Dr Dov L.Boros Department of Immunology and Microbiology, WSU School of Medicine, 540 E. Canfield Avenue, Detroit, MI 48201, USA. E-mail: firstname.lastname@example.org
In murine schistosomiasis mansoni the worm egg-induced granulomatous inflammation is bi-phasic: an initial Th1 type is subsequently switched to a Th2 type response. Analysis of the cellular, molecular base of the Th1-associated response (5–6 weeks post infection) revealed mRNA messages for interleukin (IL)-12 p40, IL-12Rβ2 and interferon (IFN)-γ in the granulomatous livers. When the Th2 type granulomas matured (8 weeks post infection) message expression weakened or became extinct. Macrophages of the Th1 type granulomas produced maximal amounts of IL-12, but production diminished in the mature granulomas. A similar pattern of IL-12 responsiveness of granuloma lymphocytes was observed. In vitro IL-12 production by Th1 type granuloma macrophages was enhanced by tumour necrosis factor (TNF)-α and IFNγ, whereas lymphocyte IL-12 responsiveness was boosted only by TNF-α. Both systems were down-regulated by IL-4 and IL-10 cytokines. Treatment of mice with anti-IL-10 monoclonal antibodies (MoAb) between 6 and 7 weeks of the infection enhanced mRNA expression for IFN-γ and IL-12Rβ2, but not for IL-12 p40. It is concluded that IL-12 and IL-12R expression and function regulate the Th1 phase of the liver granulomatous response. This phase is cross-regulated by type-2 cytokines especially IL-10.
In murine schistosomiasis mansoni the CD4+ T-helper lymphocytes initiate and maintain the granulomatous inflammatory response [ 1–3]. Accumulated evidence indicates that three T-helper subsets can participate in the generation of circumovum granulomas. Parasite eggs injected subcutaneously (s.c.) or intravenously (i.v.) stimulated an early Th0 followed by a Th2 type response in the draining lymph node of naïve mice [ 4]. In infected mice lymphocytes of the early developing liver granulomas (5–6 weeks of infection) secrete high levels of IFN-γ and minimal levels of IL-4 consonant with an induced Th1 type response [ 5]. With the maturation of the granuloma (7–8 weeks post infection) IFN-γ production drops sharply, whereas IL-4, IL-5 secretion is high, indicative of a Th2 response [ 6, 7]. A soluble egg antigen (SEA)-specific granulomagenic CD4+ Th1-type lymphocyte has been also identified [ 8]. Moreover, cytokine profiles of the mature granuloma demonstrated by mRNA expression [ 9, 10] or immunocytochemical staining [ 11] showed both type 1 and type 2 cytokine presence in the lesions and pluripotential Th0 type cells have been cloned from the mature liver granulomas [ 12]. IFN-γ has shown to down-regulate, whereas IL-4 to enhance the granulomatous response [ 13, 14]. Moreover, IL-4 deficient infected mice develop smaller granulomas with less deposited collagen [ 15–17]. Therefore, the evolution and regulation of Th-subset responsiveness has an important bearing on the intensity of the inflammatory response and subsequent fibrosis.
Recently IL-12 has been shown to be a major cytokine active in the differentiation and expansion of Th1 lymphocytes [ 18, 19]. Message for the p40 chain of IL-12 and cytokine production have been found during the early growth of egg or SEA-coated bead granulomas in the lungs [ 20, 21]. Repeated injections of rIL-12 to egg primed infected mice [ 22] or animals that received combined rIL-12 and anti Th2 type antibody treatment which prolonged the Th1 phase of granuloma development [ 23] significantly diminished liver-granuloma growth. Though a role for IL-12 in the granulomatous process has been documented, as yet no information is available on the dynamics of endogenous IL-12 production, regulation and IL-12R expression during the evolution of liver granulomas in infected mice. The present study examined the cellular, molecular basis of IL-12 cytokine and cytokine receptor production/expression and their relation to the Th1-Th2 cytokine switch that occurs during the evolution of liver granulomas.
MATERIALS AND METHODS
Animals and infection
Female CBA/Jk mice (Jackson Laboratories, Bar Harbor, ME, USA), 6–8-week-old, were used throughout the study. Mice were infected s.c. with 25 cercariae of a Puerto Rican strain of S. mansoni.
Hybridoma clones for αIL-12 (C15.1 and C15.6 kindly provided by G. Trinchieri, Wistar Institute, Philadelphia, PA, USA) [ 24], αIFN-γ (R4 6A2), αIL-10 (JES2A5), αThy 1.2 (30-H12), and αDNP were grown in complete medium: RPMI-1640 with 10% fetal bovine serum (FBS) (Gibco/BRL, Grand Island, NY, USA), 1% glutamine-penicillin-streptomycin, 20 m m HEPES, 2 m m sodium pyruvate, 50 μm 2-mercaptoethanol (Sigma Chemical Company, St. Louis, MO, USA). For ascites fluid production, CBA/Jk mice were primed with 0.2 ml Freund's incomplete adjuvant i.p., irradiated with 500 rads on day 3, and injected i.p. with 5 × 106 hybridoma cells on day 4. Ascites fluid was collected 3–7 days later and purified by ammonium–sulfate precipitation and dialysis against PBS. Antibodies used in culture or ELISA were further purified by thiophilic gel affinity chromatography (Pierce, Rockford, IL, USA) and sterile filtered. Purity was assessed by SDS-PAGE. Low-Tox rabbit complement was purchased from Accurate Chemical (Westbury, NY, USA).
SEA was prepared from homogenized S. mansoni eggs harvested from mice infected i.p. with 200 cercariae, as previously described [ 25].
Granuloma cells were isolated from dispersed lesions as previously described [ 2]. Macrophages were removed by adherence to plastic tissue culture plates during a 90-minute incubation in complete medium at 37 °C, 5% CO2. Nonadherent cells were washed, suspended in fresh complete medium, and enriched for T cells with nylon–wool columns. Lymphocyte subsets were depleted by incubating aliquots of cells with the appropriate antibody for 45 min at 4 °C, followed by 60 min at 37 °C with complement. Cells were washed at least twice between treatments and suspended at 3 × 106/ml with 20 μg/ml SEA for 48 h at 37 °C, 5% CO2 to generate culture supernatants. Adherent cells were harvested by incubation in cell-dissociation solution (Sigma) for 10 min, followed by vigorous washing with Hank's balanced salt solution (HBSS). Residual cells were removed with a cell scraper, combined with the washes, and suspended in complete medium. Cells were cultured with 1 μg/ml lipopolysaccharide (LPS) (055:B5, Sigma) for 24 h at 37 °C, 5% CO2 to generate culture supernatants. Adherent cells consisted of 70% macrophages by cytospin analysis, the remainder were neutrophils, with a few eosinophils and fibroblasts. For each experiment, single cell granuloma suspensions were prepared from at least three mice. Cell viability was at least 95% as determined by trypan blue exclusion.
Recombinant murine cytokines were purchased from the following sources: IFN-γ and TNF-α, Genentech (San Francisco, CA, USA); IL-4, Peprotech (Rocky Hills, NJ, USA); IL-10, (Sigma); IL-12, R & D Systems (Minneapolis, MN, USA).
IL-12 bioactivity was measured by the induction of IFN-γ [ 26]. Macrophage supernatants and IL-12 standards were cultured with normal splenocytes (2 × 106/ml) and phytohemagglutinin (PHA) (5 ng/ml) in 96-well plates in duplicate. Specificity was verified by the inclusion of control wells containing αIL-12 (C15.1 and C15.6) or αDNP MoAbs. Supernatants were harvested at 48 h and tested for IFN-γ by ELISA. An IL-12 capture assay was used to measure cytokine effects on IL-12 production. Prior to culture, assay plates were incubated at 4 °C overnight with anti IL-12 MoAb (C15.1; 100 μg/ml in 0.1 m carbonate buffer, pH 9.6, 50 μl/well), washed thrice with HBSS, blocked for 1 h with complete medium at 37 °C, and again washed thrice with HBSS. Plates were incubated at 4 °C overnight with test-macrophage supernatants, then washed five times with HBSS before culture of normal splenocytes as described above.
IL-12 receptor function
IFN-γ production in response to IL-12 was used to determine IL-12 receptor display [ 27]. Enriched granuloma T cells (∼70% pure) were cultured at 2 × 106/ml with 20 μg/ml SEA, and 5 n m rIL-12. Controls included PHA stimulation (5 ng/ml), IL-12 specificity (inclusion of anti IL-12 MoAbs cocktail at 10 ng/ml) and T-cell dependence (depletion with Thy 1.2 and complement). Cytokine effects on IL-12 responsiveness were analyzed by preculturing purified granuloma T cells with 5 μg/ml PHA and 1.0, 0.1 or 0.01 ng/ml cytokine for 48 h. Cells were then harvested, washed, counted and cultured with SEA and rIL-12. Supernatants were collected after 48 h and analyzed for IFN-γ content.
IFN-γ was measured by ELISA using R4–6A2 as the coating antibody, biotinylated XMG1.2 (Pharmingen, San Diego, CA, USA) as the detecting antibody, and streptavidin-alkaline phosphatase. Assays were developed with nitrophenyl diamine diethanolamine and optical densities read at 405 nm. Standards were used for each assay with dilutions of rIFN-γ.
In vivo treatments
Mice were injected i.p. with 500 μg anti IL-10 or anti dinitrophenyl (DNP) (isotype matched control) MoAbs three times weekly between 6 and 7 or 6–8 weeks post infection (days 42,44,46, and 48, or additionally days 50, 52, and 54). At the end of the test period (day 49 or 56), livers were excised and snap frozen for total RNA extraction.
Approximately 100 mg of liver from each animal was homogenized in 2 ml of TrIzol (Gibco/BRL) and total RNA was isolated according to the manufacturer's suggested protocol. RNA samples were washed twice in 80% ethanol-20% diethyl pyrocarbonated (DEPC) treated distilled H2O, briefly air-dried, and dissolved in 100% formamide. Concentration and purity (A260/A280 ≥ 1.9) were determined spectrophotometrically and portions were set at 1 mg/ml. All samples were stored at − 70 °C until analyzed. Reverse transcription of 1 μg total RNA was performed in a 30-μl reaction volume using oligo dT primer and M-MLV enzyme (Gibco/BRL). The reaction was diluted to 100 μl final volume with sterile distilled H2O upon completion. PCR primers were synthesized according to published sequences by DNAgency (Malvern, PA, USA). After cDNA standardization by cyclophilin, samples were tested for IFN-γ, IL-12Rβ1, IL-12Rβ2, IL-12 p35, and IL-12 p40. All PCR reactions were run in 30 μl volumes (5 μl cDNA, 5 μl primer mix, and 20 μl PCR Supermix (Gibco/BRL) for the indicated number of cycles (94 °C, 40 s; Ta, 20 s; 72 °C, 40 s) with increased initial denaturation (94 °C, 3 min) and final extension (72 °C, 7 min) times. Final MgCl2 concentration was adjusted to 1.5 m m. Primer sequences are shown in Table 1. All reactions were in the linear range of amplification. Results were visualized by electrophoresis on 1% agarose gel with ethidium–bromide staining. Stained bands were measured by densitometry. Densitometry values were compiled as a percentage of the cyclophilin value. Statistics were calculated on groups of six mice using the arcsine transformation for percentage data and compared using a one-tailed Student's t-test.
Table 1. Primer sequences and conditions used for the RT-PCR assay *Numbers in square brackets refer to quotations in References.
Differences between experimental and control groups were analyzed by Student's t-test; P < 0.05 was deemed significant.
Dynamics of cytokine and cytokine-receptor expression
Conditions for the RT-PCR assays are shown in Table 1.
Figure 1 shows the dynamics of expression of cytokine and cytokine-receptor messages in the livers of mice during the evolution of the granulomatous response. Message for IFN-γ was detectable already before egg deposition (4 weeks) and increased in strength throughout the development of the granulomas (5–7 weeks). However, by week 8 when granulomas attained maturity and assumed the preponderant Th2 phenotype, the IFN-γ signal became considerably weaker. At week 4 of the infection before egg deposition started messages for the p35 and p40 chains of IL-12 were only sporadically expressed. After egg deposition the mRNA expression for the p40 chain became stronger, peaked during the period of early granuloma development (6–7 weeks) but became weak by week 8 of the infection coincident with the maturation of the Th2 type granuloma. The expression of up-regulated mRNA for the IL-12Rβ2 chain relevant to display by Th1 cells, was weak at 4 weeks and strong at 6 weeks. The signal weakened by 7 weeks and disappeared by week 8 of the infection.
IL-12 production by granuloma macrophages and its regulation
As a corollary to message expression we explored granuloma macrophage function during the evolution of hepatic granulomas. As seen in Fig. 2(A), unstimulated macrophages produced minimal amounts of the cytokine. Peak production of IL-12 in LPS-stimulated granuloma macrophage cultures occurred at 6 weeks when granulomas were about 1 week old (P < 0.05 compared with unstimulated control). In the ensuing weeks the cytokine production in LPS-stimulated monolayers diminished sharply and was minimal by 8 weeks of the infection.
Furthermore, it was of interest to examine the cytokines that regulate IL-12 production. As Fig. 2(B) illustrates macrophages isolated from 6 weeks infection Th1 type granulomas upon stimulus with optimal doses (1.0 ng/ml) of IFN-γ produced enhanced amounts of IL-12. The TNF-α-mediated stimulation was weaker. In contrast, macrophages cultured with IL-4 or IL-10 showed suppressed IL-12 production, the latter being significantly more efficient (P < 0.05) versus IL-4 suppressor cytokine. Macrophages from 8 week infection granulomas upon joint IFN-γ–LPS stimuli showed enhanced IL-12 production, whereas TNF-α stimulus was ineffective. Monolayers cultured with anti-IL-10 but not with anti-DNP MoAb also showed some enhanced cytokine production indicating that IL-10 actively suppressed production of the cytokine ( Fig. 2C).
IL-12 responsiveness of granuloma lymphocytes
Subsequently, a correlation between diminished focal production of IL-12 and IL-12 responsiveness of granuloma lymphocytes, was sought. Responsiveness was assessed indirectly by the assay of IFN-γ production of IL-12 stimulated cells [ 27]. At 6 weeks of the infection SEA-stimulated nonadherent granuloma cells (∼70% lymphocyte content) produced low amounts of IFN-γ. This may be attributed to the effect of focally produced IL-10 [ 28]. Stimulus by PHA alone could not enhance IL-12 responsiveness. In the presence of rIL-12 but without SEA, IFN-γ production was enhanced ( Fig. 3A). However, when the antigenic stimulus and rIL-12 were combined a synergistic effect was observed and T cells showed greatly enhanced responsiveness expressed by elevated IFN-γ levels. Similar response was seen in the PHA-stimulated cultures ( Fig. 3B). In contrast, cells procured from the mature Th2 type granulomas (8 weeks of infection) did not respond to SEA, rIL-12 or PHA stimuli but produced a low but significant (four-fold) enhancement in IL-12 responsiveness to the combined SEA and rIL-12 or PHA and rIL-12 stimuli ( Fig. 3A,B hatched bars). Because IL-12 responsiveness to SEA and rIL-12 stimuli was abrogated by added anti-IL-12 MoAb whereas control anti-DNP–MoAb was ineffective it appears that responsiveness expressed by IFN-γ production was boosted by exogenous rIL-12. Finally, deletion of Thy1.2+ cells from the cell cultures abrogated responsiveness to rIL-12 indicating that Th cells responded to the cytokine mediated stimulus ( Fig. 3C). At 6 weeks of the infection granuloma lymphocytes precultured with IFN-γ or TNF-α showed no enhanced IFN-γ production to the former, but increased production to the latter cytokine. Similar cultures exposed to 1.0 ng/ml IL-4 and IL-10 cytokines showed diminished (∼50%) responsiveness to rIL-12 ( Fig. 3D).
The effect of neutralized IL-10 action on the restoration of the Th1 response
Based on this and previous in vivo [ 28] observations we examined the role in vivo of IL-10 on the expression of the Th1 response during granuloma development. As seen in Fig. 4(A) anti-IL-10 MoAb treatment given between 6 and 7 weeks, corresponding to the early development of the liver granuloma significantly enhanced the expression of IFN-γ and IL-12Rβ2 messages ( Table 2). However, treatment did not boost the expression of the message for the p40 chain of IL-12. A more sustained treatment with anti-IL-10MoAb that continued until the mature Th2 stage of the granuloma (6–8 weeks) significantly enhanced only IFN-γ message expression. Again, no effect was seen on the expression of message for the p40 chain of IL-12 ( Fig. 4B, Table 2).
Table 2. Densitometry* of bands shown in Fig. 4 * Band density is expressed as a ratio compared with the cyclophilin message which is taken as 100%; †Significant P < 0.05.
The present study analyzed the cellular and molecular processes during the switch from Th1 to Th2 responsiveness in the egg-induced hepatic granuloma development of S. mansoni-infected mice [ 5–7]. At 6 weeks of the infection the evolving granulomas are composed mainly of macrophages and the lesional lymphocytes produce IFN-γ consistent with a Th1 type of response [ 5]. As a corollary now we found at 6 weeks in granulomatous livers maximal expression of mRNA message for the inducible p40 chain of IL-12 and strong IL-12 production by granuloma macrophages. Because the IL-12 signal is necessary for Th1 cell maturation/differentiation [ 18, 19] the intralesional production of IL-12 confirms the Th1 stage of early hepatic granuloma development. This is also supported by the presence of the strong mRNA message for IFN-γ in the granulomatous liver tissue. This observation agrees with previous ones that showed peak IL-12 p40 message expression or IL-12 production during the early stage of schistosome egg or SEA-bead-induced lung granuloma development [ 20, 21]. In vitro experiments presented here demonstrated that at 6 weeks macrophages of the early hepatic granuloma stimulated by LPS produced maximal amounts of IL-12. In subsequent weeks the production sharply declined. The discrepancy at 7 weeks between message expression for IL-12 p40 in liver homogenates and lack of IL-12 production by granuloma macrophages may be explained by message production by hepatic Kupffer cells. IL-12 production could be enhanced by IFN-γ and to a lesser degree by TNF-α. However, the responsiveness varied according to the age of the lesion. Whereas the IFN-γ-enhanced cytokine production by macrophages derived from both the evolving (at 6 weeks) and mature (at 8 weeks) granulomas, the TNF-α was effective only in the enhancement of the former but not the latter mature granuloma macrophages. This is consistent with previous observations that showed the early involvement of TNF-α in the generation of the liver granulomas [ 29] even in the absence of functional T cells [ 30]. Whether unresponsiveness to TNF-α indicates a change in receptor display/function in macrophages of the mature granulomas needs to be further examined. The up-regulation of the IL-12 production by IFN-γ and TNF-α has been similarly shown in cultures of peritoneal [ 21] splenic [ 31] bone marrow-derived [ 32] macrophage cultures.
The early Th1 stage of granuloma development and IFN-γ production lasts for about 1 week [ 5, 23], and from the 7th week of the infection it switches over to increased production of IL-4, IL-5, IL-10 indicative of the Th2 stage of granuloma growth [ 6, 7, 23]. Therefore in the present experiment it was intriguing to find that the expression of the IL-12 message as well as the IL-12 production by granuloma macrophages sharply decreased from the 6 week infection point and onwards. That this down-regulation is attributable to focally released type 2 cytokine action is supported by data showing suppressed IL-12 production by granuloma macrophages cocultured with IL-10 cytokine ( Fig. 2B). These data complement previous observations that showed IL-10-mediated suppression of IFN-γ production and/or granuloma formation [ 33, 34] in IL-10, anti-IL-10 MoAb treated, or IL-10 deficient mice [ 35]. It is noteworthy, that in the 8 week infected mice the strongly curtailed IL-12 production by granuloma macrophages could be boosted by IFN-γ and anti-IL-10Ab indicating that cytokine production appeared to be under the control of endogenous IL-10 ( Fig. 2C).
A corollary for the switch in Th subset activity is shown here in the changed responsiveness to IL-12 stimulus. At the evolving Th1 stage of the hepatic granuloma (6 weeks) strong messages were expressed in the liver for IFN-γ, IL-12 p40 and the β2 chain of IL-12R. A similar correlation was reported recently in message expression during primary granuloma growth in the lungs of egg-injected mice [ 36]. Moreover SEA-stimulated granuloma lymphocytes responded strongly to rIL-12 indicating the presence of high affinity IL-12R on their surface [ 28]. It is of note, that responsiveness to rIL-12 of granuloma lymphocytes could be enhanced in vitro by TNF-α rather than IFN-γ. In fact, higher concentrations of added IFN-γ appeared to be inhibitory. This emphasizes the complexity in the interaction of granuloma lymphocytes with intralesional cytokine signals. Upon lesion maturation when the predominant cytokine milieu changed to the Th2 type, message expression for IL-12Rβ2 disappeared and responsiveness to IL-12 sharply diminished. The β2 chain of IL-12R is necessary for IL-12 signalling and is expressed on Th1 but not Th2 cells [ 37]. Thus the extinction of the message for the β2 chain and the concomitant sharp reduction in responsiveness to IL-12 results in the cessation of Th1-cell differentiation and curtailed IFN-γ production with concomitant enlargement of the Th2 CD4+ cell compartment within the mature granuloma. Detection of the IL-12Rβ1 message in the liver of 8 week infected mice may be indicative of the presence of Th2 cells [ 19]. That elevated Th2 cell activity and cytokine production could cross-regulate IL-12R expression and IL-12 responsiveness is shown in experiments where down-regulation of receptor activity by both IL-4 and IL-10 cytokines could be shown for lymphocytes of the evolving (6 week) granuloma ( Fig. 3).
Based on data presented here, one can construct the events that initiate and maintain the two stages of granuloma development. In the liver macrophages that converge around eggs lodged in the presinusoidal capillaries are likely to participate in the early inflammatory response. Macrophages are activated by innate or adaptive immune stimuli (IFN-γ and TNF-α) to secrete IL-12. Antigen-activated CD4+ lymphocytes attracted to the site of incipient inflammation display the β2 chain of IL-12R, become receptive for the IL-12 signal and undergo maturation/differentiation into Th1 type effector cells that produce high levels of IFN-γ. This comprises the predominantly Th1 stage of early granuloma development [ 5]. This stage is down-regulated by the production of IL-10 within the granulomas [ 34]. With the appearance of additional Th2 cells, eosinophils, mast cells, etc., IL-4 and IL-10 cytokines are focally secreted [ 38, 39]. As we show in this study IL-10 down-regulates IL-12 message expression/production, IFN-γ as well as IL-12Rβ2 expression and thus participates in the switch from Th1 to the Th2 phase of hepatic granuloma growth. This entails the development of the large eosinophil-rich florid granulomas with enhanced deposition of extracellular matrix and collagen. This Th2 phase is not permanently fixed, because experimental manipulations such as neutralization of Th2 cytokine activity with concomitant exogenous rIL-12 treatment [ 23] could generate a Th1 response presumably from the existing pool of Th0 lymphocytes [ 12]. The present experiments indicated that such in vivo manipulation is successful as shown by the partial restoration of IL-12Rβ2 expression and IL-12 responsiveness of the T lymphocytes, and appearance of IFN-γ message.
In sum, the foregoing experiments demonstrate that focal IL-12 production and IL-12R2β-expressing Th1 cells are the major factors involved at the early stage of the hepatic granulomatous inflammatory process in infected mice. Regulation by type 2 cytokines especially IL-10 [ 40], of IL-12 and IL-12R transcription shifts the response to Th2 cell activity that influences both the intensity of inflammation and the pathology of the disease.
This work was supported by Public Health Service grant AI-12913 from the National Institute of Allergy and Infectious Diseases.
Schistosome life stages or materials for this work were supplied through NIH-NIAID contract N01-AI55270.