1326
INFECTION IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS Association with a Serum Inhibitor of Complement-Derived Chemotactic Activity H. DANIEL PEREZ, RICHARD I. ANDRON, and IRA M. GOLDSTEIN
We have found subnormal amounts of chemotactic activity in zymosan-treated sera from 13 of 29 patients with systemic lupus erythematosus (SLE). As an explanation for this abnormality, the presence of a uniquely specific, heat-stable inhibitor of complement (C5)derived chemotactic activity has been documented in sera from 11 of these patients. Sera from 2 other patients contained elevated levels of nonspecific, heat-labile chemotactic factor inactivator (CFI) activity. The serum from 1 patient contained the heat-stable inhibitor as well as elevated levels of CFI. Patients with SLE whose sera contained the heat-stable inhibitor had more active disease clinically, but otherwise they were indistinguishable from patients without the inhibitor. When patients with the heat-stable inhibitor improved clinically, this usually was accompanied by a decrease in serum inhibitory activity. Only one episode of bacterial infection was observed among 16 patients with SLE whose sera yielded normal amounts of chemotactic activity after treatment with zymosan. In contrast, 7 of 11 patients with SLE whose sera contained the heat-stable From the Department of Medicine, Division of Rheumatology, New York University Medical Center, New York, NY 10016. Supported by grants from the National Institutes of Health (AM-18531, AM-11949, GM-2321 I, and HL-19721). the National Foundation-March of Dimes, and the National Science Foundation (76-0562 I). H. Daniel Perez, MD: Instructor in Medicine, recipient of NIAMDD Clinical Investigator Award (AM-00463); Richard I. Andron, MD: Fellow, Division of Rheumatology; Ira M. Goldstein, MD, FACP Associate Professor of Medicine, recipient of a Career Scientist Award from the Irma T. Hirschl Trust. Address reprint requests to H. Daniel Perez, MD,Department of Medicine, New York University Medical Center, 550 First Avenue, New York, NY 10016. Submitted for publication March 6, 1979; accepted in revised form July 14, 1979.
Arthritis and Rheumatism, Vol. 22, No. 12 (December 1979)
inhibitor suffered serious bacterial infections. The presence of this heat-stable inhibitor in sera from some patients with SLE may contribute, in part, to their increased susceptibility to infection. Patients with systemic lupus erythematosus (SLE) appear to be particularly susceptible to infections caused by common pyogenic microorganisms (1-4). Whereas modem therapy with adrenal corticosteroids is undoubtedly a contributing factor (3,4), there is some evidence that the disease per se is associated with abnormalities of host defenses against infection (4- 10). We have recently reported one such abnormality involving complement-derived chemotactic activity for polymorphonuclear leukocytes (PMN) (1 1). We found that sera from 5 of 11 patients with SLE, when activated with zymosan, failed to attract normal human PMN comparably to zymosan-activated control serum. Incubation of normal PMN with sera from those patients did not affect either the random motility of the cells or their ability to respond to chemoattractants. Serum from 1 patient, however, did contain elevated levels of a previously described nonspecific, heat-labile chemotactic factor inactivator (CFI) (12). Sera from the other 4 patients contained a previously unrecognized inhibitor of complement-derived chemotactic activity. The inhibitor was heat-stable and uniquely specific. It acted reversibly only on chemotactic peptides derived from the fifth component of complement (C5). It had no effect on the chemotactic activity exhibited by various concentrations (all suboptimal) of either the bacterial chemotactic factor from Escherichiu coli or the chemotactic synthetic peptide, N-formylmethionylleucylphenylalanine. Too few patients were examined to de-
INFECTION IN SLE PATIENTS
termine whether t h e presence of this heat-stable inhibitor was associated with any specific clinical o r laboratory parameters of disease activity. Consequently, we have extended our studies and now report findings in 29 patients with SLE.
MATERIALS AND METHODS Patient population. Twenty-nine patients with well documented SLE who were followed by the Rheumatology Division at New York University Medical Center were studied. The only criteria used to select patients for study were the availability of samples of blood for laboratory testing and the absence of severe impairment of renal function (blood urea nitrogen greater than 35 mg/dl or serum creatinine greater than 2.0 mg/dl). The criteria for diagnosis of SLE were those of the American Rheumatism Association (13). The patients ranged in age from 18 to 56 years (mean 34 years) and all were women. Duration of disease ranged from 6 months to 23 years (mean 6 years). The extent of disease activity in these patients was assessed according to the criteria of Rothfield and Pace (14). Grade 0 was assigned to patients without clinical evidence of disease activity; Grade 1 to patients with disease activity involving one organ system without fever; Grade 2 to patients with active involvement of one system with fever or involvement of more than one system; and Grade 3 to patients with involvement of at least two systems with fever. Remissions were classified as either partial or complete. Partial remission was defined as a transition from Grade 2 or Grade 3 disease activity to Grade 1. Complete remission corresponded to Grade 0. Eighteen patients with rheumatoid arthritis (RA) were also studied. All were seropositive and met the criteria for the diagnosis of classic rheumatoid arthritis established by the American Rheumatism Association (15). There were 10 women and 8 men in this group, ranging in age from 42 to 65 years (mean 55 years). Duration of disease ranged between 1 and 21 years (mean 8 years). None of these patients had abnormal renal function and none was receiving therapy with corticosteroids. Two additional groups of patients were studied. The first consisted of 17 patients admitted to New York University Medical Center for treatment of systemic bacterial infections. All infections were documented by blood cultures and were similar to those observed in patients with SLE. None of these patients had clinical or laboratory evidence of either rheumatic or hematologic disease. There were 8 men and 9 women in this group, ranging in age from 23 to 80 years (mean 52 years). The second group included 21 healthy volunteers, 18 women and 3 men, ranging in age from 20 to 40 years (mean 29 years). Preparation of leukocyte suspensions and sera. Human PMN were isolated from venous blood by dextran sedimentation, as described previously (1 1). Cell pellets were washed once with isotonic saline and finally suspended in 10 mM phosphate-buffered saline (Grand Island Biological Co., Grand Island, New York) supplemented with 0.6 mM CaCl,, 1.0 mM MgCl,, and 2% (wt/vol) bovine serum albumin. This buffer was adjusted to pH 7.4 and was used throughout. Cell
I327
suspensions contained approximately 85% PMN. Sera were obtained from clotted blood after centrifugation and either . used fresh or stored in aliquots at -7OOC. Chemotactic factors. Chemotactic activity was generated in sera by adding zymosan (1.O mg/ml, ICN Nutritional Biochemicals Division, International Chemical and Nuclear Corp., Cleveland, Ohio) (1 1). After 15 minutes of incubation at 37OC, the zymosan-sera suspension was centrifuged at 3000g for 10 minutes, and the particle-free supernate was used directly (diluted with buffer as indicated for each experiment). Chemotactic activity in zymosan-treated sera is attributable primarily to C5-derived peptides (16). Partially purified C5derived chemotactic peptides were isolated as previously described (1 1) by molecular sieve chromatography of zymosanactivated normal sera. Other chemoattractants included the synthetic peptide, N-Formylmethionylleucylphenylalanine (N-formyl-metleu-phe; Peninsula Laboratories, Inc., San Carlos, California) (17) and a bacterial chemotactic factor from E coli. The latter was prepared according to the method described by Ward et a1 (18). PMN chemotaxis. PMN random motility and directed migration (chemotaxis) were assessed as desctibed previously (1 1) by employing the “leading front” method of Zigmond and Hirsch (19). Modified Boyden chambers containing cells and chemoattractants were incubated at 37°C for 45 minutes in an atmosphere of 5% C02 and 100% humidity. The response of PMN either to buffer alone (random motility) or to chemotactic stimuli was measured as the distance the leading front of cells migrated into the filter &m/45 minutes). Triplicate chambers were employed in each experiment, and 10 fields were examined in each filter. For experiments in which unactivated sera were added to the lower compartments of the modified Boyden chambers, concentrations were selected that did not by themselves enhance PMN migration in excess of that observed when albumin-containingbuffer was employed alone (20).
RESULTS Chemotactic activity in zymosan-treated SLE serum. Sera obtained from 29 patients with SLE were activated with zymosan and examined for the presence of chemotactic activity. Zymosan-treated sera from 13 patients failed to attract PMN comparably to zymosantreated normal sera (Table 1). Whereas it was possible that the subnormal amounts of chemotactic activity in zymosan-treated sera from these 13 patients resulted from abnormalities involving the complement system, this possibility was not explored further. We have demonstrated previously (1 1) that decreased chemotactic activity in some zymosan-treated SLE sera could not be attributed either to modestly depressed levels of serum complement or to failure of zymosan to activate the alternative complement pathway. Consequently, all sera were examined for the presence of heat-stable inhibitory activity directed specifically toward CS-derived
PEREZ ET AL
1328
Table 1. Decreased chemotactic activity in zymosan-treated SLE serum Stimulus
PMN migration m / 4 5 min)*
Buffer alone (random motility) Zymosan-treated normal sera? Zymosan-treated SLE sera:? DS LM PG RJ LN JH EL AJ MC
99.6 f 2.9 135.9 f 2.2 113.5 f 1.6 112.1 f 3.1 107.3 f 2.7 106.0 2.0 112.4 2.6 114.5 f 2.8 109.7 f 2.3 108.1 f 2.2 109.5 f 2.3 111.1 f 3 . 5 110.2 f 2.0 113.3 f 4.0 123.4 f 3.6
* *
cs
AT MR
IV
* Results represent the mean f SEM of three experiments using polymorphonuclear leukocytes (PMN) from different normal donors. t Zymosan-treated sera were used at a concentration of 2.0% (vol/ vol).
chemotactic peptides as well as for elevated levels of heat-labile CFI activity. Fresh and heated (56°C for 30 minutes) sera were incubated with suboptimal concentrations of CS-derived peptides, the bacterial chemotactic factor from E cdi, and the chemotactic synthetic peptide, N-formyl-met-leu-phe. Heat-stable inhibitory activity directed specifically toward CS-derived chemotactic peptides was found in sera from 11 patients.
Heated sera from each of these patients inhibited the chemotactic activity generated in zymosan-treated normal serum (Table 2). Only 3 patients with SLE were found to have elevated levels of heat-labile CFI activity. Only unheated sera from these 3 patients inhibited the chemotactic activity of the bacterial factor and N-formyl-met-leu-phe (Table 3). Serum from 1 of these 3 patients also contained the heat-stable inhibitor. Neither the heat-stable inhibitor nor elevated levels of CFI activity (not shown) were detected in sera from 18 patients with RA, 17 non-SLE patients with systemic bacterial infection, or 2 1 healthy volunteers (Figure 1). Clinical and laboratory findings in patients with SLE. The patients with SLE were divided into two groups based on whether their sera, after activation with zymosan, contained normal amounts of chemotactic activity (Table 4). Assignment to either group was determined by the results obtained by examining the initial sample of serum available from eeach patient. Both groups proved to be comparable in age and duration of disease. Furthermore, there were no significant differences when erythrocyte sedimentation rates, levels of C3 and C4 (determined immunochemically) (21), and therapy with corticosteroids were compared. The two groups did differ, however, with respect to the clinical activity of their SLE. Based on the criteria of Rothfield and Pace (14), the SLE patients with abnormal chemotactic activity in their zymosan-treated sera also tended
Table 2. Effect of normal and SLE sera on C5-derived chemotactic activity Stimulus Buffer alone (random motility) Zymosan-treated normal sera heated normal serum$ + heated SLE serum:$ DS LM PG RJ
+
LN JH EL AJ MC
cs
AT
PMN migration m / 4 5 min)*
Percent inhibition of chemotactic activity?
100.1 f 1.3 137.0 f 2.0 136.1 f 3.1
-
116.4 f 1.1 111.1 f 3 . 7 110.8 f 4.0 106.0 f 2.4 116. I f 1.0 117.7 f 1.2 110.5 f 4.0 107.7 f 2.3 113.6 f 1.9 113.9 f 1.3 112.7 f 1.0
66.8% 70.0% 71.0% 83.8% 66.5% 52.2% 72.5% 79.2% 63.3% 62.5% 65.7%
* Results represent the mean f SEM of five experiments in which PMN from different normal donors was used.
t Net PMN chemotaxis (minus random motility) in the presence of SLE sera/net PMN chemotaxis in the presence of normal sera X 100. $ Zymosan-treated sera were preincubated ( I : I, vol/vol) for 30 minutes at 37°C with heated (56OC for 30 minutes) normal or SLE sera. The final concentration of zymosan-treated sera was adjusted to 2.0% (vol/vol).
1329
INFECTION IN SLE PATIENTS
Table 3. Elevated levels of chemotactic factor inactivator (CFI) activity in sera from three patients with SLE ~
~~
% inhibition of chemotacticactivity by SLE sera?
Chemotactic stimulus* Zymosan-treated normal sera (2.055, vol/vol) + SLE sera + heated SLE sera Bacterial chemotactic factor (5.0%, vol/vol) + SLE sera + heated SLE sera N-formyl-met-leu-phe (1 x lo-%) + SLE sera + heated SLE sera
MR
IV
EL*
35 None
34 None
63 60$
86 None
35 None
41 None
60
44
80
None
None
None
Chemotactic stimuli were incubated with equal volumes of fresh or heated (56OC for 30 minutes) SLE sera at 37°C for 30 minutes. t PMN chemotaxis @/45 min) in the presence of SLE sera/PMN chemotaxis @/45 min) in the absence of SLE sera X 100. Patient EL serum contained heat-labile CFI activity as well as the heat-stable inhibitor of CS-derived chemotactic peptides.
*
to have more active disease clinically. Seven of 13 were assigned to Grade 3, 4 to Grade 2, and 2 to Grade 1. Seven of the 11 patients in this group whose sera contained the heat-stable inhibitor had Grade 3 clinical activity. One of the patients with heat-labile CFI activity
100
7
dKuML SLE (291
Rheumatoid Arthritis (181
-
Figure 1. Presence of a heat-stable inhibitor of complement (CS)-derived chemotactic activity in sera from some patients with SLE. Zymosan-treated normal sera were incubated at 37°C for 30 minutes with equal volumes of heated (56°C for 30 minutes) SLE, non-SLE, RA,and normal sera. The final concentration of zymosan-treated sera was 2.0% (vol/vol). Results are expressed as percent inhibition of the chemotactic activity observed in zymosan-treated sera alone (see Table 2).
in the serum was classifled as having Grade 2 clinical activity, whereas the other 2 were assigned to Grade 1. In contrast, of the 16 patients with normal chemotactic activity in their zymosan-treated sera, only 3 fulfilled the criteria for Grade 3 activity. Twelve of these patients were classified as Grade 2 and l as Grade l. Only among the group of 11 patients with the heat-stable inhibitor was the incidence of Grade 3 clinical activity in, analysis with creased significantly (P~ 0 . 0 5 chi-square Yates' correction) (22). Serial determinations of inhibitory activity were performed in sera from 21 of the 29 patients with SLE. Interestingly, of 10 patients who were initially free of the heat-stable inhibitor, none was found to acquire the inhibitor despite changes in the activity of their disease (28 determinations in 10 patients). In contrast, in all of the patients in whom the heat-stable inhibitor was present initially, inhibitory activity either diminished markedly or disappeared completely when partial or complete remission of their disease was achieved (Figure 2). Sera from patients with active disease (Grade 2 or 3) inhibited the chemotactic activity in zymosantreated normal skra by 68.5 & 2.5%. This decreased to 21.2 f 15.6% with complete or partial remission (P ~ 0 . 0 1 Student's , t test). In 3 patients, maximum inhibitory activity appeared in sera within 1 week of an episode of infection and coincidentally with an exacerbation of their SLE. The inhibitor subsequently disappeared from the sera of these patients when their disease activity was controlled with corticosteroid therapy. Two of these patients were not receiving corticosteroids, and the third was receiving 20 mg of prednisone
PEREZ ET AL
1330
Table 4. Clinical and laboratory findings in patients with SLE
Chemotactic activity in zymosan-treated sera* ~~
Age, years Duration of disease, years Erythrocyte sedimentation ratet Serum C3, mg/dl$ Serum C4, mg/dl$ Prednisone therapy, &day Disease activity (number of patients) Grade 0 Grade I Grade 2 Grade 3
~
Abnormal
Normal
34.5 f 11.0 2.4 f 3.1 71.4 f 37.2 85.3 f 48.7 20.6 f 13.8 19.9 f 25. I
34.6 f 10.7 4.1 f 4.9 59.9 f 36.8 82.0 f 36.7 3 I .3 f 22.4 3 I .2 f 23.9
0 2 4 7
0 I 12 3
Values are mean f SD.
t Westergren (mm/hour).
+ Normal values for C3 = 93 f 21; C4 = 26 f 8. daily when infection and maximum inhibitory activity were documented. Elevated levels of heat-labile CFI activity in sera from 3 patients remained constant in multiple determinations and did not appear to correlate with either disease activity or episodes of bacterial infection. Interestingly, all 3 patients with elevated levels of CFI in their serum were anergic (23-26). Incidence of infections. Episodes of bacterial infection were recorded in both groups of patients. All infections were documented clinically and by bacteriologic cultures, and all responded to appropriate antimicrobial therapy. Presumed viral infections were not tabulated due to insufficient documentation by culture or serology. Only those infections that occurred within 2 weeks of examining sera (initial or followup samples) for chemotactic activity were considered. Bacterial infections were documented in 7 of the 13 patients whose sera yielded abnormal chemotactic activity after treatment with zymosan (Table 5). Sera from all 7 patients contained the heat-stable inhibitor of C5-derived chemotactic activity. Serum from 1 patient also contained elevated levels of CFI (Table 3). Of these 7 patients, 3 were not receiving therapy with corticosteroids at the time that infection was documented. Two other patients were receiving less than 20 mg of prednisone daily, while the remaining 2 were receiving 40 mg and 60 mg of prednisone a day, respectively. No patients were receiving therapy with other immunosuppressive or cytotoxic agents. The 4 patients in this group in whom bacterial infections were not documented were receiving therapy with corticosteroids in doses ranging from 20 to 60 mg of prednisone daily. Serum levels of
complement (C3 and/or C4) were moderately reduced in only 3 patients at the time that infection was documented. Only one episode of infection was recorded among the group of SLE patients whose sera contained normal chemotactic activity after treatment with zymosan (P
INFECTION IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS Association with a Serum Inhibitor of Complement-Derived Chemotactic Activity H. DANIEL PEREZ, RICHARD I. ANDRON, and IRA M. GOLDSTEIN
We have found subnormal amounts of chemotactic activity in zymosan-treated sera from 13 of 29 patients with systemic lupus erythematosus (SLE). As an explanation for this abnormality, the presence of a uniquely specific, heat-stable inhibitor of complement (C5)derived chemotactic activity has been documented in sera from 11 of these patients. Sera from 2 other patients contained elevated levels of nonspecific, heat-labile chemotactic factor inactivator (CFI) activity. The serum from 1 patient contained the heat-stable inhibitor as well as elevated levels of CFI. Patients with SLE whose sera contained the heat-stable inhibitor had more active disease clinically, but otherwise they were indistinguishable from patients without the inhibitor. When patients with the heat-stable inhibitor improved clinically, this usually was accompanied by a decrease in serum inhibitory activity. Only one episode of bacterial infection was observed among 16 patients with SLE whose sera yielded normal amounts of chemotactic activity after treatment with zymosan. In contrast, 7 of 11 patients with SLE whose sera contained the heat-stable From the Department of Medicine, Division of Rheumatology, New York University Medical Center, New York, NY 10016. Supported by grants from the National Institutes of Health (AM-18531, AM-11949, GM-2321 I, and HL-19721). the National Foundation-March of Dimes, and the National Science Foundation (76-0562 I). H. Daniel Perez, MD: Instructor in Medicine, recipient of NIAMDD Clinical Investigator Award (AM-00463); Richard I. Andron, MD: Fellow, Division of Rheumatology; Ira M. Goldstein, MD, FACP Associate Professor of Medicine, recipient of a Career Scientist Award from the Irma T. Hirschl Trust. Address reprint requests to H. Daniel Perez, MD,Department of Medicine, New York University Medical Center, 550 First Avenue, New York, NY 10016. Submitted for publication March 6, 1979; accepted in revised form July 14, 1979.
Arthritis and Rheumatism, Vol. 22, No. 12 (December 1979)
inhibitor suffered serious bacterial infections. The presence of this heat-stable inhibitor in sera from some patients with SLE may contribute, in part, to their increased susceptibility to infection. Patients with systemic lupus erythematosus (SLE) appear to be particularly susceptible to infections caused by common pyogenic microorganisms (1-4). Whereas modem therapy with adrenal corticosteroids is undoubtedly a contributing factor (3,4), there is some evidence that the disease per se is associated with abnormalities of host defenses against infection (4- 10). We have recently reported one such abnormality involving complement-derived chemotactic activity for polymorphonuclear leukocytes (PMN) (1 1). We found that sera from 5 of 11 patients with SLE, when activated with zymosan, failed to attract normal human PMN comparably to zymosan-activated control serum. Incubation of normal PMN with sera from those patients did not affect either the random motility of the cells or their ability to respond to chemoattractants. Serum from 1 patient, however, did contain elevated levels of a previously described nonspecific, heat-labile chemotactic factor inactivator (CFI) (12). Sera from the other 4 patients contained a previously unrecognized inhibitor of complement-derived chemotactic activity. The inhibitor was heat-stable and uniquely specific. It acted reversibly only on chemotactic peptides derived from the fifth component of complement (C5). It had no effect on the chemotactic activity exhibited by various concentrations (all suboptimal) of either the bacterial chemotactic factor from Escherichiu coli or the chemotactic synthetic peptide, N-formylmethionylleucylphenylalanine. Too few patients were examined to de-
INFECTION IN SLE PATIENTS
termine whether t h e presence of this heat-stable inhibitor was associated with any specific clinical o r laboratory parameters of disease activity. Consequently, we have extended our studies and now report findings in 29 patients with SLE.
MATERIALS AND METHODS Patient population. Twenty-nine patients with well documented SLE who were followed by the Rheumatology Division at New York University Medical Center were studied. The only criteria used to select patients for study were the availability of samples of blood for laboratory testing and the absence of severe impairment of renal function (blood urea nitrogen greater than 35 mg/dl or serum creatinine greater than 2.0 mg/dl). The criteria for diagnosis of SLE were those of the American Rheumatism Association (13). The patients ranged in age from 18 to 56 years (mean 34 years) and all were women. Duration of disease ranged from 6 months to 23 years (mean 6 years). The extent of disease activity in these patients was assessed according to the criteria of Rothfield and Pace (14). Grade 0 was assigned to patients without clinical evidence of disease activity; Grade 1 to patients with disease activity involving one organ system without fever; Grade 2 to patients with active involvement of one system with fever or involvement of more than one system; and Grade 3 to patients with involvement of at least two systems with fever. Remissions were classified as either partial or complete. Partial remission was defined as a transition from Grade 2 or Grade 3 disease activity to Grade 1. Complete remission corresponded to Grade 0. Eighteen patients with rheumatoid arthritis (RA) were also studied. All were seropositive and met the criteria for the diagnosis of classic rheumatoid arthritis established by the American Rheumatism Association (15). There were 10 women and 8 men in this group, ranging in age from 42 to 65 years (mean 55 years). Duration of disease ranged between 1 and 21 years (mean 8 years). None of these patients had abnormal renal function and none was receiving therapy with corticosteroids. Two additional groups of patients were studied. The first consisted of 17 patients admitted to New York University Medical Center for treatment of systemic bacterial infections. All infections were documented by blood cultures and were similar to those observed in patients with SLE. None of these patients had clinical or laboratory evidence of either rheumatic or hematologic disease. There were 8 men and 9 women in this group, ranging in age from 23 to 80 years (mean 52 years). The second group included 21 healthy volunteers, 18 women and 3 men, ranging in age from 20 to 40 years (mean 29 years). Preparation of leukocyte suspensions and sera. Human PMN were isolated from venous blood by dextran sedimentation, as described previously (1 1). Cell pellets were washed once with isotonic saline and finally suspended in 10 mM phosphate-buffered saline (Grand Island Biological Co., Grand Island, New York) supplemented with 0.6 mM CaCl,, 1.0 mM MgCl,, and 2% (wt/vol) bovine serum albumin. This buffer was adjusted to pH 7.4 and was used throughout. Cell
I327
suspensions contained approximately 85% PMN. Sera were obtained from clotted blood after centrifugation and either . used fresh or stored in aliquots at -7OOC. Chemotactic factors. Chemotactic activity was generated in sera by adding zymosan (1.O mg/ml, ICN Nutritional Biochemicals Division, International Chemical and Nuclear Corp., Cleveland, Ohio) (1 1). After 15 minutes of incubation at 37OC, the zymosan-sera suspension was centrifuged at 3000g for 10 minutes, and the particle-free supernate was used directly (diluted with buffer as indicated for each experiment). Chemotactic activity in zymosan-treated sera is attributable primarily to C5-derived peptides (16). Partially purified C5derived chemotactic peptides were isolated as previously described (1 1) by molecular sieve chromatography of zymosanactivated normal sera. Other chemoattractants included the synthetic peptide, N-Formylmethionylleucylphenylalanine (N-formyl-metleu-phe; Peninsula Laboratories, Inc., San Carlos, California) (17) and a bacterial chemotactic factor from E coli. The latter was prepared according to the method described by Ward et a1 (18). PMN chemotaxis. PMN random motility and directed migration (chemotaxis) were assessed as desctibed previously (1 1) by employing the “leading front” method of Zigmond and Hirsch (19). Modified Boyden chambers containing cells and chemoattractants were incubated at 37°C for 45 minutes in an atmosphere of 5% C02 and 100% humidity. The response of PMN either to buffer alone (random motility) or to chemotactic stimuli was measured as the distance the leading front of cells migrated into the filter &m/45 minutes). Triplicate chambers were employed in each experiment, and 10 fields were examined in each filter. For experiments in which unactivated sera were added to the lower compartments of the modified Boyden chambers, concentrations were selected that did not by themselves enhance PMN migration in excess of that observed when albumin-containingbuffer was employed alone (20).
RESULTS Chemotactic activity in zymosan-treated SLE serum. Sera obtained from 29 patients with SLE were activated with zymosan and examined for the presence of chemotactic activity. Zymosan-treated sera from 13 patients failed to attract PMN comparably to zymosantreated normal sera (Table 1). Whereas it was possible that the subnormal amounts of chemotactic activity in zymosan-treated sera from these 13 patients resulted from abnormalities involving the complement system, this possibility was not explored further. We have demonstrated previously (1 1) that decreased chemotactic activity in some zymosan-treated SLE sera could not be attributed either to modestly depressed levels of serum complement or to failure of zymosan to activate the alternative complement pathway. Consequently, all sera were examined for the presence of heat-stable inhibitory activity directed specifically toward CS-derived
PEREZ ET AL
1328
Table 1. Decreased chemotactic activity in zymosan-treated SLE serum Stimulus
PMN migration m / 4 5 min)*
Buffer alone (random motility) Zymosan-treated normal sera? Zymosan-treated SLE sera:? DS LM PG RJ LN JH EL AJ MC
99.6 f 2.9 135.9 f 2.2 113.5 f 1.6 112.1 f 3.1 107.3 f 2.7 106.0 2.0 112.4 2.6 114.5 f 2.8 109.7 f 2.3 108.1 f 2.2 109.5 f 2.3 111.1 f 3 . 5 110.2 f 2.0 113.3 f 4.0 123.4 f 3.6
* *
cs
AT MR
IV
* Results represent the mean f SEM of three experiments using polymorphonuclear leukocytes (PMN) from different normal donors. t Zymosan-treated sera were used at a concentration of 2.0% (vol/ vol).
chemotactic peptides as well as for elevated levels of heat-labile CFI activity. Fresh and heated (56°C for 30 minutes) sera were incubated with suboptimal concentrations of CS-derived peptides, the bacterial chemotactic factor from E cdi, and the chemotactic synthetic peptide, N-formyl-met-leu-phe. Heat-stable inhibitory activity directed specifically toward CS-derived chemotactic peptides was found in sera from 11 patients.
Heated sera from each of these patients inhibited the chemotactic activity generated in zymosan-treated normal serum (Table 2). Only 3 patients with SLE were found to have elevated levels of heat-labile CFI activity. Only unheated sera from these 3 patients inhibited the chemotactic activity of the bacterial factor and N-formyl-met-leu-phe (Table 3). Serum from 1 of these 3 patients also contained the heat-stable inhibitor. Neither the heat-stable inhibitor nor elevated levels of CFI activity (not shown) were detected in sera from 18 patients with RA, 17 non-SLE patients with systemic bacterial infection, or 2 1 healthy volunteers (Figure 1). Clinical and laboratory findings in patients with SLE. The patients with SLE were divided into two groups based on whether their sera, after activation with zymosan, contained normal amounts of chemotactic activity (Table 4). Assignment to either group was determined by the results obtained by examining the initial sample of serum available from eeach patient. Both groups proved to be comparable in age and duration of disease. Furthermore, there were no significant differences when erythrocyte sedimentation rates, levels of C3 and C4 (determined immunochemically) (21), and therapy with corticosteroids were compared. The two groups did differ, however, with respect to the clinical activity of their SLE. Based on the criteria of Rothfield and Pace (14), the SLE patients with abnormal chemotactic activity in their zymosan-treated sera also tended
Table 2. Effect of normal and SLE sera on C5-derived chemotactic activity Stimulus Buffer alone (random motility) Zymosan-treated normal sera heated normal serum$ + heated SLE serum:$ DS LM PG RJ
+
LN JH EL AJ MC
cs
AT
PMN migration m / 4 5 min)*
Percent inhibition of chemotactic activity?
100.1 f 1.3 137.0 f 2.0 136.1 f 3.1
-
116.4 f 1.1 111.1 f 3 . 7 110.8 f 4.0 106.0 f 2.4 116. I f 1.0 117.7 f 1.2 110.5 f 4.0 107.7 f 2.3 113.6 f 1.9 113.9 f 1.3 112.7 f 1.0
66.8% 70.0% 71.0% 83.8% 66.5% 52.2% 72.5% 79.2% 63.3% 62.5% 65.7%
* Results represent the mean f SEM of five experiments in which PMN from different normal donors was used.
t Net PMN chemotaxis (minus random motility) in the presence of SLE sera/net PMN chemotaxis in the presence of normal sera X 100. $ Zymosan-treated sera were preincubated ( I : I, vol/vol) for 30 minutes at 37°C with heated (56OC for 30 minutes) normal or SLE sera. The final concentration of zymosan-treated sera was adjusted to 2.0% (vol/vol).
1329
INFECTION IN SLE PATIENTS
Table 3. Elevated levels of chemotactic factor inactivator (CFI) activity in sera from three patients with SLE ~
~~
% inhibition of chemotacticactivity by SLE sera?
Chemotactic stimulus* Zymosan-treated normal sera (2.055, vol/vol) + SLE sera + heated SLE sera Bacterial chemotactic factor (5.0%, vol/vol) + SLE sera + heated SLE sera N-formyl-met-leu-phe (1 x lo-%) + SLE sera + heated SLE sera
MR
IV
EL*
35 None
34 None
63 60$
86 None
35 None
41 None
60
44
80
None
None
None
Chemotactic stimuli were incubated with equal volumes of fresh or heated (56OC for 30 minutes) SLE sera at 37°C for 30 minutes. t PMN chemotaxis @/45 min) in the presence of SLE sera/PMN chemotaxis @/45 min) in the absence of SLE sera X 100. Patient EL serum contained heat-labile CFI activity as well as the heat-stable inhibitor of CS-derived chemotactic peptides.
*
to have more active disease clinically. Seven of 13 were assigned to Grade 3, 4 to Grade 2, and 2 to Grade 1. Seven of the 11 patients in this group whose sera contained the heat-stable inhibitor had Grade 3 clinical activity. One of the patients with heat-labile CFI activity
100
7
dKuML SLE (291
Rheumatoid Arthritis (181
-
Figure 1. Presence of a heat-stable inhibitor of complement (CS)-derived chemotactic activity in sera from some patients with SLE. Zymosan-treated normal sera were incubated at 37°C for 30 minutes with equal volumes of heated (56°C for 30 minutes) SLE, non-SLE, RA,and normal sera. The final concentration of zymosan-treated sera was 2.0% (vol/vol). Results are expressed as percent inhibition of the chemotactic activity observed in zymosan-treated sera alone (see Table 2).
in the serum was classifled as having Grade 2 clinical activity, whereas the other 2 were assigned to Grade 1. In contrast, of the 16 patients with normal chemotactic activity in their zymosan-treated sera, only 3 fulfilled the criteria for Grade 3 activity. Twelve of these patients were classified as Grade 2 and l as Grade l. Only among the group of 11 patients with the heat-stable inhibitor was the incidence of Grade 3 clinical activity in, analysis with creased significantly (P~ 0 . 0 5 chi-square Yates' correction) (22). Serial determinations of inhibitory activity were performed in sera from 21 of the 29 patients with SLE. Interestingly, of 10 patients who were initially free of the heat-stable inhibitor, none was found to acquire the inhibitor despite changes in the activity of their disease (28 determinations in 10 patients). In contrast, in all of the patients in whom the heat-stable inhibitor was present initially, inhibitory activity either diminished markedly or disappeared completely when partial or complete remission of their disease was achieved (Figure 2). Sera from patients with active disease (Grade 2 or 3) inhibited the chemotactic activity in zymosantreated normal skra by 68.5 & 2.5%. This decreased to 21.2 f 15.6% with complete or partial remission (P ~ 0 . 0 1 Student's , t test). In 3 patients, maximum inhibitory activity appeared in sera within 1 week of an episode of infection and coincidentally with an exacerbation of their SLE. The inhibitor subsequently disappeared from the sera of these patients when their disease activity was controlled with corticosteroid therapy. Two of these patients were not receiving corticosteroids, and the third was receiving 20 mg of prednisone
PEREZ ET AL
1330
Table 4. Clinical and laboratory findings in patients with SLE
Chemotactic activity in zymosan-treated sera* ~~
Age, years Duration of disease, years Erythrocyte sedimentation ratet Serum C3, mg/dl$ Serum C4, mg/dl$ Prednisone therapy, &day Disease activity (number of patients) Grade 0 Grade I Grade 2 Grade 3
~
Abnormal
Normal
34.5 f 11.0 2.4 f 3.1 71.4 f 37.2 85.3 f 48.7 20.6 f 13.8 19.9 f 25. I
34.6 f 10.7 4.1 f 4.9 59.9 f 36.8 82.0 f 36.7 3 I .3 f 22.4 3 I .2 f 23.9
0 2 4 7
0 I 12 3
Values are mean f SD.
t Westergren (mm/hour).
+ Normal values for C3 = 93 f 21; C4 = 26 f 8. daily when infection and maximum inhibitory activity were documented. Elevated levels of heat-labile CFI activity in sera from 3 patients remained constant in multiple determinations and did not appear to correlate with either disease activity or episodes of bacterial infection. Interestingly, all 3 patients with elevated levels of CFI in their serum were anergic (23-26). Incidence of infections. Episodes of bacterial infection were recorded in both groups of patients. All infections were documented clinically and by bacteriologic cultures, and all responded to appropriate antimicrobial therapy. Presumed viral infections were not tabulated due to insufficient documentation by culture or serology. Only those infections that occurred within 2 weeks of examining sera (initial or followup samples) for chemotactic activity were considered. Bacterial infections were documented in 7 of the 13 patients whose sera yielded abnormal chemotactic activity after treatment with zymosan (Table 5). Sera from all 7 patients contained the heat-stable inhibitor of C5-derived chemotactic activity. Serum from 1 patient also contained elevated levels of CFI (Table 3). Of these 7 patients, 3 were not receiving therapy with corticosteroids at the time that infection was documented. Two other patients were receiving less than 20 mg of prednisone daily, while the remaining 2 were receiving 40 mg and 60 mg of prednisone a day, respectively. No patients were receiving therapy with other immunosuppressive or cytotoxic agents. The 4 patients in this group in whom bacterial infections were not documented were receiving therapy with corticosteroids in doses ranging from 20 to 60 mg of prednisone daily. Serum levels of
complement (C3 and/or C4) were moderately reduced in only 3 patients at the time that infection was documented. Only one episode of infection was recorded among the group of SLE patients whose sera contained normal chemotactic activity after treatment with zymosan (P
