Neutrophil Mobilization Induced by Complement ... - Europe PMC

4 downloads 321 Views 1MB Size Report
Complement-de- rived components capable of releasing marrow my- ... 50 N. Medical Drive, Salt Lake City, UT 84132. 623 ..... Myeloid cells were recovered in increased quanti- ties from ... hard,2 and C3d,g in an adult rat model.4 The C3d-K.
American Journal of Pathology, Vol. 133, No. 3, December 1988 Copyright ©) American Association of Pathologists

Neutrophil Mobilization Induced by Complement Fragments During Experimental Group B Streptococcal (GBS) Infection From the Departments of Pathology and Pediatrics, University of Utah School ofMedicine, Salt Lake City, Utah

ANN 0. SHIGEOKA, MD, R.JOSEPH GOBEL, MS, JARMILAJANATOVA, PhD, and HARRY R. HILL, MD

Degradation products of the third component of complement have been reported to have the ability to mobilize leukocytes from the marrow and induce leukocytosis. The effect of C3d,g preparations on neutrophil responses in a neonatal rat model of group B streptococcal infection in which neutrophil mobilization from the marrow is inadequate has been evaluated. Dimeric and monomeric fragments of C3d,g were isolated from human serum; the identity of the C3d,g preparations was confirmed by SDS-PAGE, Western blotting, and N-terminal amino acid sequencing. Uninfected neonatal rats responded to intraperitoneal injection of C3d,g with a peripheral blood neutrophilia at 30 minutes and 4 hours after inoculation. C3d,g,

which lacks intrinsic chemotactic activity, enhanced the local accumulation of neutrophils in the peritoneal cavity of infected, but not uninfected, neonatal rats. In addition, myeloid cell release from the marrow ofisolated femurs of neonatal rats receiving C3d,g was significantly enhanced. Thus, the effect of C3d,g in this model was to mobilize marrow cells and induce peripheral leukocytosis. Chemotactic factors released at the site of infection then resulted in the local accumulation of these inflammatory cells. Complement-derived components capable of releasing marrow myeloid elements may play a major role in determining the outcome of bacterial infection in the immature host. (AmJ Pathol 1988, 133:623-629)

FRAGMENTS OF THE THIRD (C3) and fifth (C5) components of complement are critical in recruiting the cellular components of the acute inflammatory response to sites of bacterial invasion. Ward and colleagues1 have shown that the major complement-derived chemoattractants for polymorphonuclear leukocytes (PMNs) are related to C5, while others2-6 have indicated that degradation products ofC3b, including C3d,g, and C3d-K, may cause the release of PMNs from marrow myeloid reserves. We7'9 have suggested that delayed release of these granulocyte reserves may contribute to the failure of human infants and neonatal rats to accumulate sufficient PMNs at the local site of bacterial infection. In the present studies, we have evaluated the effects of a C3 fragment (C3d,g) on the release of myeloid cells from uninfected and group B streptococcal infected neonatal rats. This fragment is physiologically derived from C3b after its degradation to iC3b by Factor I. The cleavage of iC3b yields C3d,g and C3c. We report that C3d,g preparations are effective in mobilizing marrow myeloid cells. Moreover,

administration of C3d,g dimers to experimental animals resulted in enhanced accumulation of PMNs at the site of group B streptococcal inoculation. Thus, fragments of the third component of complement, including C3d,g, are critical to the mobilization of these cells from the marrow and the enhancement of their accumulation at the site of bacterial invasion.

Materials and Methods Serum Blood was drawn from three healthy adult donors using Vacutainers (Becton-Dickinson #6432) and alSupported by grants AI 13150 and Al 19094 from the National Institute of Health. Accepted for publication August 9, 1988. Address reprint requests to Dr. Harry R. Hill, Department of Pathology, University of Utah School of Medicine, 50 N. Medical Drive, Salt Lake City, UT 84132.

623

624

SHIGEOKA ET AL

lowed to clot for 30 minutes at room temperature (approximately 22 C). Serum was separated by centrifugation at 665g for 15 minutes at 4 C and placed on ice. The serum from each donor was processed immediately to obtain three individual C3d,g containing eluates, each of which was evaluated separately for induction of leukocytosis in our neonatal rat model.

Generation of C3d,g Fragments Fresh serum was incubated with cobra venom factor (Cordis Laboratories, Miami, FL) and aggregated IgG to fully activate both complement pathways. C3 was fully activated after a 4-hour incubation at 37 C yielding iC3b and other fragments. An additional incubation for 8 hours at 30 C, and 14 hours at room temperature achieved complete processing of iC3b to C3c and C3d,g fragments by a modification of previous methods.'0 The modifications consisted of the elimination of N-ethylmaleimide, which was found to decrease the yield of C3d,g dimers, and the omission of erythrocytes that were not required for complete conversion of iC3b to C3d,g and C3c. Fifteen-milliliter aliquots of the complement-activated serum were quick-frozen in a mixture of solid CO2 and acetone, then stored at -20 C. C3 fragments were isolated by

immunoaffinity chromatography. The immunoaffinity (IA) matrices were prepared by previously described techniques" using goat polyclonal anti-human C3 (Cappel/Cooper Biomedical, Malvern, PA), and rabbit anti-C3d antibodies raised against an elastase-generated C3d (a gift from Dr. B. F. Tack'2). The anti-human C3 antibody was shown previously to react with the C3c antigen ofC3.'0 Thus, it adsorbs C3c and any remaining C3, C3b, or iC3b. IgG fractions were prepared from each antiserum by caprylic acid precipitation at pH 5.0,13 dialyzed against the coupling buffer at pH 6.8, and then covalently linked to cyanogen bromide activated Sepharose 4B (Sigma Chemical Co., St. Louis, MO). The total amount of IgG immobilized per ml of Sepharose was 36 mg for anti-C3 and 20 mg for anti-C3d,g IA beads. The binding capacity per ml for the respective IA was approximately 3 mg of C3, and 0.4 mg of C3d containing fragments. Pepsin treatment was used to convert anti-C3 IgG-IA to F(ab')2-IA." Both anti-C3 and anti-C3d IA were pre-equilibrated with equilibration buffer (EQB) (50 mM Na P04 with 250 mM NaCl, pH 7.4) containing 0.02% azide. These IA preparations were analyzed by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) as described previously. '0"

AJP * December 1988

Isolation of C3d,g and C3c Fragments

The activated serum was applied to the pre-equilibrated anti-C3-IA at a flow rate of 10 ml/sq cm/h, which was directly connected to the anti-C3d-IA column. Both columns were washed with EQB after the serum application. The eluate was collected in 0.5 ml aliquots and protein content measured by absorbance at 280 nm. When the absorbance of undiluted fractions approached zero, columns were disconnected. To elute nonimmunologically bound proteins such as albumin, each column was washed with 3M NaCl in EQB. The columns were then washed with EQB to decrease the NaCl concentration. The majority of each antigen was eluted using aqueous 1 M acetic acid (pH 2.5), and adjusted to pH 7.4 with 2M Tris-Base (Sigma). Both columns were promptly washed with EQB to decrease the effect of low pH on the antibodies and Sepharose matrix, thus re-equilibrating the IA columns. After SDS-PAGE analysis, the eluates from the columns were exhaustively dialyzed against 200-fold volumes of phosphate-buffered saline (PBS) (pH 7.4) with 3 changes at 4 C, using Spectrapor dialysis tubing (Spectrum Medical Industries, Inc., Los Angeles, CA) with a cutoffof 3500 daltons. Pools were concentrated by ultrafiltration over an Amicon PM 10 membrane. When required, remaining albumin was removed by gel filtration on Sephacryl S-200. Separation of C3d,g dimers and monomers from the C3d,g eluates was performed using SDS-PAGE in the apparatus produced by Savant Instruments, Inc. (Farmingdale, NY), which consists of a 6 cm resolving gel (10% polyacrylamide) and 2 cm stacking gel layered in a 1.5 X 11 cm tube.'4 The C3d,g eluate (5 mg in 2.5 ml PBS) was mixed with 1.25 ml of sample buffer containing 0.3% SDS and 30% glycerol. Other conditions were as described for C3c and C4c isolation.'0 Each eluate was examined by absorbance at 220 nm and 280 nm and using analytic SDS-PAGE. The preparations were then exhaustively dialyzed against PBS, sterilized by passage through a 0.22 ,u Millipore filter, and stored below 5 C. Endotoxin contamination, assayed by Limulus test using the Endotect kit (Schwarz/Mann, Cleveland, OH), was not detected. For N-terminal sequence analysis, samples were dialyzed in the 3500 dalton Spectrapor tubing against 50 mM NH4HCO3. Aliquots of the eluates and purified monomeric and dimeric C3d,g used for in vivo assays were examined by SDS-PAGE under reduced and nonreduced conditions and by immunoelectrophoresis. In addition, Ouchterlony double immunodiffusion was performed using the anti-C3d described above, commercial anti-

NEUTROPHIL MOBILIZATION BY COMPLEMENT

Vol. 133 oNo. 3

625

kDo

Figure 1-SDS polyacrylamide gel electrophoresis of C3d,g preparations obtained from activated human serum. Contaminants of albumin and C3c were detected by Ouchterlony analysis of the eluate represented in lane 1. Lane 2 contains a 10,000 dalton fragment; lanes 6 and 7 the monomeric C3d,g; and lane 8 is the dimenc C3d,g preparation. The 27,000 dalton fragment in lanes 5 and 6 is C3d related by Westem blot and SDS analysis with radiolabeled anti-C3d antibody. The approximately 68,000 dalton fragment has not been specifically identified, but is also C3d related by analysis using anti-C3d antibody.

0.6

-

0.5

-

804010-

E 0.40

1

4

0.2* 0.1

2 3 4 5 6 7 8 I = STARTING MATERIAL

a

I Ill 1 2 3456 7

100

C3c and anti-C3d (Dako Corp., Santa Barbara, CA). There was no cross-reaction between anti-C3c and anti-C3d as defined by immunoelectrophoresis in our studies or those of Davis and coworkers.'5 Western blot analyses used polyclonal anti-human C3 and monoclonal anti-C3d antibodies. None of the C3d,g eluates contained C3c. Figure 1 depicts an SDSPAGE under nonreduced conditions of a representative C3d,g eluate from the first serum fractionated. Contaminants of albumin and C3c (as determined by Ouchterlony analysis) were detected in lane one. Lane two contained a 10,000 dalton fragment that was blocked at the N-terminus and, therefore, could not be sequenced. Lanes 5 and 7 contained the monomeric C3d,g while the 80 kd fragment was shown to consist of two disulfide bonded C3d,g monomers by N-terminal sequence analysis. Further studies were performed using two additional C3d,g eluates derived from two separate serum donors. Dimeric C3d,g and monomeric C3d,g preparations were obtained by further fractionation of the second and third eluates. These purified dimeric and monomeric fractions did not contain the 10 kd fragment or other C3d,g-related fragments on SDS-PAGE analysis. The residues in the C3d,g preparations were present in the sequence EGVQKEDIP, which corresponds to the sequence for C3d,g reported by Davis and coworkers.'5 In contrast, the residues from the C3d-K nonapeptide were not detected. Neonatal Rat Model The neonatal rat model of GBS infection used to determine the effect of the C3d,g preparations on leukocyte release from the bone marrow and on perito-

120

140 160 FRACTION NUMBER

180

200

neal accumulation of neutrophils has been described previously.7 "61'8 The protocol was approved by the University of Utah committee for laboratory animal research. Outbred Sprague-Dawley rats were infected at 3-4 days of age, randomizing animals from at least three litters to each group. Human isolates of type III GBS for inoculation were cultured in Todd-Hewitt broth (Difco Laboratories, Detroit, MI) at 37 C for 16 hours, washed three times in sterile PBS, and adjusted to an optical density of 0.9 at a wavelength of 620 nm (Spectronic 20; Bausch & Lomb, Rochester, NY). The bacterial suspensions contained approximately 5 x 108 colony-forming units per ml. Rats were injected intraperitoneally with a total volume of 50 ,il of a C3d,g preparation (whole eluate, monomers or dimers) at 1.0-3.0 x 10-9 moles/kg and/or type III GBS at a dose of 106 bacteria/g body weight diluted in sterile PBS. Groups of GBS-infected animals treated or untreated with C3d,g were followed for mortality in each experiment. As expected, in the absence of typespecific antibody7"16-'8 all animals in these experiments died of infection; the time to death varied among experiments but was generally less than 24 hours. Animals were anesthetized by ether inhalation at 30 minutes, 3-4 hours, 7-8 hours, and 24 hours before peritoneal lavage and isolation of femurs for cell sampling. As described previously, peritoneal lavages were performed using six equal volumes of heparinized PBS after weighing each animal. The lavage fluid was mixed with Hank's balanced salt solution containing 0.1% gel (HBSS-gel), centrifuged, and resuspended in HBSS-gel. White blood cell counts and differentials were performed by Coulter counter or hemocytometer, and Wright's stained smears. Periph-

SHIGEOKA ET AL

626

AJP o December 1988

1 20 >_

a

100

0

E 80

o z

0

/t

60

60

40

GBS 106/gm(n-S)

p