Preferential Inhibition of Primary Granule Release from Bovine Neutrophils by a Brucella Abortus Extract TIMOTHY A

Home , Degranulation

INFECTION AND IMMUNITY, Apr. 1986, p. 285-292 Vol. 52, No. 1 0019-9567/86/040285-08$02.00/0 Copyright © 1986, American Society for Microbiology Preferential Inhibition of Primary Granule Release from Bovine Neutrophils by a Brucella abortus Extract TIMOTHY A. BERTRAM,lt* PETER C. CANNING,' AND JAMES A. ROTH2 National Animal Disease Center, Agricultural Research Service, U.S. Department ofAgriculture,' and Department of Veterinary Microbiology and Preventive Medicine, Iowa State University,2 Ames, Iowa 50010 Received 16 August 1985/Accepted 2 January 1986 A low-molecular-weight Brucella abortus extract (a nucleotidelike material) inhibited zymosan-elicited neutrophil degranulation and trichloroacetic acid-precipitable protein iodination (a measure of myeloperoxidase and H202 release from neutrophilic leukocytes). Inhibition of neutrophil function was directly related to the concentration of the Brucella extract. The extract preferentially inhibited degranulation of primary (azurophilic or peroxidase positive) granules and had limited inhibition of secondary (specific or peroxidase negative) granule release but did not inhibit opsonized zymosan ingestion. Inhibition of protein iodination closely paralleled that of primary granule release but was unrelated to inhibition of secondary granule release. These results suggest that B. abortus has a component which is capable of inhibiting release of myeloperoxidase by dose-dependent preferential inhibition of primary granule release from bovine neutrophilic leukocytes.

The ability of neutrophils to phagocytize and then kill (proteases, cationic proteins, phospholipase A2) pathways is ingested bacteria is critical for mammalian resistance to necessary for microbicidal activity by the neutrophil (1, 28). pathogenic microbes. Facultative intracellular bacteria such Neutrophil degranulation is inhibited by exogenous cAMP, as Mycobacterium tuberculosis, Brucella abortus, and cAMP analogs, and various agents which increase intracel- Listeria monocytogenes have evolved mechanisms to cir- lular cAMP levels (10, 15, 17, 18). Most studies have shown cumvent these processes. A wide variety of virulence factors that primary and secondary granule release is inhibited by which alter phagocytic cell function have been discussed (9). these materials. Included with these factors are a cell wall component of B. Secretion of both primary and secondary granules is abortus (20, 21) and sulfolipids of M. tuberculosis (12, 13). promoted by opsonized zymosan (16). Secondary granules The demonstrated effector molecule is frequently secreted or of rabbit neutrophils degranulate as early as 30 s after located on the bacterial surface which is first exposed to the phagocytosis, and primary granules degranulate 1 to 3 min phagocytic cell. after phagocytosis (3). Preferential degranulation of second- B. abortus is ingested by neutrophils but does not stimu- ary granules has been demonstrated with phorbol myristate late the oxidative burst (20). It inhibits neutrophil degranula- acetate or the ionophore A23187 (30). The purpose of this tion and is resistant to some granule-released antimicrobial study was to examine the ability of a nucleotidelike material enzymes (24). However, B. abortus is sensitive to the isolated from B. abortus (7) to inhibit bovine neutrophil myeloperoxidase-H202-halide system (20). A cell wall- degranulation and to determine whether there was any associated lipopolysaccharide has been proposed as the cell preferential inhibition of primary or secondary granule re- wall component that is the key virulence factor which lease from neutrophilic leukocytes. inhibits neutrophil degranulation (20, 21). Recently, a nucleotidelike compound has been isolated from the surface MATERIALS AND METHODS of virulent B. abortus and is capable of inhibiting protein iodination (a measure of the myeloperoxidase-H202-halide Preparation ofB. abortus and Brucella extract. Bacteria for pathway) by bovine neutrophils (7). The protein iodination extract isolation were prepared as previously described (7). reaction is dependent on a number of neutrophil processes B. abortus 2308 (a highly virulent, smooth form of bovine including phagocytosis, production of H202, release of origin) was grown in a fermentor for 48 h at 37°C. The myeloperoxidase from primary granules, and the presence of bacteria were harvested by filtration and washed three times halide ions (19). The B. abortus extract did not inhibit with sterile 0.85% saline by centrifugation. They were heat Staphylococcus aureus phagocytosis or superoxide anion killed at 65°C for 60 min and pelleted by centrifugation at production, but it does inhibit the iodination reaction. Since 18,000 x g for 20 min. Supematant from the heat-killed adequate amounts of protein and iodide were present, it was bacteria was collected and separated into two fractions by speculated that the reduced iodination was a result of molecular filtration (YM2 filters; Amicon Corp., Lexington, Brucella extract-mediated inhibition of degranulation. Mass.). The YM-2 filter has a nominal molecular mass cutoff For intracellular degranulation to occur, the membranes of of 1,000 daltons, a clear water flow of 0.02 to 0.035 mlI/cm2 the per min at a pressure of 3.8 kg/cm2, and a solute rejection of phagosome and lysosome must fuse. Release of granule 40% for sucrose, 65% for raffinose, and 80% for inulin enzymes for the activation of oxygen-dependent (myeloper- (Amicon). Material retained by the filter was considered as oxidase-H202-halide system) or oxygen-independent having a molecular mass greater than 1,000 daltons, and the filtrate was arbitrarily designated as less than 1,000 daltons. * Corresponding author. The filtered extract was separated by high-performance t Present address: Department of Veterinary Pathobiology, Uni- liquid chromatography to remove salts and lyophilized to versity of Illinois, Urbana, IL 61801. remove volatile buffer components. The remaining dried 285 286 BERTRAM ET AL. INFECT. IMMUN.

material was weighed directly, and the active components TABLE 1. Abbreviations used to describe morphometric analysis contained no detectable lipids or proteins (7). The extract of neutrophilic leukocytes passing through the 1,000-dalton molecular mass ifiter was Abbrevia- Definition tested at three concentrations. tion Neutrophil leukocyte preparation. Four adult Holstein- Vv Relative volume percent of the given neutrophil structure Friesian steers, serologically negative for B. abortus, were expressed as a percentage of the given neutrophil, used as a source of polymorphonuclear leukocytes (PMNs) e.g., Vvag, Vvsg, etc. for this study. Bovine PMNs were isolated as previously described (25). Peripheral blood was collected in acid- V . Absolute volume of the neutrophil or organelle ex- citrate-glucose solution and centrifuged. The resulting pressed in cubic micrometers, e.g., Vag, vsg, etc. plasma and buffy-coat cells were aspirated and discarded, ag.... Primary granules (peroxidase positive) and erythrocytes present in the packed cell layer were lysed sg.... Secondary and large granules (peroxidase negative) tg.... All granules (peroxidase positive and peroxidase nega- with hypotonic phosphate-buffered, deionized water. tive) ,Isotonicity was restored to the remaining cells (90% or pl.... Phagolysosomes containing zymosan greater being granulocytes) by addition of phosphate- TC. . . Total points, area, or volume for the neutrophilic leuko- buffered (0.0132 M, pH 7.2) 2.7% saline (PBS). Granulocytes cyte minus the respective measurement for the were washed, resuspended in PBS to a concentration of 5 x zymosan-containing phagolysosomes 107 cells per ml, and held at room temperature for approximately 1 h before use. PMN function test. The iodination assay (19, 25) was conducted in duplicate, and the average of duplicate values idase-positive granules contained an electron-dense enzyme- was used for calculations. The assay (a measure of the substrate precipitate. myeloperoxidase-H202-halide system of PMNs) was per- Morphometric analysis. Analysis was conducted as previ- formed in capped polystyrene tubes (12 by 75 mm). The ously described (4, 6, 29). Six to 10 blocks were made, and standard reaction mixture contained 0.05 ml of PMNs (2.5 x 1 block was randomly chosen from each duplicate sample. 106 cells), 0.05 ,uCi of 125I (carrier-free; Amersham Corp., Ten photomicrographs were obtained by random systematic Arlington Heights, Ill.), 40 nmol ofNal, 0.05 ml ofpreopson- subsampling at a magnification of x6,900. Photographs were ized zymosan (10 mg/ml), and 0.3 ml of Earle balanced salt made at a final magnification of x 18,000. For morphometric solution. Results were standardized and expressed as analysis, a point count lattice (121 points) was superimposed nanomoles of Nal per 107 PMNs per hour. One hundred on photographs, and points falling on phagolysosome, nu- microliters of each of the three different Brucella extract cleus, and primary (peroxidase positive) and secondary concentrations was added to the standard reaction mixture, (peroxidase negative) granules were recorded. Points on resulting in final Brucella fraction concentrations of 0.04, other structures within the neutrophil were also recorded for 0.16, and 0.32 mg/ml (wt/vol). The approximate number of determining total cell points. The percentage of the bacteria required to produce each concentration was 2.5 x neutrophil occupied by each of these structures was calcu- 109/ml (0.04 mg/ml), 1 x 1010/ml (0.16 mg/ml), and 2 x lated as previously described (4, 6, 29). Ifthe total cell points 1010/ml (0.32 mg/ml). This extract was collected from killed (excluding those on phagolysosomes) are used as the refer- cells only, but B. abortus may be able to release the ence compartment, point-to-point ratios of a given organelle materials present in the extract without cell lysis and death to the total cell will be the relative volume percent (Vv) of (7). To determine the total possible degranulation and iodina- that organelle within that cell (6, 29). Abbreviations used to tion in the test system, 100 ,il of PBS was added to the describe morphometric data are found in Table 1. mixture instead of Brucella extract. To determine the mini- To determine absolute areas for absolute volume determi- mum level of iodination and degranulation in the system, no nation, neutrophil and nuclear profile areas were measured zymosan was added to the reaction mixture (unstimulated by semiautomatic, computerized morphometry (Bioquant II; neutrophils). The reaction in each test group was allowed to R&M Biometrics, Nashville, Tenn.). Since neutrophils were proceed for 20 min at 37°C, and the amount oftrichloroacetic induced to become spherical, mensuration coefficients can acid-precipitable radioactivity was determined. Results were be used to calculate the actual volume of a neutrophil from expressed as nanomoles of Nal per 107 PMNs per hour. the cross-sectional area (28) by the following equations: Sample preparation for electron microscopy. Duplicate samples for electron microscopic examination were pre- Ax (1) pared simultaneously with samples for the iodination assays. 4 Sodium iodide was used in place of Na'125I for electron and microscopic samples. At the end of the 20-min incubation VTCD3 period, cold (4°C) 2.5% glutaraldehyde in 0.1 M cacodylate VTC = 6 (2) buffer was added to the reaction mixture, and fixation was allowed to proceed for 3 h; Neutrophils were fixed while in where Ax is the cross-sectional area of the neutrophil, D is suspension and at 4°C so microtubules would depolymerize the diameter of the neutrophil, and VTC is the neutrophil and the cells would become spherical. Samples were then volume. By multiplying the actual neutrophil volume (VTc) postfixed in 1.0% OS04 for 30 min, dehydrated in graded by the volume percent (Vv) of a given organelle, the actual alcohols, cleared in propylene oxide, embedded, and poly- volume (V) for that organelle can be determined (4, 6, 29). merized in resin. Ultrathin sections were cut at 70 to 90 nm The relative volume percent of primary granules (Vv) and stained with uranyl acetate and lead citrate. was determined on peroxidase-labeled, unstained thin sec- Identification of primary granules was done as previously tions. These sections were then stained with uranyl acetate reported (22). Neutrophils were fixed in 2.5% glutaralde- and lead citrate, and the relative volume percents of all hyde, washed in cacodylate buffer, and suspended in the granules (Vvtd and of the zymosan-containing phagosomes peroxidase substrate of Graham and Karnovsky (14). Perox- (Vvpi) were determined. The relative volume of secondary VOL. 52, 1986 INHIBITION OF NEUTROPHIL DEGRANULATION 287

TABLE 2. Iodination and granule volumes of neutrophils exposed to Brucella extracta Iodination (nmol of Vol (,m3) Group NaIl/107 PMN V.g:V,S ratiof per h)b Vtgc Vagd vsgC Unstimulated 3.2 ± 0.1 49.8 ± 3.7 9.8 ± 1.5 40.1 ± 1.5 0.24:1 Saline + zymosan 31.8 ± 3.3 32.1 ± 4.7 7.2 ± 1.4 24.9 ± 3.3 0.28:1 Brucella extract + zymosan at: 0.04 mg/ml 28.5 ± 3.2 31.1 ± 2.8 5.4 ± 0.3 25.6 ± 2.6 0.21:1 0.16 mg/ml 20.4 ± 0.4' 40.3 ± 6.1 9.1 ± 0.69 31.1 ± 5.5 0.29:1 0.32 mg/ml 18.6 ± 1.39 42.8 ± 3.3' 11.5 ± 1.3' 31.3 ± 2.4 0.37:1 a Each value is the mean ± the standard error of the mean for four animals, each conducted in b Slope, -41.1 (P < 0.01). duplicate. c Total granule volume; slope = 38.1 (P < 0.05). d Vt., Vag, Primary granule (peroxidase positive) volume; slope = 16.9 (P < 0.01). ' VSg, Secondary granule (peroxidase negative) volume; slope = 21.2 (not significant). f Slope not determined. ' P < 0.05. granules was calculated by: granule volume (Vtg) was 10% ofthe neutrophil volume (VTc) (Table 2). The volume ratio of primary (Vd) to secondary Vvtg - VVag = Vvsg (3) (Vsg) granules was greater in neutrophils exposed to saline where Vvsg is the relative volume of secondary granules. and zymosan than that seen in unstimulated neutrophils Inhibited degranulation of all granules was calculated as a (Table 2). Approximately 36% of the volume of all granules percentage of PBS-treated neutrophil by: (Vtg) was released from neutrophils in this group. Most ofthe Vtg(Vtg/s) released granule volume was the result of secondary granule [(Vtgp- Vtg/s)/Vtg/s] x 100 (4) release (Table 2). Neutrophils exposed to the Brucella extract and opsonized zymosan had significant (P < 0.002) where Vtg/p is the Vtg of the given principal group. Inhibited concentration-dependent inhibition of primary granule re- degranulation of primary granules was calculated as a per- lease, and inhibition of primary granule release was centage of PBS-treated neutrophil Vn(VaWjs) by: greatest [Vaglp - Va/s)lVasl] X 100 (5) where V,,gp is the Vag of a given principal group. Inhibited degranulation of secondary granules was calculated as a percentage of PBS-treated neutrophil Vsg by: v-e PEROXIDASE (G)GRANULES (Vsgjs) O-NO PEROXIDASE (-) GRANULES [(Vsgp-vsg5s)/Vsgts] x 100 (6) C/) - TOTAL Of BOTH GRANULES where Vg5p is the Vsg of any given principal group. The ratio "' 80 of primary to secondary granule volumes was calculated by: LAJ Vag,PlVsg/p (7) I60 Statistical analysis. Linear regression analysis was used to statistically determine the influence of Brucella fraction -c concentration on protein iodination, granule volume, and = 40 percent inhibition of neutrophil degranulation. All slopes U- were tested against the hypothesis that there was no differ- ence from zero. Student's t test was used to statistically determine differences between groups in the volume of O 20 zymosan-containing phagolysosomes and primary and sec- m ondary granules. 0 RESULTS Inhibition of granule release. Unstimulated neutrophilic -20 leukocytes were considered to have the maximal granule volume possible for animals in these experiments and to 0 0.04 0.16 0.32 have released no granules. Granule volume (Vt5) represented mg/ml 26% of the neutrophil volume (VTc), with most of that being FIG. 1. Percent inhibition of primary (peroxidase positive), sec- peroxidase-negative (secondary) granules (Table 2). Neutro- ondary (peroxidase negative), and total neutrophil granule release. phils exposed to saline and opsonized zymosan were con- The slopes for inhibition of total neutrophil granule release sidered to have the maximal amount of degranulation possi- (157.2%/mg per ml of Brucella extract) and secondary granule ble with our test system and 0.0% inhibition of granule release (104.4%o/mg per ml of Brucella extract) are not statistically release. The number of granule profiles was markedly re- significant; that of primary granule release (344.3%o/mg per ml of duced compared with unstimulated neutrophils, and the total Brucella extract) is significant at P < 0.002. 288 BERTRAM ET AL. INFECT. IMMUN.

FIG. 2. Unstimulated bovine neutrophilic leukocyte peroxidase labeled but without uranyl acetate and lead citrate staining. The neutrophil is round with few short cytoplasmic projections. The nucleus (N) is bilobed and composed primarily of heterochromatin. Primary or azurophilic (peroxidase positive) granules (AG) have a punctate, electron-dense precipitate from the peroxidase-substrate reaction. Secondary or specific (peroxidase negative) granules (SG) are large, round, ovoid to oblong, and slightly electron dense. Bar = 1.0 ,um.

in neutrophils treated with 0.32 mg ofBrucella extract per ml tions (Fig. 2). Granules were located in and around the and zymosan (Fig. 1). centrosome and separated from the plasma membrane by a Morphometry of neutrophilic leukocyte granules. Neutro- thin rim of organelle-free cytoplasm. In peroxidase-labeled, philic leukocytes exposed to the lowest extract concentra- unstained sections, peroxidase-positive (primary) granules tion had granule volumes similar to those in neutrophils had a punctate, electron-dense precipitate which was con- exposed to saline and zymosan (Table 2). The ratio of fined to the granule cross-sectional profile (Fig. 3). These primary (Vag) to secondary (Vsg) granule volume in neutro- granules were variably sized, round to oblong, and randomly phils exposed to 0.16 mg of Brucella extract per ml was distributed in the cell. The average cut surface area (mean ± greater than that in unstimulated neutrophils. The volume of the standard error of the mean) of neutrophils was 40.7 ± 1.1 primary granules (Vag) was greater (P < 0.05), but the .m2 and 4.5 ± 0.3 [Lm2 for the nucleus, which gave the volume of secondary granules (Vsg) was not different (P > mensuration values in Table 3. 0.05) from that in neutrophils treated with saline and Neutrophilic leukocytes exposed to saline and opsonized zymosan (Table 2). Neutrophils exposed to 0.32 mg of zymosan were round to ovoid and had multiple pseudopodia. Brucella extract per ml had a greater Vtg than saline-treated The average neutrophil profile area was 42.8 ± 2.3 p.m2, and neutrophils but less than that in unstimulated neutrophils. the nucleus profile area was 4.1 ± 0.1 p.m2 (Table 3). One to / The volume of primary (Vag) granules was significantly (P < several phagolysosomes were in most neutrophils, and each 0.05) greater, but the volume of secondary granules (Vsg) was contained one to several zymosan particles. The mean not significantly (P > 0.05) different from that of neutrophils volume of phagolysosomes (Vpl) was not statistically greater treated with saline and zymosan. The volume ratio of than that of those from Brucella extract-treated neutrophils primary (Vag) to secondary (Vsg) granules in this group was (Table 3). greater than that from neutrophils treated with 0.16 mg of Neutrophils exposed to the lowest concentration (0.04 Brucella extract per ml and zymosan (Table 2). mg/ml) of Brucella extract had a neutrophil cross-sectional Neutrophilic leukocyte morphology. Unstimulated neutro- area of 44.3 + 2.4 p.m2 and a nuclear cross-sectional area of philic leukocytes were round with few cytoplasmic projec- 4.4 ± 0.7 p.m2 (Table 3). Neutrophils-ad--h-multiple VOL. 52, 1986 INHIBITION OF NEUTROPHIL DEGRANULATION 289

FIG. 3. Unstimulated bovine neutrophilic leukocyte, peroxidase labeled but without uranyl acetate or lead citrate staining. Primary or azurophilic (peroxidase positive) granules (AG) are round to ovoid, smaller than secondary granules, and have a punctate electron density. Both primary and secondary granules are randomly distributed. Bar = 1 ,um. pseudopodia and reduced granulation (Fig. 4). Phagolyso- 0.41). Some granules had a flocculent matrix and a cross- somes containing one to several zymosan particles had a sectional profile similar to that of primary granules (Fig. 5). volume (Vpl) of 3.8 ,um3, which was not significantly greater Protein iodination. Neutrophilic leukocytes exposed to than that in the other groups. saline and opsonized zymosan were considered to have the Neutrophils exposed to 0.16 mg ofBrucella extract per ml maximal amount of protein iodination (myeloperoxidase- and zymosan had an average neutrophil cross-sectional area H202-halide activity) possible with our test system. A con- of 44.4 + 2.4 ,um2 and a nuclear cross-sectional area of 4.3 + centration-dependent decrease in iodination was present in 0.6 p,m2 (Table 3). One to several zymosan-containing neutrophils exposed to the Brucella extract, and this de- phagolysosomes were present in most neutrophils. The crease paralleled the inhibition of primary granule release. volume of zymosan-containing phagolysosomes (Vpl) was At the two highest concentrations of extract, protein iodina- not statistically different from that in the other groups (Table tion was significantly less than that in neutrophils exposed to 3). Granules were polarized, with most located away from saline and zymosan only. and a few juxtaposed to phagolysosomes. Neutrophils exposed to 0.32 mg of Brucella extract per ml DISCUSSION and zymosan had an average cross-sectional area of 45.7 + Preferential inhibition of primary granule release from 0.8 p.m2 and a nuclear cross-sectional area of 4.0 ± 0.3 p.m2 bovine neutrophils by Brucella extract was dose dependent. (Table 3). Phagolysosome volume (Vpl) was not significantly The influence of Brucella extract concentration on granule different (Table 3) from that in the other groups, and there volume, the ratio of primary to secondary granule volume, was no significant concentration effect on this volume (P > and the inhibited release of primary granules suggests that 290 BERTRAM ET AL. INFECT. IMMUN.

FIG. 4. Bovine neutrophilic leukocyte exposed to 0.04 mg of B. abortus extract per ml and stained with uranyl acetate and lead citrate. The cell is round with few thick pseudopodia. The phagolysosome contains a zymosan particle (Z). Cytoplasmic granules are reduced in number (compare with Fig. 5), and few granules are primary; most are secondary (specific) granules. Granules are separated from the phagosome by a zone of microfilaments (M). preferential inhibition of neutrophil degranulation was dose (11). The increase in cellular cyclic nucleotides is seen only dependent. Phorbol myristate acetate and ionophore A23187 after a period of phorbol myristate acetate-neutrophil incu- cause preferential release of neutrophil secondary granules bation (11). Activity of the Brucella extract used in the (11, 30) by their ability to increase cellular levels of cGMP present study has been attributed to nucleotidelike com-

TABLE 3. Mensuration values for neutrophils exposed to Brucella extracta Diameter (>.m) of: Vol (pm3) of: Group Neutrophil Phagolysosome Neutrophil Phagolysosome (DTC) (Dpl) (VTC) (VPI) Unstimulated 7.2 ± 0.1 0.0 193.5 ± 8.0 0.0 Saline + zymosan 7.4 ± 0.3 1.97 ± 0.1 213.7 ± 14.0 4.1 ± 0.7 Brucella extract + zymosan at: 0.04 mg/ml 7.5 ± 0.2 1.75 ± 0.3 216.9 ± 16.5 3.8 ± 2.2 0.16 mg/ml 7.6 ± 0.2 1.81 ± 0.1 226.9 ± 15.0 3.1 ± 0.2 0.32 mg/ml 7.6 ± 0.1 1.73 ± 0.1 227.3 ± 15.9 2.8 ± 0.5 a Each value is the mean ± the standard error of the mean for four animals, each conducted in duplicate. VOL. 52, 1986 INHIBITION OF NEUTROPHIL DEGRANULATION 291

Vi~~~~

F. - e 8,% 4,Y ,

FIG. 5. Bovine neutrophilic leukocyte exposed to 0.32 mg of B. abortus extract per ml and stained with uranyl acetate and lead citrate. A zymosan particle (Z) is focally digested. The cell has more primary granules than the cell in Fig. 4. Numerous flocculent granules (F) correspond to primary granules. Some granules are juxtaposed to the phagosome. pounds (7), and in the present study its inhibitory effects activity of the granule enzymes. These and other (11, 30) occurred without a preincubation period with the peutrophil. observations support the notion that there is selective con- This immediate effect on neutrophils suggests the Brucella trol of neutrophil degranulation, yet most inhibitors of fraction works at a different cellular location than phorbol degranulation block both primary and secondary granule myristate acetate or directly substitutes for a nucleotide that release (10, 15, 17, 18, 31). The preferential inhibition of may control neutrophil granule release. Purine nucleotides primary granule release by Brucella extract was concentra- and modified analogs inhibit zymosan-stimulated lysosomal tion dependent. This suggests that the extract inhibits a secretion from macrophages (23). This Brucella fraction may different pathway from that which controls secondary gran- be partially responsible for the inhibition of neutrophil ule release or has a dose-crpendent effect on the same degranulation seen with live B. abortus (20). pathway that controls release of both granules from neutro- Reduced neutrophil degranulation and preferential inhibi- phils. tion of primary granule release are not explained by inhibi- B. abortus is phagocytized by neutrophils without activation of zymosan phagocytosis. Phagolysosomal volumes tion of the hexose monophosphate pathway (20) and inhibits were reduced in a dose-dependent manner for each of the neutrophil degranulation (24). A cell wall component, possi- Brucella-treated groups, but this reduction was not statisti- bly lipopolysaccharide, was proposed as the mediator of cally significant. Furthermore, Brucella extract does not these effects on human and guinea pig neutrophils (24). The inhibit ingestion of Staphylococcus aureus by bovine neu- concentrations of Brucella fraction used in our study have trophils (7). Inhibition of protein iodination could be the been demonstrated to inhibit the protein iodination reaction result of inhibited neutrophil myeloperoxidase activity, but of bovine neutrophils, and this same activity has been this is not likely since similar concentrations of Brucella demonstrated with live B. abortus (7). Results of the present extract did not reduce in vitro peroxidase activity (7). study demonstrate that this fraction also preferentially inhib- Preferential inhibition of primary granule release suggests its neutrophil primary granule release. These studies indicate that different mechanisms control release of the two granule that B. abortus has at least two components which inhibit types from bovine neutrophils. Granules are released from neutrophil degranulation: a lipopolysaccharide (20, 21) and a rabbit neutrophils at different times during the process of nucleotidelike material. Similar results as those obtained in degranulation (3). The sequential release is coordinated with the present study could be obtained with a lipopolysaccha- a pH change in the phagosome which provides optimal ride preparation if it were contaminated with the nucleotide- 292 BERTRAM ET AL. INFECT. IMMUN. like material. The B. abortus extract used in the present 13. Goren, M. B., P. D. Hart, M. R. Young, and J. A. Armstrong. study did not contain lipopolysaccharide or lipid A (7). 1976. Prevention of phagosome-lysosome fusion in cultured Which, if any, of these Brucella fractions is the most macrophages by sulfatides of Mycobacterium tuberculosis. significant biologically was not established. Since the Bru- Proc. Natl. Acad. Sci. USA 73:2510-2514. cella extract used in the present study has been demon- 14. Graham, R. C., and M. J. Karnovsky. 1966. The early stages of in extracellular media of absorption of injected horseradish peroxidase in the proximal strated live B. abortus (7), it would tubules of mouse kidney: ultrastructural cytochemistry by a likely reduce the ability of neutrophils to kill B. abortus if it new technique. J. Histochem. Cytochem. 14:291-302. were released in vivo. 15. Hawkins, D. 1974. Neutrophilic leukocytes in immunologic Protein iodination is an in vitro assay for the myeloperox- reactions in vitro. III. Pharmacologic modulation of lysosomal idase-H202-halide antibacterial system of the neutrophilic constituent release. Clin. Immunol. Immunopathol. 2:141-152. leukocyte (5, 7, 8, 27). In the conditions used in the present 16. Henson, P. H. 1971. The immunologic release of constituents study, this assay closely correlates with primary (peroxidase from neutrophil leukocytes. I. The role of antibody and com- positive) granule release from bovine neutrophils. This sug- plement on nonphagocytizable surfaces or phagocytizable sur- gests that primary granule release is a rate-limiting step in faces. J. Immunol. 107:1535-1546. the myeloperoxidase-H202-halide 17. Hoffstein, S., R. B. Zurier, and G. Weissmann. 1974. Mecha- pathway. This also sug- nisms of lysosomal enzyme release from human leukocytes. III. gests that secondary and large (2) (peroxidase negative) Quantitative morphologic evidence for an effect of cyclic nucle- granules of the bovine neutrophil do not participate in the otides and colchicine on degranulation. Clin. Immunol. Immu- protein iodination reaction. lodination can be inhibited with- nopathol. 3:201-217. out inhibition of nitroblue tetrazolium reduction (26), sug- 18. Ignarro, C. J., and S. Y. Cech. 1976. Bidirectional regulation of gesting that the rate-limiting step in the iodination reaction is lysosomal enzyme secretion and phagocytosis in htuman neutro- myeloperoxidase (from primary granules) and not oxidative phils by guanosine 3'5'-monophosphate and adenosine 3'5'- metabolism. By combining these two assays in our study, we monophosphate. Proc. Soc. Exp. Biol. Med. 151:448-452. demonstrated that a B. abortus component is capable of 19. Klebanoff, S. J., and R. A. Clark. 1977. Iodination by human polymorphonuclear leukocytes; a reevaluation. J. Lab. Clin. preferentially inhibiting primary granule release. Med. 89:675-686. 20. Kreutzer, D. L., L. A. Dreyfus, and D. C. Robertson. 1979. LITERATURE CITED Interaction of polymorphonuclear leukocytes with smooth and rough strains of Brucella abortus. Infect. Immun. 23:737-742. 1. Babior, B. M. 1978. Oxygen-dependent killing by phagocytes. 21. Kreutzer, D. L., and D. C. Robertson. 1979. Surface macromol- N. Engl. J. Med. 298:659-668. ecules and virulence in intracellular parasitism: comparison of 2. Baggiolini, M., U. Horisberger, R. Gennaro, and B. Dewald. cell envelope components of smooth and rough strains of 1985. Identification of three types of granules in neutrophils of Brucella abortus. Infect. Immun. 23:819-828. ruminants. Ultrastructure of circulating and maturing cells. 22. McCall, C. E., I. Katayama, R. S. Cotran, and L. M. Finland. Lab. Invest. 52:151-158. 1969. Lysosomal and ultrastructural changes in human "toxic" 3. Bainton, D. F. 1973. Sequential degranulation of the two types neutrophils during bacterial infection. J. Exp. Med. of polymorphonuclear leukocyte granules during phagocytosis 129:267-293. of microorganisms. J. Cell Biol. 58:249-264. 23. Riches, D., J. L. Watkins, P. M. Henson, and D. R. Stanworth. 4. Bertram, T. A., and F. L. Coignoul. 1982. Morphometry of 1985. Regulation of macrophage lysosomal secretion by adeno- equine neutrophils isolated at different temperatures. Vet. sine, adenosine phosphate esters, and related structural ana- Pathol. 19:534-543. logues of adenosine. J. Leukocyte Biol. 37:545-557. 5. Bertram, T. A., and A. E. Jensen. 1984. Responses of equine 24. Riley, L. K., and D. C. Robertson. 1984. Ingestion and intracel- neutrophils to contagious equine metritis organism and its lular survival of Brucella abortus in human and bovine polymor- lipopolysaccharides. Am. J. Vet. Res. 45:1099-1104. phonuclear leukocytes. Infect. Immun. 46:224-230. 6. Bolender, R. P. 1978. Correlation of morphometry and stereol- 25. Roth, J. A., and M. L. Kaeberle. 1981. Evaluation of bovine ogy with biochemical analysis of cell fractions. Int. Rev. Cytol. polymorphonuclear leukocyte function. Vet. Immunol. Immu- 55:247-289. nopathol. 2:157-174. 7. Canning, P. C., J. A. Roth, L. B. Tabatabai, and B. L. Deyoe. 26. Roth, J. A., M. L. Kaeberle, and W. H. Hsu. 1982. Effect of 1985. Isolation of components of Brucella abortus responsible estradiol and progesterone on lymphocyte and neutrophil func- for inhibition of function in bovine neutrophils. J. Infect. Dis. tions in steers. Infect. Immun. 35:997-1002. 152:913-921. 27. Schwartz, A., H. Sugge, T. W. Ritter, and E. Fernandez- 8. Coignoul, F. L., T. A. Bertram, and N. F. Cheville. 1984. Repollet. 1983. Direct determination of cell diameter, surface Functional and ultrastructural changes in neutrophils from area, and volume with an electronic volume-sensing flow mares and foals experimentally inoculated with a respiratory cytometer. Cytometry 3:456-458. tract strain of equine herpesvirus-1. Am. J. Vet. Res. 28. Spitznagel, J. K. 1975. Mechanisms of killing by polymorpho- 45:1972-1975. nuclear leukocytes, p. 209-214. In D. Schlessinger (ed.), Micro- 9. Denson, P., and G. Mandell. 1980. Phagocyte strategy vs. biology-1975. American Society for Microbiology, Washing- microbial tactics. Rev. Infect. Dis. 2:817-838. ton, D.C. 10. Goldstein, I., S. Hoffstein, J. Gallin, and G. Weissmann. 1973. 29. Weibel, E. R. 1969. Stereological principles for morphometry in Mechanisms of lysosomal enzyme release from human leuko- electron microscopic cytoldgy. Int. Rev. Cytol.. 26:235-302. cytes: microtubule assembly and membrane fusion induced by a 30. Wright, D. G., D. A. Bralove, and J. I. Galin. 1977. The component of complement. Proc. Natl. Acad. Sci. USA differential mobilization of human neutrophil granules. Effects 70:2916-2920. of phorbol myristate acetate and ionophore A23187. Am. J. 11. Goldstein, I. M., S. T. Hoffstein, and G. Weissmann. 1975. Pathol. 87:273-281. Mechanisms of lysosomal enzyme release from human polymor- 31. Zurier, R. B., G. Weissmann, S. Hoffstein, D. Kammerman, and phonuclear leukocytes. J. Cell Biol. 66:647-652. H. H. Tai. 1974. Mechanisms of lysosomal enzyme release from 12. Goren, M. B. 1977. Phagocyte lysosomes; interactions with human leukocytes. II. Effects of cAMP and cGMP, autonomic infectious agents, phagosomes, and experimental perturbations agonists, and agents which affect microtubule function. J. Clin. in function. Annu. Rev. Microbiol. 3:507-533. Invest. 53:297-309.