Cephapirin (20 mg/kg of body weight, IV) was administered before and after 3 doses of probenecid (25, 50, or 75 mg/kg, intragastrically, at 12-hour intervals) to 2 mares. Clearance and apparent volume of distribution, based on area under the curve, were negatively correlated with probenecid dose. Clearance of cephapirin was decreased by approximately 50% by administration of 50 mg of probenecid/kg. Serum, synovial fluid, peritoneal fluid, CSF, urinary and endometrial concentrations of cephapirin were determined after 5 doses of cephapirin (20 mg/kg, IM, at 12-hour intervals) without and with concurrently administered probenecid (50 mg/kg, intragastrically) to 6 mares, including the 2 mares given cephapirin, IV. Highest mean serum cephapirin concentrations were 16.1 +/- 2.16 micrograms/ml at 0.5 hour after the 5th cephapirin dose [postinjection (initial) hour (PIH) 48.5] in mares not given probenecid and 23.7 +/- 1.30 micrograms/ml at 1.5 hours after the 5th cephapirin dose (PIH 49.5) in mares given probenecid. Mean peak peritoneal fluid and synovial fluid cephapirin concentrations were 6.2 +/- 0.57 micrograms/ml and 6.6 +/- 0.58 micrograms/ml, respectively, without probenecid administration and 12.3 +/- 0.46 micrograms/ml and 10 +/- 0.78 micrograms/ml, respectively, with concurrent probenecid administration. Mean trough cephapirin concentrations for peritoneal and synovial fluids in mares given probenecid were 2 to 3 times higher than trough concentrations in mares not given probenecid. Overall mean cephapirin concentrations were significantly higher for serum, peritoneal fluid, synovial fluid, and endometrium when probenecid was administered concurrently with cephapirin (P less than 0.01). Cephapirin was not detected in CSF samples. Overall mean urinary cephapirin concentrations (2.47 mg/ml without concurrent probenecid administration and 3.06 mg/ml with concurrent probenecid) were not significantly different (P greater than 0.05). Mean trough serum probenecid concentration was 61.2 +/- 5.28 micrograms/ml. Highest serum probenecid concentration was 148.8 +/- 5.97 micrograms/ml, 2 hours after the 5th cephapirin dose (PIH 50). Probenecid administration increased serum, synovial fluid, peritoneal fluid, and endometrial concentrations of cephapirin in mares.

Twenty-one mixed-breed pony foals, reared and maintained under parasite-free conditions, were used to test the efficacy of ivermectin in oral drench and paste formulations (200 microgram/kg) against 11-day-old migrating larvae of Parascaris equorum. Three replicates of 4 foals and 3 replicates of 3 foals were formed on the basis of age. Foals in replicates of 4 were randomly allocated to be indicators, or to receive vehicle (control) or ivermectin paste or ivermectin liquid. Foals in replicates of 3 were randomly allocated to receive vehicle or ivermectin paste or ivermectin liquid. The recovery of larvae from the lungs, liver, and small intestines of the indicator foals showed that 99.9% of the larvae were in the lungs 11 days after inoculation (day 0 of treatment). The recoveries of larvae from lungs and small intestines of controls at 25 days after inoculation indicated that all larvae had migrated to the small intestine by this time. The mean length of larvae recovered from the lungs (11 days after inoculation) was 0.87 mm; the mean length of those recovered from the small intestine (25 days after inoculation) was 3.65 mm. Using larvae recovered from small intestinal contents for calculations, ivermectin in both formulations was 100% effective against 11-day P equorum (P less than 0.01, compared with control group geometric mean of 1498.4).

Twenty-one mixed-breed pony foals, reared and maintained under parasite-free conditions, were used to test the efficacy of ivermectin in oral drench and paste formulations (200 microgram/kg) against 11-day-old migrating larvae of Parascaris equorum. Three replicates of 4 foals and 3 replicates of 3 foals were formed on the basis of age. Foals in replicates of 4 were randomly allocated to be indicators, or to receive vehicle (control) or ivermectin paste or ivermectin liquid. Foals in replicates of 3 were randomly allocated to receive vehicle or ivermectin paste or ivermectin liquid. The recovery of larvae from the lungs, liver, and small intestines of the indicator foals showed that 99.9% of the larvae were in the lungs 11 days after inoculation (day 0 of treatment). The recoveries of larvae from lungs and small intestines of controls at 25 days after inoculation indicated that all larvae had migrated to the small intestine by this time. The mean length of larvae recovered from the lungs (11 days after inoculation) was 0.87 mm; the mean length of those recovered from the small intestine (25 days after inoculation) was 3.65 mm. Using larvae recovered from small intestinal contents for calculations, ivermectin in both formulations was 100% effective against 11-day P equorum (P less than 0.01, compared with control group geometric mean of 1498.4).

Pharmacologic effects of alpha-methylfentanyl and 3-methylfentanyl, analogs of fentanyl, were investigated in mares. The ability of an 125I-labeled fentanyl radioimmunoassay (125I-RIA) to detect these methylated fentanyl analogs in individual and pooled urine samples from horses was evaluated. Also, the ability of 7 fentanyl antibodies to react with fentanyl and fentanyl derivatives (sufentanil, alfentanil, and carfentanil) was investigated. Mares were studied in a locomotor test to determine the amount of stimulation methylated fentanyl analogs might induce. Two mares each were given alpha-methylfentanyl at 1, 2, 4, 8, or 13 microgram/kg of body weight, IV, or 3-methylfentanyl at 0.4, 0.7, or 1 microgram/kg IV. The cross-reactivity of sufentanil, alfentanil, carfentanil, alpha-methylfentanyl, and 3-methylfentanyl with 7 fentanyl antibodies was studied, using the 125I-RIA. All fentanyl analogs, with the exception of alfentanil, cross-reacted well with a C1 antibody raised to fentanyl. Less satisfactory cross-reactivity was determined with 6 other antibodies raised to fentanyl derivatives. When the C1 antibody was combined with an iodinated analog to fentanyl, good detectability of alpha-methylfentanyl and 3-methylfentanyl, in terms of fentanyl equivalents, was obtained from urine samples of dosed mares. The ability of the 125I-RIA to detect methylated fentanyl analogs in forensic urine samples pooled in groups of up to 20 samples was evaluated. When these methylated analogs were administered to mares in doses that induced measurable locomotor stimulation, the analog's presence was readily detected in individual or pooled samples.

Pharmacologic effects of alpha-methylfentanyl and 3-methylfentanyl, analogs of fentanyl, were investigated in mares. The ability of an 125I-labeled fentanyl radioimmunoassay (125I-RIA) to detect these methylated fentanyl analogs in individual and pooled urine samples from horses was evaluated. Also, the ability of 7 fentanyl antibodies to react with fentanyl and fentanyl derivatives (sufentanil, alfentanil, and carfentanil) was investigated. Mares were studied in a locomotor test to determine the amount of stimulation methylated fentanyl analogs might induce. Two mares each were given alpha-methylfentanyl at 1, 2, 4, 8, or 13 microgram/kg of body weight, IV, or 3-methylfentanyl at 0.4, 0.7, or 1 microgram/kg IV. The cross-reactivity of sufentanil, alfentanil, carfentanil, alpha-methylfentanyl, and 3-methylfentanyl with 7 fentanyl antibodies was studied, using the 125I-RIA. All fentanyl analogs, with the exception of alfentanil, cross-reacted well with a C1 antibody raised to fentanyl. Less satisfactory cross-reactivity was determined with 6 other antibodies raised to fentanyl derivatives. When the C1 antibody was combined with an iodinated analog to fentanyl, good detectability of alpha-methylfentanyl and 3-methylfentanyl, in terms of fentanyl equivalents, was obtained from urine samples of dosed mares. The ability of the 125I-RIA to detect methylated fentanyl analogs in forensic urine samples pooled in groups of up to 20 samples was evaluated. When these methylated analogs were administered to mares in doses that induced measurable locomotor stimulation, the analog's presence was readily detected in individual or pooled samples.

Ten commercially available parathormone (PTH) assays were competitively validated, using dilutional parallelism, intra-assay and interassay coefficients of variation, and sensitivity and measured responses of 2 dogs to calcium and EDTA infusions. A 2-site immunoradiometric assay for intact human-PTH was superior to the others for estimating canine-PTH, met the criteria for validity, and was further investigated. A series of sample-handling studies was performed. Serum and plasma samples stored at 24 C lost 15% (n = 5; P less than 0.05) of PTH between 2 and 24 hours. This did not occur at 6 C. The mean PTH concentration of sera from blood samples clotted at 24 C was 6% (P less than 0.05) higher than equivalent EDTA samples. Serum samples stored at 6 and 37 C deteriorated 35% and 100% (n = 5; P less than 0.05), respectively, after 1 week, whereas samples stored at -20 and -70 C for 4 weeks did not deteriorate. There was no significant deterioration of PTH in samples frozen (-40 C) and thawed up to 7 times (n = 5). Parathyroid function testing was investigated by use of 2-hour infusions of disodium EDTA (25 mg/kg/h), 10-minute infusions of calcium gluconate (3 mg of elemental calcium/kg/10 min), and physiologic saline controls (n = 8). Renal function was monitored before and after EDTA infusion by exogenous creatinine clearance. Infusion of disodium EDTA increased mean PTH concentration from 67 (time 0) to 317 and 235 pg/ml at 90 and 180 minutes, respectively (P less than 0.001). Infusion of calcium gluconate decreased mean PTH concentration from 84 (time 0) to 14 and 12 pg/ml at 15 and 60 minutes, respectively (P less than 0.005). There were no observable side effects of the infusions in normal conscious dogs and no differences in exogenous creatinine clearance after EDTA infusion.

Ten commercially available parathormone (PTH) assays were competitively validated, using dilutional parallelism, intra-assay and interassay coefficients of variation, and sensitivity and measured responses of 2 dogs to calcium and EDTA infusions. A 2-site immunoradiometric assay for intact human-PTH was superior to the others for estimating canine-PTH, met the criteria for validity, and was further investigated. A series of sample-handling studies was performed. Serum and plasma samples stored at 24 C lost 15% (n = 5; P less than 0.05) of PTH between 2 and 24 hours. This did not occur at 6 C. The mean PTH concentration of sera from blood samples clotted at 24 C was 6% (P less than 0.05) higher than equivalent EDTA samples. Serum samples stored at 6 and 37 C deteriorated 35% and 100% (n = 5; P less than 0.05), respectively, after 1 week, whereas samples stored at -20 and -70 C for 4 weeks did not deteriorate. There was no significant deterioration of PTH in samples frozen (-40 C) and thawed up to 7 times (n = 5). Parathyroid function testing was investigated by use of 2-hour infusions of disodium EDTA (25 mg/kg/h), 10-minute infusions of calcium gluconate (3 mg of elemental calcium/kg/10 min), and physiologic saline controls (n = 8). Renal function was monitored before and after EDTA infusion by exogenous creatinine clearance. Infusion of disodium EDTA increased mean PTH concentration from 67 (time 0) to 317 and 235 pg/ml at 90 and 180 minutes, respectively (P less than 0.001). Infusion of calcium gluconate decreased mean PTH concentration from 84 (time 0) to 14 and 12 pg/ml at 15 and 60 minutes, respectively (P less than 0.005). There were no observable side effects of the infusions in normal conscious dogs and no differences in exogenous creatinine clearance after EDTA infusion.

The vasculature of 22 small colons from dead adult ponies was perfused with latex or barium sulphate solution. The vascular anatomy was studied by use of dissection and alkali digestion of the latex specimens and microangiography of the barium sulphate-perfused specimens. The small colon is supplied by the caudal mesentric artery. The left colic artery arises from the caudal mesenteric artery, which then becomes the cranial rectal artery. Branches from the left colic and cranial rectal arteries form anastomosing arcades that become narrower distally along the length of the small colon. From these arcades arise terminal arteries, which enter the small colon wall and give rise to a subserosal, an intermuscular, and a large submucosal plexus, with frequent anastomoses between them. The venous drainage closely parallels the arterial supply, except near to its origin from the portal vein, when the left colic vein and caudal mesentric vein are separate from the corresponding arteries.

The vasculature of 22 small colons from dead adult ponies was perfused with latex or barium sulphate solution. The vascular anatomy was studied by use of dissection and alkali digestion of the latex specimens and microangiography of the barium sulphate-perfused specimens. The small colon is supplied by the caudal mesentric artery. The left colic artery arises from the caudal mesenteric artery, which then becomes the cranial rectal artery. Branches from the left colic and cranial rectal arteries form anastomosing arcades that become narrower distally along the length of the small colon. From these arcades arise terminal arteries, which enter the small colon wall and give rise to a subserosal, an intermuscular, and a large submucosal plexus, with frequent anastomoses between them. The venous drainage closely parallels the arterial supply, except near to its origin from the portal vein, when the left colic vein and caudal mesentric vein are separate from the corresponding arteries.

The in vitro and in vivo binding of a monoclonal antibody (MAB) that recognizes a tumor-associated carbohydrate antigen was studied in dogs. Monoclonal antibody 155H.7 was raised in response to innoculation of mice with beta-galactose(1-3)betaN-acetylgalactosamine conjugated to human serum albumin. Avidin-biotin-complex immunohistochemical staining of cryostat sections of normal and neoplastic canine tissue specimens revealed heterogenous binding of MAB 155H.7 to the cells of many canine mammary and lung carcinomas and homogenous staining of may sarcomas, including osteogenic sarcoma. In addition, there was variable staining of a variety of normal tissues including some ductual epithelium, peripheral nerve fibers, and some endothelial cells and fibroblasts. Immunoscintigraphy with 131I-labeled MAB 155H.7 was used to study the in vitro distribution of the antibody. The 131I-labeled MAB 155.H.7 was administered to 1 clinically normal dog, 7 dogs with osteogenic sarcoma, 1 dog with undifferentiated sarcoma, and 2 dogs with mammary tumor. Scintigraphy revealed concentration of radioactivity in 8 of 10 tumor sites within 24 hours after MAB administration. The ratio of 131I in tumor sites to 131I in the surrounding normal tissues, compared with the similar ratio of 99mTc-labeled erythrocytes ranged from 1.1 to 4.3 in tumor vs normal tissue with a mean value of 2, confirming tumor localization of the radiolabeled MAB in excess of that associated with enhanced tumor vascularization.

The in vitro and in vivo binding of a monoclonal antibody (MAB) that recognizes a tumor-associated carbohydrate antigen was studied in dogs. Monoclonal antibody 155H.7 was raised in response to inoculation of mice with beta-galactose(1-3)betaN-acetylgalactosamine conjugated to human serum albumin. Avidin-biotin-complex immunohistochemical staining of cryostat sections of normal and neoplastic canine tissue specimens revealed heterogenous binding of MAB 155H.7 to the cells of many canine mammary and lung carcinomas and homogenous staining of many sarcomas, including osteogenic sarcoma. In addition, there was variable staining of a variety of normal tissues including some ductual epithelium, peripheral nerve fibers, and some endothelial cells and fibroblasts. Immunoscintigraphy with 131I-labeled MAB 155H.7 was used to study the in vitro distribution of the antibody. The 131I-labeled MAB 155.H.7 was administered to 1 clinically normal dog, 7 dogs with osteogenic sarcoma, 1 dog with undifferentiated sarcoma, and 2 dogs with mammary tumor. Scintigraphy revealed concentration of radioactivity in 8 of 10 tumor sites within 24 hours after MAB administration. The ratio of 131I in tumor sites to 131I in the surrounding normal tissues, compared with the similar ratio of 99mTc-labeled erythrocytes ranged from 1.1 to 4.3 in tumor vs normal tissue with a mean value of 2, confirming tumor localization of the radiolabeled MAB in excess of that associated with enhanced tumor vascularization.

Thirty-eight tracheobronchial aspirates (TBA) were collected from twenty 1 to 6-month-old foals, which were free of clinical signs of respiratory tract or other infectious disease. We collected TBA from 9 of the foals 3 times when they were approximately 8, 16, and 24 weeks old. Aspirates were examined cytologically after staining with modified Wright-Giemsa, Gram, toluidine blue, and prussian blue stains. Aerobic bacterial culturing was performed on all aspirates. Of the 20 initial TBA, 4 (20%) were normal cytologically on the basis of previously defined criteria for TBA from clinically normal horses, 6 (30%) had a high percentage of eosinophils (> 5%), 8 (40%) were classified as indicative of subacute inflammation, and 2 (10%) were classified as indicative of acute inflammation. Nine (45%) were positive for mast cells and none were positive for hemosiderin-laden macrophages (hemosiderophages). Of the 9 foals from which samples were collected at 16 and 24 weeks of age, results were similar, except for an increase in the number of TBA classified as indicative of chronic inflammation (33% and 22% respectively) and the number positive for hemosiderophages (33% and 88%, respectively). One TBA was considered nondiagnostic because of pharyngeal contamination. Culturing of 12 of the 37 aspirates (32%) yielded a potential microbial pathogen. Only 2 were positive cultures from the same foal. The following organisms were isolated: beta-hemolytic Streptococci spp (4), Actinobacillus/Pasteurella spp (4), Rhodococcus equi (2), unidentified nonenteric Gram-negative rod (1), and Escherichia coli (1). Thirty-four of the 37 aspirates (92%) yielded light growth of various organisms considered to be nonpathogenic and normal inhabitants of the upper respiratory tract. It was concluded that the presence of inflammatory cells, eosinophils, and mast cells in the tracheobronchial aspirates from clinically normal foals is a common finding. These cytologic findings were consistent in the samples collected from foals at 8, 16, and 24 weeks of age. It was also concluded that bacteria with recognized pathogenicity can be isolated from TBA from clinically normal foals and were most frequently isolated from 1- to 2-month-old foals or those with cytologic evidence of inflammation, even in the absence of clinical signs of respiratory tract disease.

Thirty-eight tracheobronchial aspirates (TBA) were collected from twenty 1 to 6-month-old foals, which were free of clinical signs of respiratory tract or other infectious disease. We collected TBA from 9 of the foals 3 times when they were approximately 8, 16, and 24 weeks old. Aspirates were examined cytologically after staining with modified Wright-Giemsa, Gram, toluidine blue, and prussian blue stains. Aerobic bacterial culturing was performed on all aspirates. Of the 20 initial TBA, 4 (20%) were normal cytologically on the basis of previously defined criteria for TBA from clinically normal horses, 6 (30%) had a high percentage of eosinophils (> 5%), 8 (40%) were classified as indicative of subacute inflammation, and 2 (10%) were classified as indicative of acute inflammation. Nine (45%) were positive for mast cells and none were positive for hemosiderin-laden macrophages (hemosiderophages). Of the 9 foals from which samples were collected at 16 and 24 weeks of age, results were similar, except for an increase in the number of TBA classified as indicative of chronic inflammation (33% and 22% respectively) and the number positive for hemosiderophages (33% and 88%, respectively). One TBA was considered nondiagnostic because of pharyngeal contamination. Culturing of 12 of the 37 aspirates (32%) yielded a potential microbial pathogen. Only 2 were positive cultures from the same foal. The following organisms were isolated: beta-hemolytic Streptococci spp (4), Actinobacillus/Pasteurella spp (4), Rhodococcus equi (2), unidentified nonenteric Gram-negative rod (1), and Escherichia coli (1). Thirty-four of the 37 aspirates (92%) yielded light growth of various organisms considered to be nonpathogenic and normal inhabitants of the upper respiratory tract. It was concluded that the presence of inflammatory cells, eosinophils, and mast cells in the tracheobronchial aspirates from clinically normal foals is a common finding. These cytologic findings were consistent in the samples collected from foals at 8, 16, and 24 weeks of age. It was also concluded that bacteria with recognized pathogenicity can be isolated from TBA from clinically normal foals and were most frequently isolated from 1- to 2-month-old foals or those with cytologic evidence of inflammation, even in the absence of clinical signs of respiratory tract disease.

Salivary glands from males of 3 Dermacentor species (D andersoni, D variabilis and D occidentalis) that were infected with either the Virginia or Idaho isolate of Anaplasma marginale as nymphs or adults were examined for colonies of A marginale by use of light and electron microscopy. Prior to dissection of salivary glands, exposed ticks were held at 25 C for 15 to 18 days, followed by a 3-day incubation at 37 C. Ticks of 2 species transmitted A marginale to calves; the third tick species was confirmed infected by demonstration of typical colonies in tick gut cells, but transmission was not attempted; Colonies of A marginale were seen with light microscopy in salivary glands of all 3 species of ticks; they were located in acinar cells that contained simple granules. Colonies varied morphologically from small, compact ones to larger structures that contained distinct organisms and often were adjacent to the host cell nucleus. Electron microscopy confirmed that the colonies were rickettsial organisms. Morphologic features of A marginale varied and included reticulated forms, forms with electron-dense centers, and small particles; these various forms were similar to those described previously in midgut epithelial cells of ticks. We believe that the organism seen within tick salivary glands may replicate in the glands before its transmission to the vertebrate host.

Salivary glands from males of 3 Dermacentor species (D andersoni, D variabilis and D occidentalis) that were infected with either the Virginia or Idaho isolate of Anaplasma marginale as nymphs or adults were examined for colonies of A marginale by use of light and electron microscopy. Prior to dissection of salivary glands, exposed ticks were held at 25 C for 15 to 18 days, followed by a 3-day incubation at 37 C. Ticks of 2 species transmitted A marginale to calves; the third tick species was confirmed infected by demonstration of typical colonies in tick gut cells, but transmission was not attempted; Colonies of A marginale were seen with light microscopy in salivary glands of all 3 species of ticks; they were located in acinar cells that contained simple granules. Colonies varied morphologically from small, compact ones to larger structures that contained distinct organisms and often were adjacent to the host cell nucleus. Electron microscopy confirmed that the colonies were rickettsial organisms. Morphologic features of A marginale varied and included reticulated forms, forms with electron-dense centers, and small particles; these various forms were similar to those described previously in midgut epithelial cells of ticks. We believe that the organism seen within tick salivary glands may replicate in the glands before its transmission to the vertebrate host.

Hemodynamic effects of spontaneous ventilation, intermittent positive-pressure ventilation (IPPV), and high-frequency oscillatory ventilation (HFOV) were compared in 6 dogs during halothane anesthesia. Anesthesia was induced with IV thiamylal Na and was maintained with halothane (end-tidal concentration, 1.09%). During placement of catheters, dogs breathed spontaneously through a conventional semiclosed anesthesia circuit. Data were collected, and dogs were mechanically ventilated, using IPPV or HFOV in random order. Ventilation was adjusted to maintain PaCO2 between 38 and 43 mm of Hg during IPPV and HFOV. Cardiac index, aortic blood pressure, and maximum rate of increase of left ventricular pressure were significantly (P less than 0.05) less during HFOV than during spontaneous ventilation, whereas right atrial and pulmonary artery pressure were significantly greater during HFOV than during spontaneous ventilation. During IPPV, only the maximum rate of increase of left ventricular pressure was significantly less than that during spontaneous ventilation.

Hemodynamic effects of spontaneous ventilation, intermittent positive-pressure ventilation (IPPV), and high-frequency oscillatory ventilation (HFOV) were compared in 6 dogs during halothane anesthesia. Anesthesia was induced with IV thiamylal Na and was maintained with halothane (end-tidal concentration, 1.09%). During placement of catheters, dogs breathed spontaneously through a conventional semiclosed anesthesia circuit. Data were collected, and dogs were mechanically ventilated, using IPPV or HFOV in random order. Ventilation was adjusted to maintain PaCO2 between 38 and 43 mm of Hg during IPPV and HFOV. Cardiac index, aortic blood pressure, and maximum rate of increase of left ventricular pressure were significantly (P less than 0.05) less during HFOV than during spontaneous ventilation, whereas right atrial and pulmonary artery pressure were significantly greater during HFOV than during spontaneous ventilation. During IPPV, only the maximum rate of increase of left ventricular pressure was significantly less than that during spontaneous ventilation.

Bovine pulmonary artery cells in cell culture were exposed to lipopolysaccharide (LPS) purified from Pasteurella haemolytica serotype A1. This resulted in severe membrane damage, which caused a time- and dose-dependent release of lactate dehydrogenase that was first detected 4 hours after exposure and reached a maximal mean release of 67% after 24 hours of exposure to 1 microgram of LPS/ml. Mean release of 51chromium followed by a similar pattern and reached a maximum of 61% following 24 hours of exposure to 10 micrograms of LPS/ml. Morphologically, endothelial cells responded to LPS by marked cell membrane retraction, the formation of numerous cytoplasmic blebs, and ruffling of the cell membrane. Subsequently, the cells became round and detached. Cell detachment reached a mean of 95% following 8 hours of exposure to 1 microgram of LPS/ml. These studies demonstrated that P haemolytica LPS is capable of causing direct damage to bovine pulmonary arterial endothelial cells, which may be important in the pathogenesis of bovine pneumonic pasteurellosis.

Bovine pulmonary artery cells in cell culture were exposed to lipopolysaccharide (LPS) purified from Pasteurella haemolytica serotype A1. This resulted in severe membrane damage, which caused a time- and dose-dependent release of lactate dehydrogenase that was first detected 4 hours after exposure and reached a maximal mean release of 67% after 24 hours of exposure to 1 microgram of LPS/ml. Mean release of 51chromium followed by a similar pattern and reached a maximum of 61% following 24 hours of exposure to 10 micrograms of LPS/ml. Morphologically, endothelial cells responded to LPS by marked cell membrane retraction, the formation of numerous cytoplasmic blebs, and ruffling of the cell membrane. Subsequently, the cells became round and detached. Cell detachment reached a mean of 95% following 8 hours of exposure to 1 microgram of LPS/ml. These studies demonstrated that P haemolytica LPS is capable of causing direct damage to bovine pulmonary arterial endothelial cells, which may be important in the pathogenesis of bovine pneumonic pasteurellosis.