Available online at www.sciencedirect.comBRAIN,

www.elsevier.com/locate/ybrbi

Brain, Behavior, and Immunity 22 (2008) 367–374

BEHAVIOR,and IMMUNITY

Brief and long maternal separations decrease corticosterone secretionin a lupus-prone strain: Dissociation from disease-related parameters

Bruna Catallani a, Beatriz Duarte Palma a, Frida Zaladek Gil b, Deborah Suchecki a,*

a Department of Psychobiology, Universidade Federal de Sao Paulo (UNIFESP), Rua Napoleao de Barros, 925,

04024-002 Vila Clementino, Sao Paulo, SP, Brazilb Department of Physiology, Universidade Federal de Sao Paulo (UNIFESP), Sao Paulo, Brazil

Received 5 June 2007; received in revised form 23 August 2007; accepted 23 August 2007Available online 24 October 2007

Abstract

Neonatal manipulations are known to alter the activity of the immune system and the hypothalamus–pituitary–adrenal (HPA) axis.This study was performed in order to examine whether brief and long maternal separations (BMS and LMS, respectively) interfere withthe onset and development of murine lupus in NZB/NZWF1 females, and to determine whether the pattern of corticosterone (CORT)secretion throughout life is associated to the expression of the disease. Maternal separation was performed daily during postnatal days 1–14, lasting 15 min in the BMS group and 3 h in the LMS group. Blood was sampled from the retro-orbital plexus on the 9th week, andevery other week, from 10th to 34th weeks of life, for detection of anti-nuclear antibodies (ANA) and anti-double-strand DNA (anti-dsDNA) antibodies, and for determination of CORT serum levels. Urine samples were collected on the 21st, 27th, 33rd and 37th weeksof life. There were no group differences in regard to disease-related parameters, but LMS females presented a tendency for late onset ofanti-dsDNA antibodies. BMS and LMS mice exhibited reduced CORT levels compared to non-manipulated (NM) animals. There was astrong negative correlation between total mean CORT concentration and onset of ANA, and a strong positive correlation between totalmean CORT concentration and life span only in the NM group. Neonatal manipulations appeared to eliminate these correlations; hence,both BMS and LMS modified basal CORT secretion and the association between glucocorticoids and immune activity in the NZB/NZWF1 mouse strain.� 2007 Elsevier Inc. All rights reserved.

Keywords: Autoimmunity; Maternal Separation; NZB/WF1 mice; HPA axis; Systemic lupus erythematosus (SLE); Early life stress; Anti-nuclear anti-bodies (ANA); Anti-double-strand DNA (anti-dsDNA) antibodies; Proteinuria

1. Introduction

Dam–pup interaction is essential for the development ofthe infants and maternal care is a main regulator of a num-ber of physiological processes in the pup (Levine et al.,1991; Rosenfeld et al., 1993; Suchecki et al., 1993). Disrup-tions of this interaction, such as brief and long maternalseparations (BMS and LMS, respectively) during the firstweeks of life are known to affect systems as the hypothala-mus–pituitary–adrenal (HPA) axis. On one hand, BMS hasbeen shown to decrease ACTH and corticosterone (CORT)

0889-1591/$ - see front matter � 2007 Elsevier Inc. All rights reserved.

doi:10.1016/j.bbi.2007.08.010

* Corresponding author. Fax: +55 11 5572 5092.E-mail address: suchecki@psicobio.epm.br (D. Suchecki).

responses to a wide variety of stressors, and rats submittedto this manipulation also present a more efficient HPA axisnegative feedback (Meaney et al., 1985, 1996). On the otherhand, LMS promotes an increase in HPA axis stressresponsiveness and a decrease in the CORT negative feed-back (Ladd et al., 2004; Meaney et al., 1996).

Glucocorticoids (GCs) play a major regulatory role onthe immune system (IS), since they act as immunomodula-tors, interfering with transcription genes involved in theinflammatory response. At physiological levels, these hor-mones induce a shift from the Th1 cytokine productionpattern (proinflammatory) to the Th2 pattern (anti-inflam-matory). Proinflammatory and anti-inflammatory cytokinebalance is essential, and loss of this balance leads to

368 B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374

numerous neoplasic, infectious, allergic, inflammatory andautoimmune diseases (AD) (reviewed in Elenkov andChrousos, 2002). A classical example of the influence ofHPA axis on AD is evidenced in the Lewis and Fischerrat strains. Lewis rats are susceptible to some inflammatorydiseases, in part due to their defective HPA axis response.Conversely, the Fischer strain displays a hyperesponsiveHPA axis and is relatively resistant to the same autoim-mune inflammatory diseases (Sternberg et al., 1989a, b).

Neonatal manipulations, which affect HPA axis respon-siveness, also alter the susceptibility to experimental aller-gic encephalomyelitis (EAE), an AD with Th1 cytokinepattern. Studies have shown that BMS increases whileLMS decreases susceptibility to EAE (Laban et al.,1995a, b). However, there seems to be some controversy,since Stephan and colleagues (2002) reported an early onsetand worsening of EAE in female Lewis rats submitted to adaily 2 h-maternal separation with a concomitant decreasein CORT levels, compared to control rats. In addition, 2 h-neonatal handling with periodic touching also aggravatesEAE in Lewis female rats (Manni et al., 1998) and rats sub-mitted to 24 h-maternal deprivation on day 9 of life exhibit,as adults, a much higher EAE clinical score and reducedLPS-induced TNF-a and increased nitrite production (asa measurement of nitric oxide production) (Teunis et al.,2002).

Systemic lupus erythematosus (SLE) is a multi-systemicAD, characterized by overproduction of Th2 cytokines,and disease remissions and bursts. The better knownimmune alteration in lupus is the production of numerousantibodies against self-antigens, including antibodiesagainst DNA, histones, nucleosomes and ribonucleopro-teins, the so-called anti-nuclear antibodies (ANA). Themain markers for SLE are the anti-double-strand DNA(anti-dsDNA) and anti-nucleosome antibodies, which areimplicated in the formation and deposition of immunecomplexes in target organs, including the kidney (Mansonand Rahman, 2006). The female-to-male incidence ratio ofthis disease is approximately 9:1, and it affects mainlywomen at the reproductive age, essentially due to the influ-ence of sex hormones (Wilder, 1995). Other factors are alsoassociated with the onset and worsening of SLE, such aspregnancy, stress and a variety of drugs (Ostensen, 1999;Peralta-Ramirez et al., 2004; Rubin, 2005).

The New Zealand Black/New Zealand White F1 (NZB/WF1) mouse strain spontaneously develops an AD thatclosely resembles immunological and clinical characteris-tics of human SLE, such as a high production of ANA,including anti-dsDNA antibodies, lymphadenopathy,arthritis, hemolytic anemia, vasculitis, and a variety of his-topathological manifestations, of which glomerulonephritisis the most prominent (Theofilopoulos and Dixon, 1985).As in humans, the incidence is higher and severity of thedisease is worse in females (Theofilopoulos and Dixon,1981). Data from our laboratory show that this strainexhibits an earlier onset of the disease when submitted tosleep deprivation (Palma et al., 2006), and that animals

exposed to this kind of stress present sustained elevatedCORT levels for several weeks afterwards (Palma et al.,2007), suggesting an important role of stress on manifesta-tion of AD and HPA axis activity in NZB/WF1 mice.

Since neonatal manipulations modify HPA axis respon-siveness and the course of AD in some animal models, wehypothesized that a spontaneous AD could also be affectedby disruptions in the dam–pup interaction. Therefore, thisstudy was performed in order to examine whether BMSand LMS modify the onset and development of murinelupus in NZB/WF1 females, and to associate the patternof corticosterone (CORT) secretion to the expression ofthe disease.

2. Materials and methods

2.1. Animals

All procedures were carried out in accordance with the guidelines onanimal care of the National Institutes of Health and were approved bythe Ethics Committee in Research of the Universidade Federal de SaoPaulo (CEP 0881/03).

Female NZB and male NZW mice were purchased from Universidadede Sao Paulo (Sao Paulo, Brazil), and then matted in the Research Labo-ratory of the Department of Psychobiology of Universidade Federal deSao Paulo to generate NZB/WF1 hybrids. Both parents remained withthe litter until weaning. After weaning, mice were housed in plastic cagesin groups of three to eight animals and kept at a constant temperature(21 ± 2 �C) and under a 12/12 h light/dark cycle (lights on at 7:00 h). Foodand water were available ad libitum. Once females are more affected bymurine lupus than males, only this gender was used in the present study(Theofilopoulos and Dixon, 1981).

2.2. Maternal separations

The day of delivery was set as post-natal day (PND) 0. Maternal sep-arations were performed daily from PND 1 to 14, from 13:00 h to 16:00 h.Whole litters were removed from the nest and placed in separate cagesduring 15 min (BMS) or 3 h (LMS) in an adjacent room. Pups were keptwarm during the separation period by heating pads set at 33 �C. The non-manipulated (NM) group was not disturbed, except for cage cleaning oncea week. Weaning was performed between days 21 and 23, resulting in threegroups: NM (n = 19), BMS (n = 15) and LMS (n = 17).

2.3. Blood sampling and antibody determination

After a rapid anesthesia by ether vapors (approximately 15 s), bloodwas sampled from the retro-orbital plexus on the 9th week, and everytwo weeks, from 10th to 34th weeks of life. After centrifugation(2500 rpm for 7 min), serum was separated and stored at �20 �C untilanalysis. Blood sampling was carried out between 9:00–11:00 h.

ANA was determined by a standard indirect immunofluoresce (IIF)technique using HEp-2 cells as the substrate (Hemagen Diagnostics,Inc., Columbia, MD, USA). Briefly, the slide with the substrate was incu-bated with diluted serum for 30 min at 37 �C, allowing autoantibodies inthe sample to bind to antigens in the substrate. Excess serum from the slidewas removed by PBS wash. Substrate was then incubated with fluorescein-conjugated (FITC) rabbit anti-mouse IgG, kindly donated by BioLab, for30 min at 37 �C. New PBS wash removed the excess of FITC and slide wascoverslipped and checked for fluorescent patterns with a fluorescentmicroscope. Positive and negative controls were added in each assay.Starting dilution of serum in PBS was set in 1/50 and was progressivelyincreased, until no antibodies were detected in the sample. Proportionalnumeric values (1, 2, 10, 20, 40 and 80) were attributed to dilutions

6 10 12 14 16 18 20 22 24 26 28 30 32

Weeks of Life

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Pro

port

ion

ofA

nim

als

Neg

ativ

e fo

r A

NA

NMBMSLMS

NMBMSLMS

NMBMSLMS

8

Fig. 1. Onset of ANA production. Each marker expresses the initial ANAmanifestation of a particular animal. Censure was set in the 34th week oflife. NM, non-manipulated (n = 19); BMS, brief maternal separation(n = 15); LMS, long maternal separation (n = 17).

B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374 369

(1/50, 1/100, 1/500, 1/1000, 1/2000 and 1/4000, respectively) when theresult was positive. Value 0 was attributed to negative samples. The othervalues were transformed to base 10 logarithm plus 1.

ANA-positive samples were investigated for the presence of anti-dsDNA antibodies, which are thought to reflect the disease activity (Schurand Sandson, 1968). The same IIF technique was applied, the only differ-ences lying on the substrate, the flagellated Crithidia luciliae (DTS—Diag-nostic & Technical Services, Randburg, South Africa) and on the serumdilution in PBS (1/5).

The analyses of the slides were made by a blind expert-reader influorescence microscope.

2.4. Measurement of proteinuria

Since renal involvement in lupus is characterized by proteinuria, urinesamples were collected by placing the animals individually in metaboliccages at 21st, 27th, 33rd and 37th weeks of age for 5–6 h. The protein con-centration was assessed by precipitation with 3% sulfosalicylic acid andmeasurements were performed in a spectophotometry (Spectronic, Gene-sys 5 Milton Roy). The microplate reader was set in 650 nm. A standardcurve was generated to obtain optical density values.

2.5. CORT determination

CORT concentrations were quantified by a double radioimmunoassaymethod, specific for rats and mice, using a commercial kit (ICN Biomed-icals, Costa Mesa, CA, USA) with a modification of the original method(5 ll of plasma and half the volume of reagents), developed by Thrivikr-aman et al. (1997). The sensitivity of the assay is 25 ng/mL, and intra-and interassay variations are, respectively, 10.3% and 7.1%. Assays wereperformed to include all groups and all time-points in a balanced order.All samples were assayed in duplicate.

2.6. Longevity study

Longevity was measured in weeks of life and so mice were examineddaily to determine age at death.

2.7. Statistical analysis

The proportion of mice producing ANA and anti-dsDNA antibodiesand the surviving proportion were estimated by the Kaplan–Meyermethod, followed by the Log-Rank test when necessary. Mice that didnot produce ANA or anti-dsDNA antibodies until the 34th week of agewere censored in the first two tests, whereas those that did not die untilthe 55th week of age were censored in the surviving proportion test.

Friedman ANOVA, followed by Wilcoxon Matched Pairs Test, wasemployed to assess ANA titration throughout the course of the study.Kruskal–Wallis ANOVAs were performed, in each week of life, in orderto evaluate differences in ANA titers among the groups.

A two-way repeated measure ANOVA was carried out for the analysisof proteinuria and CORT levels throughout life and in three time-pointsrelated to ANA production onset. Post hoc comparisons were done byBonferroni test.

Pearson and Spearman correlation tests were performed in order todetermine associations between the analyzed parameters.

In all cases, differences were considered significant at p < 0.05.

3. Results

Due to the placement of uneven number of animals percage after weaning, we carried out an initial statistical anal-ysis of the data considering the cage in each group as a fac-tor. No significant cage differences were observed in any ofthe groups in the parameters: ANA/anti-dsDNA onset,

ANA titers, corticosterone levels, proteinuria and survival(p > 0.05 for all the tests). Therefore, the number of ani-mals/cage did not influence the results.

3.1. ANA and anti-dsDNA antibodies

Analyses revealed similar proportion curves betweenNM, BMS and LMS groups, indicating that positiveANA proportion increased equally among the groups(Fig. 1). Average and standard deviation (SD) of ANAonset was 21.9 ± 5.4, 20.3 ± 5.6 and 23.8 ± 4.6 week inNM, BMS and LMS groups, respectively.

NM, BMS and LMS groups presented similar ANAtiters in each week of life. However, differences betweenANA titers throughout life were detected (v2 = 365.9,p < 0.0001). In the 16th week of life, animals displayed higherANA titers than in the 9th, 10th and 12th weeks. A progres-sive increase was observed from 18th to 30th week in relationto the preceding week (p < 0.05), while titers in the 30th, 32ndand 34th did not differ from each other (Fig. 2).

Anti-dsDNA antibody production was similar among allgroups according to the statistical analyses (v2 = 5.30,p = 0.07), although LMS showed a tendency to present a lateonset of these auto-antibodies compared to NM (p = 0.07)and BMS (p = 0.06) groups (Fig. 3). Mean ± SD of anti-dsDNA antibody onset was 26.6 ± 3.2, 24.8 ± 5.2 and29 ± 3.3 week in NM, BMS and LMS groups, respectively.

3.2. Proteinuria

No group and time effect or an interaction between thesefactors were observed, although time effect almost reachedsignificant values (F(3,9) = 3.19; p = 0.08), probably due toa time-increase in proteinuria, especially in the 37th week,when protein levels in the urine were almost 100% higherthan in the 33th week.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

9 10 12 14 16 18 20 22 24 26 28 30 32 34

Weeks ofLife

AN

A T

iters

(lo

g)

NM

BMS

LMS

*

§

Fig. 2. Log of ANA titration throughout the mice’ life span. NM, non-manipulated (n = 19); BMS, brief maternal separation (n = 15); LMS,long maternal separation (n = 17). §Different from 9th, 10th and 12thweek of life. *Different from preceding week. Data are expressed asmean ± SEM.

14 16 18 20 22 24 26 28 30 32 34 36

Weeks of Life

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Pro

port

ion

of A

nim

als

Neg

ativ

e fo

r an

ti-ds

-DN

A

NMBMSLMS

14 16 18 20 22 24 26 28 30 32 34 36

NMBMSLMS

NMBMSLMS

Fig. 3. Onset of anti-dsDNA antibody production. Each marker expressesthe initial anti-dsDNA antibody manifestation of a particular animal.Mice that did not exhibit anti-dsDNA antibodies till the 34th week of lifewere censured. NM, non-manipulated (n = 16); BMS, brief maternalseparation (n = 13); LMS, long maternal separation (n = 12).

0

50

100

150

200

250

300

350

400

9 10 12 14 16 18 20 22 24 26 28 30 32 34

Weeks of Life

CO

RT

Ser

um C

once

ntra

tions

(ng

/mL)

NM

BMSLMS

*

Fig. 4. Corticosterone serum levels throughout life. NM, non-manipu-lated (n = 19); BMS, brief maternal separation (n = 15); LMS, longmaternal separation (n = 17). *Different from NM group. Data areexpressed as mean ± SEM.

20 25 30 35 40 45 50 55 60 65 70

Weeks of Life

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Sur

viva

l Cum

ulat

ive

Pro

port

ion

NMBMSLMS

.1

NMBMSLMS

NMBMSLMS

NMBMSLMS

Fig. 5. Survival curves of NM, BMS and LMS mice. Each markerexpresses death of a particular animal. Censure was set in the 55th week oflife. NM, non-manipulated (n = 19); BMS, brief maternal separation(n = 15); LMS, long maternal separation (n = 17). +Censored data.

370 B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374

3.3. Corticosterone

ANOVA revealed a main effect of Group(F(2,34) = 9.32; p < 0.001). Post hoc test indicated thatBMS and LMS had lower CORT levels than NM group(p < 0.05 and p < 0.001, respectively). There was no effectof weeks of age or an interaction between factors (Fig. 4).

3.4. Life span

There was no difference in the groups’ survival curves.Mean ± SD of life span was 39.6 ± 8.8, 36.0 ± 6.7 and38.1 ± 8.8 week for the NM, BMS and LMS groups,respectively (Fig. 5).

3.5. CORT and ANA onset

Because chronic inflammatory disease generally modi-fies HPA axis activity, we evaluated serum CORT levelsin three-specific time-points related to ANA productiononset: samples taken immediately before (Pre-ANA), onthe week when ANA was first detected (ANA onset), andimmediately after detection (Pos-ANA). Given that ani-mals exhibited ANA onset in different moments, thesetime-points are specific for each animal.

ANOVA detected no group or condition (Pre-ANA,ANA onset and Pos-ANA) effects. There was no interac-tion between factors.

Table 1Correlations between the average CORT levels of all-time points andother disease-related parameters

CORT concentration average

NM BMS LMS

ANA onset r = �0.70a r = 0.16 r = �0.14Anti-dsDNA onset r = �0.22 r = 0.19 r = �0.53Longevity r = 0.65a r = 0.12 r = �0.10

a p < 0.01.

B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374 371

3.6. Correlations

Regardless of the group, there were strong negative cor-relations between life span and proteinuria in the 33rd(n = 36, r = �0.74; p < 0.0001) and 37th week (n = 17,r = �0.73, p < 0.001). Hence, mice showing high levels ofproteinuria in late life had a reduced life span. There wasno correlation between onset of anti-dsDNA antibodiesand proteinuria in any of the weeks.

Table 1 shows the correlations between the average ofall time-points CORT concentration and each of the otherparameters. In NM group, high average CORT levels wereassociated with early ANA onset and longer life span. BMSand LMS groups did not present these correlations. Nocorrelations between CORT and ANA titers were observedin any of the weeks (data not shown).

There was no correlation between onset of ANA or anti-dsDNA antibodies and life span in any of the groups.

4. Discussion

Neonatal manipulations did not affect onset of ANAproduction, ANA titers, proteinuria or longevity in NZB/WF1 mice. LMS group showed only a tendency to presenta late onset of anti-dsDNA antibody production in relationto the other groups. However, BMS and LMS female miceexhibited decreased levels of circulating CORT.

There is no consensus in the literature in regard to theeffect of neonatal manipulations on basal CORT levels.Some studies find no difference in resting ACTH andCORT concentrations between NM, BMS and LMS rats(Kim et al., 2005; Ladd et al., 2004; Meaney et al., 1992;Tiba et al., 2004), whereas others report decreased CORTlevels in BMS male and female rats (Panagiotaropouloset al., 2004; Papaioannou et al., 2002), and in LMS femalerats (Slotten et al., 2006). BMS and LMS C57BL/6 micepresent no alterations in basal CORT levels (Avitsuret al., 2006; Parfitt et al., 2004), but recent findings fromour group show that LMS Swiss female mice exhibit higherCORT basal levels (Kawakami et al., 2007).

The effects of stressors on central neurochemical activityvary appreciably across strains and across individualswithin a strain, though divergence in the literature is a con-stant. In the present study we found that BMS and LMSdecreased CORT secretion in an analogous way in femaleNZB/WF1 mice, but since we have not analyzed the dam’s

behavior, it is impossible to define whether this effect was aconsequence of maternal care. Despite this similaritybetween the groups, LMS females exhibited late onset ofanti-dsDNA antibody production, which did not occurwith BMS mice. Since the onset of anti-dsDNA antibodieswas not related to proteinuria, it is hard to establish theconsequence of this neonatal manipulation on diseaseprogression.

Synthetic GCs are often used to treat chronic autoim-mune inflammatory diseases such as LES, rheumatoidarthritis and asthma, since they can effectively reduce manyinflammation parameters (Sloka and Stefanelli, 2005).However, under normal physiological conditions, the bodyattempts to down-regulate inflammation by increasing nat-ural GC production (Sternberg, 2001). Nevertheless, inAD-prone strains there appears to exist a controversy inregard to the pattern of CORT secretion. For instance,Lechner et al. (1996) found no difference in CORT levelsbetween the 2nd and the 5th months of life in two lupusstrains, the MRL/lpr and the NZB/WF1, whereas Chidaet al. (2005) observed an increase in CORT levels through-out the life span of MRL/lpr mice, except when these ani-mals were exposed to chronic social isolation stress. Recentdata from our laboratory shows that NZB/WF1 femalesexhibit a rise in CORT basal levels in the 13th week of life,in addition to a further increase in response to sleep depri-vation, without any return to basal, at least in the intervalsused to assess CORT levels (Palma et al., 2007). In thepresent study, we did not find significant differences inCORT levels throughout the life span of NM mice,although a sharp increase in CORT levels were alsoobserved at 12 weeks of age. High CORT levels seem tobe a characteristic of NZBWF1 strain, since these animalsappear to display higher CORT serum levels compared toSwiss albino female mice (Palma et al., 2007). A hypothesisto explain high CORT production throughout the life spanin lupus-prone mice is that AD induces cytokine produc-tion, including IL-6, which is an important glucocorti-coid-increasing factor (GIF) (Alarcon-Riquelme et al.,1992; Elenkov and Chrousos, 2002). Interestingly, therewas no difference in CORT secretion after onset of ANAproduction, and this finding was even more evident inBMS and LMS mice, suggesting that autoantibodies didnot elicit an immediate HPA axis response. Besides, itwas reported previously that NZB/WF1 mice show lowerCORT increasing ratios after rhu IL-1a injection at twoand five months of age (Lechner et al., 1996), alluding toa diminished HPA responsiveness to immune alterations.In humans the same controversy is reported, with somestudies revealing that SLE patients have similar and otherslow cortisol levels compared to controls (Harle et al., 2006;Martins et al., 2002; Pool et al., 2004). In other AD, such asrheumatoid arthritis, the HPA axis also exhibits an inade-quate inhibitory response to ongoing inflammation, sincepatients display normal cortisol in the presence ofenhanced TNF-a, IL-1 and IL-6 levels (Crofford et al.,1997; Elenkov and Chrousos, 2002; Masi et al., 1999).

372 B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374

Th2 cells are characterized by their production of anti-inflammatory cytokines, such as IL-4, IL-10 and IL-13,being more involved in B-cell help and in support ofhumoral antigen-specific response. At physiological con-centrations, GCs inhibit Th1 and enhance Th2 cytokineproduction (Elenkov and Chrousos, 2002). Inasmuch aslupus is a Th2-pattern disease, it would be expected thathigh CORT concentrations were related to a modificationof illness development, especially by an early disease onset.As predicted, NM animals which showed higher CORTconcentrations had an earlier onset of ANA production,although higher CORT levels were also associated with alonger life span in this group. This result may be due to aworst glomerulonephritis pattern in animals with lowerCORT levels. NM group presented a negative correlationbetween CORT concentrations and proteinuria in the37th week of life, but the small number of animals at thistime-point (n = 6) did not allow statistic significance toemerge (R = �0,60; p = 0,21).

Glomerulonephritis is induced by IgG2a, which is class-switched by the Th1 cytokine, interferon c (IFN-c), in thepresence of IL-6 (Hasegawa and Hayashi, 2003; Richardset al., 2001). Female NZB/WF1 mice treated with solubleIFN-c receptor or anti-IFN-c monoclonal antibody pres-ent less histological renal damage and increased survivalrate, whereas treatment with IFN-c leads to severe glomer-ulonephritis and premature death (Ozmen et al., 1995).Moreover, IFN-c receptor deletion in NZB/WF1 andMRL/lpr mice protects the animals against developmentof glomerulonephritis (Haas et al., 1998; Schwartinget al., 1998). Since Th1/Th2 ratio may be decreased by highCORT levels, it is possible that NM females which dis-played higher CORT levels had less pronounced kidneydamage, resulting in prolonged life span.

It is important to mention that in the present study, wedid not challenge the groups with additional stress in adult-hood, and although both BMS and LMS promoted disso-ciation between CORT levels and disease parameters suchas ANA onset and longevity, most studies assess the effectsof early life events in AD-prone strains in response to astressful situation. Such dissociations were shown in recentstudies, in which BMS blunts IL-1b response to LPS inboth brain and periphery, but CORT levels after this chal-lenge do not differ between BMS and control rats (Bilboet al., 2007). In addition, studies indicate that factors otherthan CORT secretion may be involved with the activity ofimmune system in neonatal-manipulated rodents. Forinstance, maternal deprivation during 24 h in 9-day-oldpups results in the worst clinical signs of EAE, while LPSinjection induces similar CORT response in maternallydeprived and control rats (Teunis et al., 2002). Stephanet al. (2002) described an enhancement in clinical scoresand a decrease in CORT levels after the maximal clinicalsigns of EAE in LMS female rats. Imipramine treatmentreversed LMS effects on EAE without increasing CORTconcentrations (Stephan et al., 2002), suggesting thatCORT does not play the decisive role in susceptibility to

AD in these animals. Although BMS and LMS resultedin reduced CORT secretion throughout life, the course ofthe disease was similar between all the groups, suggestingthat other aspects involved in the effects of GC are alteredby these manipulations. Among these, GC receptors, bothmineralo (MR)- and glucocorticoid-receptors (GRs),rather than stress-induced CORT response could be adeterminant mechanism of this dissociation in neonatallymanipulated groups.

The similar CORT secretion between BMS and LMSfemale mice was an unexpected finding and makes difficultto hypothesize mechanisms by which interference inmother-infant relationship may alter the development ofauto-immune diseases in prone animals. Although thereis no single work in the literature assessing the immediateeffects of early life events in NZB/WF1 pups or in theirmothers, data on a different mouse strain indicate that asingle separation on PND 12 or 14, but not earlier, pro-duces a stress response in the pups and in the mother,but this effect is eliminated by repeated 15 min separations(D’Amato et al., 1992). Curiously, no change in CORTsecretion is observed when mothers are kept away fromtheir pups for 15 min for 14 days, whereas the pups showhigh CORT levels at the end of this period. On the con-trary, when mothers are removed from the nest andexposed to a psychological stressor there is an increase ofCORT secretion, but no change in hormone levels in theirpups. When adults, separated pups from stressed andunstressed mothers exhibit low anxiety behavior comparedto non-manipulated mice, indicating that increased CORTsecretion in one element of the dyad does not translate inelevated hormone levels in the other element and bothtypes of manipulation impact positively in the adult off-spring behavior (Moles et al., 2004). Naturally the resultsof the present study demand that the immediate effects ofboth LMS and BMS on CORT secretion of NZB/NZWF1litters and their mothers be investigated.

Another aspect that must be discussed is the role of cor-ticosterone coming from the mother’s milk as a mediator ofstress-related processes in the offspring. To the best of ourknowledge there are no reports in mice, but in rats, moder-ate corticosterone concentrations offered to the mothers inthe drinking water results in immediate increase of cortico-sterone secretion at PND 11 and 16, and increased hippo-campal MR and GR at PND 30 (Casolini et al., 1997). Asadults, the offspring of corticosterone-drinking mothersexhibit smaller stress response and better cognitive perfor-mance in the water Morris maze (Catalani et al., 1993).However, we are aware that these data refer to a differentspecies and, especially, to animals that do not exhibit spon-taneous immune alterations. Therefore, to determine towhich extent the corticosterone transferred from motherto pup impact on the neuro-immune balance of NZB/NZWF1 offspring, a specific study must be carried out.

It is important to emphasize that the development ofNZB/WF1 pups involves the presence of the father in thehome-cage, resulting in mothers becoming pregnant while

B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374 373

lactating. Whether or not there was an interaction betweenpregnancy and the neonatal manipulations is impossible todetermine at present; nonetheless, the onset and severity ofthe disease were very similar among the groups and resem-ble the reports from other laboratories (Lechner et al.,1996; for review see Theofilopoulos and Dixon, 1985)and from our own (Palma et al., 2006), which suggest thatthis particular family situation does not interact with theneonatal manipulations to determine the development ofthe disease in the adult offspring.

In conclusion, neonatal manipulations did not interferewith the course of disease in NZB/WF1 lupus-prone femalemice, although both BMS and LMS induced a decrease inbasal CORT concentrations throughout the animals’ lifespan and promoted a dissociation between CORT levelsand disease-related parameters such as the onset of ANAproduction and longevity, whereas in NM mice highCORT levels were correlated to an early ANA onset anda longer survival.

Acknowledgments

We thank Tome Pimentel dos Anjos for the care of theanimals, Flavia Cristaldi for IIF slide analyses, and Mariade Fatima Cavanal for her assistance in measurement ofproteinuria. Research was supported by grants from Con-selho Nacional de Desenvolvimento Cientıfico e Tecnolog-ico (CNPq) and Associacao Fundo de Incentivo aPsicofarmacologia (AFIP). B.C. was the recipient of a fel-lowship from Fundacao de Amparo a Pesquisa do Estadode Sao Paulo (FAPESP) and D.S. is the recipient of a re-search scholarship from CNPq.

References

Alarcon-Riquelme, M.E., Moller, G., Fernandez, C., 1992. Age-depen-dent responsiveness to interleukin-6 in B lymphocytes from a systemiclupus erythematosus-prone (NZB · NZW)F1 hybrid. Clin. Immunol.Immunopathol. 62, 264–269.

Avitsur, R., Hunzeker, J., Sheridan, J.F., 2006. Role of early stress in theindividual differences in host response to viral infection. Brain Behav.Immun. 20, 339–348.

Bilbo, S.D., Newsum, N.J., Sprunger, D.B., Watkins, L.R., Rudy, J.W.,Maier, S.F., 2007. Differential effects of neonatal handling on early lifeinfection-induced alterations in cognition in adulthood. Brain Behav.Immun. 21, 332–342.

Casolini, P., Cigliana, G., Alema, G.S., Ruggieri, V., Angelucci, L.,Catalani, A., 1997. Effect of increased maternal corticosterone duringlactation on hippocampal corticosteroid receptors, stress response andlearning in offspring in the early stages of life. Neuroscience 79, 1005–1012.

Catalani, A., Marinelli, M., Scaccianoce, S., Nicolai, R., Muscolo, L.A.A.,Porcu, A., Koranyi, L., Piazza, P.V., Angelucci, L., 1993. Progeny ofmothers drinking corticosterone during lactation has lower stress-induced corticosterone secretion and better cognitive performance.Brain Res. 624, 209–215.

Chida, Y., Sudo, N., Kubo, C., 2005. Social isolation stress exacerbatesautoimmune disease in MRL/lpr mice. J. Neuroimmunol. 158, 138–144.

Crofford, L.J., Kalogeras, K.T., Mastorakos, G., Magiakou, M.A., Wells,J., Kanik, K.S., Gold, P.W., Chrousos, G.P., Wilder, R.L., 1997.

Circadian relationships between interleukin (IL)-6 and hypothalamic–pituitary–adrenal axis hormones: failure of IL-6 to cause sustainedhypercortisolism in patients with early untreated rheumatoid arthritis.J. Clin. Endocrinol. Metab. 82, 1279–1283.

D’Amato, F.R., Cabib, S., Puglisi-Allegra, S., Patacchioli, F.R., Cigliana,G., Maccari, S., Angelucci, L., 1992. Effects of acute and repeatedexposure to stress on the hypothalamo–pituitary–adrenocorticalactivity in mice during postnatal development. Horm. Behav. 26,474–485.

Elenkov, I.J., Chrousos, G.P., 2002. Stress hormones, proinflammatoryand antiinflammatory cytokines, and autoimmunity. Ann. N. Y. Acad.Sci. 966, 290–303.

Haas, C., Ryffel, B., Le Hir, M., 1998. IFN-gamma receptor deletionprevents autoantibody production and glomerulonephritis in lupus-prone (NZB · NZW)F1 mice. J. Immunol. 160, 3713–3718.

Harle, P., Straub, R.H., Wiest, R., Mayer, A., Scholmerich, J., Atzeni, F.,Carrabba, M., Cutolo, M., Sarzi-Puttini, P., 2006. Increase ofsympathetic outflow measured by neuropeptide Y and decrease ofthe hypothalamic–pituitary–adrenal axis tone in patients with systemiclupus erythematosus and rheumatoid arthritis: another example ofuncoupling of response systems. Ann. Rheum. Dis. 65, 51–56.

Hasegawa, K., Hayashi, T., 2003. Synthetic CpG oligodeoxynucleotidesaccelerate the development of lupus nephritis during preactive phase inNZB x NZWF1 mice. Lupus 12, 838–845.

Kawakami, S.E., Quadros, I.M.H., Takahashi, S., Suchecki, D., 2007.Long maternal separation accelerates behavioural sensitization toethanol in female, but not in male mice. Behav. Brain Res. doi:10.1016/j.bbr.2007.06.023.

Kim, H.J., Lee, J.H., Choi, S.H., Lee, Y.S., Jahng, J.W., 2005. Fasting-induced increases of arcuate NPY mRNA and plasma corticosteroneare blunted in the rat experienced neonatal maternal separation.Neuropeptides 39, 587–594.

Laban, O., Dimitrijevic, M., von Hoersten, S., Markovic, B.M., Jankovic,B.D., 1995a. Experimental allergic encephalomyelitis in adult DA ratssubjected to neonatal handling or gentling. Brain Res. 676, 133–140.

Laban, O., Markovic, B.M., Dimitrijevic, M., Jankovic, B.D., 1995b.Maternal deprivation and early weaning modulate experimentalallergic encephalomyelitis in the rat. Brain Behav. Immun. 9, 9–19.

Ladd, C.O., Huot, R.L., Thrivikraman, K.V., Nemeroff, C.B., Plotsky,P.M., 2004. Long-term adaptations in glucocorticoid receptor andmineralocorticoid receptor mRNA and negative feedback on thehypothalamo–pituitary–adrenal axis following neonatal maternal sep-aration. Biol. Psychiatry 55, 367–375.

Lechner, O., Hu, Y., Jafarian-Tehrani, M., Dietrich, H., Schwarz, S.,Herold, M., Haour, F., Wick, G., 1996. Disturbed immunoendocrinecommunication via the hypothalamo–pituitary–adrenal axis in murinelupus. Brain Behav. Immun. 10, 337–350.

Levine, S., Huchton, D.M., Wiener, S.G., Rosenfeld, P., 1991. Timecourse of the effect of maternal deprivation on the hypothalamic–pituitary–adrenal axis in the infant rat. Dev. Psychobiol. 24, 547–558.

Manni, L., Micera, A., Pistillo, L., Aloe, L., 1998. Neonatal handling inEAE-susceptible rats alters NGF levels and mast cell distribution inthe brain. Int. J. Dev. Neurosci. 16, 1–8.

Manson, J.J., Rahman, A., 2006. Systemic lupus erythematosus. OrphanetJ. Rare Dis. 1, 6.

Martins, J.M., Alves, J., Trinca, A., Grima, B., do Vale, S., Vasconcelos,T., Riso, N., Riscado, V., da Costa, J.C., 2002. Personality, brainasymmetry, and neuroendocrine reactivity in two immune-mediateddisorders: a preliminary report. Brain Behav. Immun. 16, 383–397.

Masi, A.T., Chrousos, G.P., Bornstein, S.R., 1999. Enigmas of adrenalandrogen and glucocorticoid dissociation in premenopausal onsetrheumatoid arthritis. J. Rheumatol. 26, 247–250.

Meaney, M.J., Aitken, D.H., Bodnoff, S.R., Iny, L.J., Sapolsky, R.M.,1985. The effects of postnatal handling on the development of theglucocorticoid receptor systems and stress recovery in the rat. Prog.Neuropsychopharmacol. Biol. Psychiatry 9, 731–734.

Meaney, M.J., Aitken, D.H., Sharma, S., Viau, V., 1992. Basal ACTH,corticosterone and corticosterone-binding globulin levels over the

374 B. Catallani et al. / Brain, Behavior, and Immunity 22 (2008) 367–374

diurnal cycle, and age-related changes in hippocampal type I and typeII corticosteroid receptor binding capacity in young and aged, handledand nonhandled rats. Neuroendocrinology 55, 204–213.

Meaney, M.J., Diorio, J., Francis, D., Widdowson, J., LaPlante, P.,Caldji, C., Sharma, S., Seckl, J.R., Plotsky, P.M., 1996. Earlyenvironmental regulation of forebrain glucocorticoid receptor geneexpression: implications for adrenocortical responses to stress. Dev.Neurosci. 18, 49–72.

Moles, A., Rizzi, R., D’Amato, F.R., 2004. Postnatal stress in mice: does‘‘stressing’’ the mother have the same effect as ‘‘stressing’’ the pups?Dev. Psychobiol. 44, 230–237.

Ostensen, M., 1999. Sex hormones and pregnancy in rheumatoid arthritisand systemic lupus erythematosus. Ann. NY. Acad. Sci. 876, 131–143,Discussion 144.

Ozmen, L., Roman, D., Fountoulakis, M., Schmid, G., Ryffel, B.,Garotta, G., 1995. Experimental therapy of systemic lupus erythemat-osus: the treatment of NZB/WF1 mice with mouse soluble interferon-gamma receptor inhibits the onset of glomerulonephritis. Eur. J.Immunol. 25, 6–12.

Palma, B.D., Gabriel Jr., A., Colugnati, F.A., Tufik, S., 2006. Effects ofsleep deprivation on the development of autoimmune disease in anexperimental model of systemic lupus erythematosus. Am. J. Physiol.Regul. Integr. Comp. Physiol. 291, R1527–R1532.

Palma, B.D., Suchecki, D., Catallani, B., Tufik, S., 2007. Effect of sleepdeprivation on the corticosterone secretion in an experimental modelof autoimmune disease. Neuroimmunomodulation 14, 72–77.

Panagiotaropoulos, T., Papaioannou, A., Pondiki, S., Prokopiou, A.,Stylianopoulou, F., Gerozissis, K., 2004. Effect of neonatal handlingand sex on basal and chronic stress-induced corticosterone and leptinsecretion. Neuroendocrinology 79, 109–118.

Papaioannou, A., Gerozissis, K., Prokopiou, A., Bolaris, S., Stylianop-oulou, F., 2002. Sex differences in the effects of neonatal handling onthe animal’s response to stress and the vulnerability for depressivebehaviour. Behav. Brain Res. 129, 131–139.

Parfitt, D.B., Levin, J.K., Saltstein, K.P., Klayman, A.S., Greer, L.M.,Helmreich, D.L., 2004. Differential early rearing environments canaccentuate or attenuate the responses to stress in male C57BL/6 mice.Brain Res. 1016, 111–118.

Peralta-Ramirez, M.I., Jimenez-Alonso, J., Godoy-Garcia, J.F., Perez-Garcia, M., 2004. The effects of daily stress and stressful life events onthe clinical symptomatology of patients with lupus erythematosus.Psychosom. Med. 66, 788–794.

Pool, A.J., Whipp, B.J., Skasick, A.J., Alavi, A., Bland, J.M., Axford,J.S., 2004. Serum cortisol reduction and abnormal prolactin andCD4+/CD8+ T-cell response as a result of controlled exercise inpatients with rheumatoid arthritis and systemic lupus erythematosusdespite unaltered muscle energetics. Rheumatology (Oxford) 43, 43–48.

Richards, H.B., Satoh, M., Jennette, J.C., Croker, B.P., Yoshida, H.,Reeves, W.H., 2001. Interferon-gamma is required for lupus nephritisin mice treated with the hydrocarbon oil pristane. Kidney Int. 60,2173–2180.

Rosenfeld, P., Ekstrand, J., Olson, E., Suchecki, D., Levine, S., 1993.Maternal regulation of adrenocortical activity in the infant rat: effectsof feeding. Dev. Psychobiol. 26, 261–277.

Rubin, R.L., 2005. Drug-induced lupus. Toxicology 209, 135–147.Schur, P.H., Sandson, J., 1968. Immunologic factors and clinical activity

in systemic lupus erythematosus. N. Engl. J. Med. 278, 533–538.Schwarting, A., Wada, T., Kinoshita, K., Tesch, G., Kelley, V.R., 1998.

IFN-gamma receptor signaling is essential for the initiation, acceler-ation, and destruction of autoimmune kidney disease in MRL-Fas(lpr)mice. J. Immunol. 161, 494–503.

Sloka, J.S., Stefanelli, M., 2005. The mechanism of action of methylpred-nisolone in the treatment of multiple sclerosis. Mult. Scler. 11, 425–432.

Slotten, H.A., Kalinichev, M., Hagan, J.J., Marsden, C.A., Fone, K.C.,2006. Long-lasting changes in behavioural and neuroendocrine indicesin the rat following neonatal maternal separation: Gender-dependenteffects. Brain Res. 1097, 123–132.

Stephan, M., Straub, R.H., Breivik, T., Pabst, R., von Horsten, S., 2002.Postnatal maternal deprivation aggravates experimental autoimmuneencephalomyelitis in adult Lewis rats: reversal by chronic imipraminetreatment. Int. J. Dev. Neurosci. 20, 125–132.

Sternberg, E.M., 2001. Neuroendocrine regulation of autoimmune/inflammatory disease. J. Endocrinol. 169, 429–435.

Sternberg, E.M., Hill, J.M., Chrousos, G.P., Kamilaris, T., Listwak, S.J.,Gold, P.W., Wilder, R.L., 1989a. Inflammatory mediator-inducedhypothalamic-pituitary-adrenal axis activation is defective in strepto-coccal cell wall arthritis-susceptible Lewis rats. Proc. Natl. Acad. Sci.USA 86, 2374–2378.

Sternberg, E.M., Young 3rd, W.S., Bernardini, R., Calogero, A.E.,Chrousos, G.P., Gold, P.W., Wilder, R.L., 1989b. A central nervoussystem defect in biosynthesis of corticotropin-releasing hormone isassociated with susceptibility to streptococcal cell wall-induced arthri-tis in Lewis rats. Proc. Natl. Acad. Sci. USA 86, 4771–4775.

Suchecki, D., Rosenfeld, P., Levine, S., 1993. Maternal regulation of thehypothalamic–pituitary–adrenal axis in the infant rat: the roles offeeding and stroking. Brain Res. Dev. Brain Res. 75, 185–192.

Teunis, M.A., Heijnen, C.J., Sluyter, F., Bakker, J.M., Van Dam, A.M.,Hof, M., Cools, A.R., Kavelaars, A., 2002. Maternal deprivation ofrat pups increases clinical symptoms of experimental autoimmuneencephalomyelitis at adult age. J. Neuroimmunol. 133, 30–38.

Theofilopoulos, A.N., Dixon, F.J., 1981. Etiopathogenesis of murine SLE.Immunol. Rev. 55, 179–216.

Theofilopoulos, A.N., Dixon, F.J., 1985. Murine models of systemic lupuserythematosus. Adv. Immunol. 37, 269–390.

Thrivikraman, K.V., Su, Y., Plotsky, P.M., 1997. Patterns of Fos-immunoreactivity in the CNS induced by repeated hemorrhage inconscious rats: correlations with pituitary–adrenal axis activity. Stress2, 145–158.

Tiba, P.A., Tufik, S., Suchecki, D., 2004. Effects of maternal separation onbaseline sleep and cold stress-induced sleep rebound in adult Wistarrats. Sleep 27, 1146–1153.

Wilder, R.L., 1995. Neuroendocrine-immune system interactions andautoimmunity. Annu. Rev. Immunol. 13, 307–338.