Microbial Pathogenesis 1992 ; 13 :191-202
Effect of Haemophilus somnus on phagocytosis and hydrogen peroxide production by bovine polymorphonuclear leukocytes* Cheryl G . Pfeifer, 2 Manuel Campos, 2 Terry Beskorwayne, 2 Lorne A . Babiuk' 2 and Andrew A . Potter'
1 Canadian Bacterial Disease Network and 2 Veterinary Infectious Diseases Organization, 124 Veterinary Road, Saskatoon, Saskatchewan, Canada, S7N OWO (Received May 14, 1992 ; accepted in revised form June 30, 1992)
Pfeifer, C . G . (Veterinary Infectious Diseases Organization, 124 Veterinary Road, Saskatoon, Saskatchewan, Canada, S7N OWO), M . Campos, T. Beskorwayne, L . A . Babiuk and A . A . Potter . Effect of Haemophilus somnus on phagocytosis and hydrogen peroxide production by bovine polymorphonuclear leukocytes . Microbial Pathogenesis 1992 ; 13 :191-202 . The interactions between bovine polymorphonuclear leukocytes (PMNs) and the bacterium Haemophilus somnus are known to be complex . In this paper, we evaluated the effect of H. somnus on PMN function using a flow cytometric (FC) technique that simultaneously determined the extent of phagocytosis and hydrogen peroxide production by PMNs, as well as using conventional techniques, such as the nitroblue tetrazolium (NBT) and chemiluminescence assays, to analyse the PMN respiratory burst . Results from the FC and chemiluminescence assays demonstrated that in vitro exposure of PMNs to logarithmically growing H. somnus reduced the respiratory burst of PMNs obtained from healthy calves . However, this reduction was not detected by the NBT assay . A decrease in phagocytosis by PMNs could also be shown using the FC assay . In addition, PMNs from calves with acute Hemophilosis (i .e . exposed to H. somnus in vivo) showed reduced activity when compared to PMNs from healthy calves . These in vitro and in vivo observations indicate that the modulation of bovine PMN function by H . somnus may contribute significantly towards the pathogenesis of the disease . Key words : Haemophilus somnus ; polymorphonuclear leukocyte (PMN) ; flow cytometry ; phagocytosis ; respiratory burst ; NBT ; chemi luminescence .
Introduction The response of phagocytic cells to a bacterial pathogen is one of the most important defence mechanisms against infection . Phagocytes not only play an important role in the killing of microorganisms but provide the only means of clearing debris from invaded tissues .' - ' Haemophilus somnus, the organism responsible for the Hemophilosis disease complex in cattle, has been shown to interfere with the activities of bovine phagocytic cells and it has been suggested that the interactions between polymorphonuclear leukocytes (PMNs) and this bacteria are critical in the development of the variety of clinical manifestations associated with the disease . Support for this hypothesis is that H . somnus can inhibit protein iodination,' Staphylococcus aureus uptake' and chemiluminescence 10 by PMNs . However, the formation of formazan in the nitroblue tetra"Published with the permission of the Director of VI DO as Journal Series no . 146 . 0882-4010/92/090201 +12 $08 .00/0
© 1992 Academic Press Limited
C . G . Pfeifer et al.
zolium (NBT) assay by PMNs is not affected .' Also H. somnus has been shown to survive inside PMNs 11 and replicate in mononuclear phagocytes . 12 One of the classical responses to phagocytosis by PMNs is the respiratory burst, characterized by an increase in oxygen uptake and the production of highly reactive oxygen species ." " There are a variety of assays that are used to detect the respiratory burst in PMNs . Two commonly used techniques include the NBT assay' 1 2' and the chemiluminescence assay . 23-2' Although both assays detect the production of oxygen radicals by PMNs, the methods of detection differ . The NBT assay detects a change in colour or absorbance of NBT as it is converted to formazan during the production of superoxide anions, 18,20,22 while the chemiluminescence assay detects the emission of light that is released when the excited oxygen species return to ground state . Excited oxygen species which can be detected include superoxide anions (02) as well as hydroxyl radicals ('OH), singlet oxygen ( 1 0 2 ) and hydrogen peroxide (H 2 0 2 ) . 23,27 However, both procedures give no information regarding the status of individual cells and only one parameter of cell function (i .e . respiratory burst) may be measured at a time . While microscopy can be used to analyse the status of individual PMNs in a variation of the NBT assay, the method is not quantitative . 20 Another method used to detect the respiratory burst in phagocytes is flow cytometry (FC) which involves the use of the peroxide-sensitive dye 2',7'-dichlorofluoresceindiacetate (DCFH-DA), 2a-sa This sytem was developed to include the detection of phagocytosis by PMNs through the use of propidium iodide (PI)-labelled bacteria, therefore, it is now possible to simultaneously measure the phagocytosis and hydrogen peroxide production of neutrophils . 30 Through comparison of the results found using the FC assay with those of the NBT and chemiluminescence assays, it was determined that flow cytometric measurement of PMN parameters could be a valuable tool for studying the effects of H. somnus and other microorganisms on phagocyte function . In this report, evidence is presented to indicate that H. somnus is able to reduce the phagocytic capacity and respiratory burst potential of healthy bovine PMNs when added in vitro . Furthermore, PMNs from cattle undergoing acute infection with H. somnus had reduced activity .
Results FC analysis of cellular phagocytosis and hydrogen peroxide production Using FC it was possible to define the different cell types present in whole blood (minus the red blood cells which were lysed in order to cut down on total cell volume) . Cells were compared according to cell size and granularity in order to differentiate between the various cell types, i .e . lymphocytes, monocytes and PMNs (Fig . 1 ) . Bitmaps or gates were drawn around the cells and then the individual cells in the bitmaps were analysed for red and green fluorescence . A quadrant system was used for analysis of the graphs displaying red versus green fluorescence where quadrant 2 represented cells expressing dual fluorescence, indicating they were phagocytising bacteria and producing hydrogen peroxide . Total phagocytosis was determined by the addition of quadrants 1 and 2, while total hydrogen peroxide production was determined by the addition of quadrants 2 and 4 . Quadrant 3 represented cells that were not active and remained fluorescently negative . There was a positive correlation between phagocytosis and hydrogen peroxide production by PMNs (Fig . 1) . In general, 60-85% of the bovine PMNs routinely responded by both ingesting the Staphylococcus aureus and producing hydrogen peroxide, with only 2-20% of the PMNs considered 'non-responders .' Non-responder PMNs did not exhibit fluorescence above the background level, indicating these PMNs
Haemophilus somnus and phagocytosis
2 60- (a)
20 30 40 50 60 Green fluorescence
20 30 40
Fig . 1 . Gating of cell populations within whole blood and their fluorescence as detected by the FC assay . (a) The bitmaps (gates) drawn define the different cell types that are to be further analysed by FC . (b) Bitmap 1 represents lymphocytes and the dual colour analysis is shown, where 99% of the lymphocytes remain in the negative zone indicating they are neither phagocytising nor producing hydrogen peroxide . (c) Bitmap 2 represents blood monocytes and the dual colour analysis is shown, where 22% of the monocytes respond with phagocytosis and less than 1 % produce hydrogen peroxide . (d) Bitmap 3 represents PM Ns and the dual colour analysis is shown, where 76% of the PM Ns respond by both actively phagocytising and producing hydrogen peroxide .
were not stimulated by the presence of PI-labelled S . aureus and DCFH-DA . The wide variations in the numbers of PMNs responding reflected the differences between each of the calves, although each individual calf tended to vary less than 10% over time . Less than 25% of the bovine monocytes were found to actively engage in phagocytosis of the bacteria and no more than 2% of this population responded by producing hydrogen peroxide (Fig . 1 ) . Lymphocytes served as negative controls in all cases (Fig . 1 ) .
Effect of in vitro H . somnus treatment on the respiratory burst of PMNs as detected by various assays The chemiluminescence assay detected a decrease in the respiratory burst of PMNs from healthy calves in the presence of live, logarithmically growing H . somnus (Table 1) .
C . G . Pfeifer
Effect of H. somnus on the respiratory burst of bovine Table 1 PMNs as detected by various assays Assays detecting PMN respiratory burst" FC (% positive cells)"
Chemiluminescence (total counts)"'
-H. somnus +H . somnus
75 .8 22 .8
5 .1 X 10 7 7 .6 x 10'
NBT (0D 572 ) " 0 .166 0 .215
"All assays used heat killed S. aureus to stimulate the PMN respiratory burst response . t FC results are representative of the percentage PMNs that are actively phagocytic and producing hydrogen peroxide both in the presence of H . somnus or not . Chemiluminescence results represent the total amount of light emitted over 90 min from excited oxygen species during PMN respiratory burst . NBT results represent the absorbance value of formazan once it had been reduced in the PMN . Both the chemiluminescence and NBT assay results were averaged from two separate experiments . Duplicates were analysed for each experiment for each assay . `Whole blood was used . d % AR = percentage of PMN activity remaining in each assay as described above : (resultsof PM Ns with H. somnus) R
(results of PM Ns without H. somnus)
This decrease was expressed as the percentage of PMN activity remaining and was calculated using the following formula : total light emission of sample with H. somnus %of PMN activity remaining =
total light emission of sample without H. somnus
The reduction in the light emission was such that only about 15% of the PMNs that responded in the control were able to respond when H. somnus was present . It was not possible to determine the status of phagocytosis by PMNs, as the production of light in the assay could not be directly correlated with phagocytosis . 26 The overall decrease in the light emission by the PMNs was not seen to the same extent in the presence of stationary phase H. somnus, although the total emission was reduced by about 11 % (results not shown) . In contrast to chemiluminescence, the NBT assay did not detect a decrease in PMN oxidative metabolism in the presence of either log or stationary phase H. somnus . Furthermore, results from the NBT assay suggested that the addition of log phase H. somnus to the PMN sample increased oxygen radical formation to about 30% above the PMN control sample (Table 1 ) . The addition of H. somnus to the NBT assay in the presence of S . aureus appeared to increase the PMN respiratory burst by the same amount when H. somnus was added alone in the presence of PMNs . However, it was determined that live H. somnus itself was able to reduce the NBT to formazan without the presence of PMNs, such that the difference seen in Table 1 could have been due to the presence of H. somnus rather than an increase in PMN activity (results not shown) . Heat-killed S. aureus was not able to reduce NBT . The status of PMN phagocytosis of the bacteria was not determined . Using the FC assay it was observed that live H. somnus induced a decrease in phagocytosis and hydrogen peroxide production by healthy PMNs (Table 1 ; Fig . 2) . Logarithmically growing H. somnus reduced hydrogen peroxide production of the PMNs to about 20-30% of their normal production . Overall phagocytosis by these
Haemophilus somnus and phagocytosis
40 oft 30
o ~+ 0 10
Green fluorescence Fig . 2 . The effects of logarithmically growing H. somnus on bovine PMNs as shown by the FC assay . (a) Control PMN sample, with quadrants 1, 2, 3 and 4 containing 8, 76, 9 and 7% PMNs, respectively . (b) PMN sample treated with H . somnus, with quadrants 1, 2, 3 and 4 containing 36, 23, 36 and 5% PMNs, respectively . Note the decrease in the number of PMNs in quadrant 2 in the presence of H. somnus .
PMNs was also reduced to about 70% of the controls . The decrease in phagocytosis was not necessarily due to competition for space within the PMN by bacteria not labelled with PI, since the decrease in phagocytosis was not seen in the presence of formalin-killed, heat-killed or stationary phase H. somnus (results not shown) . The addition of the killed or stationary phase H. somnus to the PMNs costimulated with S . aureus resulted in an increase in the numbers of phagocytic PMNs, rather than the decrease seen with the addition of actively growing H . somnus . Effect of H . somnus infection in vivo on PMN function in vitro The FC assay was able to detect a decrease in both hydrogen peroxide production and phagocytic ability of PMNs from healthy calves following induction by the in .vitro addition of H. somnus . Therefore, this assay was selected to study the effects of H . somnus infection on PMN activity . The results involve the comparison of bovine PMN responses before and after infection with H. somnus . Calves were challenged with H . somnus using the intravenous challenge model described by Harland et al." with clinical signs seen as fever, depression, bacteraemia and polyarthritis . Blood samples for FC determinations of PMN function were collected at 2, 4 and 7 days after the calves were challenged with H. somnus and analysed using the parameters established in the in vitro FC assay . In these experiments, Dulbecco's phosphate-buffered saline (PBSg) was substituted for H. somnus in the microtitre wells . In all instances where H. somnus infection induced clinical disease, a decrease in phagocytosis and hydrogen peroxide production by PMNs was observed indicating that the PMNs were being affected in vivo by H . somnus . Table 2 illustrates the decrease in PMN function seen throughout the trial period . Between days 1 and 4, transient bacteraemia was present, but by day 7 the disease was more localized in the tissues, where myocarditis and arthritis of the joints were prevalent . The reduction in PMN activity was correlated with acute clinical disease, as animals that had been challenged with H. somnus but did not exhibit clear signs of Hemophilosis showed variable results (results not shown) .
C . G . Pfeifer
Table 2 PMNs
The effect of H . somnus infection on bovine
PMN activity remaining (days after challenge with H. somnus) b Animal
A B C D E
103 .7 102 .7 104 .7 88 .9' 83 .5
84 .7 66 .4` 71 .3` 77 .8` 67 .3`
74 .6' 82 .2` 79 .0` 92 .8 Dead
PMN activity remaining (results of PMNs afterchallenge with
(results of PM Ns before challenge with H. somnus )
PMN activity refers to their phagocytic and hydrogen peroxide producing capabilities as determined by the FC assay . 'Cattle were infected with H. somnus on day 0 . Four control samples from each animal were taken as the baseline for activity and averaged from two separate days before challenge with H. somnus . Duplicates were analysed at each time point . `Results were found to be significantly different based on the 95% confidence intervals between each individual's control and test sample results .
Discussion This paper describes a sensitive FC technique that is able to simultaneously evaluate bovine PMN phagocytic activities and the production of hydrogen peroxide (indicative of the respiratory burst) . Results from these experiments showed that exposure of bovine PMNs to H. somnus, either in vitro and in vivo, decreased their activity . The FC assay described in this report was a refinement of a previously published method 30 and was relatively quick and simple to perform . Preliminary work indicated only a minimal amount of blood was necessary and that cell purification was not necessary, thereby reducing the amount of labour required to attain the results . This work was substantiated in the paper by Hasui et al." The use of whole blood further ensured that PMN manipulation was minimal and allowed the analysis of other cell types such as monocytes and lymphocytes . The number of samples processed at one time and the number of cells recovered for analysis, particularly monocytes, were increased when microtitre plates were used instead of test-tubes . Advantages of this system also included simultaneous measurement of four cellular parameters (cell size, cell granularity, phagocytosis and hydrogen peroxide production) and the ability to process large numbers of cells in a short period of time while examining each of the cells individually . Timing of the assay tended to be restrictive, however, as the samples had to be read on the flow cytometer within 10 h of being processed . If left for more than 18 h (at 4°C in the dark) the DCF fluorescence of the positive PMNs decreased significantly while the non-responder cells appeared to become more highly fluorescent . Phagocytosis or PI fluorescence appeared stable (data not shown) . The NBT and chemiluminescence assays were done in tubes which were bulkier than microtitre plates and it was not possible to scale down the assays . However, the NBT assay samples remained stable for up to a week at 4°C (if covered to avoid evaporation) and could be analysed later if required, while the chemiluminescence assay results were generated as the assay components were mixed . Whole blood could not be used in the NBT assay but could be used in the chemiluminescence assay . A
Haemophilus somnus and phagocytosis
decrease in the amount of light emitted was noted in the presence of whole blood from that of isolated cells in the chemiluminescence assay, and it has been reported by Glette et a/. 25 that erythrocytes, free haemoglobin and high protein concentration reduce the chemiluminescence of PMNs . In spite of this, the overall pattern of emitted light seen through chemiluminescence did not change . Also, neither procedure gave any information regarding the status of individual cells and only one parameter of cell function could be measured at a time . In the FC assay, DCF fluorescence was a direct indication of hydrogen peroxide production during the oxidative burst of a PMN 28,36 and PI fluorescence was a direct indication of phagocytosis of bacteria by a PMN, 30,37 however, the amount of actual killing by a PMN was not determined . PI was determined to be stable over a range of pH, unlike fluorescein isothiocyanate (FITC), which has been shown to have a lowered fluorescence at low pH . Hence, PI fluorescence remained unchanged within the low pH environment of the phagolysosomes . Interference by adhering (non-phagocytised) S. aureus was also eliminated, therefore, the red fluorescence detected emanated from PI-labelled bacteria that were phagocytised . It is important to note that the sensitivity of the FC assay required careful monitoring of the assay components to ensure day to day uniformity . The ratio of S. aureus to PMN affected how many bacteria were ingested by a single PMN and a lower ratio would result in less S. aureus per PMN, which in turn would result in a lower fluorescence per PMN . Therefore, samples done between a number of animals in a group and between assays done on different days would vary slightly as absolute numbers of PMNs were not constant but deviated within a range . Conflicting reports 3031 however, exist for the effect erythrocytes have on DCF fluorescence in the PMN, it should be noted that both groups recommended the use of whole blood over isolated cells . External stresses such as the animal's environment and disease exposure may also have played some part in how the PMNs responded ." The chemiluminescence assay was the most sensitive to changes involving the respiratory burst in PMNs, presumably because of its ability to detect many oxidative species . It was able to detect decreases in the respiratory burst of the PMN in the presence of H. somnus alone and in conjunction with S . aureus. These decreases were noted using both whole blood and isolated cells, although the reactions were accentuated in the latter . Logarithmically growing H . somnus were found to inhibit the PMN respiratory burst to a much greater extent than stationary phase H. somnus . The chemiluminescence assay represents the most commonly used assay to detect the respiratory burst of PMNs in conjunction with H. somnus . The NBT assay appeared to be the most limited in its ability to detect irregularities in the PMN oxidative response to log phase H. somnus, especially once it was determined that the bacterium itself was able to reduce NBT to a small extent . Sample and Czuprynski 10 stated that superoxide generation, as detected by the cytochrome c reduction assay, was slightly decreased in the presence of live H . somnus, yet not to the same extent as the decrease in generation of hydrogen peroxide . Using the NBT assay, however, an increase in the superoxide production or respiratory burst was seen, not a decrease . These results appear to indicate that the reduction of tetrazolium salts to formazan could be accomplished by enzymatic pathways independent of the respiratory burst cascade found in the PMN . The FC assay proved to be very sensitive to changes in PMN responses when exposed to H . somnus . It paralleled the sensitivity of the chemiluminescence assay in its ability to detect respiratory burst anomalies, and surpassed both assays with its versatility in simultaneously detecting phagocytosis and in distinguishing between cell types . The FC assay detected a decrease in the respiratory burst as well as in
C . G . Pfeifer et al.
phagocytosis by PMNs in the presence of both S . aureus and live H. somnus . The decrease in activity was not due to competition by H . somnus for S . aureus receptors on the PMN since formalin-killed, heat-killed and stationary phase H . somnus had very little effect on phagocytosis and actually tended to stimulate bacterial ingestion by PMNs . Therefore, it appeared that a component of logarithmically growing H. somnus was required for suppression of both phagocytosis and hydrogen peroxide production . This observation was particularly relevant when one considers that PMNs are the main cell type that respond to a H. somnus infection .' - " The ability of H. somnus to interfere with the defence mechanisms of the PMN may contribute to its ability to survive within the PMN, 11 which would in turn protect the bacterium from other host defences, such as neutralizing antibodies and other opsonins as well as previously activated phagocytes . Survival within PMNs may also be seen as a mechanism that enables the bacterium to invade various host tissue, as the bacterium may be transported across cell barriers when the infected PMN leaves the vascular environment . The presence of H. somnus within bovine PMNs may also partially explain why the bacterium is difficult to isolate during the bacteraemic stages or within the infected tissues, other than having fastidious media requirements . As well, mixed infections are commonly seen in the natural course of the disease 9-10 and it may be possible that other microorganisms are able to take advantage of the lowered activity of the PMNs and are able to establish themselves within various tissues . However, the ability of H. somnus to interfere with the respiratory burst of phagocytic cells is probably not sufficient for intracellular survival and it is likely that other virulence determinants play a role in this process . The FC assay was also able to detect a decrease in the activity of PMNs exposed to H. somnus during an acute infection . However, the reduction seen during infection was less than that seen when H . somnus was added directly to the wells in vitro . Factors contributing to the differences observed between the in vitro and in vivo results may include : the length of time that circulating PMNs were exposed to H . somnus (i .e . during transient bacteraemia) ; the ratio of bacteria to PMN within the animal ; the potential unavailability of affected PMNs to the assay due to their infiltration into infected tissues ; and the effects of fever and cytokines over the period of infection on both the PMNs and H. somnus . Nevertheless, the observed decrease in PMN activity during acute Hemophilosis could be of significance in the pathogenesis of this disease, as mentioned previously . It appears that, whether directly or indirectly, H . somnus affects the ability of the PMNs to mount an effective defence on the part of the host, resulting in tissue damage in various organs and, in some cases, in the death of the animal . This multifaceted assay could serve as the basis for further studies elucidating the effect of bacterial infections on phagocyte cell function .
Materials and methods Reagents and solutions . Ethylenediamine-tetraacetic acid (EDTA) and sodium azide were purchased from J . T . Baker Chemical Co . (Phillipsburg, NJ) and Fisher Scientific Co . (Fair Lawn, NJ), respectively . Absolute ethanol (Aristar grade), Tris [2-Amino-2(hydroxymethyl)1,3-propanediol], concentrated hydrochloric acid, ammonium chloride, dimethyl sulphoxide (DMSO-Analar grade), disodium phosphate, formalin, gelatin, glucose and sodium chloride were purchased from BDH Ltd (Toronto, ON) and the trypticase soy broth powder was bought from Baltimore Biological Laboratory (BBL, division of Becton, Dickinson & Co ., Toronto, ON) . DCFH-DA (2',7'-dichlorofluorescein diacetate) (Kodak, Rochester, NY) was made at a stock concentration of 25 mm in absolute ethanol and stored at 4°C in the dark . Luminol (5-amino2,3-dihydro-1,4-phthalzinedione), lysostaphin (from Staphylococcus staphylolyticus) and PI were purchased from Sigma (St Louis, MO) . Luminol was made at a stock concentration of
Haemophilus somnus and phagocytosis
0 .55 nn in dimethyl sulphoxide and stored at room temperature in the dark, while lysostaphin was made as a 50-times concentrated stock (2200 U/ml) and stored at -20°C . Propidium iodide was made up immediately prior to use . Bacto brain-heart infusion (Difco, Detroit, MI) was made up to a 10-times stock (stored at -20°C) which was then dialysed against 10 volumes of distilled water for 30 h at 4°C . The dialysate (D-BHI) was collected, autoclaved and supplemented with 1 µl/ml of a 2% thiamine-monophosphate solution and 100 hl/ml of a 1 M Tris-HCI solution, pH 8 .0, to yield D-BHITT. Assay diluents included PBSg (calcium and magnesium-free Dulbecco's phosphate-buffered saline containing 5 mm glucose and 0 .1% gelatin) and Hanks buffered salts solution (HBSS) . A range of citric acid-phosphatase buffers were prepared using Mcllvaine's recipe ." Equipment. The flow cytometry assays were performed in Nunclon 96-microwell plates (Nunc-Intermed, Denmark) . NBT assays were performed in 13x100 mm disposable culture tubes (Baxter-Canlab, Toronto, ON) . The chemiluminescence assays were performed in plastic 55 x 12 mm (3 .5 ml) tubes by Sarstedt (Germany) that were stored in the dark . Absorbance readings of NBT assay samples were measured with an Ultrospec 4050-LKB Biochrom (Fisher Scientific Ltd) . Adams Nutator (Clay Adams, Div . of Becton, Dickinson & Co ., Persippany, NJ) was used to agitate microtitre plate samples . Propidium iodide fluorescence was determined using a Spectro Fluorometer Model 430 (G . K . Turner & Associates) . Flow cytometry studies employed the EPICS C fluorescence activated cell sorter (Coulter Electronics of Canada Ltd) with a 5 W argon laser light source . Diagnostic checks on the FC machine were performed daily using Immuno-check fluorospheres (Epics DIVISION, Coulter Corporation, Hialeah, FL) . Chemiluminescence readings were taken on a Picolite Model 6500 Luminometer (United Technologies Packard, Downers Grove, IL) in conjunction with the PC-DAAS : Phagocytosis Data Acquisition and Analysis System (UTP) . Bacteria . S. aureus strain SA492 was stored at -70°C in skim milk and cultured aerobically to stationary phase in trypticase soy broth for 20 h at 37°C, and was referred to as S. aureus . Haemophilus somnus strain HS25 was isolated from a pneumonic calf lung 39 and stored at -70°C in egg yolk, and was referred to as H . somnus . A sample of this frozen culture was spread onto no . 2-Blood Agar (Gibmar, Calgary, AB) and grown for 24 h at 37°C, 5% CO 2 . Plates were then stored for up to 3 days at 4°C . From this plate, a loopful of H . somnus colonies were transferred to D-BHITT broth and grown aerobically for 10 h at 37°C . This late log-phase culture was then diluted by 1 /5 into fresh media and grown to OD,, (, = 0 .400, which represented a logarithmically growing culture . Where indicated in the text, logarithmically growing H. somnus was killed by treating with 0 .1% formalin for 18 h at 37°C or by heating at 65°C for 30 min . Killed cells were washed in PBSg and stored at -20 ° C . Stationary phase H . somnus cultures were aerobically grown for 36 h in 5 ml of D-BHITT . P/-labelling and opsonization of S . aureus . PI-labelling of S . aureus was carried out according to the method of Hasui et al.," except that a 0 .5% solution of PI was used instead of a 5% solution and an extra wash step was added after labelling . This stock of PI-labelled S . aureus was stored at -70°C in 500 µl aliquots . In the case of non-labelled S . aureus, the 30 min incubation with PI was changed to 30 min with saline . Staphylococcus aureus was opsonized using anti-S . aureus bovine convalescent serum, which helped the PMNs to more efficiently phagocytise labelled S . aureus . Equal volumes of stock S . aureus and serum were mixed and incubated together for 1 5 min at 37°C . The bacteria and sera were separated by centrifugation, the serum removed and the bacteria resuspended to the 25-times stock concentration in PBSg . Blood. In all experiments, blood was collected in Vacutainer tubes (Becton, Dickinson & Co ., Rutherford, NJ) containing 143 USP units of sodium heparin (Sigma Chemical Co .) . Assays were performed within 90 min after bleeding, with the blood samples being placed on ice 30 min after bleeding . Blood was taken from healthy beef calves ranging in age from 2-5 months old . Blood was also taken from calves infected with H. somnus . 35 Antibody titres to H. somnus were monitored in each animal and all animals were found to be non-immune at the time the blood was taken . Blood samples were also taken from human male volunteers for comparison purposes in optimizing the FC assay . Cell isolation . Heparinized blood was centrifuged at 1000xg for 20 min and the top layer of mononuclear cells was removed . The red blood cells in the lower layer were then lysed using
C . G . Pfeifer et al.
ammonium chloride lysis 40 at 37°C . Remaining cells were resuspended in HBSS to 5X10 6 live cells/ml as determined through trypan blue exclusion . Suspensions routinely contained 85% PMNs as determined by Geimsa-Wright staining .
FC phagocytosis and hydrogen peroxide detection assay (FC assay) . Up to 30 min before use, 25 mm DCFH-DA was diluted to 0 .3 mm in PBSg and kept in the dark . To each microtitre well, 70 pl opsonized PI-labelled S . aureus, 20 pl 0.3 mm DCFH-DA, 20 pl H. somnus and 100 pl whole blood were added and mixed by pipetting . PBSg was used for the controls . Plates were then incubated for 60 min at 37°C with gentle mixing . After incubation, 20 pl 300 mm EDTA were added to each well and plates were centrifuged at 1000xg for 20 min . Ammonium chloride was used at 37°C to lyse red blood cells in the wells ." After lysis, 200 111 lysostaphin (44 U/ml) were incubated with cells for 30 min at 37°C to destroy any non-phagocytised S. aureus remaining on cell surfaces . As a final step, plates were placed on ice for 1 min before resuspending remaining leukocytes in PBSg containing 0 .5% formalin and 5 mm sodium azide . Duplicates were performed for each sample . Cell populations for FC analysis were selected by gating the desired population based on 43 Bitmaps were their size (forward angle light scatter) and granularity (90° light scatter) ." drawn to select for PMNs, monocytes and lymphocytes (Fig . 1) and 5000 individual cells of each type were routinely examined for each sample . Light at a wavelength of 488 nm was used to excite the fluorescent molecules . Red fluorescence from P1 was collected through a 610630 nm long pass filter . Green fluorescence from DCF was collected through a 550 nm band absorbance filter in combination with a 525 nm dichroic mirror . Phagocytosis and hydrogen peroxide production by cells were analysed further using the EPICS .CS System and Reproman ver . 1 .6 computer programs . A quadrant structure was used to distinguish positive from negative cells (Fig . 1) . Quadrant 1 represented PMNs actively phagocytising PI-labelled bacteria ; quadrant 2 represented PMNs both actively phagocytising PI-labelled bacteria and producing hydrogen peroxide; quadrant 3 represented non-active PMNs ; and quadrant 4 represented PMNs that produced hydrogen peroxide but have not phagocytised the PI-labelled bacteria . All samples were analysed using fluorescence compensation to omit the overlap of fluorescence seen between PI and DCF emissions ." ," Fluorescence of PI-labelled S . aureus was also measured over a pH range of 2 .6-8 .0 to ensure the stability of PI fluorescence under the conditions of low pH . Samples were excited with a wavelength of 488 nm and their emissions measured at a wavelength of 630 nm on the Spectro Fluorometer . Stability of PI fluorescence was found between pH 3 .0-8 .0, indicating that detection of engulfed PI-labelled bacteria would not be affected by a decrease in pH within the PMN, as lysosomal pH generally does not fall below 3 .5 . 44 This is in contrast to the dye FITC, which has been found to have pH-dependent 4546 fluorescence .
Chemiluminescence assay. At the start of the first reading of each tube, 700 pl opsonized S . aureus and 200 pI H. somnus were added to tubes containing 15 1d luminol and 1 ml whole heparinized blood or isolated cells . HBSS was substituted for bacteria in the controls . Chemiluminescence readings were measured on the Picolite luminometer every 3 min for 10 s each over a period of 90 min . Constant stirring and a temperature of 37°C were maintained throughout the entire assay . Duplicates were performed for each sample .
Nitroblue tetrazolium assay . Into each tube, 175 Eil opsonized S . aureus, 50 Ill H . somnus, 250 pl isolated cells and 500 pl 0 .1% NBT (made in HBSS) were mixed and incubated for 45 min at 37°C . Bacteria were replaced with HBSS in control samples . The reactions were stopped by the addition of 1 ml of 0 .1 M HCl and cells were then centrifuged at 1000xg for 5 min . The pellets were washed once in saline and resuspended in 1 ml DMSO . The samples were sonicated using the Sonicator Vibra-cell with microtip (Sonics and Materials Inc ., Danbury, CN) set at a 30 s pulse, 50% duty-set . The tubes were then heated for 1 h at 80°C and the absorbances read at 572 nm . Duplicates were performed for each sample .
This work was supported by the Canadian Bacterial Disease Network and Agriculture Canada/ NSERC Research Partnership Grant No . UGP 0045697 with the W . Garfield Weston Foundation .
Haemophilus somnus and phagocytosis
References 1 . Babior BM . Oxygen-dependent microbial killing by phagocytes . N Engl J Med 1978 ; 298 : 659-68, 721-5 . 2 . Beaman L, Beaman BL . The role of oxygen and its derivatives in microbial pathogenesis and host defense . Ann Rev Microbiol 1984 ; 38 : 27-48 . 3 . Kaufmann SHE, Reddehase MJ . Infection of phagocytic cells . Curr Opin Immunol 1989 ; 2 : 43-9 . 4 . Klebanoff SJ . Antimicrobial mechanisms in neutrophilic polymorphonuclear leukocytes . Sem Haematol 1975 ; 12 : 117-41 . 5 . Leijh P, van den Barselaar M, van Zwet T, Dubbeldeman-Rempt I, van Furth R . Kinetics of phagocytosis of Staphylococcus aureus and Escherichia coli by human granulocytes . Immunol 1979 ; 37 : 453-65 . 6 . Muller-Eberhard HJ . Innate immunity . Curr Opin Immunol 1989 ; 2 : 3-4 . 7 . Root R, Metcalf N, Chance B . H 2 0 2 release from human granulocytes during phagocytosis . J Clin Invest 1975 ; 55 : 945-55 . 8 . Chiang YW, Kaeberle ML, Roth JA . Identification of suppressive components in 'Haemophilus somnus' fractions which inhibit bovine polymorphonuclear leukocyte function . Infect Immun 1986 ; 52 : 792-7 . 9 . Hubbard RD, Kaeberle ML, Roth JA, Chiang YW . Haemophilus somnus-induced interference with bovine neutrophil functions . Vet Microbiol 1986 ; 12 : 77-85 . 10 . Sample AK, Czuprynski CJ . Elimination of hydrogen peroxide by Haemophi/us somnus, a catalasenegative pathogen of cattle . Infect Immun 1991 ; 59 : 2239-44 . 11 . Czuprynski CJ, Hamilton H . Bovine neutrophils ingest but do not kill Haemophi/us somnus in vitro . Infect Immun 1985 ; 50 : 431-6 . 12 . Lederer JA, Brown JF, Czuprynski CJ . 'Haemophilus somnus', a facultative intracellular pathogen of bovine mononuclear phagocytes . Infect Immun 1987 ; 55 : 381-7 . 13 . Rosen H, Klebanoff S . Bactericidal activity of a superoxide anion-generating system . J Exp Med 1979 ; 149 : 27-39 . 14 . Rossi F, Romeo D, Patriarca P . Mechanism of phagocytosis-associated oxidative metabolism in polymorphonuclear leucocytes and macrophages . J Retic Soc 1972 ; 12 : 127-49 . 15 . Sawyer DW, Donowitz GR, Mandell GL . Polymorphonuclear neutrophils : an effective antimicrobial force . Rev Infect Dis 1989 ; 11 : S1532-44 . 16 . van der Valk P, Herman CJ . Biology of disease-leukocyte functions . Lab Invest 1987 ; 57 : 127-37 . 17 . Ward PA, Marks RM . The acute inflammatory reaction . Curr Opin Immunol 1989 ; 2 : 5-9 . 18 . Baehner RL, Boxer LA, Davis J . The biochemical basis of nitroblue tetrazolium reduction in normal human and chronic granulomatous disease polymorphonuclear leukocytes . Blood 1976 ; 48 : 309-13 . 19 . Boxer LA, Hedley-Whyte ET, Stossel TP . Neutrophil actin dysfunction and abnormal neutrophil behavior . N Engl J Med 1974 ; 291 : 1093-9 . 20 . Park BH . Infection and nitroblue-tetrazolium reduction by neutrophils a diagnostic aid . Lancet 1968 ; is 533-4 . 21 . Stossel TP . Evaluation of opsonic and leukocyte function with a spectrophotometric test in patients with infection and with phagocytic disorders . Blood 1973, 42 : 121-30 . 22 . Stossel TP . Quantitative studies of phagocytosis kinetic effects of cations and heat-labile opsonin . J Cell Biol 1973,58 : 346-56 . 23 . DeChatelet LR, Shirley PS . Chemiluminescence of human neutrophils induced by soluble stimuli : effect of divalent cations . Infect Immun 1982 ; 35 : 206-12 . 24 . Faden H, Maciejewski N . Whole blood luminol-dependent chemiluminescence. J Retic Soc 1981 ; 30 : 219-26 . 25 . Glette J, Solberg CO, Lehmann V . Factors influencing human polymorphonuclear leukocyte chemiluminescence . Acta Path Microbiol Scand 1982; Sect C, 90 : 91-5 . 26 . Robinson P, Wakefield D, Breit SN, Easter JF, Penny R . Chemiluminescent response to pathogenic organisms : normal human polymorphonuclear leukocytes . Infect Immun 1984 ; 43 : 744-52 . 27 . Welch WD . Correlation between measurements of the luminol-dependent chemiluminescence response and bacterial susceptibility to phagocytosis . Infect Immun 1980 ; 30: 370-4 . 28 . Bass D, Parce J, Dechatelet L, Szejda P, Seeds M, Thomas M . Flow cytometric studies of oxidative formation by neutrophils : a graded response to membrane stimulation . J Immunol 1983; 130: 1910-17 . 29 . Hafeman D, McConnell H, Gray J, Dean P . Neutrophil activation monitored by flow cytometry : stimulation by phorbol diester is an all-or-none event . Science 1982 ; 215 : 673-5 . 30 . Hasui M, Hirabayashi Y, Kobayashi Y . Simultaneous measurement by flow cytometry of phagocytosis and hydrogen peroxide production of neutrophils in whole blood . J Immunol Methods 1989 ; 117 : 53-8 . 31 . Hirabayashi Y, Taniuchi S, Kobayashi Y . A quantitative assay of oxidative metabolism by neutrophils in whole blood using flow cytometry . J Immunol Methods 1985 ; 82 : 253-9 . 32 . Rouahi N, Levallois C, Favier F, Balmes J, Mani J . Flow cytometric analysis of oxidative product formation in phytohemagglutinin-stimulated ethanol-treated immune mononuclear cells . Drug Alcohol Depend 1989; 23 : 55-62 . 33 . Salgar SK, Paape MJ, Alston-Mills B, Miller RH . Flow cytometric study of oxidative burst activity in bovine neutrophils . Am J Vet Res 1991 ; 52 : 1201--7 . 34 . Szejda P, Parce W, Seeds M, Bass D . Flow cytometric quantification of oxidative product formations by polymorphonuclear leukocytes during phagocytosis . J Immunol 1984 ; 133 : 3303-7 .
C . G . Pfeifer et al.
35 . Harland RJ, Potter AA, Schuh JCL . Development of an intravenous challenge model for Haemophi/us somnus disease in beef calves . Conf Res Work Admin Dis 71 51 1990 ; 29 : 6 . 36 . Brandt R, Keston A . Synthesis of diacetyldichlorofluorescein : a stable reagent for fluorometric analysis . Anal Biochem 1965 ; 11 : 6-9 . 37 . Saad A, Hageltorn M . Flow cytometric characterization of bovine neutrophil phagocytosis of fluorescent bacteria and zymosan particles . Acta Vet Scand 1985, 266 : 289-307 . 38 . Dawson RMC, Elliot DC, Elliot WH, Jones KM (eds) . Data for biochemical research . 2nd edn . Oxford : Clarendon Press, 1969 ; 484 . 39 . Theisen M, Potter AA . Cloning, sequencing, expression, and functional studies of a 15,000 molecular weight Haemophi/us somnus antigen similar to Escherichia co/i ribosomal protein S9 . J Bacteriol 1992 ; 174 : 17-23 . 40 . Wahl SM, Rosenstreich DL, Oppenheim JJ . Separation of human lymphocytes by E rosette sedimentation . In : Bloom BR, David JR, eds . In vitro methods in cell-mediated and tumor immunity . New York : Academic Press, 1976 ; 231-40 . 41 . Haynes JL . Principles of flow cytometry . Cytometry 1988 ; 3 : 7-17 . 42 . Shapiro H . Practical flow cytometry . 2nd edn . New York : Alan R . Liss, 1988 . 43 . Sucic M, Kolevska T, Kopjar B et al. Accuracy of routine flow-cytometric bitmap selection for three leukocyte populations . Cytometry 1989 ; 10 : 442-7 . 44 . Mims CA . The pathogenesis of infectious disease . 3rd edn . Toronto : Academic Press, 1987 . 45 . Cantinieaux B, Hariga C, Courtoy P, Hupin J, Fondu P . Staphylococcus aureus phagocytosis : a new cytofluorometric method using FITC and paraformaldehyde . J Immunol Methods 1989 ; 121 : 203-8 . 46 . Orpegen Corporation . Phagotest ° --test kit for the determination of the phagocytic activity of monocytes and granulocytes in whole blood . Instructions of use . Heidelberg, Germany : ORPEGEN, Med .Molekularbiologische Forschungsgesellschaft mbH, 1990 ; 1--16 .