Influence of murine infection on plasma and bronchoalveolar lavage fluid eosinophil numbers and its correlation with cytokine levels

Influence of murine Toxocara canis infection on plasma and

bronchoalveolar lavage fluid eosinophil numbers and its

correlation with cytokine levels

Ney Roner Pecinali a, Rachel N. Gomes a, Fabio C. Amendoeira a,Augusto C.M. Pereira Bastos c, Maria J.Q.A. Martins c, Claudia S. Pegado b,Otılio M. Pereira Bastos c, Patrıcia T. Bozza a, Hugo C. Castro-Faria-Neto a,*

a Laboratorio de Imunofarmacologia, Departamento de Fisiologia e Farmacodinamica e Fundacao Oswaldo Cruz,

Av. Brasil 4365, CEP 21045-900 Rio de Janeiro, RJ, Brazilb Laboratorio de Anatomia Patologica, Fundacao Educacional Serra dos Orgaos, Teresopolis, Brazil

c Laboratorio de Parasitologia, Departamento de Microbiologia e Parasitologia,

Universidade Federal Fluminense, Niteroi, Brazil

Accepted 5 June 2005

Abstract

Toxocara canis is a nematode of the Ascaridae family that normally parasites the small intestine of canid species. Humans are

accidentally infected upon ingestion of embryonated eggs, and can manifest several clinical alterations such as fever,

hepatomegaly, splenomegaly, respiratory symptoms, muscle pain and anorexia. In the present work, we investigated the

kinetics of tissue distribution of L2 larva in lungs, liver, kidney, brain, skeletal muscle and myocardium. Also, we analyzed the

blood and bronchoalveolar lavage fluid (BAL) for levels of IL-6, IFN-g, eotaxin and Regulated on Activation Normal T Cell

Expressed and Secreted (RANTES) in experimental murine T. canis infection. We observed liver, lung and kidney lesions

correlated to larva migration as early as the first day of infection. After the seventh post-infection day, larva could also be

detected in brain, skeletal muscle and heart, as an indicator of biphasic migration pattern. Increased inflammatory activity was

detected in BAL and plasma of infected animals, as was an intense eosinophil migration associated with an increase in the levels

of all the cytokines studied. In conclusion, our results establish a tight correlation between tissue lesions caused by larva

migration and increased plasma levels of pro-inflammatory and eosinophil chemotactic cytokines. Thus, murine T. canis

infection may prove to be useful in understanding the role of cytokines in infection.

# 2005 Elsevier B.V. All rights reserved.

Keywords: VLM; Chemokine; Cytokine; Mouse; Toxocara canis; Eosinophils

www.elsevier.com/locate/vetpar

Veterinary Parasitology 134 (2005) 121–130

* Corresponding author. Tel.: +55 21 2598 4492x222; fax: +55 21 2590 9490x213.

E-mail address: [email protected] (H.C. Castro-Faria-Neto).

0304-4017/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.vetpar.2005.06.022

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130122

1. Introduction

Toxocara canis is a nematode of the family

Ascaridae that normally parasitizes the small intestine

of canid species. Humans are accidentally infected

upon ingestion of embryonated eggs, consequently

harbor third-stage larva in their tissues, and thus

become paratenic hosts. The persistence or migration

of T. canis in deep tissues of humans causes a

condition known as Visceral Larva Migrans (VLM)

(Beaver, 1952). In most cases, the human infection by

Toxocara larva does not show clinical manifestations

and evolves in an asymptomatic manner. However, in

some cases, it can cause several alterations such as

fever, hepatomegaly, splenomegaly, respiratory symp-

toms, muscle pain and anorexia (Schantz and Glick-

man, 1983). Ocular manifestations are found in

isolated cases, and probably result from a mild

parasite burden.

Previous works indicate the existence of two

distinct phases of larva migration. The first phase is a

visceral one affecting the liver, lungs and kidneys.

This phase starts at the first day after infection and is

usually over by the seventh day. The second is a

myotropic–neurotropic phase affecting the brain,

skeletal and cardiac muscles. The second phase

usually starts at the 7th day after infection, peaking

on the 14th day (Buijs et al., 1994; Carter, 1992;

Helwigh et al., 1999; Kusama et al., 1995; Parsons

et al., 1993; Piergili Fioretti et al., 1989).

Eosinophilia has been described by several authors

as one of the most outstanding characteristic of the

VLM in naturally infected humans and experimental

models of infection (Arango, 1998; Beaver, 1952;

Kayes et al., 1987;Kayes andOaks, 1980;Meeusen and

Balic, 2000; Parsons et al., 1993; Roig et al., 1992;

Rothenberg, 1998; Sugane and Oshima, 1984). VLM

eosinophilia peaks at the fourth day after experimental

infection (Okada et al., 1996) and it has been correlated

with the production of interleukin-5 (IL-5) by Th2

CD4+ lymphocytes in the lungs as well as with elevated

production of IgE (Kusamaet al., 1995;Takamoto et al.,

1995; Takamoto and Sugane, 1993). In addition to IL-5,

leucotrienes (LT) and platelet activating factor (PAF)

seem also to play an important role in the eosinophil

accumulation in the lungs of T. canis-infected rats.

Overall, the literature suggests a predominant Th2

response related to VLM eosinophilia (Arango, 1998;

Beaver, 1952; Kayes et al., 1987; Kayes and Oaks,

1980; Meeusen and Balic, 2000; Parsons et al., 1993;

Roig et al., 1992; Rothenberg, 1998; Sugane and

Oshima, 1984), but several authors suggest the

formation of granuloma which is tightly associated

with a Th1 response (Buijs et al., 1994; Carter, 1992;

Helwigh et al., 1999; Kusama et al., 1995; Parsons and

Grieve, 1990; Piergili Fioretti et al., 1989). Another

point of discrepancy is related to the day of maximum

inflammatory reaction, as the available reports are

inconclusive, usually locating the peak reaction

between the 7th and the 20th day after infection (Buijs

et al., 1994; Dent et al., 1999; Kayes et al., 1987; Kayes

and Oaks, 1980; Kusama et al., 1995; Okada et al.,

1996; Parsons et al., 1993; Schaffer et al., 1992; Sugane

and Oshima, 1984; Takamoto and Sugane, 1993).

In the present work, we investigated the levels of

IFN-g and eotaxin as well IL-6 and Regulated on

Activation Normal T Cell Expressed and Secreted

(RANTES) in experimental infection by T. canis in

mice, aiming to validate these parameters for

determination of tissue damage in this parasitic

disease. In this respect, IL-6 was chosen because it

is a well-characterized marker of inflammatory

response in several diseases such as arthritis and

sepsis (Bozza et al., 2004; Nishimoto and Kishimoto,

2004), while RANTES was previously shown to be

important on eosinophilic responses (Lampinen et al.,

2004). Our results establish a tight correlation between

tissue lesions caused by larval migration and the

plasma levels of IL-6, IFN-g, eotaxin and RANTES in

this model of VLM.

2. Materials and methods

2.1. Animals

Male BALB/c mice (from the Oswaldo Cruz

Foundation breeding unit) 6–8 weeks old, weighing

20–25 g, were used. Previous to the beginning of the

experiments, feces from all mice were collected and

submitted to parasitologic examination and only those

mice from colonies with negative parasitologic results

were used in the following experiments. The animals

were divided into two groups: the first one was

composed of animals infected with T. canis (n = 80)

and the second one consisted of animals administered

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130 123

with saline (control group, n = 80). Micewere killed in

a CO2 chamber 1, 4, 7, 11, 15, 21, 30 and 45 days after

infection. Ten animals were examined in each group at

each time point. Both groups were maintained with

free access to food and water and kept at 25–28 8Cwith controlled 12-h light:12-h dark cycle. This study

received prior approval from the Oswaldo Cruz

Institute’s Animal Welfare Committee.

2.2. Toxocara canis infection

Adult T. canis specimens were collected from feces

of naturally infected and thiabendazol-treated puppies.

After identification, T. canis females were washed in

physiologic solution to eliminate the fecalmaterial, and

underwent hysterectomy in order to obtain viable T.

canis eggs (Fenoy et al., 1987). T. canis eggs were

washed in 1% sodium hydrochloride solution and were

centrifuged at 2000 rpm in a microcentrifuge for 3 min

(Bowman et al., 1987). The sediment was suspended in

H2SO4 (0.1N) and incubated for 40 days at room

temperature in an Erlenmeyer flask under natural

illumination (Kayes and Oaks, 1980). The eggs were

counted in Neubauer chamber and the embryonated

eggs were used to infect the animals. Mice were

infected by oral gavage with approximately 1000

embryonated eggs. The control group received the same

volume of sterile physiologic solution.

2.3. Cellular analysis in the blood and

bronchoalveolar lavage fluid (BAL)

Total and differential cell counts were performed in

peripheral blood samples that were collected from the

caudal vein of mice immediately before the killing of

the animals. Total counts were performed in Neubauer

chambers after diluting the blood with 2% acetic acid,

whereas differential cell counting was performed in

blood smears stained with May–Grunwald–Giemsa

dye. The bronchoalveolar lavage was performed after

isolating the trachea by blunt dissection. A small-

caliber tube was inserted and secured in the airway.

Three volumes of 1.0 mL of PBS with 3.7% sodium

citrate were then instilled and gently aspirated and

pooled. In every instillation/aspiration cycle, the same

volume (1.0 mL) was recovered from each animal.

Total cell counts were performed in Neubauer

chambers after diluting the BAL fluid with 2% acetic

acid, whereas differential cell counting was performed

in cytosmears stained with May–Grunwald–Giemsa

dye. The BAL fluid supernatants were separated by

centrifugation and were stored at �70 8C for cytokine

determinations.

2.4. Histological examination

Tissues from control and infected animals were

obtained immediately after killing and used for

histological analyses. The following organs were

removed and washed in physiologic solution: heart,

right lung, liver, right kidney, right thigh skeletalmuscle

and brain. These tissues were fragmented and fixed in

10% buffered neutral formaldehyde. After seven days,

the tissues were processed in paraffin blocks by cutting

serial sections with 5 mm thickness for each organ from

each mouse and were prepared and stained with

hematoxylin and eosin (HE). Four sections from each

tissue were randomly selected and examined.

2.5. Measurements of cytokines levels in the blood

and in the BAL

Lung concentrations of IL-6, IFN-g, eotaxin and

RANTES were determined using BAL fluid super-

natants. The plasma cytokine levels were analyzed in

the blood obtained from the abdominal aorta of mice.

The blood was collected with a syringe containing

3.7% sodium citrate in PBS, immediately after

sacrifice and before performing the BAL. The blood

was centrifuged (2000 rpm for 5 min), to separate the

plasma. The plasma was divided in two aliquots that

were used to measure cytokine levels. Commercially

available enzyme-linked immunosorbent assay

(ELISA) kits were used for measurement of murine

IFN-g, IL-6, eotaxin and RANTES (R&D Systems,

Minneapolis, MN) in cell-free BAL or plasma,

according to the manufacturer’s instructions. Al the

measures were performed in duplicate.

2.6. Documentation and statistical analysis

The microphotographs were obtained with a digital

camera model PMC35B (OLYMPUS), coupled to an

optical microscope, model BX-60 (OLYMPUS). Data

were represented as mean � standard error of mean

(S.E.M.). The statistical analysis involving only two

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130124

groups was done using Student’s T-test. ANOVA

followed by Neuman–Keuls Student’s test was used to

compare three or more groups, and significance was

assessed at the P < 0.05 level of confidence.

3. Results

3.1. Kinetics of tissue larva distribution and

leukocyte alterations

The histological examination of the tissues revealed

the presence of the T. canis larva and inflammatory

lesions in all the organs studied. The kinetics of tissue

larva distribution followed the two phases profile

previously described before (Buijs et al., 1994; Carter,

1992; Helwigh et al., 1999; Kusama et al., 1995;

Parsons and Grieve, 1990; Piergili Fioretti et al., 1989).

Fig. 1 shows the absolute percentage of larva in the

different tissues by day post-infection (PI). We

observed that in the liver, lung and kidney, the larvae

presence ismore evident from the 1st to the 15th day PI.

On the other hand, in the brain, skeletal muscle and

heart, the presence of the larva was more evident after

the 11th day PI. As expected, no sign of larva presence

was detected in control animals.

Liver cuts obtained between the first and seventh day

PI disclosed scattered necrosis with an inflammatory

infiltrate represented predominantly by mononuclear

cells among which T. canis larvae were observed (data

not shown). Moreover, we can observe the presence of

Fig. 1. Kinetics of tissue larva distribution in T. canis-infected

BALB/c mice. Each bar represents percentage of animals with

the presence of L2 larvae in the different tissues on the day post-

infection (n = 10).

eosinophilic bodies among the preserved cells. In

sinusoids lumen, the presence of larvae was frequent.

Hepatic portal spaces had perivascular mononuclear

and eosinophilic inflammatory infiltrate. On the 11th

and the 15th day PI, the lobular structurewas preserved,

but with intense congestion of sinusoids and intralob-

ular veins (data not shown). The hepatocytes presented

diffuse vacuolization and Kupfer cells showed hyper-

trophy. Focal infiltration of lymphocytes and eosino-

phils was also observed in the portal spaces. Starting at

the 21st day PI, eosinophil infiltrates were detected as

granulomatous structures (data not shown).

In the lung, histological cuts obtained between the

first and seventh day PI revealed a normal pleura, with

intense capillary congestion in the alveolar septum, as

well as extensive areas of disrupted alveoli with

eosinophilic inflammatory infiltrate and some poly-

morphonuclear cells around the bronchioles (data not

shown). Bronchial and bronchiolar structures exhibited

the same type of inflammatory infiltrate with fragments

of T. canis larvae. Between the 11th and 30th day PI, we

observed eosinophilic granulomatous structures.

The kidneys of animals killed between the 1st and

21st day PI showed glomerular congestion while

tubular structures exhibited discrete epithelium

vacuolization and presence of eosinophilic and

amorphous material in lumen. Intense vascular

congestion was frequently observed within the

interstices. Importantly, larval presence was more

frequently seen at the 15th day PI (data not shown).

The heart of animals killed between the 11th and

30th day after infection revealed focal lymphocytes

infiltrated in the endocardium. The myocardium

frequently exhibited eosinophilic infiltrate and vas-

cular congestion with frequent larva presence (data not

shown). In skeletal muscle, we observed vascular

congestion and intense inflammatory eosinophilic

infiltrate, with presence of fragments of T. canis larva

in the muscular fibers. Moreover, in the central

nervous system, the meninges were thickened and

contained a large number of eosinophils and some

lymphocytes. Ependymarium covering epithelium

was preserved despite the intraventricular hemorrhage

and frequent larva presence (data not shown).

Differential blood leukocyte counts revealed a

significant eosinophilia from the 4th up to the 45th

day PI, ranging from 4 to 21% and reaching a peak of

3850 � 233 cells/mm3 on the 11th day (Fig. 2). We

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130 125

Fig. 2. Peripheral blood eosinophil counts in T. canis-infected

BALB/c mice. Each bar represents mean � standard error from

10 animals. The asterisks indicate statistical significance

(P < 0.05) when compared to uninfected controls.

also observed a significant increase in lymphocyte

numbers in the blood of infected animals from the 11th

up to the 21st day PI, reaching a peak of 11,235�341 cells/mm3 on the 11th day (data not shown).

Eosinophil counts were also increased in the BAL

fluid of infected animals from the 4th to the 30th day

PI, showing a peak at the 11th day PI (Fig. 3). The

maximum absolute number of eosinophils was

11.5 � 1.2 � 103 cells/mm3 for the control group,

whereas this value reached 152 � 12.43 � 103 cells/

mm3 in infected animals.

3.2. Plasma and BAL levels of cytokines

Analysis of the plasma obtained from animals killed

between the 4th and the 15th day PI revealed a marked

Fig. 3. BAL eosinophil counts in T. canis-infected BALB/c mice.

Each bar represents mean � standard error from 10 animals. The

asterisks indicate statistical significance (P < 0.05) when compared

to uninfected controls.

increase in plasma IL-6 concentrations in infected

animals as compared to control group (Fig. 4A). Peak

plasma IL-6 concentrations (0.76� 0.07 ng/mL) were

achieved at the 11th day PI. In addition, a significant

increase in IL-6 was also detected in the BAL of

infected animals again with peak values

(1.07� 0.11 ng/mL) at the 11th day PI (Fig. 4B).

Plasma levels of IFN-g (Fig. 5A) were found

above control values between the 4th and the 21st

day PI. As it was observed with IL-6, peak

concentrations of IFN-g (2.67 � 0.13 ng/mL) were

seen at the 11th day PI. Analysis of the BAL fluid

(Fig. 5B) showed similar results, with significant

increase in the IFN-g concentrations between the 4th

and the 21st day PI with peak (0.763 � 0.024 ng/mL)

observed at the 11th day PI. However, it is important

to note that peak IFN-g concentrations observed in

the plasma were at least three times higher than in the

BAL (Fig. 5A and B).

Fig. 4. Plasma (A) and BAL (B) IL-6 levels in T. canis-infected

BALB/c mice. Each bar represents mean � standard error of from

10 animals. The asterisks indicate statistical significance (P < 0.05)

when compared to uninfected controls.

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130126

Fig. 5. Plasma (A) and BAL (B) IFN-g levels in T. canis-infected

BALB/c mice. Each bar represents mean � standard error from 10

animals. The asterisks indicate statistical significance (P < 0.05)

when compared to uninfected controls. Fig. 6. Plasma (A) and BAL (B) eotaxin levels in T. canis-infected

BALB/c mice. Each bar represents mean � standard error from 10

animals. The asterisks indicate statistical significance (P < 0.05)

when compared to uninfected controls.

eotaxin in the plasma and BAL fluid (Fig. 6A and B).

Eotaxin was significantly increased between the 4th

and the 21st day PI, peaking at the 11th day

(1160 � 129 pM/mL in the plasma and 1587 �171 pM/mL in the BAL fluid). Importantly, eotaxin

BAL concentrations were approximately 40% higher

than in the plasma (Fig. 6A and B).

As shown in Fig. 7A, RANTES plasma levels rose

significantly above control values from the 4th to the

30th day PI with peak concentration at the 11th day

(180 � 15.11 pM/mL). Analysis of RANTES in the

BAL fluid showed a significant increase from the 7th

to the 15th day PI with peak values (85.91 �15.67 pM/mL) achieved at the 11th day PI (Fig. 7B).

4. Discussion

In this work, we have shown the existence of

a direct correlation between tissue lesion induced by

T. canis larva migration and high levels of eotaxin,

RANTES, IL-6 and IFN-g in mice VLM. The

histological analysis revealed that on the first day

PI, lesions caused by larva migration, in the liver, lung

and kidney were detected. The larval presence in only

these organs was observed until the seventh day PI,

when larval presence was also observed in the brain,

skeletal muscle and heart. These results confirm

previous descriptions about the existence of two

phases in larva migration (Buijs et al., 1994; Carter,

1992; Helwigh et al., 1999; Kusama et al., 1995;

Parsons et al., 1993; Piergili Fioretti et al., 1989).

Different authors classified the first phase as a visceral

one, beginning immediately PI and prevailing up to

the seventh day. A second phase, generally referred to

as a myotropic–neurotropic phase, initiates at the

seventh day PI (Buijs et al., 1994; Carter, 1992;

Helwigh et al., 1999; Kusama et al., 1995; Parsons and

Grieve, 1990; Piergili Fioretti et al., 1989).

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130 127

Fig. 7. Plasma (A) and BAL (B) RANTES levels in T. canis-

infected BALB/c mice. Each bar represents mean � standard error

from 10 animals. The asterisks indicate statistical significance

(P < 0.05) when compared to uninfected controls.

Considering that the eggs were administered by

oral route, the detection of larva initially in the liver

could be explained by the influence of enterohepatic

circulation in transporting those larva from the

intestinal wall. Moreover, we demonstrate that the

larva appear in the lungs before they appear in heart.

This was unexpected since the larva traveling in the

blood stream would pass first through the heart before

they reach the lung, and may be explained by the

reflection coefficient of these organs, due to their

different capillary types. Different organs have

different endothelium fenestrations. For instance,

the brain microcirculation has no pores and 4 nm of

width fenestrations are characteristics of the skeletal

and cardiac muscles, whereas 20–100 nm fenestra-

tions are seen in the kidneys and a discontinuous

endothelium is a hallmark of the liver (Berne et al.,

2000). From our results, we can say that larger pores

and/or discontinuous capillary system are associated

with an earliest larval penetration in the organs.

Infected mice showed a significant leukocytosis

(data not shown), with prevalence of neutrophils seen

by the first day PI, indicating an acute inflammatory

process in these animals. At this point, eosinophil

accumulation appears slightly elevated and not

statistically significant. On the fourth day, eosinophil

numbers were increased both in the blood and in the

BAL which may be related to the defensive role of

eosinophils in helminthic infections, as those cells are

able to release granules with enzymatic components

able to inactivate and/or destroy the helminthes

(Mendes et al., 2000). This observation is in

agreement with previous reports showing that inflam-

mation and eosinophilia constitute striking classical

signals of VLM (Arango, 1998; Beaver, 1952; Kayes

et al., 1987; Meeusen and Balic, 2000; Parsons et al.,

1993; Roig et al., 1992; Rothenberg, 1998).

On the seventh day PI, the total leukocyte,

eosinophil and lymphocyte counts were significantly

elevated in peripheral blood and in the BAL. Peak

values of change in leukocyte counts were observed on

the 11th day PI and remained elevated until the 15th

day. These results, when compared to previous

findings in the literature, demonstrate that the day

of peak values in leukocyte numbers may vary

according with different factors of the host and also

the infectivity of the larva (Buijs et al., 1994; Dent

et al., 1999; Kayes et al., 1987; Kayes and Oaks, 1980;

Kusama et al., 1995; Okada et al., 1996; Parsons et al.,

1993; Sugane and Oshima, 1984; Takamoto and

Sugane, 1993). In all paratenic hosts, migrating T.

canis larva neither molt to the next stage nor increase

in size. Nevertheless, those larva are metabolically

active and put out both excretory and secretory

materials in addition to shedding epicuticular sub-

stances into the extracellular matrices of the host

(Kayes, 1997). Larva products can, in turn, activate the

host’s immune cells to secrete inflammatory mediators

such as cytokines and chemokines (Faccioli et al.,

1997), leading to an increased number of leukocytes.

The number of total leukocytes and eosinophils

decreases after the 21st day PI, indicating the

reduction of the systemic and the local inflammatory

response and pointing to the resolution of the process.

Increased eosinophil counts in the blood and BAL

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130128

persisted until the 30th day, and by the 45th day blood

eosinophilia was the only parameter still significantly

above control.

The concentrations of IL-6, IFN-g, eotaxin and

RANTES in the plasma and BAL of T. canis infected

mice were measured and compared to the inflamma-

tory reaction. It is important to note that among these

cytokines, only IFN-g had been described in literature

in association with the T. canis infection. In this work,

in addition to IFN-g, we also observed significant

increases in IL-6, eotaxin and RANTES levels from

the 4th day PI, peaking at the 11th day and decreasing

thereafter. This pattern is in close association with the

presence of the larva in the tissues studied.

Variations in IL-6 levels had not previously been

correlated with T. canis infection. Increased IL-6

levels were noted concomitantly with an increase in

total leukocytes and eosinophils number, reinforcing

the hypothesis of an efficient antigenic presentation

followed by early and significant inflammatory

reaction triggered by parasite antigens and possibly

suggesting that VLM induces an inflammatory

reaction with a mixed cytokine profile (Viola and

Rao, 1999). Classically, IL-6 is a cytokine with

important prognostic value in bacterial sepsis.

Although the pathophysiologic role of IL-6 in this

syndrome is still controversial, IL-6 has been proposed

as an important cytokine biomarker in sepsis due to its

slow and stable plasma kinetics and its good

correlation with the intensity of the inflammatory

response (Gogos et al., 2000; Kox et al., 2000). In fact,

IL-6 production is observed not only during bacterial

infection, but also during the inflammatory response

of non-infectious origin. Therefore, IL-6 must be more

suitably described as a marker for inflammation

(Nishimoto and Kishimoto, 2004). Despite the fact

that VLM is described as an infection that triggers

predominantly Th2-type responses (Del Prete et al.,

1991; Meeusen and Balic, 2000), the significant

increase in IFN-g levels, a Th1-type cytokine,

reinforces our observation that VLM triggers a mixed

cytokine response which is in agreement with the

findings of Del Prete et al. (1995). Importantly, IFN-g

can also be secreted by eosinophils (Fruh and Yang,

1999; Woerly et al., 1999) and the formation of

eosinophilic granulomas can be regarded as a

manifestation of an active Th1 response during T.

canis infection (Kayes, 1997; Meeusen and Balic,

2000; Ovington and Behm, 1997; Piergili Fioretti

et al., 1989).

Eotaxin is an important chemotactic factor involved

in selective eosinophil recruitment to the inflammatory

site and togetherwith IL-5 it seems to play an important

role in bronchial asthma (Conroy et al., 1997; Teran,

2000). To our knowledge, this is the first report

demonstrating increased eotaxin levels during T. canis

infection; however, the presence of this cytokine was

postulated earlier based on the similarities between

VLM and asthma, which in some cases may cause

confusion in the diagnosis (Roig et al., 1992). In this

work, we have shown a significant rise in plasma and

BALeotaxin concentrations from the 7th to the 21st day

PI. This increase is correlated to the largest number of

lesions observed in histological analyses and the larval

presence in the examined organs.

RANTES is a chemokine that binds to theCCR3 and

CCR1 receptors leading to activation of eosinophils

(Gangur and Oppenheim, 2000). As with eotaxin,

increased levels of RANTES have never previously

been reported in a T. canis infection. Our experiments

demonstrate that a significant increase in RANTES

levels is detected concomitantly with increased con-

centrations of eotaxin, as well as with increased

eosinophil counts and the observation of eosinophils in

the tissue. Taken together, these observations indicate a

marked cytokine response aimed to amplify eosinophil

recruitment and activation that is consistent with the

histological findings in T. canis infection.

In conclusion, our results establish a tight correla-

tion between tissue lesions caused by larva migration

with the plasma cytokine production. Moreover, we

describe eotaxin and RANTES as potential factors

responsible for the marked eosinophilic response that

is a hallmark of this infection.

Acknowledgements

This work was supported by a grant from FAPERJ,

CNPq and PAPES.

References

Arango, C.A., 1998. Visceral larva migrans and the hypereosino-

philia syndrome. S. Med. J. 91, 882–883.

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130 129

Beaver, P., 1952. Toxocarosis (visceral larva migrans) in relation to

tropical eosinophilia. In: Bulletin de la Societe de Pathologie

Exotique, pp. 555–576.

Berne, R., Levy, M., Koeppen, B., Stanton, B., 2000. Microcircu-

lacao e Linfaticos. In: Koogan, G. (Ed.), Fisiologia. Berne R.M.

& Levy M.N., Rio de Janeiro.

Bowman, D.D., Mika-Grieve, M., Grieve, R.B., 1987. Circulating

excretory–secretory antigen levels and specific antibody

responses in mice infected with Toxocara canis. Am. J. Trop.

Med. Hyg. 36, 75–82.

Bozza, F.A., Gomes, R.N., Japiassu, A.M., Soares, M., Castro-Faria-

Neto, H.C., Bozza, P.T., Bozza, M.T., 2004. Macrophage migra-

tion inhibitory factor levels correlate with fatal outcome in

sepsis. Shock 22, 309–313.

Buijs, J., Lokhorst, W.H., Robinson, J., Nijkamp, F.P., 1994. Tox-

ocara canis-induced murine pulmonary inflammation: analysis

of cells and proteins in lung tissue and bronchoalveolar lavage

fluid. Parasite Immunol. 16, 1–9.

Carter, K.C., 1992. The use of 35S-labelled Toxocara canis to

determine larval numbers in tissues. J. Helminthol. 66, 279–287.

Conroy, D.M., Humbles, A.A., Rankin, S.M., Palframan, R.T.,

Collins, P.D., Griffiths-Johnson, D.A., Jose, P.J., Williams,

T.J., 1997. The role of the eosinophil-selective chemokine,

eotaxin, in allergic and non-allergic airways inflammation.

Mem. Inst. Oswaldo Cruz 92 (Suppl. 2), 183–191.

Del Prete, G., De Carli, M., Almerigogna, F., Daniel, C.K., D’Elios,

M.M., Zancuoghi, G., Vinante, F., Pizzolo, G., Romagnani, S.,

1995. Preferential expression of CD30 by human CD4+ T cells

producing Th2-type cytokines. FASEB J. 9, 81–86.

Del Prete, G.F., De Carli, M.,Mastromauro, C., Biagiotti, R.,Macchia,

D., Falagiani, P., Ricci, M., Romagnani, S., 1991. Purified protein

derivative ofMycobacterium tuberculosis and excretory–secretory

antigen(s) of Toxocara canis expand in vitro human T cells with

stable and opposite (type 1 T helper or type 2 T helper) profile of

cytokine production. J. Clin. Invest. 88, 346–350.

Dent, L.A., Daly, C.M., Mayrhofer, G., Zimmerman, T., Hallett, A.,

Bignold, L.P., Creaney, J., Parsons, J.C., 1999. Interleukin-5

transgenic mice show enhanced resistance to primary infections

with Nippostrongylus brasiliensis but not primary infections

with Toxocara canis. Infect. Immun. 67, 989–993.

Faccioli, L.H., Vargaftig, B.B., Medeiros, A.I., Malheiros, A., 1997.

Cytokines in the modulation of eosinophilia. Mem. Inst.

Oswaldo Cruz 92 (Suppl. 2), 109–114.

Fenoy, R., Del Hoyo, C., Guillen-Llera, J., 1987. Estudio compar-

ativo de la influencia de la luz en el embrionamento experi-

mental de los huevos de Toxocara canis y Toxocara leonine. Rev.

Iber. Parasitol. Extraordinario 173–177.

Fruh, K., Yang, Y., 1999. Antigen presentation by MHC class I and

its regulation by interferon gamma. Curr. Opin. Immunol. 11,

76–81.

Gangur, V., Oppenheim, J.J., 2000. Are chemokines essential or

secondary participants in allergic responses? Ann. Allergy

Asthma Immunol. 84, 569–579 quiz 579–581.

Gogos, C.A., Drosou, E., Bassaris, H.P., Skoutelis, A., 2000. Pro-

versus anti-inflammatory cytokine profile in patients with severe

sepsis: a marker for prognosis and future therapeutic options. J.

Infect. Dis. 181, 176–180.

Helwigh, A.B., Lind, P., Nansen, P., 1999. Visceral larva migrans:

migratory pattern of Toxocara canis in pigs. Int. J. Parasitol. 29,

559–565.

Kayes, S.G., 1997. Human toxocariasis and the visceral larva

migrans syndrome: correlative immunopathology. Chem. Immu-

nol. 66, 99–124.

Kayes, S.G., Jones, R.E., Omholt, P.E., 1987. Use of bronchoalveo-

lar lavage to compare local pulmonary immunity with the

systemic immune response of Toxocara canis-infected mice.

Infect. Immun. 55, 2132–2136.

Kayes, S.G., Oaks, J.A., 1980. Toxocara canis: T lymphocyte

function in murine visceral larva migrans and eosinophilia onset.

Exp. Parasitol. 49, 47–55.

Kox, W.J., Volk, T., Kox, S.N., Volk, H.D., 2000. Immunomodula-

tory therapies in sepsis. Intensive Care Med. 26 (Suppl. 1),

S124–S128.

Kusama, Y., Takamoto, M., Kasahara, T., Takatsu, K., Nariuchi, H.,

Sugane, K., 1995. Mechanisms of eosinophilia in BALB/c-nu/+

and congenitally athymic BALB/c-nu/nu mice infected with

Toxocara canis. Immunology 84, 461–468.

Lampinen, M., Carlson, M., Hakansson, L.D., Venge, P., 2004.

Cytokine-regulated accumulation of eosinophils in inflamma-

tory disease. Allergy 59, 793–805.

Meeusen, E.N., Balic, A., 2000. Do eosinophils have a role in the

killing of helminth parasites? Parasitol. Today 16, 95–101.

Mendes, D., Camargo, M., Aun, V., Fernandes, M., Aun, W., Mello,

J., 2000. Eosinophilia. Rev. Bras. Alerg. Immunopathol. 23, 84–

91.

Nishimoto, N., Kishimoto, T., 2004. Inhibition of IL-6 for the

treatment of inflammatory diseases. Curr. Opin. Pharmacol. 4,

386–391.

Okada, K., Fujimoto, K., Kubo, K., Sekiguchi,M., Sugane, K., 1996.

Eosinophil chemotactic activity in bronchoalveolar lavage fluid

obtained from Toxocara canis-infected rats. Clin. Immunol.

Immunopathol. 78, 256–262.

Ovington, K.S., Behm, C.A., 1997. The enigmatic eosinophil:

investigation of the biological role of eosinophils in parasitic

helminth infection. Mem. Inst. Oswaldo Cruz 92 (Suppl. 2), 93–

104.

Parsons, J.C., Coffman, R.L., Grieve, R.B., 1993. Antibody to

interleukin 5 prevents blood and tissue eosinophilia but not liver

trapping in murine larval toxocariasis. Parasite Immunol. 15,

501–508.

Parsons, J.C., Grieve, R.B., 1990. Kinetics of liver trapping of

infective larvae in murine toxocariasis. J. Parasitol. 76, 529–

536.

Piergili Fioretti, D., Moretti, A., Mughetti, L., Bruschi, F., 1989.

Eosinophilia, granuloma formation, migratory behaviour of

second stage larvae in murine Toxocara canis infection. Effect

of the inoculum size. Parassitologia 31, 153–166.

Roig, J., Romeu, J., Riera, C., Texido, A., Domingo, C., Morera, J.,

1992. Acute eosinophilic pneumonia due to toxocariasis with

bronchoalveolar lavage findings. Chest 102, 294–296.

Rothenberg, M.E., 1998. Eosinophilia. N. Engl. J. Med. 338, 1592–

1600.

Schaffer, S.W., Dimayuga, E.R., Kayes, S.G., 1992. Development

and characterization of a model of eosinophil-mediated cardi-

N.R. Pecinali et al. / Veterinary Parasitology 134 (2005) 121–130130

omyopathy in rats infected with Toxocara canis. Am. J. Physiol.

262, H1428–H1434.

Schantz, P.M., Glickman, L.T., 1983. Ascaridos de Perros y Gatos:

un problema de salud publica y de medicina veterinaria. Bol.

Sanit. Panam. 94, 571–585.

Sugane, K., Oshima, T., 1984. Interrelationship of eosinophilia and

IgE antibody production to larval ES antigen in Toxocara canis

infected mice. Parasite Immunol. 6, 409–420.

Takamoto, M., Kusama, Y., Takatsu, K., Nariuchi, H., Sugane, K.,

1995. Occurrence of interleukin-5 production by CD4–CD8-

(double-negative) T cells in lungs of both normal and congeni-

tally athymic nude mice infected with Toxocara canis. Immu-

nology 85, 285–291.

Takamoto, M., Sugane, K., 1993. Mechanisms of eosinophilia in

Toxocara canis infected mice: in vitro production of interleukin

5 by lung cells of both normal and congenitally athymic nude

mice. Parasite Immunol. 15, 493–500.

Teran, L.M., 2000. CCL chemokines and asthma. Immunol. Today

21, 235–242.

Viola, J.P., Rao, A., 1999. Molecular regulation of cytokine gene

expression during the immune response. J. Clin. Immunol. 19,

98–108.

Woerly, G., Roger, N., Loiseau, S., Capron, M., 1999. Expression of

Th1 and Th2 immunoregulatory cytokines by human eosino-

phils. Int. Arch. Allergy Immunol. 118, 95–97.