Intestinal IFN-γ production is associated with protection from clinical signs, but not with...

ORIGINAL PAPER

Intestinal IFN-γ production is associated with protectionfrom clinical signs, but not with elimination of worms,in Echinostoma caproni infected-mice

Alba Cortes & Javier Sotillo & Carla Muñoz-Antoli &Bernard Fried & J-Guillermo Esteban & Rafael Toledo

Received: 2 December 2013 /Accepted: 4 March 2014 /Published online: 16 March 2014# Springer-Verlag Berlin Heidelberg 2014

Abstract In the present paper, we assess the relationshipbetween the expression of IFN-γ and the development ofclinical signs in Echinostoma caproni-infected mice. For thispurpose, we studied the course of the infection in three mousestrains: ICR (CD-1®) (a host of high compatibility withE. caproni), BALB/c (a prototypical Th2 strain), andBALB/c deficient for IFN-γ mice (IFN-γ−/−). Infection inICR mice is characterized by the elevated expression ofIFN-γ and iNOS in the intestine concomitantly with the lackof clinical signs. In contrast, the infection was more virulent inBALB/c and IFN-γ-deficient mice that developed a severeform of the disease together with the absence of IFN-γ ex-pression. The disease was more severe in IFNγ−/− mice inwhich the disease was lethal during the few first weeks of theinfection. The analysis of different parameters of the infectionin each host strain showed that most of the features weresimilar in the three mouse strains, suggesting the IFN-γ playsa central role in that protection against severe disease. Thus,IFN-γ seems to play a dichotomous role in the infectionfacilitating the parasite establishment, but it may also benefitmice since it protects the mice from morbidity and mortalityinduced by the parasite.

Keywords Intestinal helminth . Echinostoma caproni .

Mouse strain . IFN-gamma . iNOS . Clinical signs . Disease

Introduction

Echinostoma caproni (Trematoda: Echinostomatidae) is anintestinal trematode with no tissue phases in the definitivehost (Fried and Huffman 1996). Although E. caproni is ableto parasitize a wide range of laboratory rodent hosts, itscompatibility differs considerably between rodent species. Inhighly compatible hosts, such as ICR mice, the infectionbecomes chronic, while in hosts of low compatibility, suchas rats and jirds, the worms are expelled a few weeks post-infection (wpi) (Odaibo et al. 1988, 1989; Christensen et al.1990; Mahler et al. 1995; Toledo et al. 2004, 2009; Toledo andFried 2005; Muñoz-Antoli et al. 2007). Because of this char-acteristic, the E. caproni/rodent model has been extensivelyused to study several aspects of the host/parasite interactions,with emphasis on the host-specific components that determineearly rejection of the worms or the establishment of chronicinfections. Previous studies have shown that the early rejec-tion ofE. caproni is associatedwith the development of a localTh2/Th17 local phenotype (Sotillo et al. 2011; Trelis et al.2011). In contrast, the establishment of chronic infections inICR mice is dependent upon a local Th1 response with ele-vated production of IFN-γ (Trelis et al. 2011). The infectioninduces important inflammatory responses and a rapid in-crease of iNOS expression (Toledo et al. 2006a; Muñoz-Antoli et al. 2007; Trelis et al. 2011). These results giveIFN-γ a central role in the outcome of the infection andprovide an excellent model for the study of the effect of thiscytokine in the intestinal helminth infection.

The role of IFN-γ in the intestinal helminth infections iscontroversial. This cytokine has been commonly associatedwith the development of severe forms of disease and

A. Cortes : J. Sotillo :C. Muñoz-Antoli : J.<G. Esteban :R. Toledo (*)Departamento de Parasitología, Facultad de Farmacia, Universidadde Valencia, Av. Vicente Andrés Estellés s/n, 46100 Burjassot,Valencia, Spaine-mail: [email protected]

J. SotilloCentre for Biodiscovery and Molecular Development ofTherapeutics, James Cook University, Smithfield, Queensland,Australia

B. FriedDepartment of Biology, Lafayette College, Easton, PA 18042, USA

Parasitol Res (2014) 113:2037–2045DOI 10.1007/s00436-014-3851-7

pathology in intestinal helminth infections (Artis et al. 1999;Toledo et al. 2006a; Humphreys et al. 2008; Martinez et al.2009). The mechanisms involved in the generation of pathol-ogy seem to be related with the stimulation of inducible nitricoxide synthase (iNOS) and subsequent production of nitricoxide (Mowat 1989; Lawrence et al. 2000; Brunet 2001).However, there are several reports showing that IFN-γ andiNOS are not necessarily related to an increase in morbidity,and even these mediators may have beneficial effects. It isshown in other models that the downregulation of Th1 cyto-kines ameliorates the severity of the disease. For example, inTrichinella spiralis infections in mice, it has been shown thatthat the enteropathy is not regulated by IFN-γ, and, in con-trast, Th2 cytokines such as IL-4 or IL-13 are essential for thegeneration of pathology associated to the infection (Lawrenceet al. 1998; McDermott et al. 2005). Furthermore, it has beenproposed that the generation of proinflammatory cytokines,such as IFN-γ, during chronic Trichuris muris infections, mayplay a positive role by promoting intestinal epithelial cellapoptosis which enhances the integrity of the mucosal archi-tecture (Cliffe et al. 2007).

Considering the importance of IFN-γ in the outcome ofintestinal helminth infections, we analyzed herein the outcomeof the E. caproni infection in three mice strains with differentimmunological backgrounds to assess the factors that deter-mine the characteristics of the infection. The results show thatIFN-γ plays a central role in the final outcome of the infectionsince this cytokine is involved in the severity of the diseaseinduced by the parasite.

Materials and methods

Animals and experimental infections

A total of three strains of mice were used in this study: ICR(CD-1®) (Charles River Laboratories), BALB/c (Charles Riv-er Laboratories), and BALB/c deficient for IFN-γ mice(IFN-γ−/−) (The Jackson Laboratories). A total of 25 male 5-week old mice of each strain were each infected by stomachtube with 50 metacercariae of E. caproni. The strain ofE. caproni has been described previously (Fujino and Fried1993). Encysted metacercariae of E. caproni were removedfrom the kidneys and pericardial cavities of experimentallyinfectedBiomphalaria glabrata snails and used to infect mice.Infected mice were randomly allocated to group A (10 mice ofeach strain) and B (15 mice of each strain). Moreover, ten andfive naïve mice of each strain were used as controls for thegroups A and B, respectively. Group Awas used for the studyof host weight and mortality over the course of the infection.Animals of the group Bwere used to study worm recovery, thepathology induced by E. caproni in each host strain and theexpression of cytokines and mucins. At each 2 weeks post-

infection (wpi), five animals of each strain belonging to thegroup B were necropsied to collect the materials required. Allthe animals were maintained under conventional conditionswith food and water ad libitum. The study was approved bythe institutional committee on animal care at the University ofValencia (Spain), in compliance with the European Agree-ment of Strasbourg.

Total RNA extraction

At each 2 wpi, five mice of each strain belonging to group Bwere necropsied. Total RNAwas extracted from full-thicknesssections of ileum at the sites where the worms were located.Total RNA from each tissue obtained from rats and mice wasisolated using Rapid Total RNA Purification Kit (MarligenBiosciences Inc.) according to the manufacturer’s instructions.The cDNAwas synthesized using SuperScript III First-StrandSynthesis System (Invitrogen).

Real-time PCR and relative quantification analysis

For quantitative PCR, 40 ng total RNA reverse transcribed tocDNA was added to 10 μL of TaqMan® Universal PCRMaster Mix, No AmpErase® UNG (2x), 1 μL of the specifiedTaqMan® Gene Expression Assay, and water to a final reac-tion volume of 20 μL. Reactions were performed on the AbiPrism 7000 (Applied Biosystems), with the following thermalcycler conditions: initial setup of 2-min UNG activation at50 °C, followed by 10 min at 95 °C, and 40 cycles of 15-sdenaturation at 95 °C and 1 min of annealing/extension at60 °C each. Samples were amplified in a 96-well plate. Oneach plate, endogenous control, samples and negative controlswere analyzed in triplicate. All TaqMan® Gene Expressionprimers and probes both for inducible nitric oxide synthase(iNOS) cytokines and mucins were designed by AppliedBiosystems and offered as an Inventoried Assays. Each assaycontains two unlabeled primers and one 6-FAMTM dye-labeled, TaqMan® MGB probe.

Cycle threshold (Ct) value was calculated for each sample,housekeeping, and uninfected control. To normalize for dif-ferences in the efficiency of sample extraction or complemen-tary DNA synthesis, we used β-actin as housekeeping gene.To estimate the influence of infection in the expression levels,we used a comparative quantification method (2 − ΔΔCTmethod) (Livak and Schmittgen 2001). This method is basedon the fact that the difference in threshold cycles (ΔCt)between the gene of interest and the housekeeping gene isproportional to the relative expression level of the gene ofinterest. The fold change in the target gene was normalized toβ-actin and relative the expression at time zero (uninfectedanimals) to generate relative quantification of expressionlevels (Klein 2002).

2038 Parasitol Res (2014) 113:2037–2045

Histology

Histopathological responses to E. caproni infections in theleum of mouse were evaluated at each 2 wpi on the animals ofthe group B. At each time post-infection, five mice of eachstrain from group B were necropsied, and ileal sections ofabout 0.7 cm in length were obtained from each animal andfixed in 4 % buffered formalin. After embedding in paraffinwax, serial 4-μm sections were cut from each tissue block.Intestinal sections were stained with hematoxylin-eosin,Alcian Blue, and Toluidine Blue. All cell counts wereexpressed as the number of cells per villus-crypt unit(VCU). Results are expressed as the mean number of cellsper VCU ± standard error.

Statistical analysis

To determine the statistical significance of the effect of timepost-infection and the infection with E. caproni in the differentparameters studied in eachmouse strain, a two-factor ANOVAwith interactions was used. As independent variables wereused wpi and infected versus non-infected or wpi and mousestrain as indicated in each case. Moreover, when a significantinteraction was detected, the Bonferroni t test of the differencebetweenmeans was performed as post hoc analysis to confirmthe significance of the differences. Differences in mean valueswere considered statistically significant at p<0.05. Prior toanalyses, data were log transformed to achieve normality andverified by the Anderson-Darling Test.

Results

Host mortality

None of the ICR and BALB/c mice died during the experiment.In contrast, all of the IFNγ−/− mice died before the end of theexperiment at 6 wpi. At 2 and 4 wpi, two mice (20 %) died ateach time point. A total of 3 (30 %) IFN-γ-deficient mice diedat 4 wpi. The remaining three animals (30 %) died at 5 wpi.

Worm recovery

Theweekly worm recovery expressed in percentage in the ICR,BALB/c, and IFNγ−/−mouse strains is shown in Fig. 1. In ICRand BALB/c mice, worms were recovered at each week of thestudy, and the number of worms collected weekly ranged from32 to 50 (39.3±9.9) and 26 to 50 (38.1±9.0), worms/mouse,respectively. In the IFN-γ-deficient mice, worm recovery couldonly be studied at 2 and 4 wpi, since at 6 wpi all the mice weredied. The number of worms collected weekly ranged from 37to 45 (40.2±3.1) worms/mouse. Two-factor ANOVA usingwpi and mouse strain as independent variables showed no

significant differences related neither in relation to the wpinor the mouse strain.

Pathology and clinical signs

Results of body weight evaluation of the three mouse strainsover the course of the experiments are shown in Fig. 2. Theinitial average body weight of the ICR mouse was 28.6±1.8.The initial average weight of the other strains was lower, whichcorresponds with strains with BALB background. The initialaverage weights were 19.1±1.6 and 19.0±1.9 in BALB/c andIFN-γ−/− mice, respectively. All these weights correspond tonormal body weight for these strains at 5 weeks. The infectedICR mice showed normal development. No significant differ-ences between infected and non-infected animals were detectedthroughout the entire experiment. In contrast, in BALB/c strain,significant differences were found. Infected mice showed asimilar weight gain to controls during the first 2 wpi. Fromthe third wpi, a decreased trend was observed in the infectedBALB/c mice that remained until the end of the experiment.Bodyweight decreased from 20.1±1.7 g at 3 wpi to 17.9±2.9 gat 6 wpi in E. caproni-infected mice, while non-infected con-trols increased their weight up to 24.6±3.3 g in the sameinterval. The weight of infected mice was significantly lowerthan of the controls at 4, 5, and 6 wpi (p<0.01) using wpi andinfection status as independent variables. The general pattern incontrol versus infected IFN-γ-deficient mice was similar to thatobserved in BALB/c mice, though the loss of weight started1 week before in IFN-γ-deficient mice. The body weightdecreased from 21.1±2.1 g at 2 wpi to 18.7±2.1 g at 6 wpi inE. caproni-infected mice. In contrast, the non-infected controlsincreased their weight up to 24.2±2.7 g at the same interval.The weight of infected mice was significantly lower than incontrols at 3 and 4 wpi (p<0.01).

In addition to the reduced body weight, infected BALB/cmice and IFN-γ−/− mice showed another clinical signs.BALB/c mice showed evident rough coat plated, hackled hair,diarrhea, perianal irritation, and inflammation, and theylooked emaciated from 2 to 3 wpi. Moreover, they wereirritable, frightened, trembling, and evasive, and their activity

40

60

80

100

2 4 6

Wor

m r

ecov

ery

(%)

Weeks post-infection

ICR

BALB/c

IFN- -/ -γ

Fig. 1 Echinostoma caproni worm recovery at increasing age in infec-tions in different mouse strains

Parasitol Res (2014) 113:2037–2045 2039

was reduced in relation to control animals. These signs wereprogressively aggravated over the course of the experiment.The evolution of the infected IFN-γ-deficient mice was sim-ilar though the signs appeared 1 week before, and they weremore pronounced. In contrast, the infected ICR mice did notshow signs of illness during the experiment. Gross examina-tion of the intestines of infected ICR mice only showedmarked dilations in the areas where the worms were attachedas it is common in E. caproni infection (Toledo et al. 2006a, b;Muñoz-Antoli et al. 2007). In the BALB/c mice and IFN-γ−/−

mice, the intestines also showed the same dilations in relationto the presence of the adult worms, but they also were mark-edly emaciated and degraded, and the intestinal walls were

extremely thin and in some cases almost transparent. Further-more, the mucosa was almost completely destroyed.

Infected mice of the three strains showed erosion at the tipof the villi, crypt hyperplasia, and increased mitotic rates.Moreover, a significant increase in the length of the villi andthe crypts with respect to the respective controls (p<0.001)was observed in the infected ICR and BALB/c mice. Interest-ingly, crypt hyperplasia also occurred in the IFN-γ−/− mice(p<0.001), but it was not accompanied by an increase in thevillus length in relation to the control animals (Fig. 3). In thethree strains, the villus/crypt ratio was lower in the infectedanimals, indicating a predominance of the increase in thelength of the crypts (p<0.001).

We also studied the variations in the populations of gobletcells and mast cells in each mouse strain as a consequence of theinfection since these cells are considered of importance in thepathology induced by E. caproni (Toledo et al. 2006b). Thepattern was similar for each strain and characterized by anincrease of both cell populations in the first weeks of the infection(Fig. 4). Significant time-related changes were observed(p<0.01) for each strain though the dynamics were almostidentical in all the three strains after application of the two-factor ANOVA using wpi and host strain as independentvariables.

To obtain further information about the inflammatory re-sponses induced by E. caproni, we studied the expression ofiNOS in each mouse strain by real-time PCR, and the results areshown in Fig. 5. Increased expression of iNOSwas only detectedin the intestine of experimentally infected ICR mice. A markedincrease was observed from 4 wpi, and the values progressivelyincreased at 6 wpi. Two-factor ANOVA and Bonferroni t test onthe intestinal expression levels of iNOS in ICR mice and ratsover time after infection clearly showed that the values observedin infected ICR mice were significantly higher than those incontrols for each week from 4 wpi (p<0.0001). In contrast, inthe infected BALB/c and IFN-γ−/−mice, the expression of iNOSwas significantly downregulated (p<0.001) in regard to therespective controls during all the weeks that comprised the study.Moreover, application of two-factor ANOVA using wpi and hoststrain as independent variables showed that the expression ofiNOS in ICRmice was significantly higher than in the other twostrains at 2, 4, and 6 wpi (p<0.0001).

Expression of mucins in the intestine of infected mice

We also analyzed the changes in the expression of mucingenes induced by E. caproni infection in each mouse strain.To do this, the levels of gel forming mucin Muc2 and cellsurface mucins Muc1, Muc3, Muc4, Muc13, Muc16, andMuc20 expression were determined using RT-PCR real timein the mice of group B.

The pattern of mucin expression was almost identical in theinfected animals of the three strains of mice. In fact,

20

30

40

50

0 1 2 3 4 5 6

Wei

ght

(g)


20

30

40

0 1 2 3 4 5 6

Wei

ght

(g)


a

b

c

* * *

**

10

20

30

0 1 2 3 4 5 6

Wei

ght

(g)


Fig. 2 Weight curves of ICR (a), BALB/c (b), and IFN-γ-deficient (c)infected (closed circles) and control (open circles) mice infected withEchinostoma caproni over the course of the infection. Vertical barsrepresent standard error

2040 Parasitol Res (2014) 113:2037–2045

application of the two-factor ANOVA using wpi and hoststrain as independent variables showed significant time-related differences in the infectedmice, but changes in relationto the host strain were limited and isolated (Fig. 6). Mostrelevant time-related changes were an increase in the expres-sion of Muc2 and especially of Muc4 (p<0.01) in the threestrains during the experiment. In relation to the mouse strain,we observed isolated changes in the expression of Muc2,Muc3, Muc4, Muc13, and Muc16 in the infected mice(p<0.05) (Fig. 6).

To analyze further the effect of infection on the expressionof mucins in each mouse strain, we also applied the two-factorANOVA using wpi and infection as independent variables,and the results were similar for each host strain. Muc2 andMuc4 expression showed a significant increase with respect tothe respective control animals for each of the three strains(p<0.05). Muc16 also showed a transient increase in the threestrains, but it was transient and only was observed at 2 wpi(p<0.05). Finally, Muc1 expression showed significant in-creases with respect to controls for the three strains at 2 and6 wpi (p<0.05).

Cytokine expression

The infectionwith E. caproni affected in a different manner theexpression of cytokines in the ileum of each mouse strain(Fig. 7). In the infected BALB/c mice, the expression of themost of the cytokines did not change or was found to bedownregulated during the experiment. In this mouse strain, a

clear cytokine response after infection was not observed, andonly significant increases in IL-6 were detected. Moreover,decreases in several cytokines such as IL-10, IL-17, eotaxin/CCL11, and IFN-γ were found. ICR mice developed a stronglocal Th1 response characterized by the elevated levels of IFN-γobserved during the complete course of the experiment. Only aslight increase of the IL-6 and eotaxin/CCL11 levels was alsodetected. In relation to the IFN-γ expression in infected BALB/cmice, this cytokine was found to be downregulated during eachweek of the experiment. The infected IFN-γ−/− mice developeda weak Th2 response characterized by a weak upregulation ofIL-13. Moreover, slight increases of IL-6 were observed. Theexpression of other cytokines, such as IL-17, was found to bedownregulated with respect to the control animals.

Comparison of the cytokine expression over the course ofthe infection between the infected mice of the three strains(wpi and host strain as independent variables) showed signif-icant differences for IL-13 (upregulated in IFN-γ-deficientmice; p<0.001), IL-17 (downregulated in BALB/c and mice;p<0.001), eotaxin/CCL11 (downregulated in BALB/c andIFN-γ−/− mice; p<0.001), and IFN-γ (upregulated in ICRmice; p<0.0001) (Fig. 7).

Discussion

The ICR (CD-1®) mouse is considered a host of high com-patibility for E. caproni in which chronic infections with highworm recoveries are developed. In other hosts such as rats, the

0

200

400

600

0

1

2

3

4

5

Len

gth

(m

)

Villous/C

rypt Ratio

ICR

BALB/c

IFN- -/-

Villi0 wpi

Villi4 wpi

Crypt0 wpi

Crypt4 wpi

0 wpi 4 wpi

* *

* * * * * * γ

Fig. 3 Villus length, crypt length,and villus crypt ratio of ICR,BALB/c, and IFN-γ-deficientmice over the course of theinfection with Echinostomacaproni. Vertical bars representthe standard error. Asterisksrepresent significant differencesbetween control and infectedmice for each mouse strain(p<0.05)

0

50

100

150

200

250

0 2 4 6

Gob

let

cells


0

2

4

6

0 2 4 6

Mas

t ce

lls


a b

ICR

BALB/c

IFN- -/ -γ

Fig. 4 Kinetics of goblet (a) andmast (b) cells in ICR, BALB/c,and IFN-γ-deficient mice infectedwith Echinostoma caproni.Vertical bars represent thestandard error

Parasitol Res (2014) 113:2037–2045 2041

worm recoveries are lower, and parasites are rapidly rejected(Toledo et al. 2004, 2009; Toledo and Fried 2005; Muñoz-Antoli et al. 2007). Recent studies have shown that the localprofile of cytokines produced in response to the infection isessential in the course of the E. caproni infection and thedevelopment of a marked local Th1 response with elevatedlevels of IFN-γ and inflammation that were considered as themost relevant feature determining the development of chronic

infections with high worm recoveries (Sotillo et al. 2011;Trelis et al. 2011). To further clarify the role of IFN-γ in thecourse of E. caproni infections, we initially contrasted theimpact of E. caproni infection in ICR mice and in a prototyp-ical Th2 mouse strain, BALB/c mice.

The susceptibility of both strains to E. caproni infectionwas similar since the worm recoveries over the course of theinfection were similar. However, relevant differences in rela-tion to the clinical signs induced by E. caproniwere observed.Apparently, ICR mice did not show evidences of illness afterinfection, and the general aspect was similar to that of thecontrol ICR mice. Moreover, the weight curves were almostidentical to that in ICR control mice. Gross pathology inducedby E. caproni infection in ICR mice was similar to thatdescribed previously (Simonsen et al. 1989; Toledo et al.2006a; Muñoz-Antoli et al. 2007). In contrast, the infectedBALB/c mice looked emaciated from 2 wpi and developedclinical signs such as rough coat plates, diarrhea, perianalirritation, and loss of weight. The weight of the infectedBALB/c mice was significantly lower than in the respectivecontrol from 4 wpi.

To determine the factors causing these differences, weinvestigated mechanisms that play a role in the protectionagainst the virulence induced by the parasite, such as gobletand mast cell populations or mucins. However, the resultsobtained suggest that they are not associated with the

2 4 60

1

2

20406080


RQ

ICR

BALB/c

IFN- -/ -γ

Fig. 5 Expression of induced nitric oxide synthase mRNA in the ileumof infected ICR, BALB/c, and IFN-γ-deficient mice. The relative quan-tities (RQ) of induced nitric oxide synthase expression are shown afternormalization with β-actin and standardization of the relative amountagainst day 0 sample. Vertical bars represent the standard error. Asterisksrepresent significant differences between control and infected mice foreach mouse strain (p<0.05)

0

5

10

15

2 4 6

RQ


Muc1

0

2

4

6

8

2 4 6

RQ


Muc2

0

20

40

60

80

2 4 6

RQ


Muc4

0

0.5

1

1.5

2 4 6

RQ


Muc13

0

5

10

15

2 4 6

RQ


Muc16

0

2

4

6

8

2 4 6

RQ


Muc20

0

1

2

3

2 4 6

RQ


Muc3

ICR

BALB/c

IFN- -/ -

* *

**

*

γ

Fig. 6 Expression of mRNA of several mucin genes in the ileum ofinfected ICR, BALB/c, and IFN-γ-deficient mice. The relative quantities(RQ) of mucin genes are shown after normalization with β-actin andstandardization of the relative amount against day 0 sample.

Vertical bars represent the standard error. Asterisks indicated theweeks at which significant differences between the differentstrains were detected (p<0.05)

2042 Parasitol Res (2014) 113:2037–2045

increased severity of the disease observed in BALB/c mice.Goblet cells and mast cells increased in both strains, and thepattern of mucin expression was almost identical. Significantincreases in the expression of Muc2 andMuc4 together with asignificant decrease in the expression of Muc16 at 4 wpi weredetected in both strains, suggesting that it is not linkedwith thedifferences in the clinical signs.

Probably, the most striking differences between ICR andBALB/c mice that can be associated with the regulation ofclinical signs are related with the expression of local cyto-kines. The cytokine response in BALB/c mice was weak, andnone of the cytokines examined showed significant increasesover the course of the experiment. In contrast, infected ICRmice developed a strong local Th1 response, and the levels ofIFN-γ were elevated during the experiment. Another inflam-matory mediator, such as eotaxin/CCL11, was also upregulat-ed in ICR mice. Eotaxins have been implicated in the gener-ation of intestinal inflammation and the regulation of eosino-phil accumulation (Fulkerson and Rothenberg 2008). Howev-er, the role of eotaxin/CCL11 appears to be secondary since inE. caproni-infected rat upregulation of this mediator alsooccurred concomitantly with the absence of IFN-γ expres-sion, clinical signs, and inflammation (Toledo et al. 2006a;Trelis et al. 2011). Thus, IFN-γ was considered the maincausative factor of the inflammatory responses and the devel-opment of E. caproni chronic infections in mice (Muñoz-Antoli et al. 2007; Sotillo et al. 2011). Our results confirmthat IFN-γ appears to be essential for the development of localinflammatory responses as demonstrated by the levels ofiNOS expression in each host strain that correlated with the

levels of IFN-γ. Moreover, the present study supports thehypothesis that IFN-γ is not essential for E. caproni wormrecoveries but may play a major role in the protection againstthe disease and clinical signs by modulating its severity. In ourexperiment, the lack of clinical signs was correlated withelevated levels of this cytokine in ICR mice, whereasBALB/c mice developed a severe disease form in the absenceof local IFN- γ.

To confirm the role of IFN-γ in the protection againstsevere disease in E. caproni infection, we also studied theeffects of the infection in an IFN-γ-deficient mouse strain.Although, the parameters of the infection in IFN-γ−/− micewere similar to those observed in ICR and BALB/c mice (i.e.,worm recoveries, cell populations, or pattern of mucin expres-sion), this strain showed similar symptoms to that of BALB/cmice but were more intensified and appeared earlier. More-over, the host mortality was elevated and started at 2 wpi, andat 5 wpi all the infected IFN-γ−/− mice had died. These resultsstrongly support the fact that IFN-γ modulates the virulenceof E. caproni in mice and protects against the clinical signsand the host mortality induced by the parasite. The low levelor lack of IFN- γ expression in BALB/c and IFN-γ-deficientmice appears be the cause of the harmful effect of E. caproniinfection in these strains. The basal levels of IFN-γ expressionobserved in BALB/c mice may explain the lower intensity ofthe clinical signs and mortality observed with respect toIFN-γ-deficient mice. The resistance against severe diseaseobserved in ICR mice can be attributed to the marked upreg-ulation of IFN- γ and, and also probably, the consequentincreased levels of iNOS.

0

2

4

6

2 4 6

RQ


IL-10

0

1

2

2 4 6

RQ


IL-5

0

10

20

30

2 4 6

RQ


IL-6

0

1

2

3

2 4 6

RQ


IL-4

0

1

2

3

2 4 6

RQ


IL-23

0

3

6

9

2 4 6

RQ


IL-13

0

1

2

2 4 6R

Q


IL-17

0

2

4

6

8

2 4 6

RQ


Eotaxin/CCL11

IFN-


RQ

0

1

2

3

2 4 6

RQ


TGF-

0

0.5

1

1.5

2 4 6

RQ


TNF-

2 4 60

1

40

120

nd nd nd

nd nd nd

ICR

BALB/c

IFN- -/ -

***

** ** * **

*

β

γ

γ α

Fig. 7 Expression of cytokine mRNA in the intestinal tissue of infectedICR, BALB/c, and IFN-γ-deficient mice. The relative quantities (RQ) ofcytokine genes are shown after normalization with β-actin and standard-ization of the relative amount against day 0 sample. Vertical bars

represent the standard error. Asterisks indicated the weeks at whichsignificant differences between the different strains were detected(p<0.05) (nd not determined)

Parasitol Res (2014) 113:2037–2045 2043

The fact that IFN-γ and iNOS expression provides protec-tion against the disease and clinical signs is of interest sinceupregulation of this cytokine and production of nitric oxidehave been often associated with severe form of intestinaldisease (Mowat 1989; Smith 1989; Kühn et al. 1993; Krämeret al. 1995; Mizoguchi et al. 1996; Artis et al. 1999; Lawrenceet al. 2000; Keklikoglu et al. 2008). However, an increasingnumber of reports show that IFN-γ may have beneficialeffects. It has been shown that the modulation toward a Th1phenotype with upregulation of IFN-γ may ameliorate theseverity of the disease, though in most of the cases this isrelated to the resolution of the infections which did not occurin our study (Falcone et al. 1998; Castilow et al. 2008; Katoet al. 1999; Chen et al 2011). Recently, a study associated theabsence of IFN-γ to high levels of enteropathy in Eimeriafalciformis infections, but the pathology was mediated byTh17 cytokines (IL-17A and IL-22) generated in the absenceof IFN-γ (Stange et al. 2012). Lawrence et al. (1998) showedthat enteropathy in Trichinella spiralis-infected mice is exclu-sively mediated by TNF-α under regulation by IL-4 and Th2.In contrast, in our experiment the development of Th1 re-sponses in ICR mice neither implied the rejection of the adultworms nor a Th17- or Th2-driven response. In fact, thedevelopment of Th2 responses in IFN-γ-deficient mice didnot implicate protection against severe disease.

It is difficult to ascertain the manner in which IFN-γ and/oriNOS protect the host from the morbidity and mortality inducedby E. caproni. These responses can restrict the ability ofE. caproni to induce severe or even lethal disease in ICR mice.In contrast, E. caproni-infected BALB/c and IFN-γ−/− micedeveloped severe disease in the absence IFN-γ expression.Furthermore, it has been shown that IFN-γ responses duringTrichuris muris infections plays a positive role for the host bypromoting epithelial cell apoptosis (Cliffe et al. 2007). Theproduction of pro-inflammatory cytokines allows for the main-tenance of intestinal homeostasis in an attempt to counter thedysregulation of epithelial architecture induced by the parasite(Cliffe et al. 2007). Our results indicate that IFN-γ could playadditional roles in the maintenance of mucosal homeostasis.IFN-γ could play a critical regulatory role in the componentprocesses underlying crypt-villus homeostasis in E. caproni in-fections. The lack of IFN-γ induced crypt hyperplasia in theabsence of changes in the villus length which suggest an alteredrostro-caudal gradient affecting the homeostasis of the crypt-villus axis. Aspects such as cell differentiation, migration, orpromoting lineage specific maturation may be disrupted. It iswell known that alterations in these processes are determiningfactors in the generation of pathology in intestinal diseases (Liet al. 2007). In E. caproni infections, IFN-γ could participate inthe preservation of the mucosal architecture preventing a severeform of disease as shown in ICR mice. In contrast, the infectionin IFN-γ-deficient mice the infection becomes lethal due to adisruption of the mucosal homeostasis.

This reveals a novel and intriguing interplay betweenIFN-γ and pathogenic processes in intestinal helminths infec-tions in which this mediator plays a dichotomous role: on onehand, IFN-γ plays a negative role for the host since inhibitsthe generation of the protective Th2 responses and also byinducing inflammation but, on the other hand, IFN-γ is ben-eficial since it protects the host from morbidity and mortality.In this context, an essential role of IFN-γ enhancing thehomeostasis of the mucosa after E. caproni infection issuggested.

In summary, using several strains of mice, we have shownthat severity of the disease induced by E. caproni is dependentupon the capacity of express IFN-γ and probably iNOS.Although these mediators do not contribute to parasite clear-ance and, probably, favor the establishment of chronic infec-tions, they may protect against the injurious effects of theparasite on the host preventing the development of the mor-bidity and lethality. Furthermore, it is also suggested that thedevelopment of inflammatory responses in helminths infec-tions is not necessarily associated with severe forms ofdisease.

Acknowledgments This work was supported by the ProjectsPROMETEO/2009/081 from Conselleria d’Educació, GeneralitatValenciana (Valencia, Spain) and INV-AE13-136845 de la Universitat deValencia (Valencia, Spain). This work has been carried out while the firstauthor (A.C.) was a recipient of a pre-doctoral fellowship from theVicerectorat d’Investigació i Política Científic de la Universitat de València(Valencia, Spain). This research complies with the current laws for animalhealth research in Spain.

References

Artis D, Potten CS, Else KJ, Finkelman FD, Grencis RK (1999) Trichurismuris: host intestinal epithelial cell hyperproliferation during chron-ic infection is regulated by interferon-gamma. Exp Parasitol 92:144–153

Brunet LR (2001) Nitric oxide in parasitic infections. Int Immunopharmacol1:1457–1467

Castilow EM, Olson MR, Meyerholz DK, Varga SM (2008) Differentialrole of gamma interferon in inhibiting pulmonary eosinophilia andexacerbating systemic disease in fusion protein-immunized miceundergoing challenge infection with respiratory syncytial virus. JVirol 82:2196–21207

Chen N, Bellone CJ, Schriewer J, Owens G, Fredrickson T, Parker S,Buller RM (2011) Poxvirus interleukin-4 expression overcomesinherent resistance and vaccine-induced immunity: pathogenesis,prophylaxis, and antiviral therapy. Virology 409:328–337

Christensen NO, Simonsen PE, Odaibo AB, Mahler H (1990)Establishment, survival and fecundity of Echinostoma caproni(Trematoda) infections in hamsters and jirds. J Helminthol SocWash 57:104–107

Cliffe LJ, Potten CS, Booth CE, Grencis RK (2007) An increase inepithelial cell apoptosis is associated with chronic intestinal nema-tode infection. Infect Immun 74:1556–1564

Falcone M, Rajan AJ, Bloom BR, Brosnan CF (1998) A critical role forIL-4 in regulating disease severity in experimental allergic

2044 Parasitol Res (2014) 113:2037–2045

encephalomyelitis as demonstrated in IL-4-deficient C57BL/6 miceand BALB/c mice. J Immunol 160:4822–4830

Fried B, Huffman JE (1996) The biology of the intestinal trematodeEchinostoma caproni. Adv Parasitol 38:311–368

Fujino T, Fried B (1993) Expulsion of Echinostoma trivolvis (Cort 1914)Kanev 1985 and retention of E. caproni Richard 1964 (Trematoda,Echinostomatidae) in C3H mice: pathological, ultrastructural, andcytochemical effects on the host intestine. Parasitol Res 79:286–292

Fulkerson PC, Rothenberg ME (2008) origin, regulation and physiolog-ical function of intestinal eosinophils. Best Pract Res ClinGastroenterol 22:411–423

Humphreys NE, Xu D, Hepworth MR, Liew FY, Grencis RK (2008) IL-33, a potent inducer of adaptive immunity to intestinal nematodes. JImmunol 180:2443–2449

Kato Y, Manabe T, Tanaka Y, Mochizuki H (1999) Effect of an orallyactive Th1/Th2 balance modulator, M50367, on IgE production,eosinophilia, and airway hyperresponsiveness in mice. J Immunol162:7470–7479

Keklikoglu N, KorayM, Kocaelli H, Akinci S (2008) iNOS expression inoral and gastrointestinal tract mucosa. Dig Dis Sci 53:1437–1442

Klein D (2002) Quantification using real-time PCR technology: applica-tions and limitations. Trends Mol Med 8:257–260

Krämer S, Schimpl A, Hünig T (1995) Immunopathology of interleukin(IL) 2-deficient mice: thymus dependence and suppression bythymus-dependent cells with an intact IL-2 gene. J Exp Med 182:1769–1776

Kühn R, Löhler J, Rennick D, Rajewsky K,Müller W (1993) Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75:263–274

Lawrence CE, Paterson JC, Higgins LM, MacDonald TT, Kennedy MW,Garside P (1998) IL-4-regulated enteropathy in an intestinal nema-tode infection. Eur J Immunol 28:2672–2684

Lawrence CE, Paterson JC, Wei XQ, Liew FY, Garside P, Kennedy MW(2000) Nitric oxide mediates intestinal pathology but not immuneexpulsion during Trichinella spiralis infection in mice. J Immunol164:4229–4234

Li P, Lin JE, Chervoneva I, Schulz S, Waldman SA, Pitari GM (2007)Homeostatic control of the crypt-villus axis by the bacterial entero-toxin receptor guanylyl cyclase C restricts the proliferating compart-ment in intestine. Am J Pathol 171:1847–1858

Livak KJ, Schmittgen TD (2001) Analysis of relative gene expressiondata using real-time quantitative PCR and the 2-ΔΔCT method.Methods 25:402–408

Mahler H, Christensen NO, Hindsbo O (1995) Studies on the reproduc-tive capacity of Echinostoma caproni (Trematoda) in hamsters andjirds. Int J Parasitol 25:705–710

Martinez FO, Helming L, Gordon S (2009) Alternative activation ofmacrophages: an immunologic functional perspective. Annu RevImmunol 27:451–483

McDermott JR, Humphreys NE, Forman SP, Donaldson DD, Grencis RK(2005) Intraepithelial NK cell-derived IL-13 induces intestinal pathol-ogy associated with nematode infection. J Immunol 175:3207–3213

Mizoguchi A, Mizoguchi E, Chiba C, Spiekermann GM, Tonegawa S,Nagler-Anderson C, Bhan AK (1996) Cytokine imbalance andautoantibody production in T cell receptor-alpha mutant mice withinflammatory bowel disease. J Exp Med 183:847–856

Mowat AM (1989) Antibodies to IFN-gamma prevent immunologicallymediated intestinal damage in murine graft-versus-host reaction.Immunology 68:18–23

Muñoz-Antoli C, Sotillo J, Monteagudo C, Fried B, Marcilla A, Toledo R(2007) Development and pathology of Echinostoma caproni inexperimentally infected mice. J Parasitol 93:854–859

Odaibo AB, Christensen NØ, Ukoli FMA (1988) Establishment, survivaland fecundity in Echinostoma caproni (Trematoda) infections inNMRI mice. Proc Helminthol Soc Wash 55:265–269

Odaibo AB, Christensen NØ, Ukoli FMA (1989) Further studies on thepopulation regulation in Echinostoma caproni infections in NMRImice. Proc Helminthol Soc Wash 56:192–198

Simonsen PE, Bindseil E, Køie M (1989) Echinostoma caproni in mice:studies on the attachment site of an intestinal trematode. Int JParasitol 19:561–566

Smith NC (1989) The role of free oxygen radicals in the expulsion ofprimary infections ofNippostrongylus brasiliensis. Parasitol Res 75:423–438

Sotillo J, Trelis M, Cortes A, Fried B, Marcilla A, Esteban JG,Toledo R (2011) Th17 responses in Echinostoma caproni infec-tions in hosts of high and low compatibility. Exp Parasitol 129:307–311

Stange J, Hepworth MR, Rausch S, Zajic L, Kühl AA, Uyttenhove C,Renauld JC, Hartmann S, Lucius R (2012) IL-22 mediates hostdefense against an intestinal intracellular parasite in the absence ofIFN-γ at the cost of Th17-driven immunopathology. J Immunol188:2410–2418

Toledo R, Fried B (2005) Echinostomes as experimental models forinteractions between adult parasites and vertebrate hosts. TrendsParasitol 21:251–254

Toledo R, Espert A, Carpena I, Muñoz-Antoli C, Fried B, Esteban JG(2004) The comparative development of Echinostoma caproni(Trematoda: Echinostomatidae) adults in experimentally infectedhamsters and rats. Parasitol Res 93:439–444

Toledo R, Esteban JG, Fried B (2006a) Immunology and pathology ofintestinal trematodes in their definitive hosts. Adv Parasitol 63:285–365

Toledo R, Monteagudo C, Espert A, Fried B, Esteban JG, Marcilla A(2006b) Echinostoma caproni: intestinal pathology in the goldenhamster, a highly compatible host, and the Wistar rat, a less com-patible host. Exp Parasitol 112:164–171

Toledo R, Esteban JG, Fried B (2009) Recent advances in the biology ofechinostomes. Adv Parasitol 69:147–204

Trelis M, Sotillo J, Monteagudo C, Fried B, Marcilla A, Esteban JG,Toledo R (2011) Echinostoma caproni (Trematoda): differentialin vivo cytokine responses in high and low compatible hosts. ExpParasitol 127:387–397

Parasitol Res (2014) 113:2037–2045 2045

Intestinal IFN-γ production is associated with protection from clinical signs, but not with...

Documents

Transcript of Intestinal IFN-γ production is associated with protection from clinical signs, but not with...