Pediat. Res. 11: 677-()1\0 (1977) Chemotaxis neutrophil complement newborn monocytes phagocytes
Cellular and Humoral Components of Monocyte and Neutrophil Chen1otaxis in Cord Blood
SAYITA G . PAHWA.""'' RAJENDRA PAHWA. ELENA GRII\IES. AND E LIZA13 ETII SI\IITII\VICK Departlllt'lll of Pediatrics aml/nmrwwlogy, Memorial Sloa/1-1\ellcrillg Ca11ca Centa, Nell' York, New York , USA
Summary experiment, blood from a healthy adult was tested simultane ously. 1\lonoqte and polymorphonuclear neutrophil (J>I\IN) chemo taxis was studied in cord blood from healthv term infants. 1\Jono ISOLATION OF CELLS c;yte chemotaxis was normal to increased ( 115-126%) whereas PI\IN chemotaxis was decreased (79%) in comparison with that Mononuclear leukocytes were isolated by density gradient of healthy adult l'ontrol subjects. Generation of chemotactic centrifugation on a sodium mctrizoatc-Ficoll solution (Lympho factors from cord sera was impaired, being 55% of that gener prep. Nyegard and Co., Oslo) (5). The cells were washed three ated by J)(Wied normal human serum (I'NIIS). Cord serum was times and resuspended in RPI\11 (Gibco) supplemented with less inhibitory than pooled adult human serum fur adult mono penicillin 50 units, streptomycin 50 Jlg . and glutamine 2 ml\1/ml. qtes when the cells were suspended in HI % serum and tested for As simultaneous analysis of monocytcs by myelopcroxidasc stain chemotaxis. No inhibition of chemotactic factors by either cord and Wright stain were in close agreement. the percentage of or adult sera was observed. The dissociation of chemotactic monocytcs was routinely determined by a myelopcroxidasc stain response of the two diiTerent phagocytic cells may represent a ( 13 ); Wright stain was done to exclude contamination by gra nu protecth·e mechanism whereby one cell can compensate for a locytes. The final cell concentration was adjusted to I .5 x I on defect in the res1lonse of the other. monocytcs/ml. To isolate polymorphonuclear ncutrophils, the granulocyte Speculation rich. red blood cell sediment from the Lymphoprep gradient was resuspended in RPM I I ()40 ; the red blood cells were then Simultaneous anal_ysis of monocyte and neutrophil chemotaxis scdimentcd with 6 % dextran in saline solution (Abbott). Rc LI may reveal a defect restricted to one cell type. blood cells remaining in the leukocyte-rich supernatant were lysed with hypotonic saline . The leukocytes were then washed three times and resuspended in minimum essential medium The human newborn infant is uniquely susceptible to infection (MEM , Gibco) at a concentration of 1.0 x 10'; PMN's/ ml. with a variety of infectious agents (bacterial, viral. fungal, and protozoal) which often results in disseminated disease. Many CII Et\tOTACTIC STI1\IULI studies of the defense mechanisms in newborn infants have been done, including studies of cell-mediated and humoral immunity Lymplzocytc-dcril'cd Clzmzolllctic Factor ( LDCF). This lym ( 12, II-i), neutrophil function (lJ, 20. 23). inflammatory response phokinc was prepared by concanavalin A (I 0 1-lg/ml) stimulation (R, 10), and the complement system (2, 17). Little information is of lymphocytes. Cultures containing 2 X I on lymphocytcs/ ml in available about the monocyte, a cell of great importance in both RPMI 1640 without serum we re incubated at 37o in 5 % the afferent and efferent limb of the immune response. Evidence SI5 % humiLiificd air for 48 hr. After centrifugation , the superna in mice (I), rats (3), anti rabbits (2-1) suggests that the neona te tant Ouid was stored in aliquots at -70° until usctl . LDCF was has a deficiency in macrophage function . In humans, e valuation used as the chemotactic stimulus fur monocytes in the following of the newborn infant's inflammatory response using the Rebuck proportion: 30% LDCF, 40% RPMI 1640, and 30% Hank's skin window showed a ddaycd and diminished accumulation of balanced salt solution with Ca ++ , Mg+ +,and I '.:o gelatin (!lOSS- monocytcs (8), suggesting a possible defect in monocyte migra G). tion. Two recent studies on monocyte chemotaxis give conflict Actimtcd serum (AS) . Pooled normal human serum (PNHS) ing results. One, done in 1- to 5-day-old infants, reports de was stored in aliquots at -700 and thawed just before usc . The creased chemotaxis ( 15) and the other, done in cord blood, serum was activated with endotoxin (Esclzcriclzia coli 026:136, reports normal chemotaxis (16). Difco), 300 1-lg/ml at 37° for 60 min, followed by inactivation at The investigations reported here were designed to cvalualc 56° for 30 min. The proportions used were serum 0 . 1 ml, the cellular and humoral components of monocyte chemotaxis in endotoxin 0.1 ml, HBSS-G 1.0 ml, and MEM 0.8 ml, which the newborn infant, using cord blooLI . Previously, Miller (21) gave a final serum concentration of 5 %. This stimulus was tested had repo rted an impairment of neutrophil chemotaxis and in against both monocytcs and PMN's. the generation uf serum chemotactic factors in 3- to 5-day-old Appropriate controls, without any chemotactic stimuli, were infants. The present study was expanded to include neutrophil tested simultaneously . These consisted of 40% medium (RPM I chemotaxis (I) to sec whether the defects described in newborn for monocytcs and MEM for PMN's) and 60% HBSS-G. infants were also present in corLI blood, anLI (2) to sec whether the responses of monocytcs and ncutrophils resembled each CH EMOTAXIS ASSAY other. The test procedure was a modification of the method of Snyderman c/ at. (26). All assays were done in triplicate. MATERIALS AND METHODS Modified Boyden chemotactic chambers (Blind-Well, Nc uro Heparinized (20 units/ml) cord blood, obtained from healthy probc) were used. The wells of the chambers were filled with 0.2 full term infants, was tested within 4 hr of delivery. In every ml of the chemotactic stimulus, delivered with a calibrated pi- 677 678 PAIIWA ET AL. pette (Finnpipette, Neuroprobe). Numbered filters (Nude eight showed a decreased response and three had a response pore) of pore size 5 J.Lm for monocytes and 3 J.LII1 for PMN's were equal to that of adult PMN's (Table 2). The mean chemotactic placed over the filled wells, taking care not to trap bubbles. The response of cord PMN's was 79% of that of adults (P < 0.05) filter retainers were then screwed in, and the upper compart I3ackground counts of cord and adult PMN's wen; similar. ments charged with 0.2 ml of the appropriate cell suspension. The chambers were incubated at 3 7o in humidified air for a GENERATION OF SERUM CHEMOTACTIC FACTOR period of 90 min for monocytes and 45 min for PMN's. After incubation, the upper compartments were aspirated and the Cord sera generated less chemotactic activity for PMN's than filters were removed and stained in Mayer's hematoxylin. Cells did PNHS. The chemotactic response of adult PMN's to acti migrating to the undersurface of the filter were counted, using vated cord sera was 55% of that seen with activated PNHS, P < an ocular grid (7 x 7 mm), with a magnification of 1 ,000 times 0.0005 (Table 3). for monocytes and 400 times for PMN's. Ten random grid fields were counted on every filter. The results were expressed as the SERUM INIIIBITORS OF CIIEI\IOTAXIS mean number of cells per ocular grid and as a percentage of the Adult mononuclear cells suspended in serum showed a lower normal. monocyte chemotactic response than that of cells suspended in medium alone. PNHS was more inhibitory than cord serum, STUDY OF SERUM FACTORS there being 51% inhibition of monocyte chemotaxis by PNIIS, Using the basic techniques described above, experiments were and 26% inhibition by cord serum (Table 4). Monocytes in cord done to evaluate the following. serum showed a consistently higher response to I 0% AS than Ability to Generate Chemotactic Factor. Cord sera and PNHS when suspended in PNIIS (P < 0.05). On the other hand, the were tested simultaneously for chemotactic activity after endo response of PMN's suspended in cord serum varied from a toxin activation, using adult PMN's as test cells. slightly decreased to a slightly increased response as compared Inhibitors of Cells. Normal monocytes and PMN's were sus with cells suspended in PNHS, but no significant differences pened in either 10% cord serum or 10% PNIIS; 10% activated were observed. Only minimal inhibition by either serum was normal serum was used as the chemotactic stimulus. seen. Cord cells similarly suspended in either PNHS or cord Jnhibitors/lnactivators of Chemotactic Stimuli. Cord serum or serum were not tested. normal serum, 5%, was added to 5% activated normal serum. This 10% serum mixture was used as the chemotactic stimulus SERUM INHIBITORS/INACTIVATORS OF CHEMOTACTIC FACTORS for normal PMN's. Similarly, cord or normal serum, 10%, was There was no evidence of any inhibitor or inactivator in added to the 30% LDCF and tested against normal monocytes. unfractionated cord or pooled normal serum to either of the chemotactic factors tested. Monocytes exhibited an increased RESULTS response to LDCF which had added cord or pooled normal human serum than when exposed to LDCF alone (Table 5). MONOCYTE CHEMOTAXIS Addition of either 5% cord or pooled adult serum to 5% The percentage of monocytes in cord mononuclear cells was consistently higher than that in adult cells: 25 ± 8% versus 16 ± Table 2. Response of cord and adult polymorphonuclear 5%, respectively. The final monocyte concentrations were ad neutroplzils ( PMN's) to endotoxin-activated normal serum (AS) justed as described above such that the chambers all received equal numbers of either cord or adult monocytes. LDCF was the Mean chemotactic response' chemotactic stimulus for 24 cord samples; AS was used as an Cord/ additional stimulus in 12 (Table I). Of those tested with LDCF, Stimu- Adult adult, 12 had an increased response, ranging from 122-158% of the Ius Cord PMN's Pt-. IN's % I' value adult values, whereas the other 12 had a normal response (85- AS 69.9 ± 22.4 (n = II) 88.2 ± 15.2 79 <0.05 115%). The mean response of the cord monocytes was signifi ControJ2 24.2 ± 6.0 2!.0 ± 8.5 114 N.S. cantly higher (126%) than the adult response (P < 0.0025). The mean response was also higher for AS ( 115% ), but not sig ' Mean cells ± I SO/grid at a magnification of 400 times. n indicates nificantly so. The mean background count (control chambers) number of cord samples studied. 2 for cord monocytes was also significantly higher (P < 0.005). No chemotactic stimulus. IIBSS-G (sec "Materials and Methods") 60%; minimum essential medium 40%. PMN CHEMOTAXIS Table 3. Serum clzemutactic activity for normal adult Eleven cord samples were tested for PMN response to chemo polymorphonuclear nel/lruphils (PM N's) tactic influence using activated serum as the stimulus. Of these, Mean chemotactic re Stimulus sponse,' Pt\IN's Table I. Response of cord and adult monocytes to chenwtactic stimuli 5% Cord serum (n = 7? Activated 41.5 ± 12.2 t\lcan chemotactic response' Not activated 18.3 :t 5.2 Cord/ 5% Pooled normal human serum Stimu- Adult mono- adult, Activated 75.1 ± 1!.3 Ius Cord monocytcs cytes % P value Not activated 17.8±5.6 LDCF' 28.9 ± 5.6 (n = 24) 22.9 ± 5.9 126 <0.0025 Control" 21 ± 5.2 AS' 33.1 ± 7.0 (n = 12) 28.7 ± 6.1 115 >0.5 Control' 6.0 ± 3.5 3.5 ± 2.2 171 <0.005 % Response to activated cord serum' 55 1 Mean cells± I SD/grid at a magnification of I ,000 times.n indicates 1 Mean cells ± I SD/grid at a magnification of 400 times. the number of cord samples tested. 2 n indicates number of cord samples tested. 2 Lymphocyte-derived chemotactic factor. 3 No chemotactic stimulus. IIBSS-G (sec "Materials and Methods") 3 Endotoxin-activated normal serum. 60%; minimum essential medium, 40%. ' No chemotactic stimulus. HBSS-G (sec "Materials and Methods") ' Mean response to activated cord serum/Mean response to activated 60%; RPM! 40%. pooled normal human serum x I 00. CHEMOTAXIS IN CORD 13LOOD 671)
Table 4. Effects of cord serum a/Ill normal serum on chemotactic Rebuck window and in vitro by the Boyden chamber have been response of normal cells 1 described (22). A similar variance seems to exist for newborn monocyte migration: namely, increased movement in vitro but Inhibition, decreased and delayeu monocyte accumulation in vivo. The Cell suspension Chemotactic response" 'lo increased mobility of cord monocytcs as compared to mono 1.1onocytes cytcs from healthy adults suggests that these cells arc activated in In 10% corJ serum 22.3:!:: 5.-lfi 2() their resting state in the neonatal period. Such a state could In 10% PNHS 1-l .fi:!:: 5.67 51 result from some as yet undefined maternal stimulus, or be In meJium alone 30.1 :!:: 3.79 scconuary to the stress of delivery. In this regard it might be of PMN's interest to compare spontaneous mobility of maternal and cord In 10% corJ serum 73.1 :!:: lOA 8 blood monocytcs. In to% PNHS 69 .8:!:: 13 .5 12 E. coli endotoxin activates the complement system mainly by In medium alone 79.6:!:: 15.6 the alternate pathway, and also by the classic pathway utilizing 1 These data rq,rescnt the mean of four experiments in which 17 cord antibody. It has been previously demonstrated that when the sera were tested. PNHS: poolcJ normal human serum. classic pathway is blocked, chemotactic factor gencrakd by 2 Mean cells per griJ at a magnification of 1,000 times for monocytcs endotoxin is kinetically slower, but reaches normal levels if anJ 400 times for polymorphonuclear neutrophils. Stimulus: I 0 % acti activation is carried out for 60 min (II). Low levels of natural vateJ PNHS. antibody to endotoxin in cord serum would prevent activation of the classic complement pathway, but it would still be possible to activate the alternate pathway and generate normal amounts of Table 5. Effect of cord and adult serum on chemotactic factors' chemotactic factor if the necessary complement components Chemotactic re were present. In our experiments, activation of serum with endotoxin was carried out for I hr, resulting in decreased chem Stimulus sponse" otactic activity in cord serum. Here, although one cannot rule Monocytes out a deficiency of endotoxin antibody, there seems little doubt 30% LDCF 22.4 :!:: 5.9 of associated complement deficiencies. Coru serum is known to 30% LDCF + 10% cord serum (11 = 5)" 2fi.l :!:: 2.5 be deficient in all components of the complement system (2). A 30% LDCF + 10% PNilS 2fi.6:!:: 3.6 similar deficiency of neonatal serum in the generation of chemo Control 3 .3 :!:: 1.8 tactic factor for PMN's has been previously described where PMN's bacteria and antigen-antibody complexes were used to generate 5 % AS 89.6 :!:: 11.3 chemotactic activity (21). 13csiucs hypocomple mcntcmia. an 5 % AS+ 5 % cord serum (11 = 9) 88 .1 :!:: 11.3 other mechanism for abnormal serum chemotactic activity is the 5 'lo AS + 5 'lo PNI-IS 90.1 :!:: 13.1 inhibition or inactivation of chemotactic factor by other serum Control-' 22.8 :!:: fi.2 factors . An inactivator of chemotactic factors for monocytes and 1 LDCF: lymphocyte-derived chemotactic factor; PNIIS: pooled nor PMN's is present in normal human serum (4), but this activity is mal human serum; AS: activated serum. not expressed in unfractionatcd serum. If there is an increased " Mean cells per grid at a magnification of I ,000 times for monocytes amount of a chemotactic factor inactivator, as happens in certain and 400 times for polymorphonuclear neutrophils. disease conditions such as Hodgkin's disease (31) and cirrhosis " 11 indicates number of cord sera tested. (19), it is active in unfractionatcd serum as well. Unfractionatcd ' No chemotactic stimulus. 1-113SS-G (sec "Materials and 1\lethods") cord serum did not inhibit either AS or LDCF, thus ruling out 60% ; RPM! or minimum essential medium, 40%. the presence of an excessive amount of chemotactic factor inacti vator or an inhibitor in the cord serum. Humoral inhibitors of chemotaxis, acting directly on cells, activated serum did not influence the PMN chemotactic re have been dcscribcu in some patients with incrcascu incidence of sponse. infections ( 19, 25, 28, 32). Most of the inhibitors described have been active for neutrophils. Recently, Snyderman (27) detected DISCUSSION an inhibitior to monocyte migration in patients who had the Buckley synuromc of increased immunoglobulin E, many aller A major aim of this study was to investigate one aspect of gies, and repeated infections (7). In the experiments dcscribcu monocyte function , chemotaxis, using cord blood cells to sec here, both cord serum and normal serum had a significant whether an inability to migrate contributed to the disseminated inhibitory effect on normal monocytes but not on ncutrophils. It infections and slower immune responses which are characteristic was interesting that the cord serum was less inhibitory to mono of the newborn infant. It has been suggested that the neonate has cytcs than was PNHS, i.e, normal monocytcs moved better when competent lymphoid cells but lacks functional macrophagcs (I, suspended in cord serum than when suspended in PNHS. This 6). could be explained if normal serum had had more spontaneous Contrary to our expectations, the cord monocyte chemotactic chemotactic activity than coru serum, whereby the former would response to two different stimuli, endotoxin-activated norm<)l "ucactivatc" the cells and thus inhibit their response in a chemo serum (AS) and lymphocyte-derived chemotactic factor tactic analysis (14, 30). However, both sera had equal sponta (LDCF), was normal to increased when compared with adult neous chemotactic activity, yet there was a difference in the monocytes. Simultaneous studies of neutrophil (PMN) chemo degree of inhibition, it being much greater for PNHS than for taxis revealed a decreased response, confirming an earlier report cord serum. This finding suggests that cord serum lacks some by Miller of a ucficicncy of the PMN chemotactic response of the thing that is present in PNHS which is inhibitory to monocytcs. human neonate (21 ). A dissociation of chemotactic function It is interesting to speculate what this substance might be: it between the two major phagocytic cells has not been reported could be a normal humoral regulator for monocyte chemotaxis, previously. This dichotomy in the response of monocytcs and which the newborn infant lacks, and which then would represent ncutrophils may represent a protective mechanism whereby an a process that is uclayed in uevelopmcnt. ,increased response of one cell type can compensate for the decreased response of the other. CONCLUSION Random mobility of cord monocytcs as assaycu by migration through a micro pore filter was also increased, whereas cord Simultaneous analysis of monocyte and neutrophil chemotaxis PMN's did not differ from normal adult PMN's. Discrepancies in cord blood revealed a normal to incrcascu monocyte and a :between neutrophil migration as measured in vivo by the decreased neutrophil chemotactic response. In addition, cord I 61{() PAIIWA Cf AL.
sera lacked an inhibitor of nwnocytc chemotaxis which was ami sites of fetal C,.,. C,. C,. and C, synthesis. J. Clin. Invest.. 52: 671 present in adult serum. No inhibitors of chemotactic factors were (JY73). IK. Leikin, S., and Oppenheim. J . J.: Differences in transformation of adult an!i). :?1 . Miller, f\1. E.: Chemotactic fum:titHl in the human neonate: llumoral anti 2. Ballow, M., Fang, F., Good, R. A .. ami Day, N. K.: Developmental aspects cellular factors. l'edial. Res .. 5: 4H7 ( J
Copyright <<) 1777 International Pediatric Research Foundation. Inc. l'rinft•d in U.S.A.