Desferrioxamine modulates chemically induced T helper 2-mediated autoimmunity in the rat

Abstract

A rise in interleukin (IL) 4-dependent immunoglobulin E (IgE) is a hallmark of the mercuric chloride (HgCl₂)-induced Th2-mediated autoimmune syndrome in the Brown Norway (BN) rat, and one of the mediators in allergic asthma in human. Oxidative stress, a potential factor related to the pathogenesis of allergy and asthma, has been shown to up-regulate IL-4 in mast cells and predispose to degranulation in vitro. However, it remains unknown whether oxidative/antioxidative imbalance plays a role in this Th2-driven model of autoimmunity in the rat. Here we show that administration of the non-sulphydryl-containing antioxidant desferrioxamine i.p. and s.c. to BN rats reduces HgCl₂-enhanced IL-4 gene expression and inhibits HgCl₂-induced Th2-mediated autoimmunity. Desferrioxamine treatment suppresses significantly IgE production and lymphoproliferation, and reduces tissue injury in the form of caecal vasculitis in the HgCl₂-induced autoimmune syndrome. These results support a role for oxidative stress in the pathogenesis of the HgCl₂-induced Th2-dominated autoimmune syndrome. This finding might have implications for understanding the mechanisms involved in Th2 cell responses as seen in allergy and asthma and thereby aid the development of new therapeutic strategies for these diseases.

INTRODUCTION

Immunoglobulin E (IgE) is a key antibody that mediates atopic or allergic conditions such as asthma, seasonal allergy and urticaria. IgE binds to receptors on the surface of effector cells, such as mast cells and basophils, and when cross-linked causes them to release bioactive mediators. In many patients, the total serum IgE level correlates roughly with disease severity. Local IgE synthesis in allergic rhinitis and asthma has been demonstrated recently [1–3].

IgE synthesis by B lymphocyte is dependent upon at least two types of signals. The first signal is delivered through cytokines produced primarily from T helper 2 (Th2) cells such as interleukin (IL)-4 and IL-13, of which IL-4 is the most important cytokine required for isotype switching to IgE and for the development of a Th2 cell response [4]. The second signal is provided by interaction of the CD40 antigen expressed on the B cell surface with its ligand (CD154) expressed on activated T cells [5]. Elevated concentrations of IgE in allergic individuals may result from the preferential activation of Th2 cells. For managing the allergic response, different immunotheraputic strategies have been developed to inhibit IgE production including anti-IgE monoclonal antibodies, IL-4 antibodies, IL-4 receptor antibodies, soluble IL-4 receptors and interferon (IFN)-γ[6].

We have been investigating an animal model of autoimmunity that is dominated by a Th2 cell type response. Administration of chemical compounds such as mercuric chloride (HgCl₂), gold salts or d-penicillamine to the Brown Norway (BN) rat induces an autoimmune syndrome characterized by the production of autoantibodies [7], a marked increase in total serum IgE concentration, generalized lymphoproliferation [8] and tissue injury in the form of a necrotizing vasculitis and inflammatory polyarthritis [9]. The induction of a necrotizing vasculitis and antineutrophil cytoplasm antibody (ANCA) of the antimyeloperoxidase (MPO) type in the rats is analogous to human systemic vasculitis [7,10]; in particular, the elevated IgE concentration provides similarities to the Churg–Strauss syndrome [11]. The syndrome in the rat has been shown to be T cell-dependent [12]. The rise in serum IgE implicates Th2 cells, as isotype switching to IgE is dependent on IL-4, a major cytokine product of Th2 cells [13].

Although the Th2-dominated autoimmune syndrome peaks about 2 weeks after the administration of HgCl₂ in the BN rat, significant vasculitis with elevated levels of IL-4 mRNA occurs in the caecum within 24–48 h of an injection of HgCl₂, a time at which there is no significant T cell infiltrate [14]. Our work has demonstrated that mast cells play a role in this early phase of vasculitis [15].

We have also found that the compounds that cause Th2 cell-driven autoimmunity in vivo generate reactive oxygen species (ROS) and enhance IL-4 expression in mast cells in vitro, that oxidative stress by H₂O₂ mimics the effect of HgCl₂ in enhancing IL-4 expression, and that antioxidants (sulphydryl and non-sulphydryl-containing) diminish the HgCl₂-induced enhancement of IL-4 [16]. These data suggest a mechanism of IL-4 transcriptional regulation by oxidative stress. We therefore propose the hypothesis that up-regulation of IL-4 and sensitization for mediator release in mast cells, caused by production of intracellular ROS, contributes to HgCl₂-induced autoimmunity. Our recent data have shown that administration of antioxidants to BN rats inhibits HgCl₂-induced early vasculitis in vivo, indicating that oxidative stress plays a role in the pathogenesis of HgCl₂-induced mast cell-mediated early vasculitis [14]. However, it is known that T cells are not involved in this early phase of vasculitis [15]. An important separate question, therefore, is whether the later Th2 cell-dependent aspects of the syndrome would also be susceptible to antioxidant regulation, either because they are induced by the earlier mast cell production of IL-4, and/or because Th2 cell activation can be modulated directly by antioxidant treatment.

The aim of this study was to extend and test further our hypothesis by determining whether the administration of the non-sulphydryl-containing antioxidant desferrioxamine to BN rats could inhibit the later HgCl₂-induced Th2-dominated autoimmune response, and to examine whether modulation of oxidative/antioxidative balance influences IL-4 expression in vivo.

MATERIALS AND METHODS

Animals

Male BN rats (250–400 g) were bred and maintained in the Biological Research Facilities of St George's Hospital Medical School (London, UK) or obtained from Harlan-Olac (Bicester, Oxon, UK), and used in aged-matched groups. All procedures were conducted in accordance with the Home Office Animal (Scientific Procedures) Act of 1986.

Experimental protocols

BN rats received HgCl₂ (Sigma, Poole, UK) s.c. at a dose of 1 mg/kg on 1–3 alternate days starting on day 0. Desferrioxamine (Ciba, Horsham, UK) was dissolved in distilled water and was injected i.p. or i.p. and s.c. at a dose of 400 mg/kg each injection (when given i.p. and s.c., 800 mg/kg in total) for a total of five doses given on consecutive days from day 0. Blood samples were taken from tail vessels at various time-points, and serum was stored at −20°C for measurement of IgE concentration. Animals were killed and scored for macroscopic evidence of vasculitis [17] at day 14 after the start of HgCl₂ by an experienced observer (D. B. G. O.) blinded to treatment group. Biopsies were taken from caecal tissues.

In another set of experiments, the rats were killed on day 4 after two doses of HgCl₂, with or without desferrioxamine, and spleen cells were prepared and stored in guanidinium thiocyanate solution at −80°C for extraction of RNA.

IgE enzyme-linked immunosorbent assay (ELISA)

Total IgE was measured by ELISA [9]. Briefly, 96-well plates (Dynex Technologies, Billingshurst, UK) were coated with monoclonal antirat IgE heavy chain (Serotec, Oxford, UK), diluted to 5 µg/ml in carbonate buffer. After blocking unoccupied binding sites with 5% skimmed milk in phosphate buffered saline (PBS), known concentrations of rat IgE κ myeloma protein (Serotec) or serum samples were added in duplicate to coated wells and singly to anti-IgE free wells. Binding was detected with alkaline phosphatase-conjugated antirat κ and λ light chain antibodies followed by p-nitrophenyl phosphate substrate (Sigma). The optical density (OD) at 405 nm was read using a Dynatech multiplate reader (Dynex Technologies). A standard curve from the OD of the known concentrations of IgE on each plate was used to calculate the IgE concentration from mean OD of the unknown test samples.

Reverse transcription-polymerase chain reaction (RT-PCR)

RNA was prepared by a method adapted from Chomczynski and Sacci [18]. cDNA was synthesized from about 5 µg of total RNA using a cDNA cycle kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's recommendations. For PCR, cDNA was made up with 25 pmol of each PCR primer, 200 µm concentration of each deoxynucleoside triphosphate (Amersham Pharmacia Biotech, Little Chalfont, UK), 1× reaction buffer and 5 U Taq polymerase (Appligene-Oncor, Ilkirch, France) for β-actin amplification or 7·5 U ampliTaq Gold (PE Applied Biosystems, Foster City, CA, USA) with 1 mm MgCl₂ for IL-4 amplification. The forward and reverse primers are as follows: β-actin 5′-ATGCCATC CTGCGTCTGGACCTGGC-3′ and 5′-AGCATTTGCGGTGC ACGATGGAGGG-3′; IL-4 5′-TGATGGGTCTCAGCCCCCA CCTTGC-3′ and 5′-CTTTCAGTGTTGTGAGCGTGGACTC-3′; IFN-γ 5′-ATGAGTGCTACACGCCGCGTCTTGG-3′ and 5′-GAGTTCATTGACAGCTTTGTGCTGG-3′. The amount of starting material (cDNA) and the number of PCR cycles were selected by pilot experiments and samples of the PCR reactions were taken at multiple points throughout the amplification process for appropriate quantification. PCR was performed with 20–40 cycles of 94°C (1 min), 60°C (2 min) and 72°C (3 min). Pre-PCR incubation at 95°C for 10 min was performed for amplification of IL-4 cDNA.

The PCR products were separated in 1·5% agarose gels containing ethidium bromide and photographed. Densitometric analysis of the bands was carried out using the Gel Imager Program (Alpha Innotech Co., San Leandro, CA, USA). Results are expressed as the ratio of the intensity of the band for IL-4 or IFN-γ to the intensity of the band for β-actin. The specificity of the PCR products was checked further by DNA sequencing using the ABI Prism Dye Terminator Cycle Sequencing Ready Reaction Kit and ABI 373 A DNA sequencer (Applied Biosystems).

Histology

Directed biopsies from rat caecal mucosa were fixed in buffered formalin, and paraffin-embedded sections were stained with haematoxylin and eosin. Scores were assigned by an observer unaware of treatment group (S. D. J. H.) according to the presence and severity of vasculitis (0, normal; 1, grade I; 2, grade II; 3, grade III; 4 grade IV) [19].

Statistics

For comparisons between groups over time, repeated-measures analysis of variance (anova) was used on log-transformed data for IgE concentrations. Pearson's r was calculated to evaluate the relationship between IgE production and spleen weight. All other comparisons were made by the non-parametric Mann–Whitney U-test.

RESULTS

Desferrioxamine suppresses IgE production in HgCl₂-induced autoimmunity

To determine whether the non-sulphydryl-containing antioxidant desferrioxamine inhibits the elevation of IgE caused by HgCl₂ treatment, groups of rats were given different doses of HgCl₂ with or without desferrioxamine, or desferrioxamine only. The serum total IgE was measured. Administration of a single dose of HgCl₂ to BN rats resulted in a marked elevation of serum IgE concentration (Fig. 1a). The level of IgE started to rise after day 6, and then rose significantly after day 9 (P = 0·0003). Injection of desferrioxamine (i.p.) for 5 days decreased the elevated concentration of IgE stimulated by HgCl₂, although these differences did not achieve statistical significance (P = 0·077, repeated-measures anova). BN rats treated with two or three doses of HgCl₂ (Fig. 1b,c, respectively) again showed increased IgE concentrations, but in this case the elevated levels of IgE were abolished by the administration of an increased dose of desferrioxamine (i.p. and s.c.) (P = 0·001, P < 0·001, respectively). No significant changes in IgE concentration were observed in rats given five doses of desferrioxamine (i.p. and s.c.) alone (Fig. 1c).

Fig. 1.

Suppression of IgE production by desferrioxamine in HgCl₂-induced autoimmunity. Serum IgE concentration (mean ± s.e.) is shown for groups of BN rats treated with (•) HgCl₂+ saline (▴) HgCl₂+ desferrioxamine, or (◊) saline + desferrioxamine. Rats were injected with (a) one dose of HgCl₂ with (n = 6) or without desferrioxamine (n = 6) (i.p.) or (b) two doses of HgCl₂ with (n = 5) or without desferrioxamine (n = 5) (i.p. and s.c.) or (c) three doses of HgCl₂ with (n = 10) or without desferrioxamine (n = 12) (i.p. and s.c.), or saline with desferrioxamine (n = 6) (i.p. and s.c.).

Desferrioxamine inhibits lymphoproliferation in HgCl₂-treated BN rats

The spleen enlarges considerably following HgCl₂ treatment, with a marked increase in B cells and germinal centre formation [19]. Both spleen weight and the spleen index (spleen weight/rat weight) were recorded to quantify lymphoproliferation. Results in Table 1 show that treatment of the rats with three doses of HgCl₂ significantly increase spleen weight and its index (P = 0·0007, P = 0·0001, respectively). Animals given both HgCl₂ and desferrioxamine showed a decrease in spleen weight and spleen index (Table 1). There was a significant difference, either in spleen weight or spleen index, between the HgCl₂+ saline and the HgCl₂+ desferrioxamine groups (P-values ranged between 0·009 and lower than 0·0001, Mann–Whitney U-test, two-tailed). The current data indicate that desferrioxamine inhibits lymphoproliferation in BN rats treated by HgCl₂. There was a significant correlation in individual rats between final IgE concentration and spleen weight (r = 0·804, P < 0·0001, n = 28).

Table 1.

Effect of desferrioxamine on spleen weight and spleen index in HgCl₂-induced autoimmunity^A

HgCl2 treatment	HgCl2+ saline	HgCl2+ desf.	Saline + saline
Spleen weight (g, mean ± s.d.)
2 doses	0·916 ± 0·078 (n = 5)	0·634 ± 0·069 (n = 5)B	n.d.
3 doses	0·935 ± 0·296 (n = 12)	0·640 ± 0·133 (n = 10)C	0·563 + 0·045 (n = 6)D
Spleen weight/rat weight (mg/g, mean ± s.d.)
2 doses	2·738 ± 0·099	2·010 ± 0·200E	n.d.
3 doses	2·988 ± 0·479	2·030 ± 0·188F	1·667 ± 0·075G

^AStatistics represent two-tailed Mann–Whitney U-test.

^BP = 0·009 versus HgCl₂+ saline.

^CP < 0·0001 versus HgCl₂+ saline.

^DP = 0·0007 versus HgCl₂+ saline.

^EP = 0·009 versus HgCl₂+ saline.

^FP < 0·0001 versus HgCl₂+ saline.

^GP = 0·0001 versus HgCl₂+ saline.

Desferrioxamine influences T cell-dependent HgCl₂-induced caecal vasculitis in BN rats

BN rats treated with HgCl₂ develop a mucosal vasculitis primarily affecting the caecum. The late vasculitis (in contrast to early mast cell-mediated vasculitis, known to be inhibited by desferrioxamine [14]) is T cell-dependent [15]. To assess whether desferrioxamine treatment could reduce the late phase of tissue injury in HgCl₂-induced autoimmunity, both microscopic and macroscopic caecal vasculitis at day 14 after start of HgCl₂ were examined. Microscopic caecal vasculitis scores were recorded by examination of biopsies either from affected areas of the caecum or from random biopsies in animals with no macroscopic evidence of vasculitis. In BN rats treated with two doses of HgCl₂, four of five rats developed microscopic vasculitis (Fig. 2a). By contrast, none of the animals treated with two doses of HgCl₂ and desferrioxamine showed microscopic vasculitis. There was a statistically significant difference between the groups (P = 0·014, Mann–Whitney U-test, two-tailed). Similarly, the microscopic vasculitis scores in rats treated with three doses of HgCl₂ and desferrioxamine were lower than those of rats given HgCl₂ and saline (Fig. 2b) (P = 0·0373).

Fig. 2.

Suppression of caecal vasculitis by desferrioxamine in HgCl₂-induced autoimmunity. Microscopic caecal vasculitis scores for individual rats and median scores are shown for (a) BN rats treated with two doses of HgCl₂ with or without desferrioxamine or (b) BN rats treated with three doses of HgCl₂ with or without desferrioxamine.

In agreement with the above results, two of five rats receiving two injections of HgCl₂ developed macroscopic caecal vasculitis (scored as 2 and 5, respectively), while none of five rats given two injections of HgCl₂ and desferrioxamine had macroscopic vasculitis. Treatment of the rats with desferrioxamine decreased the incidence and total macroscopic vasculitis scores induced by three doses of HgCl₂: median score HgCl₂+ saline = 0·5; HgCl₂+ desferrioxamine = 0; P = 0·039 (Mann–Whitney U-test, one-tailed). Taken together, these results indicate that the antioxidant desferrioxamine can reduce HgCl₂-induced late vasculitis in BN rats in vivo.

Desferrioxamine modulates splenic cytokine gene expression in HgCl₂-induced autoimmunity

It is known that there are reciprocal regulatory interaction between Th1 and Th2 cells, with IL-4 inhibiting Th1 cells and IFN-γinhibiting Th2 cells [20]. However, recent work has revealed that IL-4 secreted by CD4⁺ T cells can play a role in the development of a Th1 immune response against an infectious pathogen [21]. RT-PCR analysis for IFN-γ and IL-4 mRNA were performed. The data shown in Fig. 3 are pooled from two identical independent experiments. As reported by others [22], administration of HgCl₂ to BN rats leads to up-regulation of splenic IFN-γ expression (Fig. 3a); desferrioxamine treatment causes a non-significant inhibition of the up-regulation of IFN-γ mRNA by HgCl₂ (Fig. 3a). A significant increase of IL-4 gene expression in splenocytes was noted on day 4 following two injections of HgCl₂ (Fig. 3b). In rats treated additionally with desferrioxamine, the increase in the amount of IL-4 mRNA transcript was decreased (P = 0·0315, Mann–Whitney U-test, two-tailed).

Fig. 3.

Modulation of splenic cytokine mRNA expression in HgCl₂-induced autoimmunity. BN rats were treated with two doses of HgCl₂ in the presence or absence of desferrioxamine for 4 days and then spleen cells were prepared for mRNA and RT-PCR. Data were pooled from two identical independent experiments with a total of six animals in each group, and expressed as the ratio of the intensity of the band for IFN-γ (a) or IL-4 (b) to the intensity of the band for β-action. *P < 0·05 compared with HgCl₂ group (rats treated with HgCl₂ only).

DISCUSSION

The data presented in this study demonstrate that the antioxidant desferrioxamine can inhibit lymphoproliferation, IgE production and late caecal vasculitis in HgCl₂-induced autoimmunity in BN rats. We have also shown that desferrioxamine decreases HgCl₂-enhanced IL-4 gene expression in splenocytes in vivo, although this conclusion should be interpreted with some caution in view of the fact that we only measured IL-4 mRNA with a semiquantitative method. This study therefore provides evidence to support our hypothesis that oxidative stress plays a role in the up-regulation of IL-4 in vivo and contributes to the HgCl₂-induced Th2-driven autoimmune syndrome. The importance of this observation is that it extends our work on the in vivo regulation of mast cells to Th2 cells; whether this is due to a direct effect of antioxidants on Th2 cells, or an indirect effect via the inhibition of mast cell driven activation, remains to be determined.

One study found that only sulphydryl-containing antioxidants decreased IL-4 production and IgE synthesis [23]. However, desferrioxamine is an iron chelator which inhibits iron-catalysed hydroxyl radical formation via the Fenton reaction [24]. We have shown previously that desferrioxamine, which has affinity for trivalent cations, does not chelate Hg⁺⁺[25]. Desferrioxamine has been shown to improve cardiac endothelial dysfunction caused by ROS production [26], and to restore dilation of the coronary microcirculation and a normal matching between myocardial metabolic demand and coronary blood flow in type II diabetic patients [27]. Our previous studies demonstrated that desferrioxamine decreased HgCl₂-induced up-regulation of IL-4 mRNA expression in mast cells in vitro and inhibited HgCl₂-induced early vasculitis in this animal model [14,16]. The rise in serum IgE is a hallmark of the HgCl₂-induced autoimmunity in the experimental model. After even a single injection of HgCl₂, elevated serum IgE levels were seen in all rats with a 10–1000-fold increase (data not shown). The data showing that desferrioxamine treatment results in a significant decrease in HgCl₂-enhanced serum IgE suggests an inhibitory effect on the production of IgE in vivo, putatively via its antioxidative effect.

IL-4 is a critical cytokine required for development of Th2 responses and isotype switching to IgE. A recent study on knock-out mice has revealed that IL-4 alone can induce Th2 responses, even in the combined absence of IL-5, IL-9 and IL-13 [4]. It has been shown that IL-4 reduces the expression of inhibitory receptors on B cells, abolishes B cell suppression by CD22 and CD32 and enhances B cell activation [28]. We have found previously that oxidative stress mimicked the effect of HgCl₂ on IL-4 expression by mast cells, while antioxidants inhibited the HgCl₂-enhanced IL-4 expression in mast cells and spleen cells in vitro[16]. In addition, antioxidants can also inhibit HgCl₂-induced early vasculitis in BN rats, with a trend to a decrease in HgCl₂-enhanced IL-4 expression in mast cells in vivo[14]. Our present data have shown that desferrioxamine suppresses HgCl₂-induced Th2-mediated autoimmunity with a decrease in HgCl₂-induced IL-4 expression in spleen cells in vivo, consistent with the hypothesis that the antioxidant desferrioxamine modulates HgCl₂-induced Th2 cell responses via inhibition of IL-4 production in vivo.

We have found previously a more significant increase of IL-4 mRNA expression in the spleen of HgCl₂-treated BN rats [29]. However, this was not as obvious in the present study, and decreased only moderately by the addition of desferrioxamine. This could be explained by our experimental procedures. To minimize use of animals, a single time-point (4 days after administration of HgCl₂) was chosen in the present experiments based on published time-course data [29]. Methods to detect human and mouse IL-4 are well established, whereas there has been a lack of sensitive assays for the measurement of rat IL-4 protein. We have been unable to detect IL-4 protein levels in vivo by intracellular staining/flow cytometry and IL-4 ELISA (BD Biosciences) due to high background staining and low sensitivity, respectively.

Although IL-4 is likely to be a major cytokine modulating the HgCl₂-induced Th2-dominated autoimmunity, the initial cellular source of IL-4 required for instigation of Th2 cell development remains to be determined. We have reported previously that an antioxidant inhibits HgCl₂-enhanced IL-4 expression moderately in the mast cells in the caecum 2 days after injection of HgCl₂, a time at which there is no obvious lymphocyte infiltration or IL-4⁺ T cells present in the tissues [14]. A T cell infiltrate was evident in the caecum and lung 5 days after injection of HgCl₂[19]. Mast cells can release prestored IL-4 within minutes of stimulation and synthesize more hours after the stimulation [30], whereas IL-4 gene expression in T cells is controlled more strictly than in mast cells [31]. It is possible that mast cells could be activated and provide an initial source of IL-4 which could induce or expand T cells for an effective Th2 response.

In conclusion, our results demonstrate that the antioxidant desferrioxamine can inhibit HgCl₂-induced Th2-mediated autoimmunity in BN rats, suggesting that oxidative stress plays a role in the pathogenesis of the HgCl₂-induced Th2-driven autoimmune syndrome. This finding might have implications for understanding the processes involved in Th2 autoimmune responses such as allergy and asthma, and thereby facilitate the development of new therapeutic strategies for these diseases.