 100 μm. S Z-series reconstruction of the cells demonstrates that LCN2 staining comprised of intense dots localised within the GFAPpositive astrocyte. Majority of the LCN2-positive cells were GFAP-positive. Bar = 20 μm

Page 33 of 38

34

Fig. 6 Changes in cellular localisation of BOCT in the rat hippocampus after KA injury. Hippocampal sections were double labelled with BOCT (green) and OX-42 (red), a marker for microglia. A, B In 2 weeks post-saline injection samples, BOCT did not colocalise with OX-42 in the CA1 and CA3 regions. Similarly, at 1 day post-KA injection, BOCT did not colocalise with OX-42 at the (C) CA1 region, but colocalised with OX-42 in the (D) CA3 region. In (E, F) 3 days and (G, H) 2 weeks post-KA injection samples, LCN2 colocalised with OX-42 in both the CA1 and CA3 regions. Bar = 100 μm. Inserts in A-H show morphology of microglia labelled with OX-42. Bar for inserts = 10 μm. (l, J) Higher magnification orthogonal projections through the lesioned areas, showing that BOCT positive cells are present in OX-42 positive microglia at 3 days and 2 weeks post-KA injection. Bar = 20 μm

Fig. 7 Immunoreactivity at CA1-CA2 region in KA-treated sections. A Colocalisation studies of BOCT (red) and OX-42 (green). BOCT was localised in unlesioned pyramidal neurons (CA2 region) and upregulated in microglia of the CA1 lesioned area. B Colocalisation studies of LCN2 (green) and BOCT (red). BOCT was predominantly expressed in the intact neurons of the CA2 region, while LCN2 was upregulated in astrocytes in CA1. Bar = 50 μm

Fig. 8 Bim mRNA and protein expression upregulated after KA injury. A Real-time RT-PCR analysis of Bim mRNA expression at the right hippocampi at 1 day, 3 days and 2 weeks post-KA injection, relative to their saline-injected controls. Data are expressed as mean ± SEM, n = 4 per treatment group. Analysed by Tukey’s HSD post hoc test following two-way ANOVA, asterisks indicate significant difference (***p < 0.005). B Western blot analysis of Bim protein expression in the hippocampus at 1 day, 3 days and 2 weeks post-KA vs. -

Page 34 of 38

35

saline icv injection. C Blocking study for Bim was performed with 2 weeks post-KA-treated hippocampal lysates, R4 and R6. Blots were incubated with Bim antibody (Lane 1 and 2 from the left), Bim antibody pre-absorbed with 10x excess of blocking peptide (Lane 3 and 4) and Bim antibody pre-absorbed with 20x excess of blocking peptide (Lane 5 and 6). D Densitometric analysis of Bim protein expression. Values are normalised with the salineinjected controls at the respective time points. Data are expressed as mean ± SEM, n = 3 per treatment group. Analysed by Tukey’s HSD post hoc test following two-way ANOVA, asterisks indicate significant difference (*p < 0.05, ******p < 0.001). E – J Regional expression of Bim in 2 weeks post-KA injection lesioned hippocampus. F, H and J Sections incubated with pre-absorbed Bim antibody with 10x excess blocking peptide showed absence of positive Bim staining. Bar = 500 μm

Fig. 9 Confocal micrographs of upregulation of Bim in activated microglia after KA injury. A Negligible Bim staining was observed in the hippocampus of 2 weeks post-saline injection controls. In contrast, upregulated Bim (red) colocalised with OX-42 (green) at (B) 1 day, (C) 3 days and (D) 2 weeks post-KA injection. E Confocal micrograph of the orthogonal projection through CA1 lesioned region of the hippocampus at higher magnification demonstrate that Bim was upregulated in the OX-42-positive cells. Bar = 20 μm

Fig. 10 Confocal micrographs of upregulation of Bim in neurons at 1 day post-KA injury. Hippocampal sections were incubated with anti-Bim (red) and anti-NeuN (green) antibody. A – C No Bim staining was detected in the CA1 region of 1 day post-saline injection samples. In KA-treated sections, Bim staining colocalised with NeuN in (D – F) CA1 and (G – I) CA3 regions of the hippocampus. J Orthogonal projections through CA1 lesioned

Page 35 of 38

36

region at higher magnification demonstrate that Bim is present in the NeuN-positive cells. Bar = 20 μm

Fig. 11 Fluorescence micrographs of TUNEL assay of the lesioned hippocampus after KA injury. In situ cell death detection kit was applied to right hippocampal sections to indicate the apoptotic status. Positive TUNEL staining (green) is an indication of apoptosis. TUNELpositive cells are present in (A, B) 1 day and (C, D) 3 days post-KA injection sections but few cells were present at (E, F) 2 weeks post-KA injection sections. Bar = 500 μm. Confocal micrographs of 1 day post-KA injection sections stained with (H) TUNEL assay (green) and double labelled with (G) anti-Bim antibody (red). i Orthogonal projections through CA1 lesioned region demonstrate that Bim was present in the TUNEL-positive cells. Bar = 20 μm

Fig. 12 Interaction of LCN2 with BOCT in primary hippocampal neurons. A – C Fluorescence micrographs of primary hippocampal neurons cultured till DIV 10, fixed and double labelled with BOCT (red) and MAP2a (green) at 20x low magnification. Bar = 100 μm. D Confocal micrographs at higher magnification demonstrate that BOCT was mostly present on the surface of the cell body and on the projections of the MAP2a-positive cells. Primary hippocampal neurons were treated on DIV 10 with (E) vehicle (MES buffer), (F) rLCN2, (G) rLCN2:iron:enterochelin (rLCN2:Fe:Ent) for 24 hrs. Cells were fixed and Duolink (red) was performed followed by MAP2a immunostaining (green). Duolink staining colocalise with MAP2a in rLCN2 and rLCN2:Fe:Ent treatment. Bar = 20 μm. Bar for inserts = 20 μm. H Native gel loaded with rLCN2 protein and rLCN2:Fe:Ent was probed with antiLCN2 antibody. Upwards shift of weight (increase in weight) in the lane loaded with rLCN2:Fe:Ent, indicates the ability of rLCN2 to bind with the bacterial iron siderophore

Page 36 of 38

37

Fig. 13 Effect of rLCN2 and rLCN2:Fe:Ent treatment on Bim mRNA levels and cell survival in primary hippocampal neurons. A Primary hippocampal neurons were cultured and treated on DIV 10 with either media (untreated cells), vehicle (MES buffer), rLCN2, or rLCN2:Fe:Ent for 24 or 48 hrs. Cells were harvested for RNA and real-time RT-PCR analysis was performed. Values are normalised with the untreated cells (fold change = 1.0) at the respective time points to obtain the relative fold change. Analysed by 2-way ANOVA, follow-up one-way ANOVA for the effect of treatment at each time point. Tukey’s HSD posthoc comparisons at each time point revealed significant differences, indicated by (*p < 0.05, **p < 0.01, ***p < 0.001). Data are expressed as mean ± SEM, n = 3 per treatment group. B Primary neuronal cells were treated with 48 hrs of MES (vehicle) before 24 hrs of tunicamycin or PBS treatment. Upregulation of Bim mRNA (left panel) and decreased cell survival (right panel) was observed after tunicamycin (Tun) treatment. Bim mRNA expression was relative to untreated cells (Fold change = 1.0). Cell survival was normalised to untreated cells (100%). Analysed by Student’s t-test, asterisks indicate significant difference (*p < 0.05, **p < 0.01, ***p < 0.001). C Primary hippocampal neurons were plated on 96-well plates, cultured and pre-treated on DIV 10 with either media (untreated cells), vehicle (MES buffer), 20 µg/ml rLCN2, 50 µg/ml rLCN2, 20 µg/ml rLCN2:Fe:Ent or 50 µg/ml rLCN2:Fe:Ent, for 48 hrs. Cells were then treated with tunicamycin or vehicle (PBS) for 24 hrs. MTS assay was performed on DIV 13. Data are expressed as mean ± SEM, n = 3 per treatment group. Statistical analysis was performed with four separate 1-way ANOVA for each experimental setup to determine the effect of the pre-treatment on cell survival. Tukey’s HSD post-hoc comparisons at each condition revealed significant differences, indicated by asterisks (*p < 0.05, **p < 0.01, ***p < 0.001).

Page 37 of 38

38

Fig. 14 Putative mechanism of pro-apoptotic effect of holo-LCN2 on primary hippocampal neurons assuming the case that LCN2 and BOCT interact. (1) Holo-LCN2 (LCN2:Fe:siderophore) interacts with BOCT to be internalised in neurons. (2) Due to the low pH, iron is dissociated from the tricomplex, increasing intracellular iron concentrations. Iron regulatory genes/proteins are activated to decrease the iron levels. If free iron remains in excess, ROS is generated; increasing oxidative stress (3) As a result, Bim is upregulated. (4) Decrease in cell survival is observed as a result of holo-LCN2 treatment. Modified from Richardson (Richardson, 2005). Diagram not drawn to scale.

Page 38 of 38