Journal of Neuroimmunology 272 (2014) 35–41

Contents lists available at ScienceDirect

Journal of Neuroimmunology

j ourna l homepage: www.e lsev ie r .com/ locate / jneuro im

Neonatal Escherichia coli K1 meningitis causes learning and memoryimpairments in adulthood

Tatiana Barichello a,g,⁎, Valdemira S. Dagostim a, Jaqueline S. Generoso a, Lutiana R. Simões a,Diogo Dominguini b, Cintia Silvestre a, Monique Michels d, Márcia Carvalho Vilela e, Luciano K. Jornada b,Clarissa M. Comim c, Felipe Dal-Pizzol d, Antonio Lucio Teixeira f, João Quevedo b,g

a Laboratório de Microbiologia Experimental e Instituto Nacional de Ciência e Tecnologia Translacional em Medicina, Programa de Pós-Graduação em Ciências da Saúde,Universidade do Extremo Sul Catarinense, Criciúma, SC, Brazilb Laboratório de Neurociências, Programa de Pós-Graduação em Ciências da Saúde, Unidade Acadêmica de Ciências da Saúde, Universidade do Extremo Sul Catarinense, Criciúma, SC, Brazilc Laboratório de Neurociências Experimental, Programa de Pós-Graduação em Ciências da Saúde, Universidade do Sul de Santa Catarina, Palhoça, SC, Brazild Laboratório de Fisiopatologia Experimental, Programade Pós-Graduação emCiências da Saúde, Unidade Acadêmica de Ciências da Saúde, Universidade do Extremo Sul Catarinense, Criciúma, SC, Brazile Departamento de Biologia Animal, Universidade Federal de Viçosa, Viçosa, MG, Brazilf Laboratório Interdisciplinar de Investigação Médica, Faculdade de Medicina, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazilg Center for Experimental Models in Psychiatry, Department of Psychiatry and Behavioral Sciences, The University of Texas Medical School at Houston, Houston, TX, USA

⁎ Corresponding author at: Laboratório de MicrobUNASAU, Universidade do Extremo Sul Catarinense, 88Fax: 55 48 3443 4817.

E-mail address: tba@unesc.net (T. Barichello).

http://dx.doi.org/10.1016/j.jneuroim.2014.05.0030165-5728/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 18 March 2013Received in revised form 12 March 2014Accepted 4 May 2014

Keywords:Neonatal meningitisEscherichia coliCytokinesChemokinesMemory

Neonatal Escherichia colimeningitis continues to be an important cause of mortality and morbidity in newbornsworldwide. The aimof this studywas to investigate the cytokines/chemokines, brain-derived neurotrophic factor(BDNF) levels, blood–brain barrier integrity in neonatal rats following E. coli K1 experimental meningitisinfection and subsequent behavioural parameters in adulthood. In the hippocampus, interleukin increased at96 h, IL-6 at 12, 48 and 96 h, IL-10 at 96 h, cytokine-induced neutrophil chemoattractant-1 at 6, 12, 24, 48 and96 h, and BDNF at 48 and 96 h. In the cerebrospinal fluid, tumour necrosis factor alpha levels increased at 6,12, 24, 48 and 96 h. The BBB breakdown occurred at 12 h in the hippocampus, and at 6 h in the cortex.We evaluated behavioural parameters in adulthood: habituation to the open-field, step-down inhibitoryavoidance, object recognition, continuous multiple-trials step-down inhibitory avoidance and forcedswimming tasks. In adulthood, the animals showed habituation and aversive memory impairment.The animals needed a significant increase in the number of training periods to learn and not haddepressive-like symptoms.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Neonatal Escherichia coli K1meningitis continues to be an importantcause of mortality and morbidity in newborns worldwide (Kim, 2012).The prevalence of serious sequelae among a national cohort of 5-yearold children, born in England and Wales in 1996–7, who had neonatalmeningitis the overall incidence of serious disability was high, 25.5%in 1985 (De Louvois et al., 1991) compared to 23.5% in 1996 (DeLouvois et al., 2005). Moderate/severe disability was reported in 34%of children who had Streptococcus agalactiae meningitis, 30% who hadmeningitis due to E. coli or other gram-negative bacilli and 35% where

iologia Experimental, PPGCS,806-000 Criciúma, SC, Brazil.

meningitis was due to other bacteria (De Louvois et al., 2005). Further-more, E. coli is the second-most common cause of neonatal meningitisafter S. agalactiae in France (Gaschignard et al., 2012) and otherindustrialised countries (Bonacorsi et al., 2003). The inflammatory reac-tion is characterised by an acute purulent infection of the meninges af-fecting the pia mater, the arachnoid, and the subarachnoid space(Sellner et al., 2010; Zhu et al., 2012). Themicroorganisms can replicatewithin the subarachnoid space concomitantly with the release of thebacterial products that are highly immunogenic and can lead to an in-creased inflammatory response in the host (Barichello et al., 2013a;Sellner et al., 2010). Many brain cells, including glial cells, endothelialcells, ependymal cells, and resident macrophages, can produce cyto-kines and pro-inflammatory molecules in response to microorganisms(Kronfol and Remick, 2000). Pro-inflammatory mediators such as tu-mour necrosis factor alpha (TNF-α), interleukin-1 (IL-1), interleukin-6(IL-6), and interleukin-8 (IL-8) and the anti-inflammatory mediatorinterleukin-10 (IL-10) are involved in the pathophysiology of bacterialmeningitis (van Furth et al., 1996). In previous studies in neonatal

36 T. Barichello et al. / Journal of Neuroimmunology 272 (2014) 35–41

rats, we have shown an increase in cytokine-induced neutrophilchemoattractant-1 (CINC-1), IL-1β and IL-6 starting at 6 h and in TNF-α and IL-10 starting at 24 h in the hippocampus aftermeningitiswas in-duced by S. agalactiae (Barichello et al., 2011). In neonatal pneumococ-calmeningitis, the levels of CINC-1, IL-1β and TNF-α are increased in thehippocampus after 6 h (Barichello et al., 2012a). Thus, due to the pro-duction of cytokines, polymorphonuclear leukocytes are released fromthe blood and become activated, resulting in the production of highamounts of reactive oxygen and nitrogen species (Kastenbauer et al.,2002). This cascade leads to lipid peroxidation, DNA single-strandbreaks, mitochondrial damage and a breakdown of the blood–brain barrier (BBB) (Klein et al., 2006), thus contributing to braininjury during neonatal meningitis. Therefore, the aim of this studywas to investigate the levels of inflammatory mediators, BDNF atdifferent time points, and BBB integrity in neonatal Wistar rats sub-jected to E. coli K1 meningitis and subsequent behavioural parame-ters in adulthood.

2. Experimental procedures

2.1. Infecting organisms

The E. coli K1 strain was cultured overnight in 10 ml of Todd HewittBroth, Himedia® and then diluted in fresh medium and grown tologarithmic phase. The culture was centrifuged for 10 min at 5000 gand re-suspended in sterile saline to a concentration of 1 × 106

colony-forming units (CFU). The size of the inoculum was confirmedby quantitative cultures (Grandgirard et al., 2007; Barichello et al.,2010a).

2.2. Meningitis animal model

Neonatal maleWistar rats (15–20 g body weight) aged 3–4 postnataldays from our breeding colony were used for the experiments. All proce-dures were approved by the Animal Care and Experimentation Commit-tee of UNESC 165/2008, Brazil and were in accordance with theNational Institute of Health Guide for the Care and Use of LaboratoryAnimals (NIH Publications No. 80-23), revised in 1996. All surgical proce-dures and bacterial inoculations were performed under anaesthesiaconsisting of an intraperitoneal administration of ketamine (6.6 mg/kg),xylazine (0.3 mg/kg), and acepromazine (0.16 mg/kg) (Barichello et al.,2010b; Grandgirard et al., 2007). The animals received an intracisternal(i.c.) injection of 10 μl of sterile saline as a placebo or an equivalent vol-ume of E. coli K1 suspension. At the time of the inoculation, the animalsreceived fluid replacement and were subsequently returned to theircages (Barichello et al., 2010b; Irazuzta et al., 2008). Following their recov-ery from anaesthesia, the animals were fed by their mothers. Meningitiswas documented from a quantitative culture of 5 μl of cerebrospinalfluid (CSF) obtained by puncturing the cisterna magna (Barichello et al.,2010b).

2.3. Assessment of TNF-α, IL-6, IL-4, IL-10, CINC-1 and BDNF levels

Neonatal ratswere randomised and subjected tomeningitis by E. coliand were killed by decapitation at 6, 12, 24, 48 and 96 h (n = 6 pergroup). The hippocampus was immediately isolated on dry ice andstored at −80 °C for the analysis of TNF-α, IL-6, IL-4, IL-10, CINC-1and BDNF levels. The CSF was immediately isolated on dry ice andstored at −80 °C for the analysis of TNF-α levels (n = 4 to 6 animalsper group).

2.3.1. Assessment of TNF-α, IL-4, IL-6, IL-10 and CINC-1 concentrationsThe CSF and hippocampus were immediately isolated on dry ice and

stored at −80 °C for analyses of the cytokine levels. Brain tissue washomogenised in extraction solution (100mgof tissue per 1ml) contain-ing: 0.4 mol/l NaCl, 0.05% Tween 20, 0.5% 7 BSA, 0.1 mmol/l Phenyl

methyl sulfonil fluoride, 0.1 mmol/l benzethonium chloride, 10 mmol/lEDTA and 20 KI aprotinin, using Ultra-Turrax (Fisher Scientific,Pittsburgh, PA). Brain homogenate was centrifuged at 3000 g for10 min at 4 °C, and the supernatants were collected and stored at−20 °C. The concentration of chemokines and cytokinewas determinedusing ELISA. The supernatants of the brain tissue andCSFwere assayed inan ELISA setup using commercially available antibodies, according to theprocedures provided by the manufacturer (R&D Systems, Minneapolis,MN). The results are expressed in pg/100 mg of hippocampal tissue.The limit of detection was 20 pg/ml.

2.3.2. Assessment of BDNF expressionThe BDNF levels in the hippocampus were measured using an anti-

BDNF sandwich-ELISA according to the manufacturer's instructions(Chemicon, USA). Briefly, the hippocampus and frontal cortexwere homogenised in phosphate buffer solution (PBS) with 1 mMphenylmethylsulfonyl fluoride (PMSF) and 1 mM EGTA. Microtiterplates (96-well flat-bottom) were coated for 24 h with the samples di-luted 1:2 in a sample diluent, and the standard curve ranged from 7.8to 500 pg/ml of BNDF. The plates were then washed four times withsample diluent, and a monoclonal anti-BNDF rabbit antibody diluted1:1000 in sample diluent was added to each well and incubated for 3h at room temperature. After washing, a peroxidase-conjugated anti-rabbit antibody (diluted 1:1000) was added to each well and incubatedat room temperature for 1 h. After the addition of streptavidin-enzyme,substrate and stop solution, the amount of BDNF was determined bymeasuring the absorbance at 450 nm and was expressed as pg per mgofwet tissue protein. The standard curve demonstrates a direct relation-ship between Optical Density (OD) and BDNF concentration. Totalprotein was measured by the method of Lowry et al. (1951) (Lowryet al., 1951) using bovine serum albumin as a standard, as previouslydescribed by Frey et al. (2006).

2.4. Blood–brain barrier permeability to Evan's blue

To evaluate the BBB integrity, the animals were separated into twogroups: control and meningitis and were killed at different times at 6,12, 18, 24 and 30 h after induction of E. coli meningitis (Smith andHall, 1996). The animals were injected with 1ml of Evan's blue at 1% in-traperitoneal (i.p.) 1 h before being killed. The anaesthesia consisted ofan intraperitoneal administration of ketamine (6.6 mg/kg), xylazine(0.3 mg/kg), and acepromazine (0.16 mg/kg) (Barichello et al.,2012b). The chest was subsequently opened, and the brain wastranscardially perfused with 200 ml of saline through the left ventricleat a pressure of 100mmHguntil colourless perfusion fluidwas obtainedfrom the right atrium. The samples were weighed and placed in a 50%trichloroacetic solution. Following homogenization and centrifugation,the extracted dye was diluted with ethanol (1:3), and its fluorescencewas determined (excitation at 620 nm and emission at 680 nm) witha luminescence spectrophotometer (Hitachi 650-40, Tokyo, Japan).Calculations were based on the external standard (62.5–500 ng/ml) inthe same solvent. The Evan's blue tissue content was quantified from alinear standard line derived from known amounts of the dye, and itwas expressed per gramme of tissue (Smith and Hall, 1996).

2.5. Behavioural tasks

To evaluate the behavioural responses, the animals were separatedinto two groups: control and meningitis (n = 10 per group and 20 pereach behavioural task; n= 100). Eighteen hours after induction, the an-imals received ceftriaxone (100 mg/kg body weight given s.c./7 days)(Barichello et al., 2013b). Sixty days after inoculation, the animalswere free from infection. All blood cultures that were performed duringthis period were negative. The animals recovered their weight andgrooming habits; their blood counts returned to control levels, and thereactive protein C values were negative (data not shown). Finally, the

37T. Barichello et al. / Journal of Neuroimmunology 272 (2014) 35–41

animals were randomised and subjected to the following behaviouraltests: a) habituation to the open field task, b) step-down inhibitoryavoidance task, c) object recognition task, d) continuous multiple-trials step-down inhibitory avoidance task, and e) forced swimmingtest.

2.5.1. Open field testThe behaviour was assessed in the open field apparatus to evaluate

both locomotor and exploratory activities. The apparatus is a 40 ×60 cm open field surrounded by 50-cm high walls made of brownplywood with a frontal glass wall. The floor of the open field is dividedby black lines into 9 rectangles. Animals were gently placed on the leftrear quadrant and allowed in exploring the arena for 5min; the numberof crossings (the number of times that animals crossed the black lines,assessing the locomotor activity) and rearings (the explorationbehaviour observed in rats when placed in a new environment) wasmeasured. The behavioural test was performed by the same person(manual analyses) who was blind to the group treatment (Viannaet al., 2000).

2.5.2. Step-down inhibitory avoidance taskThe apparatus and procedures have been described in previous

reports (Quevedo et al., 1997; Roesler et al., 2003). Briefly, the trainingapparatus was a 50 × 25 × 25 cm acrylic box (Albarsch, Porto Alegre,Brazil) whose floor consisted of parallel calibre stainless steel bars(1 mm diameter) spaced 1 cm apart. A 7-cm wide and 2.5-cm highplatform was placed on the floor of the box against the left wall. In thetraining trial, the animalswere placed on the platform, and their latencyto step down onto the grid with all four paws was measured with anautomatic device. Immediately after stepping down on the grid, theanimals received a 0.4-mA, 2.0-s foot shock and were returned to theirhome cage. A retention test trial was performed 24 h after training(long-term memory) and was procedurally identical to the trainingexcept that no foot shock was presented. The retention test step-downlatency (maximum, 180 s) was used as a measure of inhibitory avoid-ance retention. Reactivity to the foot shock was evaluated in the sameapparatus used for inhibitory avoidance except that the platform wasremoved. Each animal was placed on the grid and allowed a 1-min ha-bituation period prior to the start of a series of shocks (0.5 s) thatwere delivered at 10-s intervals. Shock intensities ranged from 0.1 to0.5 mA in 0.1-mA increments. The adjustments in shock intensitywere made in accordance with each animal's response. The intensitywas raised by 1 unit when no response occurred and lowered by 1unit when a response was made. A “flinch” response was defined asthe withdrawal of one paw from the grid floor, and a “jump” responsewas defined as rapid withdrawal of three or four paws. Two measure-ments of the “flinch” threshold were recorded, and then two measure-ments of the “jump” threshold were recorded. For each animal, themean of the two scores for the “flinch” and the “jump” thresholdswere calculated (Izquierdo et al., 1998; Bevilaqua et al., 2003).

2.5.3. Object recognition taskThis task evaluates non-aversive, non-spatial memory. The apparatus

and procedures for the object recognition task have been described else-where (de Lima et al., 2005). Briefly, the task took place in a 40 × 50 cmopenfield surrounded by 50-cmhighwallsmade of plywoodwith a fron-tal glasswall. The floor of the open fieldwas divided by black lines into 12equal rectangles. All animals were submitted to a habituation sessionwhere theywere allowed to freely explore the openfield for 5min; noob-jects were placed in the box during the habituation trial. The number oftimes theblack lineswere crossed and thenumbers of rearings performedin this session were evaluated as indicators of locomotor and exploratoryactivity, respectively. At different times following habituation, trainingwas conducted by placing individual rats in the field for 5min. Two iden-tical objects (objects A1 and A2, both cubes)were positioned in two adja-cent corners, 10 cm from thewalls. In the long-term recognitionmemory

test that was given 24 h after training, the rats explored the open field for5min in the presence of one familiar (A) andonenovel (B, a pyramidwitha square-shaped base) object. All of the objects had similar texture(smooth), colour (blue), and size (weight 150–200 g) butwith distinctiveshapes. A recognition index was calculated for each animal and reportedas the ratio TB / (TA+TB) (TA= time spent exploring the familiar objectA; TB = time spent exploring the novel object B). Recognition memorywas evaluated as in the short-termmemory test. Explorationwas definedas sniffing (exploring the object 3–5 cm away from it) or touching theobject with the nose or forepaws.

2.5.4. Continuous multiple-trials step-down inhibitory avoidance taskThis task evaluates aversive memory in the test section and learning

when analysing the number of training trials required for the acquisi-tion criterion (see below). It was performed in the same step-downinhibitory avoidance apparatus; however, in the training session, theanimal was placed on the platform and received a 0.3-mA, 2.0-s footshock immediately after stepping downon the grid. This procedure con-tinued until the rat remained on the platform for 50 s. The animal wasthen returned to the home cage. The number of training trials requiredto reach the acquisition criterion of 50 s on the platform was recorded.The retention test was performed 24 h later (long-term memory)(Roesler et al., 1999).

2.5.5. Forced swimming taskThis test was conducted according to previous reports (Porsolt,

1979; Einat et al., 2001; Tuon et al., 2008) and was used as a modelfor depressive behaviour. Briefly, the test involves two exposures to acylindrical water tank in which rats cannot touch the bottom or escape.The tank is made of transparent plexiglass, 80 cm tall, 30 cm in diame-ter, and filled with water (22–23 °C) to a depth of 40 cm. The water inthe tank was changed for each rat. For the first exposure, the rats wereplaced in the water for 15 min (training session). After 24 h, the ratswere placed in the water again for a 5-min session (test session). Thetime of immobility was analysed in the test session; the rats werejudged to be immobile whenever they stopped swimming andremained floating in the water with their heads above the surface.

3. Statistics

Data were analysed for normality by the Shapiro–Wilk test and forhomogeneity using the Levene test. When the data were normal andhomogeneity of variance, parametric tests was used, not meeting thiscondition, non-parametric tests were used.

For cytokine, chemokine and BDNF level analyses were performedusing a Mann–Whitney U-test. The within-individual group compari-sons were made by Wilcoxon's tests. The statistically significant resultsare indicated by *p b 0.05, **p b 0.01 and ***p b 0.001.

The results of the BBB integrity are reported as the mean ± SD of 6animals in each group. Differences among groups were evaluatedusing analysis of variance (ANOVA) followed by the StudentNewman–Keuls post-hoc test.

Data from the inhibitory avoidance task and object recognition thecomparisons among groups were performed using a Mann–WhitneyU-test. The within-individual group comparisons were made byWilcoxon's tests. Data from the habituation to openfield task and forcedswimming test are reported as themean± SD and comparisons amonggroups were analysed by a paired Student's t-test. In addition, in openfield task, control and meningitis groups were compared with the TTest for independent samples. Data from the continuous multiplestep-down inhibitory avoidance task trials are reported as the medianand interquartile ranges and comparisons among groups were per-formed usingMann–Whitney U-test. The within-individual group com-parisons were made by Wilcoxon's tests. In all comparisons, p b 0.05indicated statistical significance. All analyses were performed using

0h 6h 12h 24h 48h 96h0

1000

2000

3000

4000 ****

* *

Control Meningitis

**

TN

F-α

(pg

/100

μL o

f C

SF

)

Fig. 2. Kinetics of TNF-α expression in CSF assessed by ELISA at 6, 12, 24, 48 and 96 h aftermeningitis induction. Results are expressed median (horizontal line) of 6 animals in eachgroup and were performed using a Mann–Whitney U-test. The within-individual groupcomparisons were made by Wilcoxon's tests. Symbols indicate statistically significantwhen compared with control group *p b 0.05, **p b 0.01.

38 T. Barichello et al. / Journal of Neuroimmunology 272 (2014) 35–41

the Statistical Package for the Social Science (SPSS) software version20.0.

4. Results

Fig. 1 illustrates the effects on IL-4, IL-6, IL-10, TNF-α, CINC-1 andBDNF levels in the hippocampus of neonatal Wistar rats subjected tomeningitis by E. coli K1. IL-4 was increased at 96 h (p b 0.01); IL-6 wasincreased at 12, 48 and 96 h (p b 0.001, p b 0.05 and p b 0.001, respec-tively); IL-10was increased at 96 h (p b 0.001); CINC-1was increased at6, 12, 24, 48 and 96 h (p b 0.001); and BDNFwas increased at 48 and 96h (p b 0.05 and p b 0.001, respectively) aftermeningitis inductionwhencompared with the control group. Fig. 2 shows that TNF-α levels in theCSF were increased at 6, 12, 24, 48 and 96 h after meningitis induction(p b 0.001, p b 0.05, p b 0.05, p b 0.001 and p b 0.01, respectively).

Figs. 3, 4 and 5 display the results of the behavioural tasks in adultWistar rats subjected to meningitis in the neonatal period by E. coliK1. In the habituation to open field task, as shown in Fig. 3, therewas a difference in the number of crossings and rearings between thetraining and test sessions in the control group, demonstrating memoryhabituation (p b 0.05). However, in the meningitis group, there was nodifference between the training and test sessions, demonstratingmemory impairment in this group (p N 0.05).

0h 6h 12h 24h 48h 96h

*

Meningitis

* * **

IL-4

(pg

/100

mg

of t

issu

e)

Control

0h 6h 12h 24h 48h 96h

Control Meningitis

**

* ***

IL-1

0 (p

g/10

0mg

of ti

ssue

)

0h 6h 12h 24h 48h 96h

Control Meningitis

** **** **

**

CIN

C-1

(pg

/100

mg

of t

issu

e)

0

500

1000

1500

0

500

1000

1500

2000

0

500

1000

1500

Fig. 1.Kinetics of IL-4, IL-6, IL-10, TNF-α, CINC-1 and BDNF expression in hippocampus assessed(horizontal line) of 6 animals in each group and were performed using a Mann–Whitney U-indicate statistically significant when compared with control group *p b 0.05, **p b 0.01 and **

In the step-down inhibitory avoidance task, Fig. 4, there were differ-ences between the training and test sessions in the control group,demonstrating aversive memory (p b 0.05). However, in themeningitis

0h 6h 12h 24h 48h 96h0

2000

4000

6000

8000 **

Control Meningitis

**

**

** **

IL-6

(pg

/100

mg

of t

issu

e)

0h 6h 12h 24h 48h 96h0

500

1000

1500

2000

Control Meningitis

**

*** **

**

TN

F-α

(pg

/100

mg

of t

issu

e)

0h 6h 12h 24h 48h 96h0

1000

2000

3000

Control Meningitis

*

*

* *

BD

NF

(pg

/100

mg

of t

issu

e)

by ELISA at 6, 12, 24, 48 and 96 h aftermeningitis induction. Results are expressedmediantest. The within-individual group comparisons were made by Wilcoxon's tests. Symbols*p b 0.001.

0

20

40

60

80

100Crossing - Training

Crossing - Test

Rearing - Training

Rearing - Test

*

*

Num

ber

of c

ross

ing

and

rear

ing

Control Meningitis

Fig. 3. Habituation to open field task in adult Wistar rats submitted to meningitis in neo-natal period. Numbers of the crossings and rearings, data are reported as mean (bar) ±SD (line) in training and test sessions and were analysed by the paired Student's t test,n = 10 animals per group. *p b 0.05 compared to training session.

0

100

200

300

400Training

Test*/#*

La

ten

cy (

sec)

Control Meningitis

Fig. 5. Continuous multiple trials step-down inhibitory avoidance task trials in adultWistar rats submitted to meningitis in neonatal period. Data are reported as median(bar) and interquartile ranges (line), and comparisons among groups were performedusing Mann–Whitney U test. The within-individual groups were analysed by Wilcoxon's,n = 10 animals per group. *p b 0.05 compared to training session. #p b 0.05 comparedto control group.

800

1000

1200

1400

1600Control

Meningitis

*A

eaka

ge (

ng/m

l)

39T. Barichello et al. / Journal of Neuroimmunology 272 (2014) 35–41

group, there were no differences between the training and test session,demonstrating impairment in aversive memory.

In the continuous multiple-trials step-down inhibitory avoidancetask, there was a significant increase in the number of training trialsrequired to reach the acquisition criterion (180 s on the platform) inthe meningitis group when compared to the control group (controlMd 1.0, q1 − q3 = 1.0–2.0; meningitis Md 3.5, q1 − q3 = 3.0–4.0;p b 0.05). The results of this task suggest that the meningitis group re-quired more stimuli to reach the acquisition criterion when comparedwith the control group, indicating a learning impairment. In the reten-tion test, Fig. 5, there were difference between the training and testsessions in both groups (p b 0.05), there were decrease latency timein the meningitis group when compared with control group (p b 0.05).

In the object recognition task, the meningitis group recognised thenovel object, i.e., they spent a significantly higher percentage of time ex-ploring the novel object. Therefore, the meningitis group presentedlong-term recognition retention memory when compared to the train-ing session (p b 0.05) (data not shown).

In the forced swimming task, no differences in immobility timewereobserved in the meningitis group when compared to the control group,suggesting a lack of depressive-like behaviour in the meningitis group(p N 0.05), (data not shown).

The BBB integrity of hippocampus (Fig. 6A) and cortex (Fig. 6B),were investigated using Evan's blue dye extravasation. We observedthe BBB breakdown, in hippocampus, within 12 h and 18 h (p b 0.05)and in the cortex it started at 6 h until 24 h after E. coli meningitisinduction (p b 0.05).

Control Meningitis0

100

200

300

400Training

Test

*

#

La

ten

cy (

sec)

Fig. 4. Step-down inhibitory avoidance task in adultWistar rats submitted tomeningitis inneonatal period. Data were reported as median (bar) and interquartile ranges (line), andcomparisons among groups were performed using Mann–Whitney U tests. The within-individual groups were analysed by Wilcoxon's tests, n = 10 animals per group.*p b 0.05 compared to training session. #p b 0.05 compared to control group.

5. Discussion

In the present study, we show the influence of neonatal meningitisinduced by E. coli K1 on kinetic cytokine/chemokine, BDNF levels inthe hippocampus at different time intervals, and the BBB integrity.Furthermore, we evaluated the learning and memory abilities of adultrats subjected to meningitis during the neonatal period.

The BBB breakdown occurred in the hippocampus at 12 h and in thecortex at 6 h after E. coli K1 meningitis induction. In previous studieswith experimental neonatal meningitis, the BBB breakdown in the hip-pocampus at 18 h and in the cortex at 12 h after pneumococcal menin-gitis (Barichello et al., 2012a), and in neonatalmeningitis by S. agalactiaethe breakdown of BBB started at 12 h in both structures (Barichello etal., 2011). Most cytokines and chemokines increased over the first 96h; however, CINC-1 began to be produced 6 h after the induction ofmeningitis. In previous studies with experimental animal model, we

0

200

400

600

*

Eva

ns B

lue

L

0

500

1000

1500

2000Control

Meningitis

*

*

*

*

B

Eva

ns B

lue

Leak

age

(ng/

ml)

6h 12h 18h 24h 30h

6h 12h 18h 24h 30h

Fig. 6.The integrity of the blood–brainbarrier in thehippocampus and cortex. (A)hippocam-pus and (B) cortex were obtained in several times at 6, 12, 18, 24 and 30 h after meningitisinduction. Results show the mean (bar) ± SD (line) of 6 animals in each group.

40 T. Barichello et al. / Journal of Neuroimmunology 272 (2014) 35–41

showed that TNF-α and CINC-1 levels increased in the hippocampuswithin the first 6 h after pneumococcal meningitis induction(Barichello et al., 2012a); CINC-1, IL-1β and IL-6 levels were increasedin the first 6 h and TNF-α at 24 h after S. agalactiae neonatal meningitisinduction (Barichello et al., 2011). In adult rats, CINC-1 levels first in-creased in jugular plasma as opposed to arterial plasma in adult ratsafter pneumococcal meningitis induction (Barichello et al., 2011).CINC-1 also seems to be relevant in the early stage of bacterial meningi-tis because of its ability of promoting leukocyte migration, which war-rants further investigation. This inflammatory response is essential forsurvival in response to infection, but it also has the capacity to causeconsiderable damage (Konsman et al., 2007). BDNF secreted by immunecells is bioactive (Kerschensteiner et al., 1999), and it has neuroprotec-tive effects on neurons following bacterialmeningitis induction (Li et al.,2007). Accordingly, BDNF levels were increased at 48 and 96 h aftermeningitis induction.

In the present study, neonatal Wistar rats showed habituation andaversive memory impairment following meningitis induction; inaddition, the animals required a significantly greater number of trainingsessions to learn in adulthood. Long-term sequelae in patients who hadneonatal meningitis are frequent, particularly neurosensorial (14–17%)andneurodevelopmental (10–17%) sequelae (Durrmeyer et al., 2012). Atotal of 1433 childrenwhowere survivors of childhood bacterialmenin-gitis were evaluated for sequelae after the time of discharge. Of thesechildren, 49.2% were reported to have 1 or more long-term sequelae.A number of reported cases, 45%, included behavioural and/or intellec-tual sequelae. Hearing loss accounted for 6.7% of sequelae, andneurologic deficits represented 14.3% (Chandran et al., 2011).

Another study of 19 cases of neonatal meningitis showed thatS. agalactiae meningitis patients performed worse in cognitive termsthan the Gram negative cases (Franco et al., 1992), although in anotherstudy, the Gram negative cases did worse (Fitzhardinge et al., 1974).

One cause of cognitive impairment could be the inflammatory reac-tion that can involve themeninges and the subarachnoid space and alsothe brain parenchyma vessels, contributing to the development of neu-ronal damage (Sellner et al., 2010). Our understanding of the patho-physiology of bacterial meningitis is mostly based on observations ofhuman cases and studies in animal models. In autopsy cases of bacterialmeningitis, damage in the CNS was characterised by hippocampal apo-ptosis in the dentate gyrus (Nau et al., 1999); the same damage was ob-served in animal models with a peak of apoptosis at 30 h aftermeningitis infection in infant rats (Grandgirard et al., 2007). This typeof damage has been associated with learning andmemory impairmentsin experimental models (Grandgirard and Leib, 2010). The innate im-mune signalling pathways through the Toll-like receptors (TLR) are crit-ical for the detection and recognition of bacterial compounds. E. coliinteracts through lipopolysaccharide with TLR4 (Rivest, 2003). This re-ceptor induces the activation of factor nuclear kappa B or mitogen-activated protein kinase pathways. This subsequently up-regulates theleukocyte populations and can express numerous proteins involved ininflammation and the immune response (Cuenca et al., 2013). Due tothe production of cytokines, polymorphonuclear leukocytes are re-leased from the blood and produce high amounts of reactive oxygenspecies (Kastenbauer et al., 2002). Reactive oxygen species result inBBB disruption, mitochondrial dysfunction that leads to the release ofapoptosis-inducing factors into the cytosol (Mitchell et al., 2004). Incontrast, the caspase-dependent pathway can be triggered from whiteblood cell walls (Mitchell et al., 2004). In a study of 54 newborns, thirtypatients had meningitis, and 24 were in the control group. There wereincreased levels of TNF-α, IL-1β and IL-6 in the CSF of all newbornswith meningitis (Krebs et al., 2005). In paediatric patients with CNS in-fections, mainly those with bacterial meningitis, BDNF levels were ele-vated in the serum and in the CSF (Morichi et al., 2013). We showedan increase in BDNF levels at 48 and 96 h after meningitis induction.As with cytokines, BDNF can be expressed from human immune cellssuch as T cells, B cells, monocytes, macrophages and astrocytes

(Kerschensteiner et al., 1999). Neurons are the main source of BDNF inthe CNS (Lewin and Barde, 1996). Furthermore, BDNF secreted by im-mune cells is bioactive (Kerschensteiner et al., 1999), and it has neuro-protective effects on neurons following bacterial meningitis induction(Li et al., 2007). BDNF protects against multiple forms of brain injury(Bifrare et al., 2005), and may rescue brain neurons from inflammatorybrain injury in bacterial meningitis (Li et al., 2007).

6. Conclusion

Although there are limitations of meningitis animal model, itcontinues to provide the backbone for pathophysiology studies. In thisstudy, we report the cytokine, BDNF profile at different times, BBBdisruption, and cognitive impairment in adulthood; however, animalmodel should be interpreted with caution before correlating with clini-cal observations.

Acknowledgements

Laboratory of Experimental Microbiology (Brazil) is one of the cen-tres of the National Institute for Translational Medicine (573671/2008-7) (INCT-TM) and one of themembers of the Center of Excellencein Applied Neurosciences of Santa Catarina (NENASC). This researchwas supported by grants from CNPq (475719/2012-3) (JQ and TB),FAPESC (No. 04/2011) (JQ and TB), FAPEMIG (ALT), Instituto Cérebro eMente (JQ and TB) and UNESC (JQ and TB). JQ and TB are CNPq ResearchFellows. JSG and LRS are holders of a CAPES studentship.

References

Barichello, T., Belarmino Jr., E., Comim, C.M., Cipriano, A.L., Generoso, J.S., Savi, G.D., Stertz,L., Kapczinski, F., Quevedo, J., 2010a. Correlation between behavioral deficits and de-creased brain-derived neurotrophic [correction of neurotrofic] factor in neonatalmeningitis. J. Neuroimmunol. 223, 73–76.

Barichello, T., Savi, G.D., Silva, G.Z., Generoso, J.S., Bellettini, G., Vuolo, F., Petronilho, F.,Feier, G., Comim, C.M., Quevedo, J., Dal-Pizzol, F., 2010b. Antibiotic therapy prevents,in part, the oxidative stress in the rat brain after meningitis induced by Streptococcuspneumoniae. Neurosci. Lett. 478, 93–96.

Barichello, T., Lemos, J.C., Generoso, J.S., Cipriano, A.L., Milioli, G.L., Marcelino, D.M., Vuolo,F., Petronilho, F., Dal-Pizzol, F., Vilela, M.C., Teixeira, A.L., 2011. Oxidative stress,cytokine/chemokine and disruption of blood–brain barrier in neonate rats aftermeningitis by Streptococcus agalactiae. Neurochem. Res. 36, 1922–1930.

Barichello, T., Fagundes, G.D., Generoso, J.S., Paula Moreira, A., Costa, C.S., Zanatta, J.R.,Simoes, L.R., Petronilho, F., Dal-Pizzol, F., Carvalho Vilela, M., Lucio Teixeira, A.,2012a. Brain–blood barrier breakdown and pro-inflammatory mediators in neonaterats submitted meningitis by Streptococcus pneumoniae. Brain Res. 1471, 162–168.

Barichello, T., Generoso, J.S., Silvestre, C., Costa, C.S., Carrodore, M.M., Cipriano, A.L.,Michelon, C.M., Petronilho, F., Dal-Pizzol, F., Vilela, M.C., Teixeira, A.L., 2012b. Circulat-ing concentrations, cerebral output of the CINC-1 and blood–brain barrier disruptionin Wistar rats after pneumococcal meningitis induction. Eur. J. Clin. Microbiol. Infect.Dis. 31 (8), 2005–2009.

Barichello, T., Fagundes, G.D., Generoso, J.S., Elias, S.G., Simoes, L.R., Teixeira, A.L., 2013a.Pathophysiology of neonatal acute bacterial meningitis. J. Med. Microbiol. 62,1781–1789.

Barichello, T., Goncalves, J.C., Generoso, J.S., Milioli, G.L., Silvestre, C., Costa, C.S., Coelho Jda,R., Comim, C.M., Quevedo, J., 2013b. Attenuation of cognitive impairment by thenonbacteriolytic antibiotic daptomycin in Wistar rats submitted to pneumococcalmeningitis. BMC Neurosci. 14, 1471–2202.

Bevilaqua, L.R., Kerr, D.S., Medina, J.H., Izquierdo, I., Cammarota, M., 2003. Inhibition ofhippocampal Jun N-terminal kinase enhances short-term memory but blocks long-termmemory formation and retrieval of an inhibitory avoidance task. Eur. J. Neurosci.17, 897–902.

Bifrare, Y.D., Kummer, J., Joss, P., Tauber, M.G., Leib, S.L., 2005. Brain-derived neurotrophicfactor protects against multiple forms of brain injury in bacterial meningitis. J. Infect.Dis. 191, 40–45.

Bonacorsi, S., Clermont, O., Houdouin, V., Cordevant, C., Brahimi, N., Marecat, A., Tinsley, C.,Nassif, X., Lange, M., Bingen, E., 2003. Molecular analysis and experimental virulenceof French and North American Escherichia coli neonatalmeningitis isolates: identifica-tion of a new virulent clone. J. Infect. Dis. 187, 1895–1906.

Chandran, A., Herbert, H., Misurski, D., Santosham, M., 2011. Long-term sequelae of child-hood bacterial meningitis: an underappreciated problem. Pediatr. Infect. Dis. J. 30,3–6.

Cuenca, A.G., Wynn, J.L., Moldawer, L.L., Levy, O., 2013. Role of innate immunity in neona-tal infection. Am. J. Perinatol. 30, 105–112.

De Lima, M.N., Laranja, D.C., Bromberg, E., Roesler, R., Schroder, N., 2005. Pre- or post-training administration of the NMDA receptor blocker MK-801 impairs object recog-nition memory in rats. Behav. Brain Res. 156, 139–143.

41T. Barichello et al. / Journal of Neuroimmunology 272 (2014) 35–41

De Louvois, J., Blackbourn, J., Hurley, R., Harvey, D., 1991. Infantile meningitis in Englandand Wales: a two year study. Arch. Dis. Child. 66, 603–607.

De Louvois, J., Halket, S., Harvey, D., 2005. Neonatal meningitis in England andWales: se-quelae at 5 years of age. Eur. J. Pediatr. 164, 730–734.

Durrmeyer, X., Cohen, R., Bingen, E., Aujard, Y., 2012. Therapeutic strategies for Escherichiacoli neonatal meningitis. Arch. Pediatr. 19 (Suppl. 3), S140–S144.

Einat, H., Clenet, F., Shaldubina, A., Belmaker, R.H., Bourin, M., 2001. The antidepressantactivity of inositol in the forced swim test involves 5-HT(2) receptors. Behav. BrainRes. 118, 77–83.

Fitzhardinge, P.M., Kazemi, M., Ramsay, M., Stern, L., 1974. Long-term sequelae ofneonatal meningitis. Dev. Med. Child Neurol. 16, 3–8.

Franco, S.M., Cornelius, V.E., Andrews, B.F., 1992. Long-term outcome of neonatal menin-gitis. Am. J. Dis. Child. 146, 567–571.

Frey, B.N., Andreazza, A.C., Cereser, K.M., Martins, M.R., Valvassori, S.S., Reus, G.Z.,Quevedo, J., Kapczinski, F., 2006. Effects of mood stabilizers on hippocampus BDNFlevels in an animal model of mania. Life Sci. 79, 281–286.

Gaschignard, J., Levy, C., Bingen, E., Cohen, R., 2012. Epidemiology of Escherichia coli neo-natal meningitis. Arch. Pediatr. 19 (Suppl. 3), S129–S134.

Grandgirard, D., Leib, S.L., 2010. Meningitis in neonates: bench to bedside. Clin. Perinatol.37, 655–676.

Grandgirard, D., Steiner, O., Tauber, M.G., Leib, S.L., 2007. An infant mouse model of braindamage in pneumococcal meningitis. Acta Neuropathol. 114, 609–617.

Irazuzta, J., Pretzlaff, R.K., Zingarelli, B., 2008. Caspases inhibition decreases neurologicalsequelae in meningitis. Crit. Care Med. 36, 1603–1606.

Izquierdo, I., Barros, D.M., Mello, E., Souza, T., De Souza, M.M., Izquierdo, L.A., Medina, J.H.,1998. Mechanisms for memory types differ. Nature 393, 635–636.

Kastenbauer, S., Koedel, U., Becker, B.F., Pfister, H.W., 2002. Pneumococcal meningitis inthe rat: evaluation of peroxynitrite scavengers for adjunctive therapy. Eur. J.Pharmacol. 449, 177–181.

Kerschensteiner, M., Gallmeier, E., Behrens, L., Leal, V.V., Misgeld, T., Klinkert, W.E.,Kolbeck, R., Hoppe, E., Oropeza-Wekerle, R.L., Bartke, I., Stadelmann, C., Lassmann,H., Wekerle, H., Hohlfeld, R., 1999. Activated human T cells, B cells, and monocytesproduce brain-derived neurotrophic factor in vitro and in inflammatory brain lesions:a neuroprotective role of inflammation? J. Exp. Med. 189, 865–870.

Kim, K.S., 2012. Current concepts on the pathogenesis of Escherichia coli meningitis:implications for therapy and prevention. Curr. Opin. Infect. Dis. 25, 273–278.

Klein, M., Koedel, U., Pfister, H.W., 2006. Oxidative stress in pneumococcal meningitis: afuture target for adjunctive therapy? Prog. Neurobiol. 80, 269–280.

Konsman, J.P., Drukarch, B., Van Dam, A.M., 2007. (Peri)vascular production and action ofpro-inflammatory cytokines in brain pathology. Clin. Sci. (Lond.) 112, 1–25.

Krebs, V.L., Okay, T.S., Okay, Y., Vaz, F.A., 2005. Tumor necrosis factor-alpha, interleukin-1beta and interleukin-6 in the cerebrospinal fluid of newborn with meningitis. Arq.Neuropsiquiatr. 63, 7–13.

Kronfol, Z., Remick, D.G., 2000. Cytokines and the brain: implications for clinical psychia-try. Am. J. Psychiatry 157, 683–694.

Lewin, G.R., Barde, Y.A., 1996. Physiology of the neurotrophins. Annu. Rev. Neurosci. 19,289–317.

Li, L., Shui, Q.X., Liang, K., Ren, H., 2007. Brain-derived neurotrophic factor rescues neuronsfrom bacterial meningitis. Pediatr. Neurol. 36, 324–329.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with theFolin phenol reagent. J. Biol. Chem. 193, 265–275.

Mitchell, L., Smith, S.H., Braun, J.S., Herzog, K.H., Weber, J.R., Tuomanen, E.I., 2004. Dualphases of apoptosis in pneumococcal meningitis. J. Infect. Dis. 190, 2039–2046.

Morichi, S., Kashiwagi, Y., Takekuma, K., Hoshika, A., Kawashima, H., 2013. Expressions ofbrain-derived neurotrophic factor (BDNF) in cerebrospinal fluid and plasma of chil-dren with meningitis and encephalitis/encephalopathy. Int. J. Neurosci. 123, 17–23.

Nau, R., Soto, A., Bruck, W., 1999. Apoptosis of neurons in the dentate gyrus in humanssuffering from bacterial meningitis. J. Neuropathol. Exp. Neurol. 58, 265–274.

Porsolt, R.D., 1979. Animal model of depression. Biomedicine 30, 139–140.Quevedo, J., Vianna, M., Zanatta, M.S., Roesler, R., Izquierdo, I., Jerusalinsky, D., Quillfeldt, J.

A., 1997. Involvement of mechanisms dependent on NMDA receptors, nitric oxideand protein kinase A in the hippocampus but not in the caudate nucleus in memory.Behav. Pharmacol. 8, 713–717.

Rivest, S., 2003. Molecular insights on the cerebral innate immune system. Brain Behav.Immun. 17, 13–19.

Roesler, R., Vianna,M.R., De-Paris, F., Quevedo, J., 1999.Memory-enhancing treatments donot reverse the impairment of inhibitory avoidance retention induced by NMDA re-ceptor blockade. Neurobiol. Learn. Mem. 72, 252–258.

Roesler, R., Schroder, N., Vianna, M.R., Quevedo, J., Bromberg, E., Kapczinski, F., Ferreira, M.B., 2003. Differential involvement of hippocampal and amygdalar NMDA receptors incontextual and aversive aspects of inhibitory avoidance memory in rats. Brain Res.975, 207–213.

Sellner, J., Täuber, M.G., Leib, S.L., 2010. Chapter 1— pathogenesis and pathophysiology ofbacterial CNS infections. In: Karen, L.R., Allan, R.T. (Eds.), Handbook of Clinical Neurol-ogy. Elsevier, pp. 1–16.

Smith, S.L., Hall, E.D., 1996. Mild pre- and posttraumatic hypothermia attenuates blood–brain barrier damage following controlled cortical impact injury in the rat. J.Neurotrauma 13, 1–9.

Tuon, L., Comim, C.M., Petronilho, F., Barichello, T., Izquierdo, I., Quevedo, J., Dal-Pizzol, F.,2008. Time-dependent behavioral recovery after sepsis in rats. Intensive Care Med.34, 1724–1731.

Van Furth, A.M., Roord, J.J., Van Furth, R., 1996. Roles of proinflammatory andanti-inflammatory cytokines in pathophysiology of bacterial meningitis and effectof adjunctive therapy. Infect. Immun. 64, 4883–4890.

Vianna, M.R., Alonso, M., Viola, H., Quevedo, J., De Paris, F., Furman, M., De Stein, M.L.,Medina, J.H., Izquierdo, I., 2000. Role of hippocampal signaling pathways in long-term memory formation of a nonassociative learning task in the rat. Learn. Mem. 7,333–340.

Zhu, M.L., Mai, J.Y., Zhu, J.H., Lin, Z.L., 2012. Clinical analysis of 31 cases of neonatal puru-lent meningitis caused by Escherichia coli. Zhongguo Dang Dai Er Ke Za Zhi 14,910–912.