Serum Chemotactic Inhibitory Activity: Heat Activation of Chemotactic Inhibition

Transcription

1 INFECTION AND IMMUNITY, Mar. 1976, p Copyright 1976 American Society for Microbiology Vol. 13, No. 3 Printed in U.SA. Serum Chemotactic Inhibitory Activity: Heat Activation of Chemotactic Inhibition DENNIS E. VAN EPPS* AND RALPH C. WILLIAMS, JR. Departments of Medicine and Microbiology, Bernalillo County Medical Center, University of New Mexico School of Medicine, Albuquerque, New Mexico Received for publication 8 September 1975 Serum chemotactic inhibitory activity (CIA) was studied in 46 patients with various systemic diseases, using a system consisting of normal human leukocytes as indicator cells and 10% fresh normal serum as a control chemotactic attractant. It was shown, as previously reported, that an association exists between CIA and skin test anergy. Heat treatment of sera at 56 C for 30 min increased both the incidence and the degree of chemotactic inhibition observed in these patients. The effects of heat treatment of sera containing CIA on other chemotactic attractants (C3a, bacteria-derived chemotactic factor [BF], and casein) are shown. Before heat treatment, some sera suppressed chemotaxis mediated by BF in the absence of suppression of normal serum-mediated chemotaxis, indicating the possible involvement of more than one system of inhibition. Multiple systems were further supported by data indicating that room temperature incubation resulted in a loss of CIA as measured by normal serum-mediated chemotaxis with no apparent decrease in the inhibition of BF-mediated chemotaxis. Separation of sera containing CIA by Sephadex G-200 showed chemotactic inhibitory activity to be increased in both the void volume and 7S regions after heat treatment, with the greatest effect seen in the void volume region. Experiments showed that heat treating before separation resulted in similar increases in both peaks, implying the presence of an antagonist to CIA. Experiments demonstrating that sera containing CIA do not suppress casein-mediated chemotaxis by means of an irreversible inactivation of chemotactic factor are included along with experiments demonstrating a cellular mode of action. The possible presence of two systems of chemotactic inhibition, one acting directly upon chemotactic factors and one interacting with the responding cell, are discussed. Serum chemotactic inhibitors have been described in patients with a variety of systemic treatment. strates a pronounced increase in CIA after heat illnesses (3, 4, 9, 12-15, 17). In previous reports we have shown the presence in serum of three MATERIALS AND METHODS inhibitors of leukocyte chemotaxis based upon Subjects. Serum was obtained from 46 patients differences in sedimentation behavior (15, 16). with various systemic diseases including osteomyelitis (1), liver disease (27), pulmonary disease (12), These inhibitors potentially may act either directly upon chemotactic factors as is the case pancreatitis (1), cellulitis (1), disseminated atypical with chemotactic factor inactivator (CFI) (2, 9, mycobacterium infection (1), and malignancy (3). No 13, 17) or at the cellular level. Serum chemotactic inhibitory activity (CIA) has been shown to as anergic on the basis of skin testing with six patients studied were receiving corticosteroids or immunosuppressive drugs. Patients were classified be transient in nature, and in many patients it antigens including mumps, purified protein derivative, candida, trichophyton, and streptokinase-strep- parallels the presence of skin test anergy (14-16). These factors appear to resemble acutephase reactants since their appearance is prom- were measured for induration at 24 and 48 h, and the todornase as previously defined (14, 15). Skin tests inent during episodes of acute illness or exacerbations of chronic disease (14-16). Additional clared the maximal skin test response. All sera were largest diameter seen at either 24 or 48 h was de- evidence indicates that in patients with cirrhosis the two larger inhibitors (6.8S and 10.7S tested initially or stored at -70 C and later tested for the presence of CIA. In addition to patient samples, sera from 13 normal adults were also tested. sedimentation coefficients) may be associated Chemotactic assay. Chemotaxis was performed with immunoglobulin A (IgA) (16). The current by using human peripheral blood leukocytes in a study expands the data on serum CIA, using a modification of the Boyden technique (1, 15). In the variety of chemotactic attractants, and demon- present study, membrane filters (Millipore Corp.) 741

2 742 VAN EPPS AND WILLIAMS with 5-,um rather than 3-j.m pore size were used. Chemotaxis was evaluated as cells per x400 highpower field (HPF) and in this assay represents greater than 95% polymorphonuclear leukocytes (PMN). All samples were run in duplicate, and results represent the mean number of cells per HPF of these duplicates. Chemotactic attractants consisted of fresh normal human serum (NS), fresh frozen NS stored at -70 C for periods no longer than 2 weeks, and bacteria-derived chemotactic factor (BF) obtained from the supernatant of a 48-h Escherichia coli (strain K-12) culture grown in Eagle minimal medium. In some experiments C3a (generously provided by Viktor Bokisch, Scripps Clinic, La Jolla, Calif.) was activated by trypsin cleavage of isolated C3 as previously described (15). In addition, some experiments used a preparation of casein-saturated minimum essential medium (Eagle) as a chemotactic attractant (100,ul/ml). Casein chemotactic factor was obtained by incubating casein in minimum essential medium (Eagle) for 2 h at 37 C with intermittent mixing. This mixture was then centrifuged to remove any particulate casein that remained, and the solution was then titrated to neutrality with 7.5% bicarbonate. Cell preparations were obtained from heparinized normal adult venous blood after plasmagel sedimentation of erythrocytes and multiple washings with Hanks balanced salt solution (HBSS) (14-16). All patients were screened for chemotactic inhibitors by using a mixture of 10% fresh NS and 10% patient serum as a chemotactic attractant. The chemotaxis of normal PMN toward this mixture was compared with that obtained by using 10% NS alone. Figure 1 demonstrates the effects of increasing concentrations of NS in the lower compartment of the Boyden chamber. Ten percent NS was chosen as a control since this point on the chemotaxis titration curve allowed for optimum sensitivity of the chemotactic assay. Increasing concentrations of NS generally resulted in an increased chemotactic response until concentrations of 30% or greater were attained, after which chemotaxis decreased. Chromatographic separation. Serum inhibitors were separated by Sephadex G-200 gel chromatography, using a column (50 by 2.5 cm) eluted with HBSS. Serum samples (0.5 ml) were applied to the column, and protein profiles were measured by optical absorption at 280 nm. Eluted peaks were concentrated by ultrafiltration. Insolubilized columns. Casein preparations were conjugated to Sepharose by the method of Mannik and Stage (10). Each column contained 25 mg of bound protein. Serum samples were applied and eluted with borate buffer (0.1 M, ph 8.0). RESULTS Increased CIA after heat treatment. Serum samples obtained from both anergic patients and normal adult controls were tested with respect to CIA both before and after heat treatment at 56 C for 30 min. In each case, 10% sample serum was added to 10% fresh NS and INFECT. IMMUN. the mixture was used as a chemotactic stimulus (14-16). Results obtained with 46 patients and 13 normal adults are shown in Table 1. Each patient was skin tested with six antigens for delayed cutaneous hypersensitivity and categorized with respect to the diameter of the largest positive skin test. Separation of patients on this basis yielded 18 patients with skin test results between 0 and 2 mm, 19 patients with skin test results between 3 and 5 mm, and 9 patients with skin test results greater than or equal to 6 mm. In the patients with skin test responses between 0 and 2 mm, 16 of 18 untreated patient sera failed to increase the chemotactic response, as would be expected with increasing concentrations of NS (Fig. 1); in fact, 11 of these 18 actually suppressed chemotaxis to less than 50% of control. If sera were heat treated at 56 C for 30 min before testing, the degree of chemotactic suppression was greatly increased so that all 18 sera failed to increase chemotaxis above control levels and 17 of 18 actually suppressed chemotaxis by more than 50%. The average effect of untreated patient serum on control chemotaxis was a 49% reduction in chemotaxis. After heat treatment, this was increased to a mean inhibition of 84%. In patients with largest skin test diameters ranging between 3 and 5 mm, the inhibitory effect was much less marked. In this case, 9 of 19 untreated patient sera failed to increase the chemotactic response above control levels, and only two of these suppressed chemotaxis by greater than 50%. After heat treatment at 56 C for 30 min, all 19 sera tested yielded less than 100% of control chemotaxis and 12 of 19 suppressed chemotaxis by more than 50%. The mean effect of untreated patient serum expressed as a percentage of control was 102% as compared with 41% after heat treatment. In the third category of patients who had skin tests of greater than or equal to 6 mm, only two of nine samples failed to increase chemotaxis before heat treatment, with only one of these below 50%. After heat treatment, five of nine failed to increase chemotaxis, with three of these below 50%. The average effect before heat treatment was a 25% increase in chemotaxis and after heat treatment a 1% increase. The remaining category consisted of serum from 13 normal adult controls. Each control was tested in conjunction with 10% fresh homologous serum, and the resulting chemotaxis was compared with the control consisting of 10% homologous serum alone. The results indicated that before heat treatment only two of 13 normal sera decreased chemotaxis and only one of these resulted in more than a 12% decrease in chemotaxis. After heat treatment, 4 of 13 failed to increase chemotaxis, and none of

3 VOL. 13, 1976 TABLE 1. SERUM CHEMOTACTIC INHIBITORY ACTIVITY 743 Effects of heat treatment on serum chemotactic inhibitory activity of sera from patients tested concomitantly for delayed-type hypersensitivity Avg no. of cells/hpf Percent of control 10% NS plus Sample 10% NS plus treated 10% heattest se- 10% NS alone 10% testsplus serum 10% NS plus C0 30 me ) rum (56 C, 30 (56 C, 30 min min) Skin tests 0-2 mma GH IN RY CT EA HR JK AM CG TR CA CJ RJ BB AJ MA MC JC Skin tests 3-5 mm LO EM ds AM EP NW MB AL CR SB BG DG GG DS LK DP MC EL RG Skin tests -6mm SM ES VS LG EJ EG dz PU dw

4 744 VAN EPPS AND WILLIAMS INFECT. IMMUN. TABLE 1-Continued Avg no. of cells/hpf Percent of control 10% NS plus Sample 10% heat- 0NSpu 10% NS plus 0NSpu 10% Ns splrusm treated test se- 10% NS alone 10% test serum 10% test serum rum (56 C, etsrm (56 C, 30 min) min) Normal controls BR CJ DV TH AB MJ EH HE DF RD MO ER I (I uj UL 0 z JA... a Skin test results recorded as diameter of largest positive skin test. t4(r- 12C % NORMAL SERJM IN LOWER CHAMBER FIG. 1. Titration curve showing the effects of increasing amounts of fresh normal human serum on PMN chemotaxis. The serum is used in the lower compartment of the chemotaxis chamber as a chemotactic attractant. The mean cell counts 1 standard deviation of five separate cell and normal serum donors are shown. these decreased chemotaxis to the 50% levels. The average effect of untreated NS on control chemotaxis was a 34% increase before heat treatment and a 41% increase after heat treatment. Statistical analysis of these data for the association of CIA and anergy was performed, using chi-square analysis of a 3-by-3 contingency table. Patient serum samples were divided into the following groups with respect to chemotaxis results: (i) samples that resulted in chemotaxis between 0 and 50% of control; (ii) samples that resulted in chemotaxis between 51 and 100% of control; and (iii) samples that increased chemotaxis above controls. These were then compared with the different categories based on skin test reactivity (O to 2, 3 to 5, or.6 mm). Analysis showed P values for the association of anergy and chemotactic inhibition to be significant (P < 0.001) for both untreated and heat-treated sera. Effect of heat treatment on chemotaxis mediated by BF, C3a, and casein. Eight sera shown to increase in CIA after heat treatment when using 10% fresh NS as a chemotactic attractant were tested both before and after heat treatment, using both 10% NS and BF as control chemotactic attractants. Increased CIA after heat treatment was not specific to the chemotactic system using NS as a chemotactic attractant. Such an increase was also demonstrated when using BF as a chemotactic stimulus, although the degree of increase was not always as great. It should be noted that in some patient sera before heat treatment, substantial suppression of chemotaxis mediated by BF was recorded in the absence of suppression of NSmediated chemotaxis (Table 2, samples 4 and 6). These results appeared to indicate possible involvement of more than one system of chemotactic suppression. In addition to experiments with BF, experiments were performed using C3a or casein as a source of chemotactic activity. Two sera were tested in conjunction with C3a, and nine sera were tested in conjunction with casein. In each

5 VOL. 13, 1976 TABLE 2. a SERUM CHEMOTACTIC INHIBITORY ACTIVITY 745 Comparative results of heat treatment of serum containing chemotactic inhibitory activity using normal serum or bacterial factor as a chemotactic attractant Chemotaxis in cells/hpf Sample 0 AofN AlofNS 10AloBF A of BF + Samp1I 100 of 10 0A NS1.d of sample 10 AI of heat- 100 il of BF 100 of sample teatl ofsheatm treated sample tetdsml (59) 14 (10) (75) 21 (20) (167) 33 (61) (116) 33 (89) (146) 21 (51) (151) 41 (80) (115) 3 (7) (24) 7 (14) (73) 3 (5) 40 8 (20) 1 (3) (97) 12 (18) (34) 10 (23) (46) 5 (4) (15) 1 (1) (132) 32 (51) (97) 20 (54) Mean (104) 15 (26) (67) 17 (36) Number in parenthesis is the percentage of control value. case samples were tested concurrently with 10% NS as a control chemotactic attractant. Results of these experiments showed that CIA was demonstrable in all cases and that after heat treatment, increases in suppression of C3a- and casein-mediated chemotaxis paralleled that observed with NS-mediated chemotaxis. These data appeared to indicate that increases in CIA after heat treatment are not specific to NS- and BF-mediated chemotaxis but include chemotaxis mediated by C3a and casein. Failure of CIA to irreversibly inactivate chemotactic factor. Berenberg and Ward (2) have demonstrated that sera containing CFI are capable of irreversibly inactivating the chemotactic activity of BF. This was possible due to the extreme heat stability of BF. These investigators demonstrated that the preincubation of BF with CFI followed by heating to 70 C for 30 min in order to denature the CFI resulted in loss of BF chemotactic activity (9). This heat treatment had no effect on BF in the absence of sera containing CFI and indicated that CFI irreversibly inactivates BF. Accordingly, we performed similar experiments using sera containing CIA in conjunction with casein, which also has extreme heat-stable characteristics. Casein-derived chemotactic activity was unaffected by the heat treatment at 80 C for 1 h, whereas CIA was markedly reduced by this treatment (Table 3). If sera containing CIA were mixed with casein for 1 h at 37 C and then tested without an 80 C heat treatment, substantial inhibition was observed. Alternatively, if casein and inhibitory sera were preincubated for 1 h at 37 C followed by an 80 C treatment for 1 h, chemotactic activity was present and approached control levels. These TABLE 3. Effects of chemotactic inhibitory serum on the chemotactic activity of casein Avg cells/ Percent in- Excpt HPF hibition Caseina Casein plus inhibitory serum heat treated at 80 C Casein plus inhibitory serum 1 99 Casein plus inhibitory serum (mixture incubated 37 C for 1 h followed by a 1-h 80 C heat treatment) a In all cases the casein preparation was heat treated at 80 C for 1 h before testing. results indicate that CIA does not irreversibly inactivate the chemotactic activity of casein, as might be expected with an inhibitor that directly inactivates chemotactic factors. Instead, it appears that CIA either exerts a cellular effect, or alternatively, reversibly binds and suppresses the chemotactic factor. Either of these interpretations is compatible with the results shown in Table 3. Attempts to absorb CIA to an insolubilized casein column. If CIA suppresses casein-mediated chemotaxis by means of reversible binding to the casein itself, insolubilized casein columns might be expected to absorb out the chemotactic inhibitory activity. Such experiments were conducted by using casein conjugated to Sepharose through the use of cyanogen bromide (10). A 0.8-ml amount of serum containing CIA was applied to the column and eluted with borate buffer. Each sample was then concentrated to its original volume and tested for CIA, using either NS or casein as a chemotactic attractant. No significant loss of

6 746 VAN EPPS AND WILLIAMS CIA was noted after passage through the insolubilized casein column. These results do not support high-affinity binding of serum chemotactic inhibitors to casein. Demonstration of cell-directed inhibition of chemotaxis. Data so far indicate that a cellular mode of action is feasible. Experiments were conducted to determine whether preincubation of PMN with sera containing CIA would result in a suppressed chemotactic response. PMN were incubated at 37 C for 15 min in a 10% solution of heat-treated serum (56 C, 30 min) containing CIA. Controls consisted of cells treated in a similar manner with HBSS or a 10% solution of heat-treated NS. Cells were washed three times with HBSS and tested for their chemotactic responsiveness, using 10% fresh NS or casein as a chemotactic attractant. Results using 10% fresh NS as a chemotactic attractant demonstrate a 37% reduction in chemotaxis when cells were preincubated in serum containing CIA (Table 4). Chemotactic suppression was greater when chemotactic inhibitory serum was present with the cells in the upper chamber throughout the assay. Suppression was more effective in two of three experiments when inhibitor was present in the upper chamber rather than the lower chamber. As shown, 10% solutions of heat-treated NS or patient serum showed minimal chemotactic activity (6 and 4 PMN/HPF, respectively), indicating that the observed suppression in chemotaxis after preincubation was not the result of a preincubation of PMN with chemotactic factor. When casein was used as a chemotactic stimulus, cells incubated in two separate heattreated serum samples containing CIA also showed suppressed chemotactic responsiveness (63 and 86% of control cells incubated in heattreated NS). Effects of heat treatment on CIA as a function of time. To ascertain the effect of prolonged heating at 56 C on CIA, sera were TABLE 4. Effects ofpreincubation of PMNs with heat-treated (56 C, 30 min) serum containing CIA Test Test samplea HPFP of Percent control sampleppmns/ Control cells Cells + 10% inhibitory serum 9 9 Cells preincubated in 10% inhibitory serum and then washed Cells preincubated in 10% NS and then washed a All serum samples were heat treated at 56 C for 30 min before testing. b HPF represents high-power field (x400). INFECT. IMMUN. heated for 2 h and sampled at various time intervals. Samples for testing consisted of 10% NS plus 10% CIA serum, and results were compared with those obtained with 10% NS alone. The effects of heating are shown in Fig. 2. This representative curve demonstrated that the increase in CIA was complete after 30 min of heat treatment. Experiments performed on four other inhibitory sera resulted in similar activation profiles. Effect of heat treatment on various chemotactic inhibitors. To ascertain the effect of heat treatment on each of the individual chemotactic inhibitors, CIA serum samples were separated by Sephadex G-200 filtration. Fractions were concentrated to the original sample volume and tested for CIA, using a mixture of 10% NS and 20% of the sample fraction. Fractions were tested both before and after heat treatment (Fig. 3). Chemotactic inhibitors were found in fractions I, II, III, and VI. The inhibitory activity seen in fraction II may have been due to overlap from fractions I and III or to small amounts of the 7S peak inhibitor carried over into fraction II. Both fraction I and III were increased in inhibitory activity after heat treatment, with the greatest increase observed in fraction I. Although the increase in fraction III activity was less than that observed in fraction I, this increase was reproducible. Inhibitory activity in fraction VI was not increased after heat treatment. Experiments performed on samples heat treated before Sephadex G-200 separation resulted in similar profiles of increased CIA in both fractions I and III. Likewise, heat treatment of fractionated sera containing CIA also showed similar profiles of heat activation using BF, with the greatest increase in inhibitory activity observed in the fraction I region (before heat treatment, fraction I = 59% inhibition, fraction III = 72% inhibition; after heat treatment, fraction I = 91% inhibition, fraction III = 79% inhibition). Effect of 22 C incubation on CIA. Since we have often observed a perplexing loss of CIA upon storage of serum even at - 70 C, an experiment was conducted to determine whether CIA decreased during incubation at room temperature. A fresh sample of serum was obtained from an anergic patient with alcoholic liver disease and was tested within 1 h for CIA, after which the serum was set aside to incubate at room temperature (22 C). Two aliquots were removed from this sample at days 0, 1, 2, and 3. One of these aliquots was heat treated at 56 C for 30 min. Both samples were then tested for CIA, using 10% BF or 10% NS as control chemotactic attractant. The data (Fig. 4) are expressed as percentage of inhibition of the NS or

7 VOL. 13, 1976 BF control. Control chemotactic values for NS and BF were as follows: day 0, BF = 50 cells/ HPF, NS = 59 cells/hpf; day 1, BF = 51 cells/ HPF; NS = 69 cells/hpf; day 2, BF = 55 cells/ HPF, NS = 75 cells/hpf; day 3, BF = 40 cells/ HPF, NS = 62 cells/hpf. Results at day 0 indicate that heat-treated and untreated serum inhibited chemotaxis toward both BF and NS to greater than 85%, with the percentage difference between untreated and treated serum being very small. By day 1 most of the inhibitory activity using NS-derived chemotactic factors was lost whereas very little change was z 0 toor 80H SERUM CHEMOTACTIC INHIBITORY ACTIVITY 747 observed using BF. This loss was maximal by day 2 whereas BF inhibition remained fairly stable throughout, with only a slight loss of activity observed by day 2. In each case, 56 C heat treatment of the test serum converted the sample to its original level of chemotactic inhibition. However, this reversible inactivation of CIA was not always effective, and older sera, which had been repeatedly frozen and thawed, irreversibly lost CIA. Differences in inhibition of BF and NS chemotaxis may imply involvement of more than one system of chemotactic suppression _-ii-z 80- \ 70 \\ ao 50- Z I FIG. 4. Effects of prolonged 22 C incubation and HOURS subsequent heat treatment on CIA activity. Samples heat treatedjust before testing (A) are compared with FIG. 2. Effect of heat treatment on CIA activity as untreated samples (A), using both BF- ( ) and a function of time. Results indicate the mean percent NS- derived chemotactic factors (-----) as control cheinhibition ± 1 standard deviation. motactic attractants. E 4 L 0 SI I I to FRACTION NUMBER FIG. 3. Effects of heat treatment on Sephadex G-200-separated chemotactic inhibitors. Solid line indicates the optical density profile at 280 nm. Hatched bars represent mean chemotactic inhibition _ I standard deviation before heat treatment; open bars represent chemotactic inhibition _ 1 standard deviation after heat treatment.

8 748 VAN EPPS AND WILLIAMS DISCUSSION Previous studies have shown CIA to be prevalent in anergic patients and to parallel the presence of skin test anergy in sequential studies on patients who could be studied at times when they were skin test negative and later skin test positive (14-16). The data presented here demonstrate that heat treatment of patient serum before testing for CIA results in an increase in inhibitory activity. Like the CIA previously described in untreated patient serum (14-16), the inhibitory activity demonstrated after heat treatment is also prevalent in anergic patients. The incidence of CIA in patients paralleled the degree of anergy, with the greatest frequency found in patients with less than 2-mm induration in skin test responses. Likewise, the mean chemotactic suppressive activity also paralleled the degree of skin reactivity, with the greatest degree of suppression present in patients with skin tests between 0 and 2 mm in diameter. It appears from these data (Table 1) that some normal control sera may contain low concentrations of CIA, but usually this activity is of a low order of magnitude, being only faintly present even after heating. Studies using BF instead of NS as a control chemotactic attractant indicated that inhibition of BF-mediated chemotaxis also increased after heat treatment, although the degree of increase was not as great and apparently inhibitors of BF-mediated chemotaxis were present in some patients when inhibitors of NS chemotaxis could not be detected (Table 2). Moreover, when isolated C3a or casein was used as chemotactic factor, increased CIA after heat treatment of test serum paralleled that observed using NS as a chemotactic attractant. Increased CIA after heat treatment was complete within 30 min, and further prolonged incubation at 56 C did not increase the inhibitory activity. In addition, if samples were incubated at room temperature, a decrease in CIA was observed when using 10% fresh NS as a chemotactic attractant. This loss of activity was reversed by a subsequent 30-min heat treatment at 56 C. These characteristics suggest that the loss of CIA may be due to the generation of heat-labile CIA suppressors or dissociation and subsequent heat-catalyzed association of chemotactic inhibitors. Heat-labile chemotactic suppressors could result from Hageman factor activation of the clotting system and the concurrent presence of chemotactic factors such as kallikrein (6, 7). These chemotactic factors could subsequently compete for, be bound by, or quantitatively exceed chemotactic inhibitors. Failure of a 22 C incubation to reduce the INFECT. IMMUN. suppression of BF-mediated chemotaxis may indicate that more than one system of chemotactic inhibition exists in these serum samples. This is further supported by the data in Table 2, indicating that the suppression of BF can occur in the absence of suppression of chemotaxis mediated by NS. Previous reports have described chemotactic inhibitors (chemotactic factor inactivator, or CFI) in similar types of patients (9), and therefore it is probable that both inhibitors existed in the serum samples tested here. The coexistence of CIA and CFI could explain the differences between inhibition of NS-mediated and BF-mediated chemotaxis. It is also possible that increased CIA after heat treatment might be due to generation of immunoglobulin aggregates similar to inhibition observed with immune complexes (11). Although such aggregates could account for increases in the CIA found in the void volume after Sephadex G-200 separation, they could not explain the increase in the 7S inhibitory activity when serum was separated after heat treatment. These results indicate that aggregated immunoglobulins alone could not be totally responsible for the increase in CIA after heat treatment, thus supporting the possible presence of a heat-labile suppressor of CIA. Such suppressors are not unprecedented; they have been referred to by Hughes et al. (Gastroenterology 66:846, A-233, 1974) and implied in a previous study (15), which demonstrated the neutralization of CIA by the addition of increasing concentrations of NS. If CIA suppression is due to heat-labile serum chemotactic factors, these data may indicate that the G-200 void volume and 7S inhibitors bind to or compete with chemotactic factors for receptor sites on the cell surface. A previous report by Kaplan et al. (7) has demonstrated PMN chemotactic activity in the leading and the trailing edge of the second peak after Sephadex G-200 gel filtration of NS and thus supports the presence of chemotactic factors in these regions. Studies reported here have dealt with an apparent increase in CIA after heat treatment. How heating affects CIA or to what extent each of these factors may be acting on soluble mediators of chemotaxis or cell surface receptors is not yet clear. Evidence at this time indicates that a cellular mode of action is at least partially responsible for chemotactic inhibition and is supported by the data in Tables 3 and 4, indicating that CIA does not irreversibly inactivate the chemotactic activity of casein and that preincubation of PMN in sera containing CIA results in suppressed chemotactic responsiveness. Furthermore, the failure of insolubilized casein-sepharose columns to remove CIA ne-

9 VOL. 13, 1976 gates a possible high-affinity binding of chemotactic inhibitors to casein. Although it is apparent from Table 4 that heat-stable cellular suppressors of chemotaxis are present in these sera, the greater suppressive effect observed when CIA was present throughout the assay again may support the presence of both celldirected and factor-directed inhibitors. Suppression at the cellular level could explain the observed universal suppression of chemotactic systems, whereas inhibition of BF-mediated chemotaxis may result from factor-directed inhibitors such as CFI (2, 9, 17). Thus, the suppression by heat-treated serum may be the result of increased cell-directed CIA, which could mask any loss of heat-labile CFI. Immune complexes or immunoglobulin aggregates may be prevalent in the sera of patients with severe systemic disease, and such aggregates may exert a suppressive effect on PMN as previously postulated by Mowat and Baum (11). The involvement of immunoglobulin in this suppression is supported by immunoadsorption and physical data previously presented (14-16) that indicate a close association between IgA and CIA in patients with liver disease. It is of interest that recent studies by Hauptman and Tomasi (5) have demonstrated the presence of IgA antibodies directed against serum albumin in patients with liver disease. These antibodies have been demonstrated in 40% of the patients studied, which compares with the data we have accumulated showing CIA in 50% of patients with liver disease (16). In our studies CIA could be removed from sera with insolubilized anti-iga immunoadsorbent columns (16). These data and recent studies by Lawrence et al. (8) indicating that PMN have IgA receptors separate from IgG receptors suggest that an IgA-PMN interaction is possible and thus is potentially a site of action for IgAassociated CIA. It is apparent from the data presented here that heat treatment at 56 C for 30 min increases serum CIA. These studies also demonstrate the presence of cell-directed inhibitors which differ from heat-labile chemotactic factor inactivators previously described by Till and Ward (13). Differences in suppression of BFmediated and NS-mediated chemotaxis indicate that many of the serum samples tested may contain both types of inhibitors. The presence of these inhibitory substances in sera from patients with severe systemic disease may represent a means of controlling the inflammatory response. SERUM CHEMOTACTIC INHIBITORY ACTIVITY 749 ACKNOWLEDGMENTS We wish to thank Linda Paxton for her excellent technical assistance. This investigation was supported by Public Health Service grant HL from the National Heart and Lung Institute and by the Arthritis Foundation. LITERATURE CITED 1. Andersen, B. R., and D. E. Van Epps Suppression of human neutrophil chemotactic activity by streptolysin 0. J. Infect. Dis. 125: Berenberg, J. L., and P. A. Ward Chemotactic factor inactivator in normal serum. J. Clin. Invest. 52: DeMeo, A., and B. R. Andersen Defective chemotaxis associated with a serum inhibitor in cirrhotic patients. N. Engl. J. Med. 286: Gewurz, H., A. R. Page, R. J. Pickering, and R. A. Good Complement activity and inflammatory neutrophil exudation in man. Int. Arch. Allergy Appl. Immunol. 32: Hauptman, S., and T. B. Tomasi, Jr Antibodies to human albumin in cirrhotic sera. J. Clin. Invest. 54: Kaplan, A. P., E. J. Goetzel, and K. F. Austen The fibrinolytic pathwlay of human plasma. V. The generation of chemotactic activity by activation of plasminogen proactivator. J. Clin. Invest. 52: Kaplan, A. P., A. B. Kay, and K. F. Austen A prealbumin activator of prekallikrein. III. Appearance of chemotactic activity for human neutrophils by the conversion of human prekallikrein to kallikrein. J. Exp. Med. 135: Lawrence, D. A., W. Weigle, and H. Spiegelberg Immunoglobulins cytophilic for human lymphocytes, monocytes and neutrophils. J. Clin. Invest. 55: Maderazo, E. G., P. A. Ward, and R. Quintiliani Defective regulation of chemotaxis in cirrhosis. J. Lab. Clin. Med. 85: Mannik, M., and D. E. Stage Antibody-agarose immunoadsorbents: complete removal of classes of immunoglobulins from serum. J. Immunol. 106: Mowat, A. G., and J. Baum Chemotaxis of polymorphonuclear leukocytes from patients with rheumatoid arthritis. J. Clin. Invest. 50: Smith, C. W., J. C. Holler, E. Dupree, A. A. Goldman, and R. A. Lord A serum inhibitor of leukotaxis in a child with recurrent infections. J. Lab. Clin. Med. 79: Till, G., and P. A. Ward Two distinct chemotactic factor inactivators in human serum. J. Immunol. 114: Van Epps, D. E., J. A. Frierson, and R. C. Williams, Jr Immunologic studies of anergic patients. Infect. Immun. 10: Van Epps, D. E., D. L. Palmer, and R. C. Williams, Jr Characterization of serum inhibitors of neutrophil chemotaxis associated with anergy. J. Immunol. 113: Van Epps, D. E., R. G. Strickland, R. C. Williams, Jr Inhibitors of leukocyte chemotaxis in alcoholic liver disease. Am. J. Med. 59: Ward, P. A., and J. L. Berenberg Defective regulation of inflammatory mediators in Hodgkin's disease. Supernormal levels of chemotactic-factor inactivator. N. Engl. J. Med. 290: Ward, P. A., and R. J. Schlegel Impaired leukotactic responsiveness in a child with recurrent infections. Lancet 2: