Systemic sclerosis (SSc) is a generalized connective tissue disorder of unknown etiology characterized by vascular lesions, immune dysregulation, and an increased deposition of extracellular matrix in the skin and in internal organs. These 3 pathogenetic processes are thought to be closely related, since endothelium, mononuclear cells, and fibroblasts interact through direct contact and via cytokines, which finally leads to fibrosis (1, 2). The initiating event that starts the pathogenetic cascade remains to be elucidated. The search for the etiology requires animal models.
University of California at Davis (UCD) line 200 chickens spontaneously develop the entire spectrum of SSc, i.e., vascular lesions, mononuclear infiltrates, fibrosis of skin and internal organs, as well as serologic abnormalities, including anti–endothelial cell antibodies (AECAs), antinuclear antibodies, anticardiolipin antibodies, and rheumatoid factors (3–6). This inherited disease begins at 1–2 weeks of age with an inflammatory stage, occurring most prominently in the comb, which becomes erythematous and swollen and is finally lost to necrosis (so-called “self-dubbing”). These early skin lesions, which resemble those of Raynaud's phenomenon in humans, proceed to a chronic stage that is characterized by excessive collagen deposition 3–4 weeks after the chicks hatch. At 6 months of age, most of the animals have internal organ involvement (7). Microvascular injury appears to be one of the earliest features of both the human and avian disease (8, 6). Recently, we demonstrated that endothelial cell apoptosis is the primary pathogenetic event underlying the occurrence of skin lesions in avian and human scleroderma, becoming apparent before any other alteration is observed (9). The programmed cell death of the endothelium appears to be induced by AECAs, as revealed by anti-immunoglobulin staining.
Autoantibodies directed against the vascular endothelium have been found to be a common serologic feature in several diseases characterized by immune-mediated damage of the vessels (10). In SSc, the proportion of AECA-positive sera ranges from 28% to 71%, depending on patient selection criteria and the laboratory method used (11). AECAs represent a heterogeneous family of autoantibodies that react with a variety of antigens on endothelial and other cells (10). AECAs have been reported to fix complement in some autoimmune diseases, such as systemic lupus erythematosus, but they fail to exhibit this effect in SSc (12). Clinically, AECAs are associated with digital infarcts and pulmonary arterial hypertension, indicating a correlation of these antibodies with the extent of the vascular involvement (13). In vitro assays have revealed that AECAs are capable not only of inducing the expression of adhesion molecules and sustaining leukocyte adhesion in SSc (14), but also of initiating apoptosis, when the endothelium was additionally exposed to mononuclear cells (15). Using human dermal microvascular endothelial cells (HDMECs) as substrate, AECAs were shown to produce apoptosis by antibody-dependent cell-mediated cytotoxicity (ADCC) in vitro (16). In contrast to these in vitro studies, there are no in vivo data concerning the cytotoxicity of AECAs in human SSc.
Based on our previous in vitro findings, the aim of the present study was to determine if transfer of AECAs into healthy chickens induces endothelial cell apoptosis and results in the development of scleroderma-like features. To consider a possible role of ADCC, additional groups of animals received activated peripheral blood lymphocytes (PBLs). We chose 2 different approaches for the transfer of AECAs: the chorionallantoic membrane (CAM) assay and the intravenous (IV) injection of AECA-positive sera into normal chicken embryos. Our findings are presented herein.
- Top of page
- MATERIALS AND METHODS
Vascular alterations in SSc, as verified by (ultra)microscopic changes and increased levels of soluble endothelial cell markers, including thrombomodulin, endothelin, or vWF, represent a very early marker in the course of disease (8, 23–25). Several in vitro studies suggest a possible role of AECAs in mediating endothelial cell damage (15, 16, 26–28), but evidence for the in vivo effects of these autoantibodies is lacking. We therefore sought to determine whether transfer of AECA-positive serum samples into healthy recipient chickens led to endothelial cell apoptosis and/or induction of SSc-like symptoms. UCD-200 chickens represented the appropriate animal model required for the acquisition of AECA-positive serum samples. Sera were transferred onto the CAMs of healthy breeds of chickens on ED 10 or, for a different approach, were injected intravenously into 13-day-old normal CC chicken embryos.
CAMs from UCD-058, NWL, and CC chicken embryos were obtained on ED 16 for microscopic evaluation. Immunofluorescence double staining revealed the presence of AECA only after transfer of UCD-200 sera, which were placed predominantly in the area of small-sized vessels. We selected the microvessels because of microscopic studies localizing the early vascular injury of SSc to the small arterioles and capillaries (8). In general, endothelial cells show tissue-dependent characteristics with regard to the expression of surface molecules, the production of prostaglandins, and in vitro growth requirements (29–31). These differences are reflected by the heterogeneity of AECAs that react with various structures on endothelial cells and fibroblasts (32). Preliminary data from studies of SSc in humans indicate a significantly higher binding activity of AECAs to the microvasculature than to the large arteries and veins (11, 16). We demonstrated that sera from all patients with early acute SSc contained IgG AECAs to HDMECs, whereas only 50% of these patients were also positive for IgG AECAs to human umbilical vein endothelial cells (HUVECs) (16). Interestingly, apoptosis induction by AECAs via ADCC was shown only in HDMECs.
Not all CAM sections examined revealed antibodies bound in vivo, since this effect depends on the topography of vascularization and the distribution of AECAs over the CAM surface after application of 0.1 ml of serum. The 0.1-ml volume of serum dropped onto the CAM is diluted ∼70 ml (calculated from the total weight of the egg), which leads to AECA concentrations below the detection limit of the IIF technique when evaluated by fluorescence microscopy.
None of the controls showed antibodies bound to the endothelium, but Ig staining of epithelial keratin-positive structures could be detected on CAM sections from all study groups, regardless of the serum transferred. This is probably due to remaining egg yolk (containing maternal Ig) on the outside of the membrane that was not entirely eliminated during dissection.
Importantly, after application of AECA-positive sera, CAMs from all strains revealed increased numbers of apoptotic endothelial cells as compared with the controls. Mainly the microvessels were affected by the cytotoxic effects of the transferred serum samples, as was expected. Nearly all CAM sections also showed apoptosis of other cell types, which were not further characterized. The latter finding is typical for the period of embryogenesis and explains the appearance of some apoptotic endothelial cells on control CAMs. Since it is not possible to discriminate microscopically between physiologically occurring cell death and exogenously induced apoptosis, evaluations were performed statistically by comparison with controls.
There are several possible mechanisms for the induction of endothelial cell apoptosis found in our experiments. The effect of a complement-mediated cytotoxicity can be excluded, since this would lead to necrosis (33). In previous experiments, we were able to show that AECA-induced endothelial cell apoptosis in vitro is accomplished by ADCC via the Fas pathway (16).
The fact that the rate of apoptosis was not further increased by the addition of IL-2–stimulated PBLs may be explained by the availability of natural killer cells from the embryos themselves. Bordron et al (34) triggered apoptosis of HUVECs by the application of purified, highly concentrated AECAs from SSc patients. Other groups of investigators have postulated the presence of peripheral blood mononuclear cells (PBMCs) as a prerequisite for the cytotoxic effect of AECAs. Penning et al (15) showed that sera from ∼20% of SSc patients were capable of causing cytotoxicity of HUVECs when cocultured with PBMCs. The mechanism was thought to be ADCC, since the activity resided in IgG fractions and the responsible effector cells were Fc receptor positive. Marks et al (27) and Holt et al (28) reported similar results with studies of HDMECs and HUVECs, respectively. In previous experiments (16), we showed that sera alone had no effect and that apoptosis induction could be observed only with microvascular endothelial cells, but not with HUVECs, as targets. We also demonstrated that IgG AECAs from SSc patients bound to both HDMECs and HUVECs, whereas apoptosis induction occurred only with AECAs to HDMECs.
These in vitro studies indicate that ADCC might be operative in the in vivo induction of apoptosis. Mature natural killer cells have been detected in ED 14 chicken embryos (35). Since the induction of AECA-dependent cellular cytotoxicity requires just 4–20 hours in vitro (Sgonc R: unpublished observations) and since harvesting of CAMs was not performed before ED 16 (i.e., 6 days after AECA transfer), this mechanism might be responsible for mediating the endothelial cell apoptosis seen in our experiments.
Similar to the results of the CAM assays, we also found an apoptosis-inducing effect of AECAs after IV injection of UCD-200 serum samples containing IL-2–stimulated PBLs into healthy CC embryos. Results of examinations for circulating AECAs in blood samples from these chickens as well as staining for AECAs bound in vivo were negative in all animals. These findings do not, however, exclude the induction of ADCC by the transferred AECAs, since concentrations of antibodies required for this process (≤50 pg/ml) (16, 36) are clearly below the detection limit of the IIF (1–100 ng/ml) when evaluated by fluorescence microscopy. An enzyme-linked immunosorbent assay for the evaluation of chicken AECAs is not available yet.
In neither the CAM assays nor the IV transfer experiments did the application of UCD-200 serum samples result in an increase in the mortality rates as compared with the controls. IV injection of AECAs was not followed by the development of SSc-like features in CC chickens.
What is the reason for the limited effect of transferred AECAs, resulting in the induction of apoptosis but not the production of a macroscopic pattern of disease? As discussed above, we hypothesize that endothelial cell apoptosis represents only a first step in the course of disease. Comparative studies of skin biopsy tissues from UCD-200 chickens and humans with SSc identified endothelial cell apoptosis as the primary pathogenetic event in the development of SSc, occurring before any other alterations, such as the formation of perivascular infiltrates (9). Endothelial cell apoptosis is not localized to the skin; it is the first visible alteration in other affected organs of UCD-200 chickens (37). Thus, we suggest that further development of disease requires factors in addition to a single transfer of 100 μl of AECA-positive serum. Whether these prerequisites consist of constant exposure to higher concentrations of AECAs, a genetically determined susceptibility of target organs to autoimmune attack (38), or other as-yet-unknown components remains to be elucidated. The single transfer of 100 μl of AECA-positive serum, which was diluted in the intra- and extravascular space, was probably not sufficient to induce an SSc-like disease in the recipients. However, chicken embryos can hardly tolerate the IV injection of larger volumes.
Although the present experiments were performed with whole sera rather than the purified IgG fraction, our study provides the first evidence in SSc that AECA-positive sera are capable of inducing endothelial cell apoptosis in vivo. In conclusion, our findings support the hypothesis that AECAs might play an important role in triggering the cascade of pathogenesis.