7) 1228–1235www.elsevier.com/locate/lifescie

Life Sciences 80 (200

Hepatic granulomas induced by Schistosoma mansoni in mice deficient forconnexin 43 present lower cell proliferation and higher collagen content

Silvia Catarina Salgado Oloris a, Marc Mesnil b, Viviane Neri de Souza Reis a, Mônica Sakai c,Patrícia Matsuzaki a, Evelise de Souza Monteiro Fonseca c, Tereza Cristina da Silva a,

José Luís Avanzo a, Idércio Luiz Sinhorini a, José Luiz Guerra a, Frederico Azevedo Costa-Pinto a,Paulo Cesar Maiorka a, Maria Lúcia Zaidan Dagli a,⁎

a Laboratory of Experimental Oncology, Department of Pathology, School of Veterinary Medicine, University of São Paulo, Brazilb Institute of Cellular Physiology and Biology, CNRS-UMR 6187, University of Poitiers, France

c Laboratory of Applied Pharmacology, Department of Pathology, School of Veterinary Medicine, University of São Paulo, Brazil

Received 26 June 2006; accepted 15 December 2006

Abstract

Granuloma formation involves a coordinated interaction between monocytes and macrophages, epithelioid cells, lymphocytes, eosinophils,neutrophils and fibroblasts. It has been established that extracellular communication via cytokines is important for the assembly of granulomas.However, the importance of gap junctions and intercellular communication to granuloma formation and development had never been assessed.Connexins are proteins that form gap junctions, and connexin 43 (Cx43) is present in macrophages, lymphoid cells, myelogenous cells, fibroblastsand others.We analyzed the effect of heterologous deletion of Gja1 (Cx43 gene) on the formation and development of hepatic granulomas induced bySchistosoma mansoni eggs. Heterozygous (Cx43+/−) and wild-type (Cx43+/+) mice were infected subcutaneously with S. mansoni cercarie andevaluated after 6, 8 and 12 weeks. Granuloma cells express Cx43, as revealed by real-time PCR in isolated granulomas, and byimmunohistochemistry. Cx43 expression was reduced in Cx43+/− mice, as expected. No differences in the average area of granulomas or numberof cells per granuloma were observed between mice of different genotypes. However, granuloma cells from Cx43+/− mice displayed a reduced indexof the proliferating cell nuclear antigen (PCNA) labeling at 8 and 12 weeks post-infection. Moreover, Cx43+/− granulomas unexpectedly presented ahigher degree of fibrosis, quantified by morphometric analysis in Sirius Red-stained slides. Our results indicate that the deletion of one allele of theCx43 gene, and possibly the reduced gap junction intercellular communication capacity (GJIC), may impair the interactions between granulomacells, reducing their proliferation and increasing their collagen content, thereby modifying the characteristics of S. mansoni granuloma in mice.© 2007 Elsevier Inc. All rights reserved.

Keywords: Gap junction; Connexin; Granuloma; Schistosoma mansoni; Fibrosis; Cell proliferation

Introduction

Granulomas are lesions that occur during chronic inflam-mation in response to various stimuli, leading to the isolationof harmful substances from the surrounding healthy tissues(Jinnouchi et al., 2003). The major histological feature of a

⁎ Corresponding author. Faculdade de Medicina Veterinária e Zootecnia,Departamento de Patologia, Av. Prof. Dr. Orlando Marques de Paiva, 87,Universidade de São Paulo, São Paulo, SP, CEP 05508-900, Brazil. Tel.: +55 113091 7712; fax: +55 11 3091 78 29.

E-mail address: mlzdagli@usp.br (M.L.Z. Dagli).

0024-3205/$ - see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.lfs.2006.12.030

granuloma is the nodular accumulation of macrophages,epithelioid cells and multinucleate giant cells. Additionally,lymphocytes, plasma cells and polymorphonuclear leukocytesmay be present (Jinnouchi et al., 2003). The formation ofgranulomatous lesions is a complex dynamic process thatleads to the inflammatory destruction or confinement of theinciting agent. This process involves recruitment of bloodmonocytes to the inflammatory sites, followed by coordinatedinteractions among lymphocytes, monocytes and macro-phages, epithelioid cells, eosinophils, neutrophils and fibro-blasts (Takahashi et al., 1999). It is known that extracellularcommunication, mediated by cytokines, is important to

Table 1Quantification of hepatic granulomas in Cx43+/− and Cx43+/+ mice 6, 8, and 12 weeks a after S. mansoni infection

Week Granuloma area (×104 μm2) b Granuloma relative area (%) c Granuloma cells/area ×10−2 d

Cx43+/+ Cx43+/− Cx43+/+ Cx43+/− Cx43+/+ Cx43+/−

6 5.05±6.14(n=6 e) 7.11±4.92(n=5) 0.29±0.6(n=6) 0.38±0.40(n=5) NA f NA8 8.89±3.28(n=8) 9.37±1.39(n=6) 3.73±3.04(n=8) 3.29±1.32(n=6) 1.13±0.11 1.08±0.1012 7.89±2.89(n=8) 7.56±1.74(n=9) 3.89±2.44(n=8) 3.98±2.41(n=9) 1.22±0.13 1.15±0.08

Data are mean±SD.a Cx43+/+ and Cx43+/− mice were infected subcutaneously with 30 cercaries of S. mansoni.b Mean size of lesions was quantified in H&E stained tissue sections.c Granuloma area/total area of slice ×100.d Number of granuloma cells per area of 6 randomly selected granulomas.e n=Number of animals.f NA = Not assessed.

1229S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

granuloma formation. However, whether direct communica-tion between cells is important to granuloma formation anddevelopment is, as of yet, unknown.

Multicellular organisms are characterized by metaboliccooperation between cells, with extensive intercellular commu-nication networks. Gap junction intercellular communication(GJIC) is intimately involved in the maintenance of tissuehomeostasis, normal cell growth, differentiation and development(Loewenstein, 1979), by its ability to regulate cell-to-cellexchange of small hydrophilic molecules and ions, less than1KD, including carbohydrates, Ca2+, cAMP and inositoltriphosphate (Simpson et al., 1977; Flagg-Newton et al., 1979).A gap junction channel is formed by a pair of hemichannelsnamed connexons. Each connexon is formed by 6 connexinmolecules. Connexins belong to a multigene family of at least 20members that share a basic structure: a cytosolicN- andC-termini,four transmembrane domains, two extracellular loops and onecytoplasmic loop (Goodenough et al., 1996; Kumar and Gilula,1996, Willecke et al., 2002). Intercellular communication via gapjunctions is relevant to the immune system, during hematopoiesis(Cancelas et al., 2000; Krenacs and Rosendaal, 1998; Montecino-Rodriguez et al., 2000), thymocyte maturation (Alves et al., 1995;Fonseca et al., 2004), and activation of monocytes andmacrophages (Eugenin et al., 2003; Jara et al., 1995).

In many diseases, coordinated interaction between cells is acrucial event in proliferation,migration, activation, differentiationand maturation of inflammatory cells. It has recently beenreported that Cx43-deficient mice present decreased developmentof atherosclerotic lesions, reduced number of atheroma-associat-ed inflammatory cells, and increased thickness of the fibrous capcovering the lesions (Kwak et al., 2003). Based on these findingswe hypothesized that GJIC may contribute to the development ofthe murine hepatic granuloma induced by Schistosoma mansonieggs. The periovular granulomatous chronic inflammatoryreaction is characterized by the changes in cell population, theprofile of released cytokines, the size of lesions and the degree offibrosis. Granulomas are reportedly larger during the early acutephase of infection and become smaller during late chronic phases(Domingo and Warren, 1968). Cx43-deficient mice (Reaumeet al., 1995) have been used in several studies as experimentalmodels of disease. These studies have clarified the functional

importance of Cx43 in physiological and pathological processes.Unfortunately, homozygous knockouts die shortly after birth,therefore heterozygous mice had to be used.

In this study, we determined the importance of connexin 43in the development of hepatic granulomas, experimentallyinduced by S. mansoni eggs, using heterozygous knockout miceof connexin 43. We present the first evidence that connexins,probably via direct cell-to-cell communication, modulategranulomatous chronic inflammatory disease.

Material and methods

Mice

Cx43+/− mice were obtained from the International Agencyfor Research on Cancer (IARC, Lyon, France). Mice weregenerated by replacing exon-2 of the Cx43 gene (Gja1) with theneomycin resistance gene (neo) (Reaume et al., 1995). Thesemice were produced originally in the C57BL/6 strain, and thebackground was changed to CD-1 by serial breeding at theIARC animal facility. Due to the perinatal lethality of Cx43knockout mice, only heterozygous (Cx43+/−) or wild-type(Cx43+/+) mice were used in this study.

Thesemice were housed under controlled conditions (22 °C±2 °C; 65±15% relative humidity; air exchange rate 15 times/h,12 h–12 h light–dark cycle), in filter-top cages, at the AnimalFacility of the Department of Pathology, School of VeterinaryMedicine, University of São Paulo. The animals received astandard pellet food (Purina Lab Chow, Curitiba, Brazil) andwater ad libitum. Animals were used in accordance to theguidelines of the Committee on Care and Use of AnimalResources of the School of Veterinary Medicine, University ofSão Paulo.

Mice genotyping by PCR

The genotyping of the Cx43+/− mice was performed by PCRanalysis of the DNA from mouse tail, as previously described(Yamakage et al., 1998). The amplicons were loaded onto anagarose gel (1.5% in TBS buffer). Typical bands of amplifiedconnexin 43 gene, Gja1 (520 bp), and neo gene (294 bp) were

Fig. 1. PCNA staining of S. mansoni-induced granulomas. (A) 8 weeks post-infection in Cx43+/+ mice, or (B) Cx43+/−mice, and (C) 12 weeks post-infection in Cx43+/+

mice, or (D) Cx43+/− mice. Arrows show PCNA-positive nuclei (objective 20×).

Fig. 2. Quantification of PCNA labeling index (LI) in S. mansoni-inducedgranulomas, 8 or 12 weeks post-infection. Data are presented as mean of PCNA-positive cells from 6 randomly selected granulomas for each mouse. Dotsrepresent individual animals, and mean values are indicated by horizontal lines;t-test, p≤0.05.

1230 S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

seen in Cx43+/− mice and only the band of Gja1 in Cx43+/+

mice.

S. mansoni infection

Eight-week-old male Cx43+/+ and Cx43+/− mice wereinfected subcutaneously with 30 cercaries of S. mansoni (BHstrain, Tropical Medicine Institute, Faculty of Medicine,University of São Paulo). Animals were weighed once weeklyuntil euthanized at 6 or 8 weeks, corresponding to the earlyacute phase, or at 12 weeks corresponding to the chronic phaseof the disease. (Borojevic et al., 1984; Oliveira et al., 2000).

Histology and morphometry of S. mansoni granulomas

Four representative fragments of liver were fixed inmethacarn for 8 h, routinely processed, embedded in paraffin,and sliced in 5 μm sections.

Histological sections of liver were stained with hematoxylinand eosin (H&E) or Sirius Red (Sigma) in saturated picric acidsolution (Andrade et al., 1999; Junqueira et al., 1979), andanalyzed in a Nikon E-800 microscope coupled with a CCDcamera (CoolSNAP-Pro color). The area of individual granu-lomas, as well as the total area of the liver sections and the areataken by granulomas per slide, was measured by computer-aided image analysis (Image Pro-Plus Media Cybernetics, Inc.-USA). The following data were thus generated: granuloma area(mean area of all granulomas in each liver section), granuloma

relative area (% represented by total granuloma area/total areaof the liver sections) and number of granuloma cells per area(total number of cells from a granuloma section/the area of therespective granuloma section) of each mouse.

Fig. 3. Collagen fibers stained with Sirius Red and observed under polarized light, within S. mansoni-induced granulomas of 8 and 12 weeks post-infection.(A) 8 weeks post-infection in Cx43+/+ mice, or (B) Cx43+/− mice, and (C) 12 weeks post-infection in Cx43+/+ mice, or (D) Cx43+/− mice. Note the increased collagenbirefringence and density of fibers in Cx43+/− mice, compared to those in Cx43+/+ mice (objective 10×).

Fig. 4. Quantification of collagen fibers stained with Sirius Red in S. mansoni-induced granulomas of 8 and 12 weeks post-infection, performed automaticallyby the software package Image Pro® Plus. Values are presented as mean area ofcollagen fibers stained with Sirius Red per granuloma area. Dots representindividual animals, and mean values are indicated by horizontal lines; Kruskal–Wallis test, p≤0.05.

1231S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

Quantification of S-phase granuloma cells positive for PCNAimmunostaining

Five granulomas were randomly selected from liverhistological sections of each animal for the quantification ofS-phase cells; 5 animals per genotype were evaluated. Primaryantibodies against PCNA were employed (1:800, overnightincubation, DAKO Corporation, CA), followed by streptavi-dine–biotin–peroxidase complex and DAB staining. All thecells of each granuloma were counted, as PCNA-positive ornegative cells. Data is presented as percentage of PCNA-positive among the total number of cells.

Quantification of collagen in granulomas

The quantification of collagen deposition in the granulomaswas performed using histological sections of liver stained withSirius Red (Andrade et al., 1999; Junqueira et al., 1979), andanalyzed under polarized light in a Nikon E-800 microscope(Tokyo, Japan). Total area of stained collagen fibers in thegranuloma was automatically quantified by image analysis. Thedata were represented as mean percentage area of red collagenfibers in the granuloma.

Connexin 43 immunostaining in S. mansoni hepatic granulomas

Immunohistochemistry was performed by incubating liversections with a polyclonal anti-Cx43 antibody (1:100, Zymed

Laboratories, CA, USA), overnight at 4 °C, followed by asecondary anti-rabbit biotinylated antibody (1:100, DAKOCorporation, CA, USA) for 2 h at room temperature. Signal

1232 S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

amplification was performed using the TSA Kit (Perkin Elmer,MA, USA) according to the manufacturer protocol. Nuclei werecounterstained with propidium iodide (1:1000, Sigma). Sectionswere analyzed under a Nikon E-800 fluorescence microscope(Tokyo, Japan).

Real-time TaqMan RT-PCR analysis

Isolated granulomas were obtained by mechanical disruptionof livers, and subsequent separation by repeated sedimentationin cold PBS (Brener and Pellegrino, 1956). Total RNA ofisolated granulomas was extracted with TRIzol Reagent(Invitrogen Life Technologies, Carlsbad, CA) according to themanufacturer protocol. The quality of RNA samples wasdetermined by electrophoresis in agarose gels and stained with

Fig. 5. Expression of Cx43 in granuloma cells. (A) 6 weeks post-infection Cx43+

(D) Cx43+/− mice; and (E) 12 weeks post-infection Cx43+/+ mice, or (F) Cx43+/− mwere stained with propidium iodide (objective 20×).

ethidium bromide; 18S and 28S bands were visualized underUV light.

Real-time PCR was performed as previously described(Avanzo et al., 2004). Briefly, total RNA was extracted fromisolated hepatic granulomas from four mice, 8 weeks afterinfection. 3 μg of total RNA were reverse-transcribed to yieldcDNA, and Gja1 expression was analyzed by real-timequantitative reverse transcriptase-PCR procedure using anABI PRISM 7000 sequence detection system (AppliedBiosystems, Foster City, CA). GAPDH was used as an internalcontrol. All primers and probes were purchased from AppliedBiosystems.

The end-point of real-time PCR analysis (the threshold cycle,or CT) is determined from a log-linear plot of the PCR signalversus the cycle number.

/+ mice, or (B) Cx43+/− mice; (C) 8 weeks post-infection in Cx43+/+ mice, orice. Arrows show connexin staining; arrowheads show S. mansoni eggs. Nuclei

1233S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

The Cx43 gene expression was normalized to GAPDH andcalculated by comparison of Cx43 gene expression in wild-typemice that was assigned a value of 1. The final analysis of geneexpression was performed according to the 2−ΔΔct method(Livak and Schmittgen, 2001).

Statistical analysis

Differences obtained from wild-type and heterozygous micewere assessed using the Student's t-test, or the Mann Whitneytest for non-parametric data, with post-hoc analysis. Compar-ison of morphometry results was performed using the Kruskal–Wallis non-parametric test, or the two-way analysis of variance(ANOVA), considering time and genotype as the sources ofvariation. All data are reported as mean±SD and the level ofsignificance considered was p≤0.05.

Results

Histopathological analysis and morphometry of S. mansonigranulomas

Histopathological analysis revealed that the constituent cellsof granulomas were predominantly mononuclear, with only afew polymorphonuclear granulocytes (eosinophils). Animals ofboth genotypes presented dense collagen fiber deposition in theextracellular matrix of the granulomas 8 and 12 weeks post-infection, particularly in Cx43+/− mice infected for 12 weeks.

Hepatic granulomas induced by S. mansoni eggs weresimilar in mean size, relative size and cellularity (Table 1), inboth genotypes. Statistical analysis via two-way ANOVAindicated that time of infection affected these results( p≤0.05). No differences in the weight gain of mice, therelative weight of spleen and thymus, or in mortality rate wereobserved in either genotype (data not shown).

Cx43+/− mice had lower in situ proliferation of granulomacells at late phases

To evaluate possible differences in the proliferation index ofgranuloma cells in Cx43+/+ or Cx43+/− mice, we performedimmunohistochemistry to detect PCNA-positive cells in 8- and12-week infected mouse groups (Fig. 1). Statistical analysis in6-week infected mice was not done due to the low incidence oflesions. However, the proliferation index of granuloma cellswas significantly lower in Cx43+/− mice at 8 and 12 weeks post-infection ( p≤0.05) (Fig. 2).

Granulomas from Cx43+/− mice had higher collagen content12 weeks post-infection

As shown in Fig. 3, 12-week infected Cx43+/− micepresented a higher degree of fibrotic areas within granulomas.Lesions from 12-week infected mice (Fig. 3C and D) were morefibrotic than those from 8-week infected mice (Fig. 3A and B).Accordingly, Sirius Red-stained collagen fibers were morepronounced in 12-week infected Cx43+/− mice ( p≤0.05)

(Fig. 4). 6-week infected mice presented thinner collagenfibers, and therefore were unsuitable for quantification (data notshown).

Analysis of Cx43 expression in granulomas

Granulomas isolated after 8 weeks of infection in mice fromboth genotypes expressed Cx43. The normalized Cx43expression in granulomas of Cx43+/+ and Cx43+/− mice was 1and 0.41, respectively. Therefore, Cx43 gene expression inCx43+/− was approximately 40% of that found in wild-typemice.

Cx43 protein was visualized by immunofluorescence andappeared as plaques inserted in the membranes of granulomacells. Cx43+/− mice (Fig. 5B, D, and F) had lower expression ofCx43 protein than Cx43+/+ mice (Fig. 5A, C, and E) at all timepoints. Cx43 protein expression was higher at late phases of thedisease, especially in Cx43+/+ mice 12 weeks after infectionwhen granulomas had an increased density of junction plaques(Fig. 5E).

Discussion

Cx43 is known to be widely distributed in the immunesystem and in bone marrow cells (Oviedo-Orta and HowardEvans, 2004; Wong et al., 2004). Macrophages are known toexpress Cx43, and this expression is associated with activationby lipopolysaccharide (LPS) (Jara et al., 1995). However, theinfluence of cell communication through gap junctions tomacrophage function is still controversial. Therefore, we used agenetically modified mouse, heterozygous to Cx43 (Cx43+/−),as a model to characterize the biological involvement of gapjunction proteins during the assembly of granulomas bymacrophages and other cell types.

Gap junction intercellular communication (GJIC) is inti-mately involved in the maintenance of tissue homeostasis(Loewenstein, 1979), and may contribute to establish granulo-ma microenvironment that mediates proliferation and differen-tiation of myeloid cells in the liver parenchyma and inperiovular granulomas in chronic murine schistosomiasis(Alvarez-Silva et al., 1993).

In our study we detected the expression of Cx43 in isolatedS. mansoni hepatic granulomas. Cx43 gene (Gja1) expression inlesions from Cx43+/− mice was reduced by 50%, confirmingprevious results of connexin expression in these mice (Avanzoet al., 2004). Data from immunohistochemistry show that Cx43forms membranous junction plaques in granuloma cells. Ourreal-time RT-PCR data show reduced expression of Cx43+/−

mice confirmed by lower levels of Cx43 protein and reduceddensity of gap junction plaques by immunohistochemisty.

To evaluate the kinetics of granuloma formation, weperformed morphometry in lesions at three different time points(6, 8 and 12 weeks after infection), corresponding to the earlyand late acute phases, and to the early chronic phase of thisprocess, respectively (Dutra et al., 1998). The dynamics ofgranuloma formation depends on the assembly of various typesof cells, their in situ proliferation, and on the generation of

1234 S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

connective tissue matrix. The granuloma mean size is a finalconsequence of the interplay between these elements. In ourexperimental design, the morphometric analysis of granulomasrevealed no differences in mean absolute and relative area, whencomparing mice of different genotypes. Cellularity of granulo-mas was also similar among animals of different genotypes.More robust results would probably be observed in a complete(homozygous) Cx43 knockout mouse, which unfortunately isunavailable. Nonetheless, important differences were detectedregarding proliferation rate of granuloma cells and degree offibrosis.

Extramedullary proliferation of macrophages in the tissuereaction to schistosomes was described previously (Stavitsky,2004; Clark et al., 1988; Lenzi et al., 1995), and might accountfor the in situ accumulation of macrophages in the absence ofovert increased proliferation and release from the bone marrow.As a general rule, based on several observations in neoplasticprocess, the reduction in GJIC capacity has been associated withincreased cell proliferation (Solan et al., 2003). Surprisingly, wefound the cell proliferation index to be reduced in granulomasfrom Cx43+/− mice 8 or 12 weeks post-infection whencompared to Cx43+/+ animals. Thus, it seems feasible that alower GJIC capacity may impair the transmission of signals thatwould otherwise modulate cell proliferation. For example,Cx32 knockout mice and Cx32 transgenic mice (Dagli et al.,2004) have delayed liver regeneration after partial hepatectomysuggesting that gap junctions are responsible by coordinatingcell growth by acting both positively and negatively to achieve anormal growth rate. (Dagli et al., 2004).

The development of hepatic granulomas is also characterizedby progressive organization of the extracellular matrix, with asignificant increase in collagen birefringence approachingchronic phase (Andrade et al., 1999). Indeed, Cx43+/− micepresent a higher level of collagen deposition in late phases ofgranuloma organization. The morbidity and mortality associat-ed with murine schistosomiasis is highly attributable to thefibrosis in affected tissues (Wynn et al., 2004). Type-2cytokines, notably IL-13, are central to the development ofhepatic fibrosis in mice, which cause blood flow impairmentand increases portal blood pressure (Chiaramonte et al., 1999a,b; Jankovic et al., 1999). Therefore, deficient cell proliferationin Cx43+/− mice might lead to a stronger induction of fibrotictissue, as an attempt to isolate S. mansoni eggs from the healthysurrounding liver. Corroborating this concept, a recent studyreported that Cx43+/− mice show decreased development ofatherosclerotic lesions, and increased thickness of the fibrouscap covering such lesions (Kwak et al., 2003). Likewise,previous studies (Qiu et al., 2003) demonstrated that Cx43knockdown reduces inflammation and accelerates the rate ofskin wound repair.

The granulomas developed in Cx43+/+ mice present moreproliferating cells and less deposition of connective tissueelements, compared to those of Cx43+/−. This could reflect anideal microenvironment, maintained at least in part by anetwork of gap junction communication, that contributes tocalcium homeostasis and dissipation of oxidative stress,therefore preventing cell death (Blanc et al., 1998). Moreover,

connexin proteins are active players in cell survival, indepen-dent of gap junction channel function (Lin et al., 2003).

Altogether, our results suggest that granuloma cells areconnected via gap junctions formed by Cx43. We conclude thatconnexin expression and gap junction formations are importantto the development and fate of granuloma in mice. We showhere that Cx43 deficiency decreases the proliferation ofgranuloma cells and therefore deposition of collagen.

Acknowledgements

Wewould like to thankMrs Cristina Conceição and Dr. PedroPaulo Chieffi (Instituto de Medicina Tropical, Faculdade deMedicina, Universidade de São Paulo, Brazil) for providing usS. mansoni cercarie, and Beth Ann Tamburini for the revisionof this article. This research is part of the Ph.D. thesis presentedby S.C.S. Oloris to the Program of Experimental andComparative Pathology, of the School of Veterinary Medicine,University of São Paulo. This work was supported by grant 02/08436-8 from the Fundação de Amparo a Pesquisa do Estado deSão Paulo (FAPESP). S.C.S. Oloris was supported by a studentfellowship from FAPESP (01/08608-0). This work is part of theInternational Cooperation Brazil/France Program CAPES/COFECUB (386/02/04).

References

Alvarez-Silva, M., da Silva, L.C., Borojevic, R., 1993. Cell membrane-associated proteoglycans mediate extramedullar myeloid proliferation ingranulomatous inflammatory reactions to schistosome eggs. Journal of CellScience 104 (Pt 2), 477–484.

Alves, L.A., Campos de Carvalho, A.C., Cirne Lima, E.O., Rocha e Souza,C.M., Dardenne, M., Spray, D.C., Savino, W., 1995. Functional gapjunctions in thymic epithelial cells are formed by connexin 43. EuropeanJournal of Immunology 25 (2), 431–437.

Andrade, G.B., Montes, G.S., Conceicao, G.M., Saldiva, P.H., 1999. Use of thePicrosirius-polarization method to age fibrotic lesions in the hepaticgranulomas produced in experimental murine schistosomiasis. Annals ofTropical Medicine and Parasitology 93 (3), 265–272.

Avanzo, J.L., Mesnil, M., Hernandez-Blazquez, F.J., Mackowiak, I.I., Mori, C.M.,da Silva, T.C., Oloris, S.C., Garate, A.P.,Massironi, S.M., Yamasaki, H., Dagli,M.L., 2004. Increased susceptibility to urethane-induced lung tumors in micewith decreased expression of connexin43.Carcinogenesis 25 (10), 1973–1982.

Blanc, E.M., Bruce-Keller, A.J., Mattson, M.P., 1998. Astrocytic gap junctionalcommunication decreases neuronal vulnerability to oxidative stress-induceddisruption of Ca2+ homeostasis and cell death. Journal of Neurochemistry 70(3), 958–970.

Borojevic, R., Nicola, M.H., Santos da Silva, C., 1984. Modulation ofmacrophage and polymorphonuclear granulocyte inflammatory reaction inexperimental murine schistosomiasis mansoni. Cellular and MolecularBiology 30 (1), 37–42.

Brener, Z., Pellegrino, J., 1956. Method for isolating schistosome granulomasfrom mouse liver. Journal of Parasitology 42 (6), 564.

Cancelas, J.A., Koevoet, W.L., de Koning, A.E., Mayen, A.E., Rombouts, E.J.,Ploemacher, R.E., 2000. Connexin-43 gap junctions are involved inmulticonnexin-expressing stromal support of hemopoietic progenitors andstem cells. Blood 96 (2), 498–505.

Chiaramonte, M.G., Donaldson, D.D., Cheever, A.W., Wynn, T.A., 1999a. AnIL-13 inhibitor blocks the development of hepatic fibrosis during a T-helpertype 2-dominated inflammatory response. Journal of Clinical Investigation104 (6), 777–785.

Chiaramonte, M.G., Schopf, L.R., Neben, T.Y., Cheever, A.W., Donaldson, D.D.,Wynn, T.A., 1999b. IL-13 is a key regulatory cytokine for Th2 cell-mediated

1235S.C.S. Oloris et al. / Life Sciences 80 (2007) 1228–1235

pulmonary granuloma formation and IgE responses induced by Schistosomamansoni eggs. Journal of Immunology 162 (2), 920–930.

Clark, C.R., Chen, B.D., Boros, D.L., 1988. Macrophage progenitor cell andcolony-stimulating factor production during granulomatous Schistosomiasismansoni in mice. Infection and Immunity 56 (10), 2680–2685.

Dagli, M.L., Yamasaki, H., Krutovskikh, V., Omori, Y., 2004. Delayed liverregeneration and increased susceptibility to chemical hepatocarcinogenesisin transgenic mice expressing a dominant-negative mutant of connexin32only in the liver. Carcinogenesis 25 (4), 483–492.

Domingo, E.O., Warren, K.S., 1968. Endogenous desensitization: changing hostgranulomatous response to schistosome eggs at different stages of infectionwith Schistosoma mansoni. American Journal of Pathology 52 (2), 369–379.

Dutra, H.S., El-Cheikh, M.C., Azevedo, S.P., Rossi, M.I., Borojevic, R., 1998.Murine schistosomiasis mansoni: experimental analysis of bone marrow andperipheral myelopoiesis. Parasitologic Research 84 (8), 668–675.

Eugenin, E.A., Branes, M.C., Berman, J.W., Saez, J.C., 2003. TNF-alpha plusIFN-gamma induce connexin43 expression and formation of gap junctionsbetween human monocytes/macrophages that enhance physiologicalresponses. Journal of Immunology 170 (3), 1320–1328.

Flagg-Newton, J., Simpson, I., Loewenstein, W.R., 1979. Permeability of thecell-to-cell membrane channels in mammalian cell junction. Science 205(4404), 404–407.

Fonseca, P.C., Nihei, O.K., Urban-Maldonado,M., Abreu, S., de Carvalho, A.C.,Spray, D.C., Savino, W., Alves, L.A., 2004. Characterization of connexin30.3 and 43 in thymocytes. Immunology Letters 94 (1–2), 65–75.

Goodenough, D.A., Goliger, J.A., Paul, D.L., 1996. Connexins, connexons, andintercellular communication. Annual Reviews in Biochemistry 65, 475–502.

Jankovic, D., Kullberg, M.C., Noben-Trauth, N., Caspar, P., Ward, J.M.,Cheever, A.W., Paul, W.E., Sher, A., 1999. Schistosome-infected IL-4receptor knockout (KO) mice, in contrast to IL-4 KO mice, fail to developgranulomatous pathology while maintaining the same lymphokine expres-sion profile. Journal of Immunology 163 (1), 337–342.

Jara, P.I., Boric, M.P., Saez, J.C., 1995. Leukocytes express connexin 43 afteractivation with lipopolysaccharide and appear to form gap junctions withendothelial cells after ischemia–reperfusion. Proceedings of the NationalAcademy of Sciences of the United States of America 92 (15), 7011–7015.

Jinnouchi, K., Terasaki, Y., Fujiyama, S., Tomita, K., Kuziel, W.A., Maeda, N.,Takahashi, K., Takeya, M., 2003. Impaired hepatic granuloma formation inmice deficient in C–C chemokine receptor 2. Journal of Pathology 200 (3),406–416.

Junqueira, L.C., Bignolas, G., Brentani, R.R., 1979. Picrosirius staining pluspolarization microscopy, a specific method for collagen detection in tissuesections. Histochemical Journal 11 (4), 447–455.

Krenacs, T., Rosendaal, M., 1998. Connexin43 gap junctions in normal,regenerating, and cultured mouse bone marrow and in human leukemias:their possible involvement in blood formation. American Journal ofPathology 152 (4), 993–1004.

Kumar, N.M., Gilula, N.B., 1996. The gap junction communication channel.Cell 84 (3), 381–388.

Kwak, B.R., Veillard, N., Pelli, G., Mulhaupt, F., James, R.W., Chanson, M.,Mach, F., 2003. Reduced connexin43 expression inhibits atheroscleroticlesion formation in low-density lipoprotein receptor-deficient mice.Circulation 107 (7), 1033–1039.

Lenzi, H.L., Lenzi, J.A., Rosman, F.C., Pelajo-Machado, M., Mota, E.M.,Panasco, M.S., Oliveira, D.N., 1995. Extramedullary hematopoiesis inmurine schistosomiasis mansoni. Memorias do Instituto Oswaldo Cruz 90(2), 169–177.

Lin, J.H., Yang, J., Liu, S., Takano, T., Wang, X., Gao, Q., Willecke, K.,Nedergaard, M., 2003. Connexin mediates gap junction-independentresistance to cellular injury. Journal of Neuroscience 23 (2), 430–441.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression datausing real-time quantitative PCR and the 2(-Delta Delta C(T)). Methodologyand Methods 25 (4), 402–408.

Loewenstein, W.R., 1979. Junctional intercellular communication and thecontrol of growth. Biochimica et Biophysica Acta 560 (1), 1–65.

Montecino-Rodriguez, E., Leathers, H., Dorshkind, K., 2000. Expression ofconnexin 43 (Cx43) is critical for normal hematopoiesis. Blood 96 (3),917–924.

Oliveira, V.R., El-Cheikh, M.C., Aguiar, A.M., Balduino, A., de Fatima, B.P.M.,Reis, L.F., Borojevic, R., 2000. Schistosoma mansoni egg-induced hepaticgranulomas in mice deficient for the interferon-gamma receptor have alteredpopulations of macrophages, lymphocytes and connective tissue cells.Microbes and Infection 2 (15), 1817–1826.

Oviedo-Orta, E., Howard Evans, W., 2004. Gap junctions and connexin-mediated communication in the immune system. Biochimica et BiophysicaActa 1662 (1–2), 102–112.

Qiu, C., Coutinho, P., Frank, S., Franke, S., Law, L.Y., Martin, P., Green, C.R.,Becker, D.L., 2003. Targeting connexin43 expression accelerates the rate ofwound repair. Current Biology 13 (19), 1697–1703.

Reaume, A.G., de Sousa, P.A., Kulkarni, S., Langille, B.L., Zhu, D., Davies, T.C.,Juneja, S.C., Kidder, G.M., Rossant, J., 1995. Cardiac malformation inneonatal mice lacking connexin43. Science 267 (5205), 1831–1834.

Simpson, I., Rose, B., Loewenstein, W.R., 1977. Size limit of moleculespermeating the junctional membrane channels. Science 195 (4275),294–296.

Solan, J.L., Fry, M.D., TenBroek, E.M., Lampe, P.D., 2003. Connexin43phosphorylation at S368 is acute during S and G2/M and in response toprotein kinase C activation. Journal of Cell Science 116 (Pt 11), 2203–2211.

Stavitsky, A.B., 2004. Regulation of granulomatous inflammation in experi-mental models of schistosomiasis. Infection and Immunity 72 (1), 1–12.

Takahashi, K., Takeya, M., Miyakawi, K., 1999. The multiple roles ofmacrophages in hepatic granuloma formation in mice. In: Tanikawa, K.,Ueno, T. (Eds.), Liver Diseases and Hepatic Sinusoidal Cells. Springer-Verlag, Tokyo, pp. 128–140.

Willecke, K., Eiberger, J., Degen, J., Eckardt, D., Romualdi, A., Guldenagel, M.,Deutsch, U., Sohl, G., 2002. Structural and functional diversity of connexingenes in the mouse and human genome. Biological Chemistry 383 (5),725–737.

Wong, C.W., Christen, T., Kwak, B.R., 2004. Connexins in leukocytes: shuttlingmessages? Cardiovascular Research 62 (2), 357–367.

Wynn, T.A., Thompson, R.W., Cheever, A.W., Mentink-Kane, M.M., 2004.Immunopathogenesis of schistosomiasis. Immunology Reviews 201,156–167.

Yamakage, K., Omori, Y., Piccoli, C., Yamasaki, H., 1998. Growth control of3T3 fibroblast cell lines established from connexin 43-deficient mice.Molecular Carcinogenesis 23 (2), 121–128.