SEARCH

SEARCH BY CITATION

Keywords:

  • Autoimmune;
  • Blood vessel;
  • Coagulation;
  • Flow cytometry

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

Background: Dogs with immune-mediated hemolytic anemia (IMHA) and certain inflammatory diseases are at high risk of developing thromboembolic disease. The presence of anti-endothelial cell autoantibodies (AECA) has been associated with an increased risk of thromboembolism in humans.

Hypothesis: AECA will be detected more often in dogs at risk of thromboembolism than in healthy control animals or dogs with diseases not associated with a higher risk of thromboembolism.

Animals: Ninety-one sick dogs and 22 healthy control dogs.

Methods: Retrospective case-controlled study. Serum was screened for the presence of AECA. Dogs were identified for the study based on the risk of thromboembolism as determined by clinical impression and the underlying disease process. Flow cytometry and normal canine endothelial cells were used to screen serum samples from sick and healthy control dogs for the presence of AECA. In addition, serum from dogs with confirmed thromboemboli was also screened for the presence of AECA by immunohistochemistry.

Results: AECA were detected in 2/91 sick dogs, both with infectious diseases, but were not found in healthy dogs. Anti-endothelial antibodies were not detected in 21 dogs with IMHA and 20 dogs with systemic inflammatory response syndrome, sepsis, or both.

Conclusions: We conclude that AECA are rarely detectable in dogs considered at high risk of thromboembolism. These findings suggest that AECA may not play an important role in the pathogenesis of thromboembolism in dogs with IMHA and other inflammatory diseases.

Thromboembolism is a major contributor to morbidity and mortality in critically ill patients in humans and animals. The presence of pulmonary and systemic thromboembolism has been detected in 29–32% of dogs with immune-mediated hemolytic anemia (IMHA), and the case fatality rate in dogs with IMHA and thromboembolism is as high as 70%.1–3 However, the pathophysiology of thromboembolism in dogs remains poorly understood. Thromboembolic complications are often associated with the development of a prothrombotic state, which may be manifested by blood stasis, hypercoagulability, and endothelial damage.4 Diseases in veterinary medicine that have been associated with an increased risk of thromboembolic complications include protein-losing nephropathies and enteropathies, neoplasia, IMHA, vasculitis, hyperadrenocorticism, sepsis, systemic inflammatory response syndrome (SIRS), disseminated intravascular coagulation (DIC), heart disease, hypothyroidism, cemented total hip replacement surgery, and prolonged recumbancy.2–8

Production of anti-endothelial cell autoantibodies (AECA) is common in a number of important autoimmune diseases of humans, including systemic lupus erythematosis, peripheral arteritis, Wegener's granulomatosis, antiphospholipid antibody syndrome, and systemic vasculitis.9–11

AECA may induce endothelial dysfunction and trigger development of thromboembolism.12 For example, AECA binding to endothelial cells can trigger endothelial cell activation and upregulation of adhesion molecules, along with secretion of proinflammatory cytokines. Endothelial cells exposed to serum containing AECA also release large amounts of tissue factor, an important factor in the coagulation cascade.11 Autoantibody binding to endothelial cell surfaces might also trigger local complement activation, leading to direct endothelial injury with subsequent platelet aggregation and clotting. Recent studies also suggest that antibody binding to endothelial cells may trigger endothelial cell exocytosis and release of von Willebrand's factor, a potent procoagulant protein.13

Thrombotic complications are common in dogs with certain immune-mediated diseases and diseases associated with systemic inflammatory states such as sepsis. Particularly in dogs with IMHA, development of pulmonary and systemic thromboembolism is typically a devastating development (IMHA).14 For example, as many as half of all IMHA-associated deaths in dogs are caused by complications related to thromboembolism.2,7,15–17 In a necropsy study of dogs that died or were euthanized due to IMHA, there was evidence of widespread embolic disease in 30% of the dogs.1 The cause of thromboembolism in dogs with IMHA and other systemic inflammatory diseases remains unknown at present, although coagulation abnormalities are common in dogs with IMHA.15

We conducted a study to determine whether AECA could be detected at a higher frequency in dogs considered to be at high risk of developing thromboemboli. Flow cytometry and a normal canine endothelial cell line were used to screen serum for the presence of AECA.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

Dogs and Sample Collection

Serum was obtained from 63 dogs hospitalized at the Colorado State University's Veterinary Teaching Hospital (CSU-VTH) and stored at −70 °C before analysis. Serum was obtained from each of these dogs within 48 hours of admission to the critical care unit at the CSU-VTH. In addition, we utilized serum samples obtained during other studies from an additional 15 dogs with positive antibody titers indicative of active infection with Ehrlichia canis, Rickettsia ricketsii, or Toxoplasma gondii, and from 13 additional dogs with IMHA. These studies were reviewed and approved by the Institutional Animal Care and Use Committee and the Institutional Review Board at Colorado State University. Dogs selected for entry into the study were those considered to be at increased risk of development of thromboembolic disease based on prior published studies and included dogs with immune-mediated disease, neoplasia, SIRS, DIC, sepsis, hyperadrenocorticism, protein-losing nephropathy, cardiovascular disease, infectious disease, or documented thromboembolic disease. Serum samples were also collected from 22 healthy adult dogs. These control samples included 14 samples obtained from dogs owned by employees and students at the CSU-VTH and 8 samples obtained from healthy dogs examined at the CSU-VTH.

Endothelial Cell Lines

Canine vascular endothelial cells (CVEC) derived from normal canine jugular veins were purchased commercially and cultured in supplemented medium at 37 °C in 5% CO2, exactly as described by the manufacturer.a The CVEC assumed typical cobblestone endothelial cell morphology in culture and were characterized by immunocytology and flow cytometry as normal endothelial cells. The CVEC were harvested from subconfluent monolayer cultures by brief trypsinization, then resuspended in complete medium for 15 minutes to inactivate trypsin,b and then washed in complete medium before being used in the flow cytometry assay.

Additional flow cytometric screening for AECA was performed using a bovine pulmonary artery endothelial cell line that was prepared from normal bovine pulmonary arteries and was kindly provided by Dr Martha Tissot, University of Colorado Health Sciences Center, Denver, CO. These cells were maintained in MEM mediumc supplemented with 10% fetal bovine serum (FBS)d and penicillin and streptomycin solution.c Their endothelial cell identity was confirmed by morphology and by Factor VIII expression (data not shown). They were prepared for flow cytometric analysis as described for CVEC.

Flow Cytometry for AECA

Serum samples were thawed and diluted 1 : 100 in a 96-well round-bottom tissue culture plate in 100 μL of fluorescence-activated cell-sorting (FACS) buffer (phosphate-buffered saline [PBS], 2% FBS, and 0.01% sodium azide). Suspensions of endothelial cells (canine vascular endothelial cells [CVEC] or bovine pulmonary aortic endothelial cells [BPAEC]) at a concentration of 2 × 104 cells per well were added to each of the prediluted wells. The serum and endothelial cell suspensions were incubated for 30 minutes at 4 °C. Next, the cells were washed twice with FACS buffer. The cells were then resuspended and incubated for 20 minutes at 4 °C with fluorescein isothiocyanate (FITC)-conjugated, rabbit anti-dog IgGe diluted 1 : 100 in FACS buffer. After staining, the cells were washed twice in FACS buffer, and then resuspended in FACS buffer and stored at 4 °C before analysis.

After immunostaining, endothelial cells were analyzed with a Cyan ADP flow cytometer.f Acquisition gates were set on live cells and 10,000–20,000 events were analyzed for each sample of endothelial cells incubated with test serum. For each experiment, negative controls included unstained endothelial cells, endothelial cells incubated with serum without the secondary antibody, and endothelial cells incubated with the secondary antibody only. A positive control included serum from a dog vaccinated against bovine endothelial cells (as part of an unrelated study), which was incubated with bovine endothelial cells and the secondary FITC anti-dog IgG antibody as described above.

Analysis of flow cytometry data was performed by Summit software.f Analysis gates were set on endothelial cells incubated with secondary antibody only and the increase in the percentage of positive cells or in the mean fluorescence intensity of the population of cells incubated with test serum plus secondary antibody was determined. For the purpose of this study, positive samples were defined as those in which the percentage of positive cells (eg, those with surface bound immunoglobulin) was 5% or greater, compared with staining with secondary antibody alone, as well as staining with serum from healthy control animals. This cutoff was used previously in a recent publication by our group examining the prevalence of anti-RBC antibodies.18

Immunohistochemistry (IHC) for the Detection of AECA

Serum samples were also screened for the presence of antibodies capable of recognizing intracellular as well as surface-expressed endothelial antigens, by IHC instead of flow cytometry. Tissues (spleen, lung, liver, and heart) obtained from a normal, purpose-bred Beagle dog (euthanized as part of an unrelated study) were collected and imbedded in OCT compound,g and were then snap-frozen in liquid nitrogen for cryosectioning. Sections were cut on a cryostat to a thickness of 5 μm and mounted on treated slides.h Tissue sections were fixed in ice-cold acetone for 4 minutes, and then incubated with 3% hydrogen peroxide for 10 minutes to block endogenous peroxidase activity. Test serum was diluted 1 : 500 in PBS with 1% bovine serum albumin and 5% normal rabbit serum and applied to tissue sections and incubated for 30 minutes at room temperature. Sections were washed and then incubated with biotinylated rabbit anti-dog IgGe for 20 minutes, then washed and incubated with streptavidin-HRP.e The slides were washed and developed with AEC enzyme substrate.i Slides were cover-slipped and evaluated by light microscopy for detection of specific IgG binding. Negative controls included sections in which the primary and secondary reagents were omitted. Positive controls included sections incubated with an antibodyj specific for CD146+ canine endothelial cells, as described previously.19

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

Canine Populations

Serum samples from 91 clinically ill dogs and 22 healthy control dogs were evaluated by flow cytometry for the presence of AECA. In the population of clinically ill dogs, there were 40 male dogs (8 intact and 32 castrated) and 51 female dogs (4 intact and 47 castrated). Within this population, there were 21 Labrador Retrievers, 19 mixed-breed dogs, 6 Golden Retrievers, 5 German Shepherds; 3 each of the following: Boxer, Maltese, Dachshund; 2 each of the following: Beagle, Rottweiler, Cocker Spaniel, Walker Hound, and Border Collie dogs; and 1 each of Dalmatian, Australian Healer, German Wirehair Pointer, Cairn Terrier, Scottish Terrier, Saint Bernard, Chihuahua, Schnauzer, Mastiff, Irish Wolfhound, Cardigan Welsh Corgi, Shi Tzu, Plotthound, Pekeingnese, Greyhound, Australian Shepherd, Rat Terrier, Fox Terrier, Keeshound, Toy Poodle, and Shetland Sheepdog dogs. The age range was 7 months to 13 years (median age 8 years) with 43% male and 56% female dogs.

Serum samples were also collected from 22 healthy adult dogs. Fourteen samples were obtained from dogs owned by employees and students at the CSU-VTH and 8 samples were obtained from healthy patients examined at the CSU-VTH. There were serum samples from 15 male dogs (1 intact and 14 castrated) and 7 female dogs (2 intact and 5 castrated). The healthy control population consisted of 7 Labrador Retrievers, 7 mixed breed dogs, 4 Golden Retrievers, and 1 each of Standard Poodle, Pug, Chesapeake Bay Retriever, and Maltese. The age range was 1–11 years (median age 4.5 years) with 64% male and 36% female dogs.

Disease Characteristics of Clinically Ill Dogs

There were 24 dogs with immune-mediated diseases, including 22 dogs with IMHA, 1 dog with inflammatory bowel disease, and 1 dog with immune-mediated pancytopenia. The next largest category was 21 dogs with sepsis/SIRS/DIC (Table 1).

Table 1.   Disease characteristics in clinically ill dogs evaluated for presence of anti-endothelial cell antibodies.
Sepsis/SIRS/DIC (n = 20)Infectious (n = 15)Neoplasia (n = 11)Cardiac (n = 8)Immune Mediated (n = 23)Renal (n = 5)Postoperative (n = 5)Central Nervous System (n = 2)Hyperadrenocorticism (n = 2)
  1. Serum samples were collected from 91 clinically ill dogs, including dogs with diseases associated with a high risk of thromboembolism. Serum samples were screened for the presence of anti-endothelial cell antibodies by flow cytometry and immunohistochemistry, as described in “Materials and methods”.

  2. SIRS, systemic inflammatory response syndrome; DIC, disseminated intravascular coagulation; GDV, gastric dialation and volvulus.

Systemic inflammatory response after cardiopulmonary bypass (n = 5)Ehrlichia canis >1 : 160 (n = 3)Hemangio-sarcoma (n = 3)Dilated cardiomyopathy with ventricular tachycardia (n = 1)Immune-mediated hemolytic anemia (n = 21)Protein-losing nephropathy (n = 2)Thoracotomy (n = 2)Unknown neurotoxin (n = 1)Pituitary-dependent hyperadrenocorticism (n = 2)
Heartworm disease (n = 3)E. canis >1 : 640 (n = 1)Lymphangio-sarcoma (n = 1)Dilated cardiomyopathy with atrial fibrillation (n = 1)Inflammatory bowel disease (n = 1)Acute renal failure (n = 2)Adrenal mass removal (n = 1)Granulomatous meningo-encephalitis (n = 1) 
Sepsis/DIC secondary to carcinomatosis (n = 1)E. canis >1 : 2,520 (n = 1)Pheochromo-cytoma (n = 1)Dilated cardiomyopathy with pericardial effusion (n = 1)Immune-mediated pancytopenia (n = 1)Chronic renal failure (n = 2)Extrahepatic shunt cellophane band ligation (n = 1)  
Tumor lysis syndrome/DIC (n = 1)Rickettsia rickettsii >1 : 40 (n = 2)Intracranial adenocarcinoma (n = 1)Mitral valve endocardiosis with atrial fibrillation (n = 2) Glomerulo-sclerosis (n = 1)Ovariohysterectomy with marked hemorrhage (n = 1)  
Multifocal thrombi secondary to lymphoma (n = 1)R. rickettsii >1 : 80 (n = 2)Mammary carcinoma (n = 1)Mitral valve endocardiosis with congestive heart failure (n = 1)     
Sepsis secondary to mammary carcinoma (n = 1)R. rickettsii >1 : 160 (n = 1)Transitional cell carcinoma (n = 1)Myocardial infarction (n = 1)     
Multifocal thrombosis secondary to gastric necrosis and GDV (n = 1)Toxoplasma gondii >1 : 256 (n = 1)Disseminated mast cell tumors (n = 1)      
SIRS secondary to pneumonia (n = 1)Toxoplasma gondii >1 : 512 (n = 3)Anal sac adenocarcinoma (n = 1)      
SIRS secondary to pulmonary fibrosis (n = 1)Toxoplasma gondii >1 : 1,024 (n = 1)Unknown intrathoracic neoplasia (n = 1)      
SIRS secondary to necrosuppurative myositis/vasculitis (n = 1)        
Idiopathic multifocal thrombosis (n = 1)        
Idiopathic vasculitis/sepsis (n = 1)        
DIC 2 days post adrenal carcinoma resection (n = 1)        

Results of Flow Cytometric Screening of Serum for AECA

AECA were not detected in serum samples from the 22 healthy mature adult dogs. As a positive control for flow cytometric detection of AECA, we also included serum from a dog vaccinated with xenogeneic (bovine aortic endothelium) endothelial cells as part of an unrelated cancer vaccine study (data not shown).

Serum samples from the 89 of 91 clinically ill dogs also did not contain detectable AECA. However, 2 samples were positive for AECA, as revealed by positive binding of canine immunoglobulins to canine endothelial cells (Fig 1). These positive samples included 1 from a dog positive for antibodies to R. ricketsii and 1 from a dog positive for antibodies to T. gondii. Thus, based on flow cytometric screening, which has been used previously in a number of studies to identify AECA in humans, none of the critically ill dogs with sepsis, SIRS, or IMHA had detectable antibodies to AECA.

image

Figure 1.   Flow cytometric detection of anti-endothelial cell antibodies in a dog with toxoplasmosis and a dog with thrombosis. Serum samples from a healthy control dog (top panel) and a dog with clinical toxoplasmosis (middle panel) and a dog with confirmed thrombosis (bottom panel) were screened for the presence of AECA using normal canine aortic endothelial cells and flow cytometry, as described in Methods. The presence of AECA in the dog with toxoplasmosis was revealed by an overall increase in the amount of IgG bound to the canine endothelial cells and by an increase in the percentage of cells strongly positive for IgG binding. The control dog and the dog with thrombosis did not have AECA.

Download figure to PowerPoint

Immunohistochemical Screening for AECA

Serum samples from 8 dogs with documented or strongly suspected thromboembolism were screened for AECA by IHC. The underlying diseases present in these 8 dogs included DIC, cardiac disease, and neoplasia (Table 2). Serum samples from 7 healthy control dogs were also screened for AECA by IHC. Tissues evaluated included lung, liver, spleen, kidney, and heart. Specific binding of IgG or IgM antibodies to endothelial cells was not detected when tissues were incubated either with serum from the 8 dogs with thromboembolic disease or with serum for 7 healthy control dogs. As a control for detection of endothelial cell antigens with IHC, there was no strong staining of endothelial cells in the large and the small blood vessels using an antibody directed against the endothelial cell antigen CD146, as described previously (data not shown).20 Thus, AECA directed against intracellular antigens were also not detected in dogs with confirmed or suspected thromboembolic disease.

Table 2.   Disease diagnosis in 8 dogs with confirmed thromboembolism.
SignalmentUnderlying DiseaseDiagnosis
  1. Serum was obtained from 8 dogs with confirmed or strongly suspected thromboembolism and screened for the presence of anti-endothelial cell antibodies by flow cytometry and immunohistochemistry, as described in “Materials and methods”.

  2. DIC, disseminated intravascular coagulation.

10Y FS Golden RetrieverDICNecropsy
1Y MI Labrador RetrieverCardiacNecropsy
5Y MC Golden RetrieverDICSurgery
8Y MC mixed breedNeoplasiaSurgery
7M MI Labrador RetrieverCardiacNecropsy
4Y FS Labrador RetrieverCardiacNecropsy
13Y FS Shi TzuNeoplasiaNecropsy
8Y FS Labrador RetrieverNeoplasiaSurgery

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

The results of this study suggest that detectable AECA are relatively rare in dogs, including dogs with diseases such as IMHA and sepsis known to be associated with a high prevalence of thromboembolic disease. In addition, AECA were also not detected in 8 dogs with confirmed thromboemboli. Thus, endothelial cell autoantibodies appear unlikely to play an important role in the pathogenesis of thromboembolism in this group of at-risk dogs. However, we cannot exclude the possibility that the detection techniques (flow cytometry and IHC) used in this study were too insensitive to detect relevant AECA concentrations. For example, AECA may be rapidly bound to endothelial cells in vivo and not circulate at detectable concentrations in dogs. However, the fact that AECA were detected in 2 dogs (1 with rickettsial infection and 1 with Toxoplasma infection) in this study suggests that flow cytometry can detect AECA in dogs.

Nonetheless, AECA have been widely detected in humans, by the same flow cytometry and IHC techniques we utilized in this study.11,21–23 For example, AECA have been detected in humans with connective tissue diseases, systemic vasculitis or arteritis, systemic immune-mediated diseases, and with accelerated coronary artery disease after cardiac transplant.9–11,21,24,25 Other approaches for screening for the presence of AECA in humans include use of cell-based ELISA, cell cytotoxicity, and Western blotting.10 However, flow cytometry appears to have the greatest sensitivity and specificity for the detection of AECA in humans.22

The results presented here suggest that detectable AECA are relatively rare in dogs and probably do not play an important role in triggering the development of thromboembolic complications. However, these conclusions are subject to several caveats. In addition to the issue of assay sensitivity, it is also possible that insufficient dogs were evaluated in this study. For example, it would also be important to include more animals with confirmed thromboembolism in a further study. In addition, in some of the animals developing acute thromboembolic diseases, such as those undergoing cardiopulmonary bypass surgery, there would not have been time to develop effective antibody responses and thus AECA would not have been expected to play a role in these animals. Also, larger numbers of animals with IMHA should be evaluated, because IMHA is associated with a very high risk of thromboembolism. However, it should be noted that in autoimmune diseases where AECA are detected in humans, the prevalence is often very high, approaching 85% in some cases.10 Thus, one would expect to have identified at least several animals positive for AECA in the 21 dogs with IMHA evaluated in this study.

We believe it would also be informative to investigate the prevalence of AECA in recently vaccinated dogs, in light of the possible association between immunization and immune-mediated diseases.3,26,27 Finally, it would also be useful to screen cats for AECA, in light of the fact that many feline vaccines inadvertently contain cellular antigens and induce antibody responses against normal cellular proteins.28,29

Footnotes

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

aAllCells LLC, Emeryville, CA

bSigma Aldrich, St Louis, MO

cGibco-InVitrogen, San Diego, CA

dHyclone, Logan, UT

eJackson ImmunoResearch, West Grove, PA

fDako-Cytomation, Fort Collins, CO

gSakura Tissue-Tek Compound, VWR International, West Chester, PA

hSuperfrost, VWR International

iVector Laboratories, Burlingame, CA

jP1H12, Chemicon International, Temecula, CA

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

The authors wish to acknowledge Drs Debra Kamstock and Doug Thamm and Ms Barb Rose for assistance with the culture of canine endothelial cells. These studies were supported in part by grants from the Morris Animal Foundation.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References