Expression of Tissue Factor in Experimental Bovine Pneumonic Pasteurellosis and Endotoxemia Javed Rashid’ Douglas J. Weiss’ Michael P. Murtaughl Ronald Bath'

Tissue factor (TF), a cell surface-associated cofactor and activator of coagulation factor Vll, has been implicated in the local and systemic activation of coagulation associated with sepsis. This study describes the pattern of TF expression in experimental bovine pneumonic pasteurellosis and endotoxemia. lmmunohistochemical techniques were used to localize TF antigen in tissue sections. Tissue factor expression was not observed in tissues from control animals. In response to Pasteurella haemolytica challenge, TF was expressed within alveolar walls, by mononuclear inflammatory cells within alveoli, and in walls of arteries, arterioles, bronchi, and bronchioles. Tissue factor was not detected in unaffected lung, liver, spleen, lymph node or kidney tissue. Administration of Escherichia coli endotoxin intravenously resulted in tissue factor expression in lung, spleen, and lymph node tissue. Results of this study indicate that TF is expressed locally at sites of inflammation and systemically in endotoxemia. Therefore, TF may be involved in coagulation events associated with pneumonic pasteurellosis.

Key Words: tissue factor, pneumonic pasteurellosis, coagulation, inflammation, endotoxemia

In trod uction

Local and/or systemic coagulopathies frequently accom- pany gram-negative At sites of inflammation, intravascular thrombosis and fibrin deposition within inflamed tissues frequently occurs. Systemic activation of the coagula- tion cascade results in the syndrome of disseminated intravascular coagulation. It is uncertain whether local inflam- mation can cause systemic activation of coagulation in the absence of endotoxemia and/or septicemia.

The lung lesions in pneumonic pasteurellosis are charac- terized by deposition of a dense fibrin meshwork within alve- oli which entraps degenerating leukocytes and erythrocytes. Ultrastructural studies of affected lung revealed degenerative changes in the alveolar wall and fibrin and platelet thrombi within alveolar capillaries which were closely associated with pulmonary intravascular macro phage^.^

Results of several studies indicate that tissue factor (TF, also termed tissue thromboplastin) may mediate fibrin forma- tion and intravascular coagulation events within inflamed

Tissue factor, a 43 kDa cell membrane-associated gly- coprotein, binds coagulation factor VII and the complex rapid- ly activates factor X, thereby initiating the common pathway of blood coagulation.8 The TF-Vlla complex also activates

coagulation factor IX, thereby activating the intrinsic coagula- tion pa th~ay .~ Expression of TF on the surface of tissue macrophages and circulating monocytes occurs within hours after onset of sepsis, consistently accelerating the clotting of blood and in some cases inducing disseminated intravascular coag~lat ion.~-~ Administration of TF to rabbits induced dis- seminated intravascular coag~lopathy,~,~ whereas experimen- tal septic shock was prevented by administration of TF-neu- tralizing antibodies to baboon^.^

Tissue factor expression within normal and inflamed tis- sues has been evaluated by immunohistochemical tech- niques. Tissue factor in normal human tissues was detected in vascular adventitia, organ capsules, epidermis, and mucosal epithelium.1° In healthy baboons, TF was detected in vascular adventitia but not in endothelial cells.” In baboons with septic shock, TF expression was increased in the spleen and possibly in alveolar septae and renal glomeruli.11 Tissue factor expression was evaluated in frozen sections of calf Iung.l2 In control lung sections, TF staining was observed within alveolar walls, perivascular connective tissue, and vas- cular smooth muscle. Some alveolar macrophages also stained positively for tissue factor. In lung sections from calves experimentally infected with Pasteurella haemolytica,

‘Department of Veterinary Pathobiology, University of Minnesota, St. Paul, MN 551 08 Veterans Administration Hospital, Minneapolis, MN

Supported in part by a grant (MINV-63-045) from the United States Department of Agriculture Continued

PAGE 198 Vol. 26, No. 4,1997 VETERINARY CLINICAL PATHOLOGY

Bovine Pneurnonic Pasteurellosis

alveolar septae, epithelial cells, and alveolar macrophages stained intensely for tissue factor.

The purpose of this study was to evaluate the sites of expression of TF in calves experimentally inoculated with F! haernolytica or given Escherichia coli endotoxin intravenously to better understand local and systemic expression of TF associated with sepsis.

Materials and Methods ANTI-BOVINE TF POLYCLONAL ANTIBODY

Purified IgG anti-TF polyclonal antibody was produced by immunizing rabbits with bovine brain TF purified 142,000-fold to hom~geneity.~.~~ The specificity of antibody for TF was con- firmed by its capacity to inhibit TF activity and by Ouchterlony double immunodiffusion.8

Animals

Healthy male Holstein calves, 2 to 4 weeks old, were obtained from the University of Minnesota Dairy Barn and were housed in an isolation facility. Calves were held for at least 7 days before initiation of experiments.

EXPERIMENTAL DESIGN

Twelve calves were assigned to three groups of four calves each. Calves were sedated with xylazine (0.10 mg/kg IV) and positioned in dorsal recumbency. A modified 12F lavage catheter with inflatable cuff was placed into the right diaphragmatic lung lobe through a midcervical transtracheal in~ision.'~ Calves were then positioned in left lateral recum- bency. Four calves (control group) were inoculated intratra- cheally with 5 ml of sterile phosphate buffered saline (PBS). Four calves were inoculated with 5 x lo9 log-phase growth of F! haernolytica (serotype A1 , strain 12296) suspended in 5 ml of PBS. Four calves were given E. coli lipopolysaccharide (serotype 026:B6, Sigma Chemical Co.:St. Louis, MO) by continuous slow intravenous infusion (0.5 vg/kg IV) over a period of 5 Calves inoculated with saline or F! haernolytica were euthanized by an overdose of sodium pen- tobarbital 24 hours after administration of inoculate. Calves given lipopolysaccharide were maintained under light surgical anesthesia (sodium pentobarbital given at approximately 2 mg/kg IV every 30 to 45 minutes), and were euthanized at 6 hours after initiation of the study. To evaluate the sites of expression of TF, samples of lung, liver, spleen, and mesen- teric lymph node were obtained. Lung samples were obtained from the center and periphery of macroscopic lesions and from corresponding areas of the left lung.

IMMUNOHISTOCHEMISTRY

In preliminary studies, lung tissue from two calves was fixed in neutral buffered 10% formalin and zinc formalin for 12,14,18 and 24 hours.Tissues fixed for 12 or 14 hours had poor morphology and fixation for 24 hours decreased the

FIG. 1 - Section of lung from a calf inoculated with /? haemolyfica demonstrating deposition of fibrin (f), degenerating leukocytes and hemorrhage (h) within alveoli and a bronchiole.

immunostaining intensity. Eighteen hours of fixation resulted in maximum reactivity and suitable tissue morphology. The quality of immunostaining was improved by fixation in neutral buffered 10% formalin as compared to zinc formalin. Based on these observations, tissues were fixed with neutral buffered 10% formalin for 18 hours. Thereafter, tissues were embedded in paraffin, sectioned at 4pm, deparaffinized, rehy- drated and washed in Tris-buffered saline (0.05 M Tris base, 0.09% sodium chloride, pH 7.6). After blocking with a 1 :25 dilution of normal goat serum, polyclonal rabbit anti-bovine TF antibody diluted 1:200 was applied and incubated overnight in a humidified chamber at 4°C. Endogenous per- oxidase activity was blocked by incubating tissue sections with 3% H,O, for 10 minutes at room temperature, followed by three washes with Tris-buffered saline. A commercial avidin-biotin immunoperoxidase kit (Vector Laboratories, Burlingame, CA), with 3'3-diaminobenzidine as the chro- mogen, was used to localize the primary antibody. The sec- tions were counterstained with Meyer's hematoxylin stain and mounted using glycerol gelatin. The immunoperoxidase stain produced a red-brown reaction product. Negative controls consisted of replacing the primary antibody with normal rab- bit serum in the primary antibody incubation step.

LABELING OF PULMONARY INTRAVASCULAR MACROPHAGES

To label pulmonary intravascular macrophages, a 1 o/o s o b tion of monastrol blue B (0.5 ml/kg, Sigma Chemical Co, St. Louis, MO) was administered by slow intravenous infusion to one calf before inoculation with F! haernolytica.i6 Monastrol blue B is a copper phthalocyanine pigment with particles of 45-60 nm in diameter which are readily phagocytized by pul- monary intravascular macrophages when administered via the jugular vein. Previous studies have shown that virtually all monastrol blue B is removed from blood in the first passage through the

Continued

VETERINARY CLINICAL PATHOLOGY Vol. 26, No. 4,1997 PAGE 199

Bovine Pneumonic Pasteurellosis

Results Lesions were firm and reddish-brown in color. Histopathologic examination of lung sections from the

No gross lung lesions were observed in control calves or in calves infected with F! haemolytica revealed lesions typical of calves given endotoxin. Calves inoculated with P haemolytica pneumonic pasteurellosis (Fig. 1). Alveoli and interlobular had gross lesions restricted to the right diphragmatic lung septa contained neutrophils, fibrin, and hemorrhage. lobe. Lesions occupied one third to two thirds of the lobe. Multifocal areas of necrosis and swirling degenerate leuko-

FIG. 2 - Sections of lung from a calf inoculated with P haemolytica and stained for tissue factor using an immunoper- oxidase immunostaining method. a) Tissue factor is expressed within an inflamed lobule but not within an adjacent unaffected lobule. b) Tissue factor staining of an arteriole within an affect- ed area of lung tissue. Tissue factor is expressed predomi- nately within the tunica adventitia with slight staining of the inti- ma and media. c) Tissue factor expression within the alveolar septae. Note linear staining of alveolar surface of alveolar sep- tae (arrow). d) Tissue factor expression by degenerating leuko- cytes within a bronchiole. e) Tissue factor expression in section of lung in which intravascular macrophages are labeled with monastrol blue. Notice the co-localization of monastrol blue and tissue factor within intravascular macrophages (arrows).

PAGE 200 Vol. 26, No. 4,1997 VETERINARY CLINICAL PATHOLOGY

Bovine Pneumonic Pasteurellosis

Discussion cytes were present. Alveolar septae were thickened and con- tained inflammatory cells.

For calves inoculated with P haemolytica, TF was observed in inflamed lung tissue but not in unaffected lung, liver, spleen, or lymph node sections (Fig. 2a). Tissue factor immunoreactivity was present within the walls of arteries, arterioles, bronchi, and bronchioles (Fig. 2b), within alveolar walls (Fig. 2c), and in individual mononuclear inflammatory cells (Fig. 2d) in all sections of inflamed lung. Tissue factor staining within blood vessels was predominately localized to the tunica adventitia, with variable expression in the tunica media and tunica intima. Within alveolar septae, TF was observed within cells with large round nuclei which resem- bled intravascular macrophages and in linear deposits along the alveolar surface of alveolar septae. To determine whether intravascular macrophages expressed tissue factor, we administered monastrol blue to label intravascular macrophages. The majority of monastrol blue-containing cells stained positively for TF (Fig. 2e).

TF EXPRESSION IN ENDOTOXEMIA

Histopathologic examination of lung, liver, spleen, and lymph node tissue 6 hours after initiation of endotoxin admin- istration revealed lesions only in lung tissue. Lung lesions were characterized by atelectasis and thickening and increased cellularity within alveolar septae (Fig. 3).

lmmunohistochemical staining revealed specific staining for TF in lung, spleen, and mesenteric lymph node tissue (Fig. 4). In lung sections, TF was localized within alveolar macrophages, alveolar septae, and arterial walls. In spleen sections, TF was localized in large mononuclear cells which resembled macrophages and in walls of the arterioles. In mesenteric lymph node sections, TF was observed in large mononuclear cells within germinal centers that resembled macrophages and in arterial walls.

lmmunohistochemical techniques were used to study the expression of TF in experimentally induced bovine pneumon- ic pasteurellosis and endotoxemia. Results of this study indi- cate that TF immunostaining was not observed in lung, liver, spleen, or lymph nodes of control calves. Results of other studies, in which frozen sections were used, observed tissue factor expression in tissue section from healthy calves, humans, and baboons.10’12 In these studies, TF was expressed by vascular adventitia, organ capsules, epidermis, and mucosal epithelium.10,11,12 In lung sections from healthy calves, TF immunostaining was also observed within alveolar septae and in some alveolar macrophages. Therefore, our failure to detect immunoreactive tissue factor in control tis-

FIG. 3 - Section of lung from a calf given endotoxin intravenously demonstrating thickening and increased cellularity within alveolar septae.

FIG. 4 - Sections of lung and mesenteric lymph node from calf given endotoxin intravenously and stained for tissue factor. a) Lung section; note tissue factor expression by mononuclear cells within an atelectatic area of lung tissue. b) Lymph node section; note expression of tissue factor within large mononuclear cells (arrows) within a germinal center.

Continued

VETERINARY CLINICAL PATHOLOGY Vol. 26, No. 4, 1997 . PAGE 201

Bovine Pneumonic Pasteurellosis

sues may have been due to formalin fixation of tissues. Tissue factor was highly expressed within alveolar septae

and lumens in inflamed lung tissue in response to t? haemolytica challenge. Expression of TF within alveolar sep- tae indicates that blood flowing through alveolar capillaries may be exposed to TF. Linear deposits of TF within alveolar septae appeared to be associated with the alveolar surface and therefore may have been produced by alveolar epithelial cells. Although immunostaining of endothelial cells within alveolar septae and larger vessels was not observed, TF was expressed by pulmonary intravascular macrophages. Because TF is a cell surface-associated molecule, expres- sion by intravascular macrophages would probably result in direct exposure of plasma coagulation factors to TF. This hypothesis is supported by electron microscopic studies of lungs from calves with induced pneumonic pasteurellosis, in which fibrin deposits and platelet aggregates were observed adjacent to intravascular macro phage^.^

Expression of TF by mononuclear inflammatory cells within intraalveolar fibrin deposits indicates that TF may be involved in the activation of coagulation within alveoli. The resultant loss of surface area for gas exchange may contribute to hypoxemia and respiratory f a i l ~ r e . ’ ~ , ’ ~ ~ ~ ~ ~ ’ ~ These observations are supported by studies in which anti-bovine TF antibody was administered to calves experimentally infected with /? haemolyfi~a.’~ Treatment with anti-bovine TF antibody markedly reduced fibrin deposition within alveoli and TF- dependent procoagulant activity in bronchoalveolar lavage fluid. Lack of TF expression in unaffected lung tissue and other organs indicates that t? haemolyfica challenge results in local but not systemic expression of TF.

Intravenous administration of endotoxin resulted in expres- sion of TF in lung, spleen and lymph nodes indicating sys- temic activation of TF procoagulant activity. lmmunostaining was limited to arterial walls and macrophages. Tissue factor expression was not observed within endothelial cells in any tissues. Intravenously administered endotoxin is cleared pri- marily by lung and liver probably by intravascular macrophages and kupffer cells.2o Therefore, intravascular macrophages may be exposed to high concentrations of endotoxin, and resultant expression of TF may activate intravascular coagulation in sepsis.

Results of several studies indicate that TF plays a central role in the activation of coagulation in sepsis.4~5~7~11~12~’9 Major ini- tiators of tissue factor expression include endotoxin, tumor necrosis factor, and interleukin-1. Endotoxins bind to cell membranes via interaction between the core polysaccharide and a specific cell receptor or by binding to lipopolysaccha- ride binding protein which binds to CD14 on the cell sur- f a ~ e . ~ ’ - ~ ~ A pertussis toxin-sensitive G-protein and protein kinase C are involved in intracellular signal tran~duction.~~ Tumor necrosis factor interacts with two cell membrane receptors and transcription is initiated through the NF-KB and bZlP Expression of TF by monocytes and macrophages has been well documented but expression by endothelial cells is controversial. Tissue factor is expressed

by endothelial cells in vitro in response to endotoxin and tumor necrosis factor but expression of TF by endothelial cells has not been demonstrated in vivo.26 Results of a recent study indicate that basal TF expression by endothelial cell cultures may result from the presence of contaminating smooth muscle cells.27

These data indicate that uncomplicated pasteurella pneu- monia induces TF expression within inflamed areas of lung but not in uninfected areas of lung or other organs within the body. Endotoxemia or septicemia appears to be required for systemic expression of TF. I

REFERENCES

1. ldell S, James KK, Levin EG, el al: Local abnormalities in coagulation and fibrinolytic pathways predispose to alveolar fibrin deposition in the adult respiratory distress syndrome. J Clin Invest 84:695-705, 1989.

2. Edwards RL, Rickles FR: The role of leukocytes in the activation of blood coagula- tion. Semin in Hematol 29202-212, 1992.

3. Whiteley LO, Maheswaran SK, Weiss DJ, et al: lmmunohistochemical localization of Pasteurella haemolytica A1 -derived endotoxin, leukotoxin and capsular polysaccharide in experimental bovine pasteurella pneumonia. Vet Pathol 27:150-161, 1991.

4. Warr T, Rao VM, Rapport SI: Disseminated intravascular coagulation in rabbits induced by administration of endotoxin or tissue factor. Blood 75:1481-1489. 1990.

5. Henry MM. Moore JN: Endotoxin-induced procoagulant activity in equine peripheral blood monocytes. Circulatory Shock 26297-309, 1988.

6. Levi M, ten Cate H, van der Poll T: Pathogenesis of disseminated intravascular coag- ulation in sepsis. J Am Vet Med Assoc 270:975-979. 1993.

7.Taylor FB Jr, Chang A, Ruf W: Lethal E. coli septic shock is prevented by blocking tis- sue factor with monoclonal antibody. Circulatory Shock 33:127-134, 1991.

8. Bach R, Nemerson Y, Konigsberg W: Purification and characterization of bovine tis- sue factor. J Biol Chem 256:8324-8331, 1981.

9. Nemerson Y: The tissue factor pathway of blood coagulation. Semin in Hematol 29,170-176, 1992.

10. Drake TA, Morrissey JM, Edgington TS. Selective cellular expression of tissue factor in human tissues.Am J Pathol 134: 1087-1097. 1989.

11. Drake TA, Cheng J, Chang Al, et al: Expression of tissue factor, thrombomodulin. and E-selectin in baboons with lethal Escherichia coli sepsis. Am J Pathol 1421458-1470, 1993.

12. Car BD, Suyemoto MM, Neilsen NR, el al: The role of leukocytes in the pathogene- sis of fibrin deposition in bovine acute lung injury. Am J Pathol 138:1191-1198, 1991.

13. Bach R, Gentry R, Nemerson Y: Factor VII binding to tissue factor in reconstituted phospholipid vesicles; induction of cooperativity by phosphatidyl serine. Biochemistry 25:4007-4020, 1986.

14. Weiss DJ, Bauer MC, Whiteley LO, et al: Changes in blood and bronchoalveolar lavage fluid components in calves with experimentally induced pneumonic pasteurellosis. Am J Vet Res 52337-344, 1991

15. Morris DD, Whitlock RH, Merryman GS: Endotoxin-induced changes in the hemo- static system in neonatal calves: The effects of antisera to a mutant E. coli (J-5). Am J Vet Res 472514-2519, 1986.

16. Longworth KE, Westgate AM, Grady MK: Development of pulmonary intravascular macrophage function in newborn lambs. J Applied Physiol 732608-2615, 1992.

17. Sibley Y, Reynolds HY: Macrophages and polymorphonuclear neutrophils in lung defense and injury. Am Rev Respir Dis 141:471-501, 1990.

18. Whiteley LO, Maheswaran SK. Weiss DJ: Pasteurella haemolytica A1 and bovine respiratory disease: Pathogenesis. J Vet Internal Med 6:l l-22, 1992.

19. Weiss DJ, Rashid J. Murtaugh M: Role of tissue factor in the pathogenesis of pneu- monic pasteurellosis. Vet Pathol 32556, 1995.

20. Maxie MG, Valli VEO, Robinson GA, et al: Studies of radioactive endotoxin I. Clearance of 51Cr-labeled endotoxin from the blood of calves. Can J Comp Med 28:347- 366,1974.

21. Car BD. Slauson DO, Suyemoto MM. et al: Expression and kinetics of induced pro- coagulant activity in bovine pulmonary alveolar macrophages Exp Lung Res 17:939-957, 1991.

22. Car BD, Suyemoto MM, Neilsen NR, et al: The role of leukocytes in the pathogene- sis of fibrin deposition in bovine acute lung injury. Am J Pathol 138:1191-1198, 1991.

23. Morrison DC, Rudback JA. Endotoxin-cell membrane interactions leading to trans- membrane signaling. Contemp Topics Mol lmmunol8:187-218, 1981.

24. Dean DF, Carrol RC, Bochler PN, et al: Endotoxin-induced signal transduction path- ways in bovine alveolar macrophages. Vet Pathol 1993; 30:476.

25. Bierhaus A, Zhang Y, Deng Y, et al. Mechanisms of the tumor necrosis factor a- mediated induction of endothelial tissue factor. J Biol Chem 2702641 9.26432, 1995.

26. Breider MA, Yang Z. Tissue factor in bovine endothelial cells induced by Pasteurella haemolytica lipopolysaccharide and interleukin-1. Vet Pathol 31 :55-60, 1994.

27. Mulder AB, Blom NR. Smit JW. et al. Basal tissue factor expression in endothelial cell cocultures is caused by contaminating smooth muscle cells. Reduction by using chy- motrypsin instead of collagenase. Thrombo Res 80:399-411, 1995.

PAGE 202 Vol. 26, No. 4,1997 VETERINARY CLINICAL PATHOLOGY