Gestational vitamin A deimmune response by decin rat offspring

Xia Liu M.D. a, Yingying Li M.D. a

Xin Li M.D. b, Yang Bi Ph.D. a, LanJie Chen Ph.D. a,*aChildren Nutrition Research Center, Ministry of EducatChongqing, CSTC2009 CA5002, Chongqing InternationalChildrens Hospital of Chongqing Medical University, Chb Institute of Pediatrics, Childrens Hospital of Chongqin

gy, Depa

nodes, Peyer patches,

signicantly decreasedel of immunoglobulin Ae interaction effects ofnicantly increased thepercentages of CD11 C

lysaccharide challengeintestinal tissue of the

une cells in offspring,es in the intestine. Thischildren and women of

Inc. All rights reserved.

Vitamin A (VA) and its metabolites play a critical role indevelopment during the gestational and the early postnatal

health and nutrition problem worldwide, especially in devel-oping countries. VAD can limit growth, cause blindness, weakeninnate and acquired host defenses, exacerbate infection,and increase the risk of death. Our previous studies havedemonstrated that supplementation with VA or with VA incombination with multiple micronutrients can improve thephysical growth levels [2], anemia status [3], and infectionmorbidity of preschool children [4]. Many studies have suggested

X.L. and Y.L. conducted the experiment and drafted the manuscript. Y.W. andX.W. established the animal model. Q.W. and X.L. detected the lymphocytes usingow cytometry. Y.B., L.L., and T.L. helped to design the study. J.C. provided fundingand overall direction. All authors read and approved the nal manuscript.

Contents lists availab

Nutrit

ww

Nutrition 30 (2014) 350357* Corresponding author. Tel.: 86-23-63630913; fax: 86-23-63622754.Introduction period [1]. VA deciency (VAD) is known to be a major publicLipopolysaccharides cytometry with samples collected from the spleen, the mesenteric lymphand intestinal intraepithelial lymphocytes.Results: Early life VAD, independent of the lipopolysaccharide challenge,serum retinol level and CD8 intestinal intraepithelial lymphocytes. The levsecretion and percentages of splenic CD4CD8 T cells were affected by ththe lipopolysaccharide challenge and VAD treatment. Gestational VAD sigpercentages of B cells in the mesenteric lymph nodes and decreased thedendritic cells and CD4CD25 T cells from the Peyer patches. The lipopoonly signicantly increased percentages of splenic CD4CD25 T cells. Thepups with VAD displayed mild inammation.Conclusions: Gestational or early life VAD decreases the numbers of immwhich may partly suppress the activities of the mucosal immune responssuggests that more attention should be given to the VA nutritional state ofreproductive age.

2014 ElsevierEarly life periodLymphocytes

chromatography, and the level of immunoglobulin A in the stool was analyzed by enzyme-linkedimmunosorbent assay. The lymphocyte immunophenotypes were evaluated with the use of owc Laboratory of Intestinal Physiology and Patholo

a r t i c l e i n f o

Article history:Received 4 February 2013Accepted 17 September 2013

Keywords:Vitamin A deciencyGestationE-mail address: jchen010@gmail.com (J. Chen).

0899-9007/$ - see front matter 2014 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.nut.2013.09.008ciency reduces the intestinalreasing the number of immune cells

, Yuting Wang M.D. a, Qinghong Wang Ph.D. b,Liu M.D. c, Xiaoping Wei M.D. a, Tingyu Li M.D. a,

ion Key Laboratory of Child Development and Disorders, Key Laboratory of Pediatrics inScience and Technology Cooperation Center for Child Development and Disorders,ongqing, P.R. Chinag Medical University, Chongqing, P.R. Chinartment of Surgery, University of Maryland School of Medicine, Baltimore, Maryland, USA

a b s t r a c t

Objectives: Vitamin A (VA) is a critical micronutrient for life, especially during growth and devel-opment. There is a close relationship between VA deciency (VAD) and the morbidity of diarrheain the clinical setting. However, the regulatory mechanisms of VA are not clearly understood.Methods: Specic-pathogenfree Wistar rats received a diet with or without VA before gestation.The offspring were submitted to an abdominal injection of Escherichia coli lipopolysaccharide. Afterthe challenge, which lasted for 12 h, the serum retinol was detected by high-performance liquidBasic nutritional investigationjournal homepage:ll rights reserved.le at ScienceDirect

ion

w.nutr i t ionjrnl .comthat a lower nutritional level of VA is correlated directly with

WINDOWSHighlight

normal (VAN) pups. The percentage changes in the various

Amresco). The concentration of total IgA in the supernatants was measured withthe use of an enzyme-linked immunosorbent assay kit (Bethyl Lab, Inc, Mont-gomery, Texas, US) according to the manufacturers instructions.

Isolation of lymphocytes from the spleen and the mesenteric lymph nodes

The spleens and the MLNs of the rats were cut into pieces with the use ofsurgical scissors on a 200-mesh metallic grid and then washed with PBS. Theindividual lymphocytes were collected via centrifugation. After washing, 3 107cells were resuspended in PBS for analysis.

Peyer patch preparation

Lymphocytes from PPs were isolated as described previously [20]. Briey, thelymphoid follicles of PPs were carefully excised from the intestine. Single lym-phocytes were released with 2 mL of FBS after disruptionwith the use of surgicalscissors on a 300-mesh metallic grid. The lymphocyte suspension was slowlyadded onto the surface of a lymphocyte separation medium for rats (Tianjing-haoyang Biological Manufacture Co., China) and centrifuged at 805 g for 15 min.

ion 3lymphocyte subsets were analyzed in the spleens of the VADoffspring after lipopolysaccharide (LPS) challenge. Second, wefocused on the study of gut-associated lymphocyte tissues(GALT)dincluding the mesenteric lymph nodes (MLNs), Peyerpatches (PPs), and IELsdto investigate the immunomodulatorymucosal immune response of the intestine. Third, we used thepups of the VAD gestational rats that were fed with the maternalmilk of VAN rats to elucidate the importance of the VA nutri-tional level during the gestational period for postnatal immunedefense. This study provides a new insight into the role thatVAD during pregnancy plays in the immunosuppression of themucosal immune response in the intestine during postnataldevelopment.

Materials and methods

Animals, diets, and the lipopolysaccharide challenge

The use and reference numbers of the animals were approved by the AnimalExperimentation Ethical Committee of Chongqing Medical University (Chongq-ing, China). Forty 6-wk-old female specic-pathogenfree Wistar rats obtainedfrom the Experimental Animals Center of the Third Military Medical University(Chongqing, China) were randomly selected and divided equally into two groupsof maternal VAN and maternal VAD. The maternal VAD rats were fed a VAD dietthat contained 400 IU/kg of VA for 4 wk to construct the VAD animal modelbefore gestation, whereas the VAN rats received VAN food that contained 6500IU/kg of VA as the control [16,17]. When the serum retinol levels of blood samplestaken from the tails of the VAD rats decreased to 1.05 mmol/L, the 20 maternalVAD rats were mated with normal males. The pregnant maternal rats were fedeither the VAD diet or the VAN diet during both the gestational and lactationperiods to maintain the retinol levels in the serum. All rats were housed at aconstant temperature of 22C to 24C with 60% relative humidity, with articialdaylight from 07:00 to 19:00 every day and with VAD or VAN food and wateravailable ad libitum. The specic-pathogenfree animal house was certied forexperimental animals. Each dam nursed eight pups from her litter during thelactation period. After weaning, the eight pups were randomly separated intotwo cages.

After a weaning period of 21 postnatal days, the pups were fed continuouslyfor 3 wk with the VAD diet and designated as the VAD group (n 60) or the VANdiet and designated as the VAN group (n 60). After birth, the VAD pups werecross-fostered to VAN dams to make the VAD-N pups (n 60) during thelactation period and then fed the VAN diet after weaning (Fig. 1). Approximatelythe pathogenesis of diarrhea [57]. However, few studies havebeen conducted on the effect of VAD on the mucosal immuneresponse of the intestine.

Increasing evidence has revealed that VAD decreases the in-testinal barrier function, thereby increasing the intestines sus-ceptibility to infection by pathogens, which may result indiarrhea [6,8]. Supplementation with the appropriate amount ofVA acts as an anti-infection agent, which reduces infant mor-tality rate [9] and the morbidity associated with diarrhea [10].Several studies have demonstrated the roles of VA and retinoicacid (RA) receptors in T-cell differentiation and of immuno-globulin A (IgA) switching and production [1115]. However, theregulatory functions of VA in the intestinal mucosa are unknown(especially Peyer patches and intestinal intraepithelial lympho-cytes [IELs]). Few researchers have studied the effects of VADfrom the beginning of gestation on the postnatal intestinalmucosal immunity.

A rat model of VAD during the gestational period was usedin the present study to investigate whether gestational VADaffected immune activity and impaired the capacity of the im-mune response in offspring after the challenge.We rst observedthe differences in the intestinal IgA levels and the pathologicchanges of the intestinal mucosa between the VAD and VA

X. Liu et al. / Nutrithalf of the total pups in each group were randomly selected to be injectedabdominally with Escherichia coli LPS (i.e., 3 mg of LPS per kilogram of animalweight). The other half of the pups were injected with approximately 0.25 to0.4 mL (i.e., about the same volume as the LPS injection) of phosphate bufferedsaline (PBS) to serve as the sham group for the LPS challenge. After the LPSchallenge had lasted 12 h, the 6-wk-old pups were anesthetized with chloralhydrate, and the blood was collected immediately from the femoral artery. Theisolated tissues of the spleen, the MLN, and the intestine were extracted andplaced into PBS until they were used in the following steps.

Detection of serum retinol

The concentration of serum retinol was determined with the use of high-performance liquid chromatography according to our previously describedmethods [18] with slight modications. Briey, 200 mL of serum was deprotei-nized with dehydrated alcohol, and then the retinol was extracted with hexaneand evaporated with nitrogen gas. The residue was dissolved in 100 mL of themobile phase mixture (methanol:water ratio of 97:3), and the entire sample wastransferred to a bottle installed in the high-performance liquid chromatographyapparatus (DGU-20 As, Shimadzu Corporation, Japan). The retinoids were sepa-rated by chromatography on an analytical column (Hypersil phenyl 120 A 5 mm,250 4.6 mm, Phenomenex, USA) via gradient elution of the mobile phase in aliquid chromatograph equipped with a 315 nm ultraviolet photodiode arraydetector.

Hematoxylin and eosin staining and enzyme-linked immunosorbent assaymeasurements

After the samples of fresh intestine were xed in 4% paraformaldehyde,dehydrated, embedded in parafn, and sectioned in accordance with standardmethods, the serial sections, which were 4-mm thick, were stained with hema-toxylin and eosin to be examined under a light microscope for pathologicobservation.

As described for the reported method of Frossard et al. [19], 100 mg of fecesfrom the ileocecum were resuspended in 1 mL of 0.01 M PBS that contained 1%fetal bovine serum (FBS) and 0.1 mM phenylmethanesulfonyl uoride (Solarbio,

Fig. 1. Schematic diagram showing the three offspring groups: VAN, VAD, and VAD-N. The VAN pups were delivered and nursed by maternal VAN rats and then fed aVAN diet for 3 wk after weaning. The VAD pups were birthed and nursed bymaternal VAD rats and fed a VAD diet for 3 wk after weaning. The VAD-N pupswere delivered from maternal VAD rats and then nursed by maternal VAN rats andfed with a VAN diet until they were 6 wk old. VAD, vitamin A deciency; VAD-N,VAD during gestation and fed the milk of VAN rats after birth; VAN, vitamin Anormal.

0 (2014) 350357 351The white occulent at the interface between the FBS and the lymphocyte sep-aration medium was collected into a new tube with a Pasteur pipette, washedtwice, and then kept in 100 mL of PBS for ow cytometry.

tholr absmon. (D)5; y

in A

ion 3Fig. 2. Effects of VAD on the levels of serum retinol, IgA secretion, and the histoparetinol levels (mmol/L) among the VAN, VAD, and VAD-N groups after the presence oeffect, independent of LPS treatment, as shown by the serum retinol levels (mmol/L) astool among the VAN, VAD, and VAD-N groups with or without LPS treatment (n 6) 3). Interaction indicates an effect of LPS treatment versus null treatment. * P < 0.0liquid chromatography; IgA, immunoglobulin A; LPS, lipopolysaccharide; VAD, vitam

X. Liu et al. / Nutrit352Isolation of intestinal intraepithelial lymphocytes

With the use of certain modications that have been described previously[21], the entire intestine was washed with 0.01 M PBS three times to remove theintestinal contents and mucus. The intestine was then incubated in ice-cold PBSthat contained 2% FBS for 2 h; the whole intestine without the PPs was washedtwice with warmed (37C) Dulbeccos modied Eagles mediumwith 2% FBS and1% dithiothreitol and then gently squeezed with ophthalmic tweezers from theserosal side to remove the loosened cells on the intestinal mucosal surface. Thecellmedium mixture was placed at room temperature for 10 min, and the su-pernatant was collected into a new tube for centrifugation. The pellet wasresuspended in FBS, and the lymphocytes were isolated with lymphocyte sepa-ration medium for detection with the use of ow cytometry.

Flow cytometric analysis

The immunophenotypes of the various lymphocytes were used to identify thespecic cell types with the following antibodies (BD Pharmingen, Heidelberg). TheT lymphocytes were stained with PECy5-anti-CD45, APC-anti-CD3, PE-anti-CD4,and FITC-anti-CD8. The B lymphocytes were stained with PECy5-anti-CD45 andFITC-anti-CD45 R, whereas the dendritic cells (DCs) were stained with PECy5-anti-CD45 and PE-anti-CD11 c. The CD4CD25 T cells were analyzed with APC-anti-CD3, PE-anti-CD4, and FITC-anti-CD25. The IELs were stained with APC-anti-CD3,FITC-anti-CD8, and PE-anti-TCRgd. The corresponding isotype-matched mono-clonal antibodies were used as negative controls. The percentages of the variouslymphocytes were determined with a BD FACSCanto II ow cytometer (BectonDickinson).

Statistics

All datawere expressed asmean SD. Signicant differences were calculatedvia two-way analysis of variance with the Bonferroni post hoc test with the use ofthe GraphPad Prism version 5.0 software package. Statistically signicant in-teractions were analyzed with the Bonferroni post hoc test between LPS and no-LPS treatments. When there was no signicant interaction, the main effects wereanalyzed with the use of the Tukey post hoc test for the three combined VAtreatment groups, and the Student t test was used for the two combined LPStreatment groups. Only the relevant comparisons of the combined groups are

VAN, vitamin A normal.ogic changes of intestinal mucosa in 6-wk-old pups. (A) The changes in the serumence of Escherichia coli LPS challenge as assessed by HPLC (n 15). (B) The VA maing the combined VAN, VAD, and VAD-N groups. (C) The levels of IgA in the intestinalHistologic examination of intestinal mucosa staining with hematoxylin and eosin (nP < 0.01; z P < 0.001; n.s. not signicant in post hoc tests. HPLC, high performancedeciency; VAD-N, VAD during gestation and fed the milk of VAN rats after birth;

0 (2014) 350357presented in the Results section [22,23]. P values of

e Tt, ind< 0ter b

ion 3pathogen invasion. The levels of IgA in the intestinal fecal ma-terial were not signicantly different in the offspring among theVAN, VAD, and VAD-N pups treatedwithout LPS (Fig. 2C). The IgAlevel in the VAN pups after the LPS challenge sharply increased to159.3 (69.28) ng/mL, which was a 10-fold increase as comparedwith the VAN rats that were not treated with LPS (P < 0.001).However, the LPS challenge only caused a 2.88-fold increase inthe IgA levels in the VADLPS pups and a 5.41-fold increase inthe VAD-NLPS pups. These values were signicantly differentwhen compared with the VANLPS group (P < 0.001 and P