An Animal Model of Autoimmune Emphysema

Although cigarette smoking has been recognized as the most important factor in the development of emphysema (1, 2), the precise mechanisms that lead to the loss of alveolar structures are not well understood. Chronic inflammation and an imbalance of protease/antiprotease activities and oxidative stress are the most frequently evoked concepts used to explain the pathobiology of emphysema (3-6), yet increased numbers of T lymphocytes infiltrating the alveolar walls of patients with emphysema (7) correlate with the extent of alveolar destruction and the severity of airflow obstruction (8, 9). Although there is an increasing number of rodent models of emphysema, which implicate complex relationships among multiple gene products in the regulation of the homeostasis of alveolar septal cells (10, 11), autoimmune mechanisms have not previously been recognized to play a role in experimental emphysema models. Recently, Agusti and coworkers (12) proposed that an acquired immune response to self- or foreign antigens may be a central component of the pathogenesis of human emphysema, yet evidence in support of such a hypothesis is missing, although a descriptive study published by Birring and colleagues (13) suggests a relationship between chronic obstructive pulmonary disease in nonsmokers and organ-specific autoimmune disease, particularly thyroid disease. The best studied paradigm of autoimmunity is type I diabetes mellitus, characterized by B- and T-cell responses (14, 15), possibly occurring as a result of exposure to a new antigen (e.g., enterovirus).

We suggest that pulmonary emphysema can - at least in part - be explained as a consequence of a failure of signals that maintain normal lung structure (16), with ensuing alveolar septal cell apoptosis and enhanced oxidative stress (17). Furthermore, vascular endothelial growth factor (VEGF), an obligatory endothelial cell (EC) survival factor (18) abundantly expressed in lung tissue (19), serves as a critical lung structure maintenance factor, because lung tissue from patients with severe emphysema shows decreased VEGF gene and protein expression (20), and chronic VEGF receptor (VEGFR) blockade causes emphysema in adult rats (21) and impaired alveolization in neonatal rats (22).

Taking into account recent novel experimental approaches to control or block tumor angiogenesis, which have relied on immunization with ECs (23, 24), DNA coding for EC growth factors (25), or receptor proteins (26), we question whether these approaches may have a collateral destructive effect on lung ECs. If so, this would support the concept that dysregulation of antibody- and cell-mediated immunity can be involved in the disruption of the lung maintenance program, alveolar cell apoptosis, and the development of emphysema (12). We therefore postulated that intraperitoneal injection of rats with xenogeneic ECs would cause a pulmonary anti-EC immune response and emphysema. Furthermore, rats injected with human umbilical vein ECs (HUVECs), but not with human pulmonary artery smooth muscle cells (HPASMCs), generate within 2 to 3 weeks antibodies against ECs and develop centrilobular emphysema associated with alveolar cell apoptosis and activation of matrix metalloproteinases (MMPs) (27). We also demonstrate that rat antihuman EC antibodies induce EC apoptosis in vitro and cause emphysema in passively immunized mice. Furthermore, passive transfer of CD4+ cells from EC-immunized rats into naive immunocompetent animals resulted in emphysema. Because immunization of athymic rats with xenogeneic ECs did not cause emphysema, this indicates that T cells participate in the immune response and emphysema development. This is the first model that provides a proof of concept for an autoimmune mechanism of EC damage in the development of emphysema.

Some of the results of these studies have been previously reported in the form of abstracts (27, 28).

METHODS

Experimental Protocols

The protocol was approved by the animal care and use committee of the University of Colorado Health Sciences Center. Adult male Sprague-Dawley rats (200 g) were injected intraperitoneally with HUVECs or HPASMCs plus adjuvant. Control rats received only adjuvant.

Antibodies

Cleaved caspase-3 (Cell Signaling Technology, Inc., Beverly, MA), CD3 and CD4 (Zymed Laboratories, San Francisco, CA), CD8, CD20, CD68 and HLA-DR (DakoCytomation, Carpinteria, CA), and VEGFR-2/ KDR (Santa Cruz Biotechnology, Inc., Santa Cruz, CA).

Cell Culture

Primary HUVEC cultures were isolated as described by Bruneel and colleagues (29). HUVEC cultures, 80% confluent, at the fourth or fifth passage were used. HPASMCs were from BioWhittaker, Inc., Walkersville, MD.

ELISA for Anti-EC Antibodies

HUVECs were plated in 96-well collagen-coated plates (Nalge Nunc Intl., Rochester, NY), grown for 24 hours, then fixed in 1% paraformaldehyde/phosphate-buffered saline (PBS). Immune-serum diluted 1:200 in bovine serum albumin/PBS was added and incubated for 2 hours at 37

Morphometry

Lungs were inflated with 0.5% low-melting agarose at a constant pressure of 25 cm H2O, fixed (30) and paraffin-embedded by standard techniques. Sections (5 µm) were stained with hematoxylin and eosin. Images were acquired with a Carl Zeiss AxioCam color camera (Carl Zeiss Vision GmbH, Hallbergmoos, Germany) and analyzed using KS300 imaging system software (Carl Zeiss Vision GmbH). Alveolar airspaces and CD4+ cell accumulation areas were measured in pixels per square micrometer.

The mean linear intercept (31) was measured as previously described (21).

Triple Immunofluorescence

Normal rat lung sections were incubated with pre- or EC-immune serum Ig (7 weeks after immunization). Bound Ig was detected with a secondary Alexa Floor-labeled goat anti-rat antibody (A-11077 and Alexa Fluor 568 goat anti-rat IgG; Molecular Probes, Eugene, OR). Microvascular endothelial cells were visualized with BS-I fluorescein-labeled lectin from Griffonia simplicifolia (Sigma); nuclei were stained with 4'-6'-diamidino-2-phenylindole dihydrochloride (DAPI; Molecular Probes).

Zymography

Electrophoresis was performed using 10% zymogram (gelatin) gels (Invitrogen Life Technologies, Inc., Carlsbad, CA).

Cell Proliferation and Cell Death Assays

Cell proliferation was assessed using the CyQuant cell proliferation assay kit (Molecular Probes). Terminal transferase dUTP nick end labeling (TUNEL) was performed as earlier described (21). Active caspase-3 in paraffin-embedded lung tissues was assessed using a rabbit polyclonal antibody to cleaved caspase-3 (Cell Signaling Technology, Inc., Beverly, MA) (21). Caspase-3/7 activity in HUVECs treated with nonimmunized and HUVEC-immunized rat serum was measured using the Apo-One homogeneous caspase-3/7 assay kit (Promega, Madison, WI). Flow cytometric analysis was performed using the Vybrant apoptosis assay kit no. 3 (Molecular Probes).

Adoptive Transfer of CD4+ T Cells

Rat spleen lymphocytes (48 hours after immunization with HUVECs) were isolated by forcing the tissue through a fine wire mesh followed hy osmotic lysis of erythrocytes and separation using a Ficoll-hypaque (Amersham Biosciences, Uppsala, Sweden) gradient. Selection of CD4+ cells used magnetic cell sorting rat CD4 magnetic microbeads (Miltenyi Biotec, Auburn, CA). A total of 1 × 10^sup 7^ CD4+ splenic cells from HUVEC-immunized or nonimmunized rats were injected into naive rats.

Statistical Analysis

Data are expressed as mean ± SEM. One-way analysis was performed with the Student-Newman-Keuls post hoc test. Statistical difference was accepted at p < 0.05.

Additional details regarding the methods are provided in the online supplement.

RESULTS

Rats Injected with HUVECs or HPASMCs Develop Anti-EC Antibodies

We injected rats with live HUVECs or HPASMCs (three animals/group) and tested for an immune response on Days 0, 1, 4, and 11, and Weeks 5, 7, and 11 after intraperitoneal injection. Intraperitoneal injection of xenogeneic cells resulted in the development of an immune response as tested by ELISA (Figures 1a and 1b). Serum from nonimmunized rats did not show any cross-reactivity with HUVECs. This finding suggests that injection of xenogeneic cells induces antibody production in the anti-EC-immunized animals. We confirmed that formalin-fixed HUVEC injection in rats also caused antibody production. On the basis of our findings and those of Wei and coworkers (24), showing that blockade of tumor angiogenesis was achieved by immunization with either live or formalin-fixed ECs, all subsequent experiments were performed using live HUVECs.

Adult Rats Injected with HUVECs but Not with HPASMCs Demonstrate Alveolar Airspace Enlargements

Rats injected with 1 × 10^sup 7^ live HUVECs (Figure 1d) showed enlarged lung alveolar airspaces with increased mean linear intercept when compared with the control rats that received only adjuvant (Figures 1c, 1d, and 1f) 3 weeks after injection. HPASMC-injected rats generated an anti-smooth muscle antibody response (Figure 1b), but they did not develop emphysema (Figures 1e and 1f) because there was no difference in the mean linear intercept between control and HPASMC-immunized rat lungs. This effect is not from HUVECs migrating into the lung and causing a local xenogeneic cell response, because we did not detect human endothelial cells in immunized rats lungs stained with a human HLA-DR antibody by immunohistochemistry at Days 1 and 4 after immunization.

Alveolar Cell Apoptosis Is Present in Lungs from EC-immunized Rats

Because our previous work showed that emphysema due to VEGFR blockade was caused by alveolar septal cell apoptosis (21), we asked whether the emphysematous lungs of the EC-immunized animals contained caspase-positive cells. Rats injected with HUVECs showed a large number of TUNEL-(Figure 2b) and caspase-3- (Figure 2d) positive alveolar septal cells, whereas rats injected with vehicle (Figures 2a and 2c) or HPASMCs (data not shown) did not.

Expression of MMPs in Lungs from Xenogeneic EC-immunized Rats

Prior studies showed an overexpression of MMP-2, MMP-8, and MMP-9 in human emphysematous lungs, suggesting that MMPs may play an important role in the pathogenesis of chronic inflammation, airway remodeling, and alveolar destruction (32, 33). Rats injected with xenogeneic EC demonstrated increased MMP-9 (sixfold increase) and MMP-2 (twofold increase) activity in their lung tissue homogenates as assessed by zymography (Figures 3a and 3b). Immunofluorescence staining showed elevated MMP-9 expression intensity in alveolar septal cells in immunized rat lungs (Figure 3d) as compared with control rat lungs (Figure 3c). High levels of MMP-9 were also detected in the immunized rat subpleural lung tissue (Figure 3, inserts in Figure 3c and 3d).

Serum from HUVEC-immunized Rats Inhibits Proliferation and Induces Apoptosis of ECs In Vitro

EC-immune serum not only inhibited EC proliferation (Figure 4a) but also induced cell death by enhancing caspase-3/7 activity (Figure 4b) by threefold when compared with preimmune rat serum. Flow cytometric analysis of annexin V and propidium iodide staining of cells showed that, after 18 hours, 33% of EC-immune serum-treated cells were apoptotic as compared with only 8% of control serum-treated cells (Figure 4c). We ruled out the involvement of complement because the addition of fresh complement alone to the cells did not have an effect on cell proliferation and did not potentiate the EC-immune serum effect (data not shown).

Serum from HUVEC-immunized Rats Contains Antibodies against VECFR-2

We have recently demonstrated that VEGFR blockade induces emphysema in rats (21). To examine whether the animals that had been immunized with xenogeneic HUVECs had developed antibodies against the VEGFR-2/KDR, we immunoprecipitated VEGFR-2 from HUVEC (Figure 4d, lanes 1, 3, and 5) and human pulmonary vascular endothelial cell (Figure 4d, lanes 2, 4, and 6) protein extracts using a rabbit polyclonal anti-VEGFR-2 antibody, followed by gradient gel separation and protein transfer to a polyvinylidene fluoride membrane. After transfer, the membrane was cut into three pieces along the molecular marker lanes: one (Figure 4d, lanes 1 and 2) was incubated with a mouse monoclonal VEGFR-2 antibody; the second, with anti-HUVEC rat serum (Figure 4d, lanes 3 and 4); and the third, with rat preimmune serum (Figure 4d, lanes 5 and 6). Both anti-EC serum and the antibody against VEGFR-2 recognized the same 210-kD band of VEGFR-2, indicating that HUVEC immunization had caused the generation of anti-VEGFR-2 antibodies, whereas preimmune serum did not (Figure 4d, lanes 5 and 6).

Serum from HUVEC-immunized Rats Recognizes Lung Capillary ECs

Triple immunofluorescence of normal rat lungs using the microvascular EC marker Griffonia symplicifolia-lectin (Figures 5a and 5e), preimmune (Figure 5b), or HUVEC-immune (Day 7 after immunization; Figure 5f) serum and DAPI (Figures 5c and 5g) was performed to assess whether the EC-immune serum recognized lung capillary ECs. The combined images show that anti-HUVEC rat serum recognized rat lung capillary endothelial cells (Figure 5h), whereas staining with preimmune rat serum was negative (Figure 5d). As with the rat lung, HUVEC-immune (Figure 5i) but not preimmune serum (Figure 5j) reacted with normal human lung septal cells.

Passive Immunization with Anti-EC Serum Causes Emphysema in Mice

To determine whether anti-EC antibodies are directly causing emphysema, we passively immunized mice (n = 3). Mice were chosen because they would require the injection of significantly less immune serum than the rat. C57BL/6J mice, when injected twice (at Days 0 and 14) intraperitoneally with purified anti-HUVEC serum (60 µl each injection; Figure 6a), but not control mice injected with the same amount of preimmune rat serum (Figure 6b) or untreated mice (Figure 6c), developed enlarged alveolar airspaces after 5 weeks of treatment, indicative of emphysema (Figure 6d). These results suggest that anti-EC antibodies suffice to cause emphysema.

Athymic Nude Rats Injected with Xenogeneic ECs Do Not Develop Emphysema

To determine whether a competent immune system is required for EC-immunization-induced emphysema, we injected HUVECs into adult athymic nude rats. Although the athymic nude rat has a normal population of bone marrow-dependent B cells, lack of T cells abolishes the signals that cause B cells to multiply and produce antibodies. Adult athymic rats did not develop emphysema (Figure 7b) as compared with adjuvant-only injected rats (Figure 7a) and did not show septal cell apoptosis following xenogeneic EC injection (data not shown). The alveolar airspace measurements were the same in lungs from immunized and nonimmunized nude rats. These findings suggest that T lymphocytes may play an important role in the development of emphysema in this model.

Xenogeneic EC Immunization Results in an Influx and Accumulation of CD4+ Cells into the Lung

Because the response to xenogeneic cells is typically mediated by CD4+ lymphocytes (34), we identified CD4+ cells in lungs from immunized and control rats using immunohistochemistry. Figure 8b shows that the number of lung tissue CD4+ cells was clearly increased at 24 hours after intraperitoneal injection, compared with control (Figure 8a); the difference persisted for at least 7 days after EC immunization (Figure 8c). There were no differences in the numbers of CD8+ (cytotoxic T cells), CD20+ (B cells), and CD68+ (macrophages) cells in immunized or control rat lungs (data not shown). There was also no difference in the number of CD68+ cells in the spleen (data not shown).

Adoptive Transfer of CD4+ Cells from Immunized Rats Causes Emphysema in Naive Animals

To examine whether antigen-specific CD4+ lymphocytes, which accumulate in the spleen in response to intraperitoneal EC immunization, can cause emphysema, we isolated CD4+ cells from the spleen at 48 hours after immunization. The purified CD4+ cell population did not contain CD8+ or B cells as tested by flow cytometry (data not shown). Naive Sprague-Dawley rats injected intraperitoneally with spleen-derived CD4+ cells from immunized rats developed emphysema at 3 weeks after CD4+ cell transfer (Figure 9b), whereas animals injected with splenocytes from nonimmunized or HPASMC-immunized animals did not (Figure 9a). Moreover, adoptive transfer of CD4+ spleen cells into secondary syngeneic rats resulted in emphysema, even though these secondary rats had not been immunized with HUVEC (data not shown). The alveolar airspace measurements were significantly increased in CD4+ cell-injected rat lungs as compared with control lungs (Figure 9c). Our data suggest that pathogenic CD4+ T lymphocytes are necessary and sufficient for breaking peripheral tolerance and causing emphysema in naive, immunocompetent rats.

DISCUSSION

This new model of emphysema in adult rats provides, for the first time, data supporting the concept of autoimmune emphysema. Our data demonstrate that intraperitoneal injection of xenogeneic ECs in immunocompetent rats causes an anti-EC humoral response, influx of CD4+ lymphocytes into the lung, apoptosis of alveolar septal cells, activation of MMPs, and emphysema. Although antiendothelial antibodies are likely participating in the alveolar septal cell apoptosis, additional injury by infiltrating lymphocytes (35) cannot be excluded. Our data support previous observations that disruption of structural integrity of alveolar septa, such as that induced by VEGFR blockade, induces emphysema (21) and that anti-EC antibodies can induce EC apoptosis (36). Both chronic VEGFR blockade and xenogeneic EC immunization in rats cause emphysema in rats without apparent damage to other organs. This finding may indicate a particular vulnerability of lung microvascular ECs and epithelial cells to disruption of survival and trophic signals, particularly related to VEGF, as VEGFR-2 is expressed not only by lung ECs but also by alveolar type II ECs (37). Earlier, Wei and coworkers (24) showed that immunization with xenogeneic EC in mice also elicits production of antibodies directed against ECs and against VEGFR-2. We show that rat anti-HUVEC antiserum induced apoptosis of HUVECs and human pulmonary vascular endothelial cells in vitro. Although it is known that VEGFR-2 antibodies induce EC apoptosis (38), as does VEGF ligand neutralization (39), antibodies directed against additional EC epitopes might be involved in our model as well as targeting of nonendothelial alveolar cells that express VEGFR-2 (37). The fact that injection of cell-free immune serum into naive mice results in emphysema in these mice indicates that anti-EC antibodies are sufficient to cause experimental emphysema.

It is likely that increased proteolytic activity-as shown zymographically-contributed to lung tissue destruction, but whether MMP activation is a consequence of apoptosis or associated with the lymphocyte response is unknown. Airway and lung parenchyma lymphocyte infiltration are well recognized in human emphysema (35), but whether the lymphocytes are cause or consequence of the human lung tissue injury remains unclear.

Because nude rats do not develop emphysema after HUVEC immunization, and transfer of CD4+ cells isolated from spleens of HUVEC-immunized rats causes emphysema in naive immunocompetent animals, we suggest that the antigen-experienced CD4+ cells, which accumulate as a consequence of the HUVEC injection, contribute to the destruction of alveolar structures. This important new finding may in part be explained by data that demonstrate that T-cell-EC interactions can result in EC apoptosis (40-42). Conversely, ECs have been described as regulators of T-cell function (43, 44), in that EC can present antigens to CD4+ T cells. In contrast to EC, fibroblasts and smooth muscle cells fail to provide the necessary stimulatory signals required for T-cell activation and proliferation (43, 45). Several articles exist that focus on CD8+ T cells in chronic obstructive pulmonary disease (35, 46, 47), whereas in our rat model we find an increase in CD4+ T cells in the lungs. However, the CD8+ reports of human chronic obstructive pulmonary disease do not rule out a significant participation of CD4+ cells in the human disease (46). In fact, our group has recently shown that severe emphysema in humans is associated with inflammation involving T lymphocytes that are composed of clonal CD4+ T-cell populations. These T cells are accumulating in the lung secondary to conventional antigen stimulation and are likely critical to the pathogenesis of severe emphysema (48).

We describe for the first time that, as in models of autoimmune diabetes (49, 50), pathogenic CD4+ cells can be sufficient for the development of emphysema, because the adoptively transferred CD4+ cells from EC-immunized animals appear to be dominant and overriding regulatory self-tolerance against pulmonary septal cells in immunocompetent animals. The concerted action of autoantibodies and T lymphocytes may produce a wide range of outcomes, such as disease (e.g., in emphysema and diabetes mellitus [14]) or protective immunity (e.g., tumor vaccines [24]). Autoreactive CD4+ cells capable of transferring disease to a secondary animal are comparable to other established models of autoimmunity, such as the 2.5 transgenic CD4 cell clone, which was generated from autoimmune nonobese, diabetic (NOD) mice (51-53). The passive transfer of 2.5 cells into NOD mice is sufficient to infiltrate pancreatic islets and cause diabetes. This is a quintessential model of autoimmunity. Therefore, the replication of disease in the second animal by a passive transfer of CD4+ cells alone in the current model makes this process most consistent with an autoimmune process. It is important to realize that a "foreign" invasion of virus or other acquired antigens may be at the source of many processes that are considered to be autoimmune (54, 55). This phenomenon is the environmental "second hit" that may be required in addition to a predisposed genotype.

In the aggregate, our findings may partially fulfill Koch's postulates for cellular immunology (56) implicating the importance of CD4 cells in experimental emphysema: (1) the disease (emphysema) develops in the presence of pathogenic CD4+ cells, (2) adoptive transfer of CD4+ cells results in emphysema, and (3) emphysema does not occur in the absence of CD4+ cells (i.e., the athymic nude rats).

At present there is no link between our autoimmune emphysema model to the human condition because anti-EC antibodies have not been described in patients with emphysema, and because there is no clear evidence that the lymphocytes that accumulate in emphysematous lungs (57) are antigen-specific. Although a connection between smoking-induced emphysema and lung tissue immune response has not been established, emphysema has been recognized in rare cases of hypersensitivity pneumonitis in nonsmokers (58). Although we cannot yet establish a link between our model and smokers' emphysema, our proof-of-concept model begs the question whether autoimmune mechanisms do play a role in human emphysema.

Conflict of Interest Statement: L.T.-S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; R.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; K.-H.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; M.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; A.S. has received a GlaxoSmithKline (GSK) Fellow research award of $40,000 received July 1, 2004, to |une 30, 2005, and it is received without any stipulation about the research and results, and GSK has no access to the results outside of journal publication; M.R.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; A.P.F. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; R.M.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; N.F.V. received $2,000 from Actelion for speaking as an invited professor at Yale University, $750 for participating in a conference sponsored by United Therapeutics, $1,000 from AstraZeneca for speaking at a AstraZeneca-sponsored meeting, and $1,000 as a consultant fee from Pfizer, and is the recipient of a research grant from GSK.

Acknowledgment: The authors thank Amy Richter for assistance with immunofluorescence staining.