Endothelial cell activation and dysfunction are key events in sepsis, leading to leukocyte adhesion, altered vasomotor tone, and increased vascular permeability.1 The endothelial-specific angiopoietin-Tie ligand receptor system is an important determinant of endothelial cell activation. Angiopoietin-1 and Ang2 are competitive ligands of the Tie2 receptor expressed by endothelial cells. Angiopoietin-1 is produced by pericytes, and binding of Ang1 to Tie2 receptors promotes vessel integrity and suppresses endothelial cell activation.2–5 Angiopoietin-2 is expressed by endothelial cells and stored in Weibel Palade bodies, from which it is released into the extracellular space upon stimulation. Binding of Ang2 to Tie2 receptors directly promotes endothelial dysfunction and inflammation.6 Angiopoeitin-2 has also been shown to mediate the breakdown of the endothelial glycocalyx in vitro, thereby indirectly contributing to an increase in vascular permeability and leukocyte adhesion.7
A marked imbalance of Ang1 and Ang2 in favor of Ang2 has been identified in critically ill humans.8,9 Associations between high plasma Ang2 concentration, disease severity, and outcome in humans with sepsis have been demonstrated in several studies.10–18
In addition to the role of angiopoietins as markers of inflammation and predictors of outcome, the circulating Ang2 concentration and Ang1:Ang2 concentration ratio could be useful for better understanding the role of the endothelium in the pathogenesis of vascular leakage conditions and could provide novel therapeutic targets.19
In the veterinary literature, no reports exist of plasma Ang2 concentrations in healthy and diseased dogs, and no dog-specific endothelial cell culture models have been developed to examine canine endothelial cell responses in vitro. This is in part due to the limited commercial availability of suitable, validated, dog-specific antibodies and reagents. Because of the close homology between canine and human Ang1 and Ang2 (97.6% and 92.3%, respectively),20 we hypothesized that antibodies against human Ang1 and Ang2 would cross-react with canine Ang1 and Ang2, allowing for the use of commercially available human ELISA kits for canine biological samples.
The purpose of the study reported here was to assess 2 commercially available human ELISA kits for measurement of Ang1 and Ang2 concentrations in canine plasma samples and to evaluate whether plasma Ang1 and Ang2 concentrations would differ between healthy dogs and dogs with bacterial sepsis. Another objective was to isolate, culture, and characterize pCAECs as a natural source of Ang2 and then use these cells for assay validation and to evaluate the effect of inflammation on Ang2 release from canine endothelial cells in vitro.
Materials and Methods
Ethics statements
The study protocol was approved by the Federal Food Safety and Veterinary Office (approval No. BLV 38/15). Owner consent was obtained for the collection of blood samples from healthy dogs. Owner consent for the use of leftover blood samples from septic dogs was obtained as part of the general hospital admission procedure. Aortas for pCAEC isolation were obtained from the bodies of dogs euthanized for various reasons that had been donated by the owners for research purposes.
Animals
Plasma samples from 10 healthy adult dogs and 10 dogs with bacterial sepsis were used for assay validation and to determine whether plasma Ang2 concentrations would differ significantly between healthy dogs and septic dogs. Healthy dogs were owned by staff and students and were deemed healthy on the basis of unremarkable findings in the medical history and on physical examination, CBC, and plasma biochemical analysis (including CRP concentration). Dogs were deemed to be septic on the basis of published clinical criteria for sepsis in dogs.21 Briefly, dogs were considered septic if they fulfilled at least 2 of the following criteria at admission to the intensive care unit: rectal temperature < 38.1° or > 39.2°C; heart rate > 120 beats/min; respiratory rate > 20 breaths/min; WBC count < 6.0 × 109 cells/L or > 16.0 × 109 cells/L; percentage of band cells > 3% of the total WBC count in conjunction with the bacteria identified on cytologic, histologic, or microbial culture evaluation of fluid or tissue samples; and intraoperative evidence of a septic focus. These dogs included those with a diagnosis of septic abdomen caused by intestinal perforation or translocation of bacteria (n = 4), pyometra (2), pyothorax (1), pyelonephritis (1), prostatic abscess (1), and pneumonia (1).
Sample collection and handling
Blood samples (2 mL) were collected from all dogs via a cephalic, saphenous, or jugular vein and 23-gauge, 1-inch needles into tubes containing lithium heparin.a Collected samples were centrifuged to separate plasma from the cellular components. Harvested plasma was then divided into aliquots and frozen at −80°C within 1 hour after collection. Plasma samples were analyzed for study purposes within 6 months after collection. Plasma CRP concentration was measured in samples from all dogs by use of a commercial assayb in accordance with the manufacturer's instructions.
Isolation, culture, characterization, and stimulation of pCAECs
Aortas were aseptically harvested from the canine cadavers within 20 minutes after euthanasia (which had been performed with an overdose of pentobarbitalc or propofold followed by KCle). Characteristics of these dogs are summarized (Supplementary Appendix S1, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.79.8.803). The ends of the aorta were tied, and the vessel was filled with collagenase-dispasef and incubated for 1 hour at 37°C. Endothelial cells were then collected by flushing the intima with DMEMg and gently rubbing it with a sterile cotton swab.
Harvested pCAECs were seeded on gelatin-coated surfaces of culture dishes and incubated at 37°C in 5% CO2 and in DMEM supplemented with bovine pituitary extract,h 1% penicillin-streptomycin, 1% l-glutamine, and 10% fetal calf serum.g Cell growth was assessed daily, and the culture medium was changed every 7 days. Endothelial cells were grown to confluence, and morphological characteristics were then assessed via light microscopy. Growth to confluence generally required 10 days when 3 × 104 pCAECs/cm2 had been seeded.
For further cell characterization, immunofluorescent staining for VE-cadherin (CD144) and vWF was performed. Endothelial cells were grown to confluence and fixed in 4% paraformaldehyde, then incubated with rabbit polyclonal anti-CD144 antibodyi diluted 1:200 in PBS solution or rabbit polyclonal anti-vWF antibodyj diluted 1:50 in PBS solution with 10% goat serum for 16 hours at 4°C. Slides were washed before incubation with the secondary antibodyk diluted 1:500 in PBS solution for 1 hour at room temperature (approx 20°C). Cell nuclei were stained with blue fluorescent bisbenzimide stain stock solution (1 mg/mL)l diluted 1:5,000 in PBS solution for 1 minute at room temperature. Immunofluorescent staining patterns for VE-cadherin and vWF were assessed, and images were obtained with a digital camera.m
For endothelial cell stimulation, 2.5 × 105 pCAECs were seeded in 25-cm2 flasks and grown to confluence. The culture supernatant was removed and replaced with fresh DMEM. The cells were then incubated with recombinant human TNF-αn (10 ng/mL of DMEM) for 24 hours. All cell culture supernatants were filteredo and frozen at −80°C. Supernatants of unstimulated pCAEC cultures and fresh medium without supplements were used as negative control substances. The Ang2 concentration of the medium was considered the baseline and therefore deduced from the final concentrations measured in supernatant from unstimulated and stimulated endothelial cells.
Preliminary assessment of canine Ang1 and Ang2 ELISA kits
We initially attempted to use commercial canine Ang1 and Ang2 ELISA kitsp to measure Ang1 and Ang2 in canine plasma samples. These solid-phase, competitive ELISAs based on polyclonal antibodies from rabbits against a recombinant full-length recombinant canine protein expressed in Escherichia coli yielded inconsistent results. Therefore, performance of the canine Ang1 and Ang2 ELISA kits was assessed by use of rcAng1q or rcAng2r as a positive control substance and canine insulin-like growth factor–binding protein,s which is a protein expressed in a similar E coli system, as a negative control substance.
Preliminary assessment of human Ang1 and Ang2 ELISA kits
The performance of 2 commercially available sandwich ELISA kitst,u based on monoclonal antibodies raised against mouse myeloma cell line NS0–expressed recombinant human Ang1 or Ang2 was subsequently assessed. Both assays are reported by the manufacturer to quantify human Ang1 or Ang2 concentrations in cell culture supernatants, serum, plasma, saliva, and milk (Ang1). Both ELISAs were performed in accordance with the manufacturer's instructions, and all measurements were performed in duplicate.
As a first step, the ability of the human ELISA kits to detect the canine proteins was assessed by use of serial dilutions of rcAng1q and rcAng2.r The correct concentration, identity, and purity of rcAng2 were verified via the Bradford method and immunoblotting.
Then, Ang1 and Ang2 concentrations were measured in canine plasma samples and supernatants from TNF-α–stimulated and unstimulated pCAEC cultures. Most canine plasma samples yielded Ang2 concentrations that exceeded the range of the standard curve when diluted to a concentration of 1:5 in accordance with the manufacturer's instructions. Samples were therefore measured at that concentration and then diluted as required (up to 50-fold) to obtain concentrations within the range of the standard curve (21 to 3,000 pg/mL) based on results for recombinant human Ang2.
To assess intra- and interassay variability, stability, spike-and-recovery results, and dilution linearity, plasma samples were diluted to a concentration of 1:5 and the measured Ang2 concentrations in diluted samples were reported. For comparison of Ang2 concentration in plasma samples from healthy and septic dogs, samples were diluted as required and the measured Ang2 concentrations were multiplied by the dilution factor to obtain the Ang2 concentration in the original sample, which was then reported. Because the human Ang1 ELISA kit failed to detect the rcAng1 fragment and Ang1 in plasma samples from 5 dogs (data not shown), this kit was not further assessed, and validation was only pursued for the human Ang2 ELISA kit.
Validation of the human Ang2 ELISA kit
Validation of the human Ang2 ELISA kit was performed following published guidelines for assay validation.22 Dilution linearity, parallelism, accuracy, precision, and reproducibility of the human Ang2 ELISA kit were determined by use of canine plasma samples, rcAng2, and supernatants from quiescent and TNF-α stimulated pCAEC cultures.
Intra- and interassay variability—Assay precision in consistent operating conditions (ie, same laboratory, technician, days, instrument, and reagent lot) was assessed by determining the intra-assay variability regarding 1 plasma sample with a low Ang2 concentration (387 pg/mL) and 1 plasma sample with a high Ang2 concentration (2,775 pg/mL) at a dilution of 1:5. Four replicates of each plasma sample were analyzed in duplicate within the same run. Interassay variability was evaluated by dividing 2 samples into 4 replicates (frozen until analyzed), which were analyzed consecutively on 3 days by use of 3 kits with the same lot number over a 1-month period.
Dilution linearity and parallelism—Dilution linearity was determined to evaluate whether canine plasma components (eg, sample matrix, complement, or heterophilic antibodies) interfered with the binding of canine Ang2 to the monoclonal anti-human Ang2 antibody of the ELISA at different dilutions. In addition to the serial dilutions of rcAng2 performed for the preliminary evaluation of the assay, serial dilutions of canine plasma samples with a broad range of Ang2 concentrations and supernatants from pCAEC cultures were made with calibrator diluent from the test kit. Furthermore, serial dilutions of 1 plasma sample from a healthy dog, spiked with supernatant from TNF-α-stimulated pCAEC cultures, were tested.
To examine whether the binding characteristics of canine Ang2 were similar to the calibrator stock solution containing human Ang2, the parallelism of the dilution curves was assessed by comparing the slopes of serial dilutions of plasma and supernatant samples with the rcAng2 standard curve.
Spike-and-recovery tests—Spike-and-recovery testing was performed to measure the accuracy of the human Ang2 ELISA kit by adding human Ang2 standard, rcAng2, or pCAEC supernatant following TNF-α stimulation to a plasma sample from a healthy dog diluted 1:5. Observed Ang2 concentrations were compared with the expected values.
Stability of canine Ang2 in plasma samples—Two canine plasma samples were diluted 1:5 with PBS solution and divided into aliquots. Afterward, Ang2 concentrations were measured before and after 1, 2, 3, and 4 freeze-thaw cycles and for 24 hours at room temperature.
Statistical analysis
Data were evaluated with the aid of statistical software.v Estimation of intra- and interassay variability was performed by calculation of coefficients of variation. Dilutional linearity was assessed through linear regression analysis to compare observed and expected concentrations of Ang2 in plasma samples and pCAEC supernatants.
For comparisons between healthy and septic dogs, continuous data (age and body weight) were assessed for normality of distribution with the Shapiro-Wilk test. Normally distributed data are reported as mean ± SD (range), and nonnormally distributed data are reported as median (IQR). Differences between healthy and septic dogs regarding plasma Ang2 and CRP concentrations were analyzed with the Kruskal-Wallis test for numeric data and the Pearson χ2 test for categorical data (dog sex). The association between plasma Ang2 and CRP concentrations was examined with the Spearman rank correlation test. Values of P < 0.05 were considered significant.
Results
Animals
No significant differences were identified between the 10 healthy dogs and 10 septic dogs regarding age, body weight, or sex distribution (Table 1). Various breeds were represented in both groups.
Characteristics and plasma CRP concentrations of 10 healthy and 10 septic dogs used for evaluation of a human Ang2 ELISA kit for measurement of plasma Ang2 concentration in dogs.
| Variable | Healthy dogs | Septic dogs | P value |
|---|---|---|---|
| Age (y) | 7.4 ± 2.1 (3.5–9.75) | 7.5 ± 2.9 (2.0–10.6) | 0.62 |
| Body weight (kg) | 28.9 ± 6 (20.3–35.8) | 24.7 ± 9.9 (4.9–42) | 0.24 |
| Reproductive status | 0.38 | ||
| Sexually intact male | 3 | 3 | — |
| Neutered male | 1 | 1 | — |
| Sexually intact female | 5 | 4 | — |
| Spayed female | 1 | 2 | — |
| Plasma CRP (mg/L) | 4.2 ± 5 (0–14.5) | 127 ± 77.5 (6.2–298.6) | < 0.001 |
Data for reproductive status represent counts; all other data represent mean ± SD (range). — = Not applicable.
pCAEC culture
Isolation of pCAECs from the cadavers of young dogs (< 1 year of age) yielded more cells, which grew faster and to higher passages than cells isolated from the cadavers of adult dogs. Although not systematically evaluated, the yield of endothelial cells appeared lower when pentobarbital had been used for euthanasia rather than a combination of propofol and KCl.
For the data reported here, pCAECs isolated from a 3-month-old Chihuahua euthanized because of a congenital urinary tract malformation were used. The harvested pCAECs were successfully grown to confluence in culture dishes and had the typical cobblestone appearance of quiescent endothelial cells. These pCAECs had positive staining at cell-cell junctions for VE-cadherin (CD144) and positive intracellular granular vWF staining consistent with the containment of vWF in cytoplasmic Weibel-Palade bodies, consistent with endothelial cells (Figure 1). Stimulation of pCAECs with TNF-α resulted in morphological changes from a cobblestone to an elongated appearance.
Representative photomicrographs of cultured pCAECs with immunofluorescent labeling for VE-cadherin (A to C) and vWF (D to F). Endothelial cells were grown to confluency and fixed in 4% paraformaldehyde, then stained for VE-cadherin with rabbit polyclonal anti-CD144 antibody or for vWF with rabbit polyclonal-anti vWF antibody. Fluorescent-labeled goat anti-rabbit IgG was used as the secondary antibody. Cell nuclei were stained with blue fluorescent bisbenzimide stain. Merged images (C and F) are overlays of the images in panels A and B and D and E. In panel F, inset represents higher (100×) magnification. Bar = 20 μm in all images.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Representative photomicrographs of cultured pCAECs with immunofluorescent labeling for VE-cadherin (A to C) and vWF (D to F). Endothelial cells were grown to confluency and fixed in 4% paraformaldehyde, then stained for VE-cadherin with rabbit polyclonal anti-CD144 antibody or for vWF with rabbit polyclonal-anti vWF antibody. Fluorescent-labeled goat anti-rabbit IgG was used as the secondary antibody. Cell nuclei were stained with blue fluorescent bisbenzimide stain. Merged images (C and F) are overlays of the images in panels A and B and D and E. In panel F, inset represents higher (100×) magnification. Bar = 20 μm in all images.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Representative photomicrographs of cultured pCAECs with immunofluorescent labeling for VE-cadherin (A to C) and vWF (D to F). Endothelial cells were grown to confluency and fixed in 4% paraformaldehyde, then stained for VE-cadherin with rabbit polyclonal anti-CD144 antibody or for vWF with rabbit polyclonal-anti vWF antibody. Fluorescent-labeled goat anti-rabbit IgG was used as the secondary antibody. Cell nuclei were stained with blue fluorescent bisbenzimide stain. Merged images (C and F) are overlays of the images in panels A and B and D and E. In panel F, inset represents higher (100×) magnification. Bar = 20 μm in all images.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Preliminary assessment of canine Ang2 ELISA kit
No difference was identified in the optical densities obtained for rcAng2 and canine insulin-like growth factor–binding protein (negative control substance) measured at various dilutions, indicating that the canine assay was unable to detect rcAng2 (Supplementary Table S1, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.79.8.803).
Validation of the human Ang2 ELISA kit
Preliminary assessment—The human Ang2 ELISA kit was able to detect rcAng2. Comparison of the standard curve established with recombinant human Ang2 provided for this ELISA revealed that results obtained for serial dilutions of the commercially available rcAng2 fragment had good linearity between concentrations of 41 and 1,597 pg/mL (Figure 2). However, the measured concentrations of rcAng2 were 4 to 5 times lower when the human standard curve was used than the expected concentration reported by the manufacturer of the rcAng2, even though the correct concentration, purity, and identity of the rcAng2 fragment were clearly confirmed via Bradford assay and western blot analysis (Supplementary Figure S1, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.79.8.803).
Observed versus expected concentrations of rcAng2 measured with a human Ang2 ELISA kit. The human Ang2 ELISA kit was able to detect rcAng2. Measured concentrations of the rcAng2 fragment were significantly lower than expected on the basis of the concentration reported by the protein manufacturer. The R2 value for the linear regression equation of this line was 0.9996.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Observed versus expected concentrations of rcAng2 measured with a human Ang2 ELISA kit. The human Ang2 ELISA kit was able to detect rcAng2. Measured concentrations of the rcAng2 fragment were significantly lower than expected on the basis of the concentration reported by the protein manufacturer. The R2 value for the linear regression equation of this line was 0.9996.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Observed versus expected concentrations of rcAng2 measured with a human Ang2 ELISA kit. The human Ang2 ELISA kit was able to detect rcAng2. Measured concentrations of the rcAng2 fragment were significantly lower than expected on the basis of the concentration reported by the protein manufacturer. The R2 value for the linear regression equation of this line was 0.9996.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Linearity, parallelism, and intra- and interassay variability—Excellent linearity and parallelism were identified between standard curves for human Ang2 and serial dilutions of canine plasma samples containing a range of Ang2 concentrations, supernatant of TNF-α–stimulated pCAECs, and canine plasma samples spiked with supernatant (Figure 3). Respective coefficients of variation reflecting intra- and interassay variability for 4 canine plasma samples with known low (mean, 570 ± 31 pg/mL) and high (mean, 2,292 ± 128 pg/mL) Ang2 concentrations assayed at a dilution of 1:5 were 5.6% (both concentrations) and 6.7% and 4.0% (low and high concentrations, respectively).
Observed versus expected canine Ang2 concentrations as measured with a human Ang ELISA kit. The Ang2 was measured in serial dilutions (1:2, 1:4, 1:8, and 1:16) of canine plasma samples (black line and triangles; R2 = 0.9983), supernatant of pCAEC cultures after cell stimulation with TNF-α (green line and squares; R2 = 0.9997), and canine plasma samples spiked with supernatant of pCAEC cultures after cell stimulation with TNF-α (pink line and circles; R2 = 0.9999). These results indicate excellent dilutional linearity and parallelism of the dilution curves.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Observed versus expected canine Ang2 concentrations as measured with a human Ang ELISA kit. The Ang2 was measured in serial dilutions (1:2, 1:4, 1:8, and 1:16) of canine plasma samples (black line and triangles; R2 = 0.9983), supernatant of pCAEC cultures after cell stimulation with TNF-α (green line and squares; R2 = 0.9997), and canine plasma samples spiked with supernatant of pCAEC cultures after cell stimulation with TNF-α (pink line and circles; R2 = 0.9999). These results indicate excellent dilutional linearity and parallelism of the dilution curves.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Observed versus expected canine Ang2 concentrations as measured with a human Ang ELISA kit. The Ang2 was measured in serial dilutions (1:2, 1:4, 1:8, and 1:16) of canine plasma samples (black line and triangles; R2 = 0.9983), supernatant of pCAEC cultures after cell stimulation with TNF-α (green line and squares; R2 = 0.9997), and canine plasma samples spiked with supernatant of pCAEC cultures after cell stimulation with TNF-α (pink line and circles; R2 = 0.9999). These results indicate excellent dilutional linearity and parallelism of the dilution curves.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Spike-and-recovery testing—Percentage recovery of human Ang2 in canine plasma samples spiked with low, medium, and high concentrations of human Ang2 ranged between 85% and 129%, and that for and rcAng2 ranged between 85% and 95%. Recovery of Ang2 from canine plasma samples spiked with low, medium, and high concentrations of supernatant of pCAEC cultures after cell stimulation with TNF-α ranged between 71% and 123% (Table 2).
Results of spike-and-recovery testing involving use of a human Ang2 ELISA kit to measure Ang2 concentrations in canine plasma samples spiked with rcAng2, human Ang2, or supernatant from TNF-α–stimulated pCAECs at various Ang2 concentrations.
| Ang2 concentration (pg/mL) | No. of dogs | Mean ± SD percentage recovery | Range in percentage recovery |
|---|---|---|---|
| rcAng2 | |||
| 625 | 2 | 93.9 ± 0.1 | 93.8–94.0 |
| 1,250 | 2 | 90.9 ± 6.4 | 86.4–95.4 |
| 2,500 | 2 | 88.1 ± 5.1 | 84.5–91.7 |
| Human Ang2 | |||
| 93.7 | 3 | 107.8 ± 21.9 | 85.4–129.1 |
| 375 | 3 | 95.1 ± 18.1 | 85.1–116 |
| 1,500 | 3 | 98.1 ± 7.7 | 89.3–103.3 |
| Supernatant | |||
| 250 | 3 | 85.9 ± 2.4 | 83.9–88.6 |
| 550 | 3 | 98.1 ± 21.8 | 84.4–123.3 |
| 1,100 | 3 | 74.6 ± 4.35 | 70.5–88.6 |
Stability of Ang2 in canine plasma samples—Results of stability testing indicated that Ang2 was stable in canine plasma samples for 4 freeze-thaw cycles and for 24 hours at room temperature (Supplementary Table S2, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.79.8.803).
Plasma Ang2 concentrations in healthy and septic dogs
The median plasma Ang2 concentration in dogs with bacterial sepsis (25.5 ng/mL; IQR, 19.9 to 69.9 ng/mL) was significantly (P = 0.002) higher than that in healthy dogs (6.7 ng/mL; IQR, 5.9 to 9.8 ng/mL). In addition, median plasma CRP concentration was significantly (P < 0.001) higher in septic dogs (117.2 mg/L; IQR, 88.7 to 158.8 mg/L) than in healthy dogs (2.7 mg/L; IQR, 0.15 to 5.9 mg/L). Plasma Ang2 concentrations were positively correlated with plasma CRP concentrations (R2 = 0.62; P < 0.001; Figure 4).
Scatterplot of the correlation between plasma Ang2 and CRP concentrations in healthy (n = 10; blue dots) and septic (10; red dots) dogs.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Scatterplot of the correlation between plasma Ang2 and CRP concentrations in healthy (n = 10; blue dots) and septic (10; red dots) dogs.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Scatterplot of the correlation between plasma Ang2 and CRP concentrations in healthy (n = 10; blue dots) and septic (10; red dots) dogs.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.803
Supernatant Ang2 concentrations
The Ang2 concentration in pooled supernatants from 3 quiescent pCAEC cultures was 149 pg/mL. Stimulation of pCAECs with TNF-α led to a marked increase in this concentration (median, 13,073 pg/mL; IQR, 9,403 to 13,711 pg/mL).
Discussion
In the study reported here, we initially attempted to measure Ang2 concentrations in canine plasma samples using a commercial canine Ang2 ELISA kit but obtained inconsistent results. According to the manufacturer, this assay is based on a polyclonal antibody against full-length rcAng2 expressed in E coli. Despite use of a polyclonal antibody with presumably a broader range of detection of Ang2-specific epitopes, the assay was unable to detect rcAng2, thereby suggesting a lack of sensitivity of the antibody. We therefore set out to determine whether commercially available human Ang1 and Ang2 ELISA kits could be suitable for detection of Ang1 and Ang2 in canine samples. The secondary goal was to isolate, culture, and characterize pCAECs, which are a natural source of Ang2. These cells were then used for assay validation and to evaluate the effect of inflammation on Ang2 release from pCAECs in vitro.
The human Ang1 ELISA was unable to detect a recombinant fragment of canine Ang1 and also failed to detect Ang1 in the tested canine plasma samples. This was unexpected, given that homology between canine and human Ang1 is reportedly high (97.6%) and indeed higher than the homology between human and canine Ang2 (92.3%).20 Because the human Ang2 ELISA kit had promising results during preliminary analyses, a decision was made to focus on validation of the Ang2 ELISA kit.
After several attempts to isolate pCAECs from aortas of adult dogs, which led to low cell yield with slow growth and limited survival, pCAECs were successfully isolated and cultured from juvenile dogs and characterized via positive linear staining for VE-cadherin at cell-cell junctions and intracytoplasmic granular staining for vWF, which are typical characteristics of endothelial cells. Consistent with results achieved with murine and human macrovascular endothelial cell cultures,23,24 we were able to show that stimulation of pCAECs with TNF-α led to endothelial activation, as demonstrated by cellular morphological changes and Ang2 release. Supernatants of stimulated cells were then used as a source of natural canine Ang2 for assay validation.
Although the dilution curve was linear for rcAng2 in the study reported here, the measured concentrations were approximately 4 to 5 times lower than expected. This low analytic sensitivity could be explained by the fact that the rcAng2 is a fragment of the full-length protein from which important epitopes may be missing, the potential for a lack of posttranslational modifications by this E coli–expressed protein fragment, and the existence of small differences between human and canine Ang2 that affect protein structure. Because of these discrepancies and limitations, the measured canine Ang2 concentrations in plasma samples and supernatant from pCAEC cultures may need to be considered as relative values.
Dilution linearity and parallelism of the evaluated human Ang2 ELISA were excellent for serial dilutions of canine plasma, supernatant of stimulated pCAEC cultures, and rcAng2 (R2 = 0.99). Spike-and-recovery testing also yielded excellent results when human Ang2 (85% to 129%) and canine recombinant Ang2 (85% to 95%) were added to canine plasma and quite acceptable results when supernatant of stimulated pCAEC cultures was added (71% to 123%). Typically, a percentage recovery value between 80% and 120% is desirable.22 Test reproducibility and precision indicated by intra- and interassay variation were excellent (range in coefficients of variation, 4.0% to 6.7%), with a value < 10% to 15% usually considered an acceptable standard.22 These findings indicated that the human Ang2 ELISA kit appropriately detected rcAng2 and Ang2 in canine plasma samples and pCAEC culture supernatants and can therefore be used to determine relative concentrations of canine Ang2 in these fluids. Consistent with reported data for human Ang2,16,25 canine Ang2 was stable in plasma samples stored at room temperature for 24 hours and following 4 freeze-thaw cycles.
Measured plasma Ang2 concentrations for healthy dogs (median, 6.7 ng/mL; IQR, 5.9 to 9.8 ng/mL) were approximately 3 times as high as those reported for healthy human volunteers.7,11,12 A larger group of healthy dogs would be needed to establish reference intervals for plasma Ang2 concentration in dogs. Plasma Ang2 concentrations in septic dogs were significantly higher than those in healthy dogs, which is consistent with reported data for humans9 and mice.26 Plasma Ang2 concentrations in septic dogs varied greatly (median, 25.5 ng/mL; IQR, 19.9 to 69.9 ng/mL), but were not drastically different from those reported for humans with sepsis (between 3.4 and 41 ng/mL, depending on the study).7,11,12 Whether associations exist among plasma Ang2 concentration, disease severity, and outcome in dogs, as reported for humans,10–18 will need to be determined for a larger number of dogs than was included in the present study.
Plasma CRP concentrations were significantly higher in septic dogs than in healthy dogs, reflecting severe systemic inflammation. These concentrations were positively correlated with plasma Ang2 concentrations. This correlation could be explained by the proinflammatory effects of Ang2 or, inversely, by the release of Ang2 from endothelial cells during inflammation.27 Whether the association between high circulating amounts of Ang2 and adverse outcomes in septic humans10–18 is primarily due to the destabilization of the endothelium or additionally depends on the proinflammatory effects of Ang2 remains to be investigated.
The human ELISA kit evaluated in the present study was able to detect canine Ang2 in plasma samples and endothelial cell supernatant. Plasma Ang2 concentration was higher in septic dogs than in healthy dogs, and Ang2 was released from cultured pCAECs following stimulation by TNF-α, similar to findings for other species. We believe that with the validation of the Ang2 ELISA and the establishment of pCAEC isolation and culture techniques, this study yielded 2 important tools for the in vivo and in vitro evaluation of Ang2 as a marker of endothelial cell activation in dogs.
Acknowledgments
Funded by a grant by the Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Bern, Bern, Switzerland.
The authors declare that there were no conflicts of interest. The authors thank Sigridur Jonsdottir for performance of the Coomassie blue staining and western blot analysis.
ABREVIATIONS
| Ang1 | Angiopoietin-1 |
| Ang2 | Angiopoietin-2 |
| CRP | C-reactive protein |
| DMEM | Dulbecco modified Eagle medium |
| IQR | Interquartile (25th to 75th percentile) range |
| pCAEC | Primary canine aortic endothelial cell |
| rcAng1 | Recombinant canine angiopoietin-1 |
| rcAng2 | Recombinant canine angiopoietin-2 |
| TNF-α | Tumor necrosis factor-α |
| VE | Vascular endothelial |
| vWF | Von Willebrand factor |
Footnotes
Microtube (1.3 mL) lithium-heparin, Sarstedt AG, Nürnberg, Germany.
Canine CRP assay, Randox Reagents, London, England.
Euthasol 40%, Virbac AG, Glattbrugg, Switzerland.
Propofol 1% MCT, Fresenius Kabi AG, Oberdorf, Switzerland.
KCl 7.45%, Braun Medical AG, Crissier, Switzerland.
Collagenase/dispase, Roche, Basel, Switzerland.
ThermoFisher Scientific, Walham, Mass.
Bovine pituitary extract, Sigma-Aldrich Chemie GmbH, Darmstadt, Germany.
Bioss Antibodies, Woburn, Mass.
Dako, Agilent Pathology Solutions, Santa Clara, Calif.
Alexa Fluor 488 goat anti-rabbit IgG, Invitrogen Corp, Eugene, Ore.
Hoechst stain solution, Sigma-Aldrich Corp, Chemie GmbH, Darmstadt, Germany.
Hamamatsu digital camera C10600 ORCA-R mounted on a Nikon 80i fluorescent microscope, Hamamatsu Photonics Ltd, Welwyn Garden City, England.
Invitrogen, Carlsbad, Calif.
Millex GV filter unit (0.22 μm), Merck Millipore, Carrigtwohill, Ireland.
Blue Gene Biotech, Shanghai, China.
rcAng1 fragment protein (molecular weight, 15.34 kDa), antibodies-online, Atlanta, Ga.
rcAng2 fragment protein (Lys274-Phe495; molecular weight, 27.8 kDa), SinoBiological, Peking, China.
SinoBiological, Peking, China.
Human Ang1 immunoassay, R&D Systems, Minneapolis, Minn.
Human Ang2 immunoassay, R&D Systems, Minneapolis, Minn.
NCSS11 statistical software, version 2016, NCSS LLC, Kaysville, Utah.
References
1. Aird WC. The role of the endothelium in severe sepsis and multiple organ dysfunction syndrome. Blood 2003;101:3765–3777.
2. Witzenbichler B, Westermann D, Knueppel S, et al. Protective role of angiopoietin-1 in endotoxic shock. Circulation 2005;111:97–105.
3. Thurston G, Rudge JS, Ioffe E, et al. The anti-inflammatory actions of angiopoietin-1. EXS 2005;233–245.
4. Thurston G, Rudge JS, Ioffe E, et al. Angiopoietin-1 protects the adult vasculature against plasma leakage. Nat Med 2000;6:460–463.
5. Fiedler U, Reiss Y, Scharpfenecker M, et al. Angiopoietin-2 sensitizes endothelial cells to TNF-alpha and has a crucial role in the induction of inflammation. Nat Med 2006;12:235–239.
6. Fiedler U, Scharpfenecker M, Koidl S, et al. The Tie-2 ligand angiopoietin-2 is stored in and rapidly released upon stimulation from endothelial cell Weibel-Palade bodies. Blood 2004;103:4150–4156.
7. Lukasz A, Hillgruber C, Oberleithner H, et al. Endothelial glycocalyx breakdown is mediated by angiopoietin-2. Cardiovasc Res 2017;113:671–680.
8. Fang Y, Li C, Shao R, et al. Prognostic significance of the angiopoietin-2/angiopoietin-1 and angiopoietin-1/Tie-2 ratios for early sepsis in an emergency department. Crit Care 2015;19:367.
9. Milam KE, Parikh SM. The angiopoietin-Tie2 signaling axis in the vascular leakage of systemic inflammation. Tissue Barriers 2015;3:e957508.
10. Siner JM, Bhandari V, Engle KM, et al. Elevated serum angiopoietin 2 levels are associated with increased mortality in sepsis. Shock 2009;31:348–353.
11. Lymperopoulou K, Velissaris D, Kotsaki A, et al. Angiopoietin-2 associations with the underlying infection and sepsis severity. Cytokine 2015;73:163–168.
12. Orfanos SE, Kotanidou A, Glynos C, et al. Angiopoietin-2 is increased in severe sepsis: correlation with inflammatory mediators. Crit Care Med 2007;35:199–206.
13. David S, Mukherjee A, Ghosh CC, et al. Angiopoietin-2 may contribute to multiple organ dysfunction and death in sepsis. Crit Care Med 2012;40:3034–3041.
14. Kümpers P, Lukasz A, David S, et al. Excess circulating angiopoietin-2 is a strong predictor of mortality in critically ill medical patients. Crit Care 2008;12:R147.
15. Calfee CS, Gallagher D, Abbott J, et al. Plasma angiopoietin-2 in clinical acute lung injury: prognostic and pathogenetic significance. Crit Care Med 2012;40:1731–1737.
16. Lukasz A, Hellpap J, Horn R, et al. Circulating angiopoietin-1 and angiopoietin-2 in critically ill patients: development and clinical application of two new immunoassays. Crit Care 2008;12:R94.
17. Davis JS, Yeo TW, Piera KA, et al. Angiopoietin-2 is increased in sepsis and inversely associated with nitric oxide-dependent microvascular reactivity. Crit Care 2010;14:R89.
18. Kranidioti H, Orfanos SE, Vaki I, et al. Angiopoietin-2 is increased in septic shock: evidence for the existence of a circulating factor stimulating its release from human monocytes. Immunol Lett 2009;125:65–71.
19. Retzlaff J, Thamm K, Ghosh CC, et al. Flunarizine suppresses endothelial angiopoietin-2 in a calcium-dependent fashion in sepsis. Sci Rep 2017;7:44113.
20. Kato Y, Asano K, Mizutani I, et al. Gene expressions of canine angiopoietin-1 and -2 in normal tissues and spontaneous tumours. Res Vet Sci 2006;81:280–286.
21. Hauptman JG, Walshaw R, Olivier NB. Evaluation of the sensitivity and specificity of diagnostic criteria for sepsis in dogs. Vet Surg 1997;26:393–397.
22. Andreasson U, Perret-Liaudet A, van Waalwijk van Doorn LJ, et al. A practical guide to immunoassay method validation. Front Neurol 2015;6:179.
23. Rondaij MG, Bierings R, Kragt A, et al. Dynamics and plasticity of Weibel-Palade bodies in endothelial cells. Arterioscler Thromb Vasc Biol 2006;26:1002–1007.
24. Chan JM, Cheung MS, Gibbs RG, et al. MRI detection of endothelial cell inflammation using targeted superparamagnetic particles of iron oxide (SPIO). Clin Transl Med 2017;6:1.
25. Parikh SM, Mammoto T, Schultz A, et al. Excess circulating angiopoietin-2 may contribute to pulmonary vascular leak in sepsis in humans. PLoS Med 2006;3:e46.
26. Lomas-Neira J, Venet F, Chung CS, et al. Neutrophil-endothelial interactions mediate angiopoietin-2-associated pulmonary endothelial cell dysfunction in indirect acute lung injury in mice. Am J Respir Cell Mol Biol 2014;50:193–200.
27. Scholz A, Plate KH, Reiss Y. Angiopoietin-2: a multifaceted cytokine that functions in both angiogenesis and inflammation. Ann N Y Acad Sci 2015;1347:45–51.
