NATURAL ANTIBACTERIAL ACTIVITY AGAINST CORD BLOOD ...

0022-1767/89/1427-25l3S02.00/0 Voi. 142, 2513-2518. No. 7.Aprll 1. 1989 Prlnted in U.S.A.

THEJOURNAL OF IMMUNOLCGY Copyright 0 1989 by The American Associationof Immunologlsts

NATURAL ANTIBACTERIAL ACTIVITY AGAINST SaZrnoneZZa typhi INHUMAN CORD BLOOD MIRIAM ROMANO,* M. TERESA DE MAGISTRIS,' LUIGI VILLA,*SANDRA NUTI,' VINCENZO DE LEO' DIANA BORASCHI,'LUCIAN0 NENCIONI,' AND ALDO TAGLIABUE'*

Cord blood mononuclear cells from normal human also observed for antibody-dependent cellular cytotoxicnewborns possess natural antibacterial (NA) activ- ity against homologous targets (9)and, in a more severe ity against Salmonella typhi, as assessed by an in form, for NK activity (10). vitro test. NA activity was significantly higher than It has often been suggested that thehigher susceptibilthat observed in PBMC from normal adult donors. ity of neonates to infections might result from the above Using fractionation on nylon wool and Percoll gra- mentioned deficiencies in some immunereactivities. dient or C-dependent killing with mAb, it was found However, neonates do not seem to be completely unrethat cells of the monocyte-macrophage series and sponsive toward the different infective agents. It is likely CD4+ T lymphocytes were capable of exerting NA that the most primitive non-specific surveillance mechactivity in newborns, in contrast with results ob- anisms may normally function in neonates toprotect the tained in adults, where the effector cell wasCD4' a host during the maturation of more specific immune T lymphocyte. The capability of expressing NA acresponses. Indeed, cells of the monocyte-macrophage setivity by CD4+ T lymphocytes from cord blood was also confirmed byflowcytometry sorting. Pretreat- ries were observed to phagocytize bacteria normally at ment of cord blood mononuclear cells with F(ab'), birth (11) andto presentAg in a way comparable to those fragments against human IgG, but not against hu- in adults (12). whereas responses to chemotactic stimuli man I g A . abrogate the NA activity. Furthermore, were impaired (11). We have recently reported that lymphocyte-dependent human IgA anti-S. typhi cannot arm CD4' lymphoantibacterial mechanisms might be important in resistcytes in cord blood. Thus it can be suggested that in newborns, the immune system still being imma- ance to gram-negative infections in experimental and in (13-1 8). Theresults obtained have ture, NA activity might be the expression a of mech- humansystems anism of defence against infections, actingas anti- shown that NA2 activity can be expressed by T lymphobody-dependent cellular cytotoxicity expressedby cytes of the CD4 subset armed with IgA (16). Inasmuch monocytes andCD4' T lymphocytes armed with pre- as IgA are present only in extremely low proportions in existing maternal IgG antibodies. This differs from cord blood,it was thought that ananalysis of NA activity NA activity of adults which is only mediated by CD4+at thislevel might be of particular interest. T lymphocytes armed byIgA. Our study therefore was performed to better understand the ontogeny of NA activity and the in vivo relevance of this newly described antibacterial mechanism. It is usually suggested that the immune system of the Indeed, NA activity was strongly expressedat the cord neonate differs from that of the adult because of the blood leveleven though with a mechanism different from relative lack of prior antigenic stimulation, resulting in that in adults. Thus, the results obtained shed further reduced immune responses.However, not allthe immune light on this antibacterialmechanism. functions are impaired in neonates to the same extent. A s far as humoral responses areconcerned, neonates at MATERIALS AND METHODS birth show high serum levels of IgG of maternal origin, Isolation of CBMC and PBMC. Blood samples were collected from but a lack of endogenously produced IgA and very low the umbilical cord of infants at birth. We studied 28 infants born IgM (1-3).despite the presence of a numberof circulating after full-term gestation: 12 were delivered by cesarian section, and IgM- and IgA-bearing B lymphocytes comparable to that the remaining 16were delivered vaginally. CBMC were isolated from of normal adults (4). However, neonatal T cells are rela- heparinized whole blood. After dilution 1/6 with PBS, 30 ml of blood were placed on 15 ml Ficoll-Hypaque (Pharmacia Fine Chemicals. tively mature in their capacityto proliferate in response Uppsala, Sweden], and centrifuged at 400 x g for 20 min a t room to mitogens and alloantigens (5, 6),whereas their ability temperature. CBMC were collected at the interface, washed twice to expresscell-mediated cytotoxicity is generally impaired with PBS, and resuspended in RPMI 1640 (GIBCO,Grand Island, NY) (7, 8).A diminished cytotoxic potential in newborns was medium supplemented with 10%heat-inactivated fetal bovine serum 2 mM L-glutaReceived for publication November 2, 1987. Accepted for publication January 9. 1989. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' Address correspondence and reprint requests to Dr. Aldo Tagliabue, Laboratory of Immunopharmacology. SclavoResearch Center. Via Fioretina. 1. 53100 Siena. Italy.

(Sera-lab,Crawley Down, West Sussex, England] and mine (GIBCO)(hereafter referred to a s complete medium). Heparinized venous blood, obtained from healthy donorswho had never been affected by typhoid fever or salmonellosis, was diluted 1/3 with PBS. and PBMC were separated a s described above. Cell separation by physical techniques. A one-step discontin-

'Abbreviations used in the paper: NA, natural antibacterial: ADCC. antibody-dependentcellular cytotoxicity:CBMC, cord blood mononuclear cells: FACS, fluorescence-activated cell sorter: PTA.PBS containing 0.05%Tween and 0.02% sodium azide.

2513

2514

ANTIBACTERIALNATURAL

uous gradient of Percoll (Pharmacia) was used as described by Colotta et al. (19). to obtain cells enriched in monocytes. CBMC or PBMC were resuspended in 5 ml isoosmotic (285 mosmol/liter] complete medium and carefully layered on top of 5 ml of 46% (v/v) Percoll in iso-osmotic complete medium in 15-ml conicaltubes (Corning Glass, Corning, NY). After centrifugation a t 500 X g for 30 min,monocytes were collected at the interface, washedin PBS, and resuspended in complete medium. To obtain populations enriched in lymphocytes, CBMC or PBMC were passed on nylon wool columns (20) and then treated with carbonyl iron and magnet (21). This procedure resulted in a population of 99%lymphocytes. Depletion of cell subpopulation by mAb and C.To characterize the phenotype of the effector cells in the antibacterial assay, lo6 CBMC were incubated for 45 min a t 4°C in 0.5 ml of complete medium with 50 p1 of the following mAb of the OK series (Ortho Diagnostic System Inc.. Raritan, NY): T3, a mouse IgG2a against a human pan-TAg (CD3);T4, a mouse IgG2b against a human inducer/ helper T cell subset (CD4); T8. a mouse IgG2a against a human suppressor/cytotoxic T cell subset (CD8);M 1, a mouse IgG2a against human monocytes (CD11);and B7, a mouse IgG2b against humanB cells (CD21).Incubation with mAb 5.9 (Technogenetics, Turin, Italy), a mouse IgG3 against a subset of human Thcells (22). wasperformed a t room temperature. A total of 5 plof mAb AB8.28 (23). a mouse IgGl against human large granular lymphocytes (CDl6). kindly provided by Dr. F. Malavasi (University of Turin, Turin,Italy) were used. After washing. cells wereresuspended in 0.5 ml of rabbit C (lot 1214, Cedarlane, Hornby, Ontario, Canada) diluted 1/2 and incubatedfor 60 min at 37°C. Then, cells were washed twice and resuspended in complete medium with or without readjusting their concentration. Cellfractionation by FACS.Non-adherent PBMC were incubated at the concentrationof 4 X lo7cells in 1ml with a-Leu 3a (i.e.. CD4) mAb fluorescein conjugated (Becton Dickinson, Mountain View, CA) for 30 min on ice in the dark. Non-adherent CBMC were incubated with a-Leu 3a mAb phycoerythrin conjuat the same concentration gated together with a-Leu 4 (i.e., CD3) mAb fluorescein conjugated (Becton Dickinson) for 30 min on ice in the dark. Cells were then washed twice in PBS with 1% BSA, adjusted to the concentrationof 4 X 106/ml and sorted with FACStar in sterile conditions. For these experiments, argon laser output was 1000 mW at 488 nm. Sorting of cells was carried out according to standardFACStar specifications with a 2.5 drop deflection criterion using 70-nmnozzle. Sorted cells had less than 1% mortality and more than 95% positivity for the marker used. Antibacterial assay. The bacterial assay was performed as described in detail(13- 18)with the useof Salmonella typhistrain Ty2 as target. Briefly, bacteria were placed in 15-ml conical tubes together with either medium or appropriately diluted antibodies and 4°C. The cell suspensions were centrifuged at 1300x g for 10 min at were then added to bacteria a t different E/T ratios, and the tubes were again centrifuged a t 5 0 0 X g for 5 min a t 4°C. To maintain optimal proportions of the reactants, the final volume of the mixture was limited to 0.3 ml, consisting of 0.1 ml of bacterial suspension, 0.1 ml of medium, and 0.1 ml of effector cells. The experimental and control tubes, which contained bacteria and medium but not cells, were then incubated at 37°C for 2 h. At the end of the incubation period, the pellets wereresuspended, thevolume was brought to 1 ml, and diluted 1/20 in medium. Then, 0.1-ml aliquots were plated onPetri dishes containing agar tryptose. After overnight incubation, CFU were counted. Usually 100 to 200 CFU/dish were scored in controlgroups.Duplicate tubes were set up for each experimental group and two Petri dishes were prepared for each tube.The percentage of antibacterial activity was calculated a s follows:

ACTIVITY IN CORD BLOOD humidified chamber touniform the electrostatic activity of the solid phase. Wells of each plate were then coated with 100 pl of LPS-W S . typhi 0901 (Difco Laboratories. Detroit. MI), 50 &well in 0.05M carbonate buffer, pH 9.6. Plates were incubated for 30min a t 37'C and overnight a t room temperature. The coating buffer was aspirated and thewells washed with 150 pl of 0.15M PBS containing 0.05%Tween and 0.02% sodium azide (PTA).To minimize non-specific adsorption of serum protein to the plastic, the wells were coated with 150 pl of a blocking solution consisting of 2% BSA in PBS, and then incubated for 2 h at 37°C. The plates were then washed three times in PTA and 100-pl serum aliquots in 0.15 M PBS containing 0.1%BSA and 0.2%sodium azide (PBS-BSA)were added to the wells after 1/50 dilution for IgA and 1gM and 1/100 for 1gG. After incubation for 1.5 h at 37°C the plateswere washed three times with PTA and 100 pl of class-specificgoat anti-human Ig conjugated with alkaline phosphatase (anti-IgA lot 23069: anti-IgG lot 25808; anti-IgM lot 27150; Cappel) were added to the wells. The antibodies conjugated to enzyme used in the assay were previously titratedagainst chromatographically purified humanantibodies (Cappel) and were diluted 1/500 for IgA, 1/1500 for IgG, and 1/1000 for 1gM. After 1.5 h incubation a t 37°C. plates were washed with PTA and 100 plof p-nitrophenyl phosphate substrate (Sigma Chemical Co.. St. Louis, MO) 1 mg/ml in 1M diethanolamine. pH 9.8. containing 1 mM MgC12 was added to eachwell The enzyme-substrate reaction which developed a t room temperature wasstopped after 30 min and optical the density of the samples was measured at 405 nm againsta blank (substrate in diethanolamine pH 9.8) on a Titertek Multiskan (Flow Laboratories Inc., McLean, VA). Controls for eachplate included wells with serum samples but no Ag and affinity purified human IgA, igG, IgM used for the assay standardization. This techniqueallowed the detection of a s little a s 1.25 ngof specific antibody per well. Each serumsample was tested in duplicate and absorbancevalues were averaged. Statistical analysis. The results areexpressed as mean, and the SE is not reported because it was usually less than 10%. Antibacterial activity was statistically analyzed by parallel line assay (25) after logarithmic transformation of the variables. RESULTS

NA activity against S. typhi expressed by PBMC was previously observed in normal adults using high E/T ratios, from 200 to 50/1 [ 16-18), in the in vitro assay. The effector cell of NA activity among PBMC was identified a s a subset ofCD4' lymphocytes (16-18). Indeed, when purified populations ofCD4' lymphocytes were obtained by flowcytometry sorting it could befound that significant antibacterial activity can be expressed at effector target ratioslower than 1/1 (Table I). When the NA activity ofCBMC in 25 newborn infants (50%male donors) wascompared with that of normal adults (80%male donors], it can be shown that newborns possessa severalfold higher NA activity than adults [Fig. 1A]. Interestingly, NA activity of newborns was not different from that of the respective mothers (CBMC from 15 of 25 newborns were directly compared with their mother's PBMC) (Fig. 1B). Experiments were then performed in a n % antibacterial activity = 100-100 X (no. CFU of experimental attempt to identify the effector cells responsible for NA tubes]/(no. CFU of control tubes without cells) Antibodies used in antibacterial assay. For blocking studies, activity in newborns. For this purpose, cellular populations enriched in monocytes or in lymphocytes were inithe following antibodies were used: F(ab'), fragments of goat antihuman serum IgA (lot 231 19. N. L. Cappel Laboratories, Cochran- tially employed. A s shown in Table 11, a significant inville. PA] and goat anti-human IgG (lot 24063 Cappel) obtained from crease inNA activity was obtained when Percoll-purified affinity chromatography purified antibodies. Effector cells were premonocytes were used; however, some activity was still treated with F(ab'), fragments for 1 ha t 4°C and thenwashed twice in PBS. For arming studies, human 1gA from serum of volunteers detectablewith purified lymphocytes. Thus it can be orally vaccinated with S . typhi mutant strainTy2 l a (Neotyf vaccine. suggested that at the cord blood level NA activity is exSclavo, Siena. Italy) wereused. IgA antibodies were purified as pressed by both monocytes and lymphocytes. previously described in details (18). To further characterize the effector cells, CBMC were ELISA. Sera obtained from cord blood and from peripheral blood. respectively, fromnewborns and their mothers, were storeda t -80°C treated with cytotoxic mAb and C before the in vitro to be used in the ELISA test. The ELlSA method was a modified antibacterial assay. A s shown in Table111, a statistically version of that described by Engvall and Perlmann (24). significant reduction in NA activity was obtained after Flat-bottomed polystyrene microtest plates (Dynatech Laboratodepletion ofCD11'of CD3' cells. In a further set of ries lnc.. Alexandria, VA) were preincubated for 30 min at 37°C in a

2515

NATURAL ANTIBACTERIAL ACTIVITY IN CORD BLOOD TABLE I NA activity against S . t y p h i of PBMC s e p a r a t e d b y FACS according to the CD4 surface marker % Antlbacterlal Activity

Effector Cells 50

0.75 0.35'

Non-adherent PBMC 71 CD4' PBMC CD4-

PBMC

15 2 5

3.1 12.5

-7

67 -3 156

0

6.2

-6

-3 768 7 -1 -2

19 0

0 69 1

-5

10

100

200

23 79' 8b

27

E/T ratio. b p C 0.01 vs non-adherentPBMC.

60

A

B

40 -

/

Flgure 1 . Mean NA activity against S . typhi ofCBMC from 2 5 neonates (V)compared to PBMC from 25 normal adults (01[ A ) or from corresponding mothers (0)(B) (15 cases).

//io O '/

20: 0 -20

6

12

25

50

100

12

200 6

25

50

100

EFFECTOR : TARGET RATIO TABLE I1 Distribution of NA actiuity againstS . typhi among CBMC separated by physical methods % Antibacterial Activity

9% of Donor

Treatment CBMC of Macrophages

1

2

Polymorphonuclear cells

Other cells

i2O 50

25

50 84 -5

None Interface Percoll46% Nylon wool plus carbonyl iron and magnet None 64.1 lnterface Percoll46% 85.9 Pellet Percoll 46%

Lymphocytes

100

200

69 85 3

69 93b 11'

ND 32.8 96 12.6 93b 8.1 91.9

1

12 1 1 94 1

2.3 62 1

76 51 27b 18

873

24 ND

28 14

E/T ratio. b p C 0.05 vs corresponding input. ' p 5 0.01 vs corresponding input. TABLE 111 Characterization of p h e n o t y p e of CBMC exerting NA activity against S . t y p h i b y d e p l e t i o nof cell subpopulations % Antibacterial Activity Donor

Treatment

CD

50"

1

2

None 59 C OKT3 + C OKT4 + C OKMl + C

61 3 27 4gb4 18 11

25 27 12 15 -1

None 92 86 C 87 59 4 +C OKT4 97 8 49 OKT8 + C 11 OKMl + C OKB721 C 80 66 16 AB8.28 + C 5.9 + c cellsTh

47 48 2 43 -4 48 12 30

+

100

200

85 93 39b 3

25b

23

44'

38

43b

57 33

86 90

.=E/T

ratio. b p s 0.05 vsc.

experiments(Table IV) performed by lysing CD3' or CD11+cells and adjusting the numberof effectors to the original E/T ratios it was shown that both T cells and macrophages can clearly express NA activity becausethe enrichment of both subsets resulted in increased activity. Furthermore, thedouble treatment with OKT3 and OKMl plus C completely abrogated the NA activity (Table IV). Interestingly, it was observed that the T effector cell

belongs to a CD4'8- subset. However, treatment with 5.9 mAb. which recognizes the helper-inducer subset among CD4' cells (22) did not affect NA activity. Furthermore, the negative results obtained in these experiments with other mAb rule outa role for B cells or LGL in this system (Table 11). A final characterization of the lymphoid effector cell among CBMC was performed by flow cytometry sorting: results in Table V confirmed that these belong to a CD4' subset of T cells. Studies in murine systems( 1 4- 15) have demonstrated that IgG and IgA with antibacterial specificity can arm lymphocytes and direct them to express ADCC against bacteria. The NA activity in human adults seemed to be attributable to a spontaneous form of IgA-dependent ADCC (16). a s found in mice. Inasmuch as newborns have no detectable IgA at birth and IgG are of maternal origin (1-3). we then investigated whether theexpression of neonatal NA activity against S . typhi could be related to a n ADCC mechanism mediated by IgG. A first evidence that NA activity at thecord blood levelmight be mediated by cytophilic antibodies came from the decrease of the antibacterial activity obtained by incubating cells at 37°C in serum freemedium for 2 h (Table VI). Then monocyteenriched CBMC were depleted of CD4' lymphocytes, and 100 @/mlof the F(ab')2 fragment of goat antibodies against human IgA or IgG were added. A s shown in Table

2516

IN CORD BLOOD

NATURALANTIBACTERIALACTIVITY

TABLE IV NA activity againstS . typhi ofCBMC after enrichment ofcell subpopulations E/T Ratio Treatment Actual before

E/T Ratio

Treatment

50

25

None C OKT3 C -52 OKMl + C OKT3 -27 + OKMl -30 C

50

-4" 1 -30 -24 -20

+

+

13 10 17b -14

10025

14 12 -28 -18

200

100

200

39 28 7b -10b

59 -1 64 11

70 58 10

66

72 82

9b

Percent anti-bacterialactivity. b p S 0.01 vs C. Significant reduction. ' p < 0.01 vsC. Significant increase. Oi

TABLE V NA activity against S . t y p h i of CBMC s e p a r a t e d b y FACS according to CD3 and CD4 surface marker % Antlbacterlal Actlvtty Effector Cells

6 . T 50 25 12.5

CBMC -7 Non-adherentCBMC -12 CBMC CD4' CD3+ -7 CBMC -12 -17 CD3+ CD4a

6 -8 -8 -3

15 57 27 -10 -8 5 -4 -5

100

200

40 2 15 1

15 26b

E/T ratio.

' p < 0.05 vs non-adherent CBMC. TABLE VI Effect of the incubation at 37°C on NA activity againstS . t y p h i b y CBMC

seems to be attributable to a spontaneous form ofIgGdependent rather than IgA-dependent ADCC. However, the addition of purified IgA to CD3' CD4' CBML was not able to increase the NA activity a s observed with CD4' PBMC from adults (Table X). Finally, when humoral immunity againstS. typhi was investigated by employing a sensitive ELISA assay, we found that neonates and theirmothers possess the same level of IgG to S. typhi-LPS, whereas newborns do not have detectableamounts of IgA and IgM against S. typhiLPS (Fig. 2). DISCUSSION

We have previously reported that lymphocytes from different anatomical sitesin the mouse express NA activity against a variety of enteric pathogens [ 13) and anti4°CwithFBSb37°Cwith FBS 20 3 15 42 33 bacterial antibodies of the IgG and IgA classes were able 6 18 13 40 42 4°C with FBS 37°C without FBS -9 6 13 15b 37"Cwith FBS 37"Cwithout FBS 2 to arm lymphocytes and increase antibacterial activity 2 10 14b 37"CwithoutFBS 37"Cwtthout FBS -21 -55 (14. 15).More recently, we also observed that PBMC from E/T ratio. normal volunteers possess NA activity against S. typhi, Fetal bovine serum. and theexistence of this activity suggested a mechanism p < 0.05 vs control 1 . of defence against infections, mediated by human cells, TABLE VI1 and acting a s ADCC expressed by CD4' T lymphocytes Effect ofpretreatment with F(ab'j,fragments against human IgA and coated with pre-existing anti-Salmonella IgA antibodies IgC on cord blood monocytes exerting NA activity againstS . typhi (16). % Antlbaclerlal Activity In a n attempt to better understand the origin of NA Experimental Group 12'50 25 100 activity, experiments were performed employing CBMC CBMC 6 13 31 31 in the invitro assay againstS . typhi. The resultsreported 49 34 80 49 Monocytes* here clearly demonstrate that CBMC from newborn in17 56 66 86 Monocytes plusF(ab'), fants possess NA activity that is higher than thatfound against IgA -12 3 -17' 18 Monocytes plus F(ab'), in normal adults, although it does not differ from that of against IgG the corresponding mothers. This indicates that despite E/T ratio. the immunedeficits at birth, some antibacterialreMononuclear cells after Percoll46%and treatment withOKT4 + C. sponses are functional and even more efficient than in p s 0.05 vsmonocytes. adults. This is in sharp contrast with other forms of VII, a significant inhibition of NA activity was observed ADCC which are impaired in neonates (9). Interestingly, it was foundthat NA activity in newborns only after treatment with F(ab'), fragment against IgG, but not against IgA. Table VI11 shows that nylon wool- is also qualitatively different from that of adults. In fact, enriched T cells which were further depleted of mono- besidesT lymphocytes, cells of the monocyte-macrocytes by treatment with OKMl + C were also inhibited in phage series can also act a s NA effectors. Furthermore, CDl 1 cells in cord blood are naturally armedby antibactheir NA activity only by exposure to F(ab'), fragments against human IgG. Table VI11 confirms that theactivity terial antibodies of the IgG class, as demonstrated by is due to CD4' cells. These results strongly suggest that inhibition experiments with F(ab'), antiIgG, and do not function by means of phagocytic mechanisms. This is IgG. but not IgA, are involved in NA activity at birth. Inasmuch as a n increase in NA activity was also ob- quite differentfrom our observationsin previous studies: that in mice, macrophages were never active served with maternal blood when compared to normal it was found adults, we examined whether the increase was dueto a by themselves, even though the addition of antibodies different class of antibody. A s shown in Table IX, pre- specific for bacteria rendered themcytotoxic [ 15). It may treatments with F(ab'), fragments against humanIg dif- be that neonatal monocytes have a higher affinity Fcy monocytes. ferently affected NA activity in adults and newborns. receptors than maternal or other adult A second point to be stressed is that also at the cord These results confirm that in adults the NA activity is NA effectors IgA-mediated; whereas in newborn infants this activity blood level, CD4' lymphocytes arethe W Antibacterial Activity

Pre-lncubatlon 4 h

Assay 2 h

1Zn

25 50 100 200

2517

NATURAL ANTIBACTERIAL ACTIVITY IN CORD BLOOD TABLE VI11 Effect ofpretreatment with F(ab'Jzfragment against human IgC a n d h u m a n IgA on cord blood lymphocytes exerting NA activity against S . t y p h i 9% of

9% Antibacterial Activity

Experlmental Group

LY

CBMC 15.1 Non-adherent cellsb plus OKM 1 C plus OKMl + C + F(ab'),to 1gA ND plus OKMl + C + F(ab'),to 1gC ND ulus OKMl + C OKT4 C

+

+

+

82.5

1.4

89.2

8.3 ND ND ND 18.0

75.5

100

Macrophages Other

LGL

cells

92

25a

1 .o5 8 2.0

0.5 ND ND 0

50

ND 6.5

46 12 313 6 -10 -10

19 20 -9 -5

37' -11'

-3'

E/T ratio. Mononuclear cells after nylon wool and carbonyl iron plus magnet. ' p s 0 . 0 5 vs corresponding control. a

TABLE IX may be induced by subclinical infections or by crossE f f e c t o f p r e t r e a t m e n t w f t h F ( a b ' J 2 f r a g m e n t s a g a i n s t h u m a n I g A areactive nd bacteria of the endogenous flora. Later, bacterial IgC O R p e r f p h e r a l blood and cord blood cells exerting NA activity infections begin to interact with thehost mucosa, where against S . typhi

IgA is the most represented class of antibody. This may cause the switch to IgA-ADCC in adults and the major 100 50 25" expression of this defence mechanism will take place at 30 59 85 PBMCb the intestinal level rather than at theperipheral one, a s 3 2 4' plus F(ab'), againstIgA observed in mice (1 4, 15). 76 41 90 plus F(ab'), against1gG To explain the loss of monocyte-dependent ADCC, it 74 54 58 CBMCd may be suggested that monocytes, although highly effi9 50 plus F(ab'), 7against IgA3 7 -20 15 14' plus F(ab'), againstIgG cient in bacterial elimination, are more likely to cause inflammatory reactions harmful for the host. In fact, a E/T ratio. PBMC from mother. variety of inflammatory soluble factors can be released ' p s 0 . 0 5 vs corresponding untreated cells. by activated macrophages in response to different stimuli Cord blood mononuclear cells from newborn. including bacteriallipopolysaccharide (26). Inasmuchas, among T cells as in adult peripheral blood. A s for mono- on the contrary, T cells do not releaseinflammatory factors, the expression of their antibacterial activity is cytes, IgG antibodies act as humoral arm also in the antibacterial ADCC mediated by T lymphocytes in new- expected to be potentially less harmfulfor the host. A s yet, no information is available on T cells with F c ~ borns. This is in contrastwith previous studies in adults (15-18) and here with PBMC from mothers, where only receptor in mothers during pregnancy and in neonates, IgA-driven ADCC could be observed against S . typhi. whereas it is known that in adult blood the majority of Inasmuch a s IgA and IgM are present at the cord blood Ta cells are of the T4 subset (27). Inasmucha s binding level in extremely low amounts, the observation that of IgA to FcR has been shown in experimental studies antibacterial ADCC is IgG-driven is not completely un- (28.29)to promote IgA switching throughsoluble factors. expected. The analysis of the levels of antibodies to S . it is conceivable that IgA taken in by the newborn through typhl LPS in cord bloodfurther confirmed this point, also the mother's milk may play a role in promoting the switch indicating the existence of a n equilibrium between ma- from IgG- to IgA-driven ADCC. Finally, it is of interest that IgA-driven ADCC is internal and neonatalblood levels of IgG anti-LPS. What remains a matter of speculation is why and when creased in mothers'blood when compared to apopulation a s total IgA are not increased during development antibacterial IgG-ADCCby mono- of normal donors. Inasmuch cytes ceases to functionand CD4' lymphocyte-mediated at the end of pregnancy (30).whereas IgA anti LPS are ADCC switches from IgG to IgA. It can be hypothesized decreased at this time (M. Romano, unpublished obserthat IgG-mediated NA activity of both monocytes and vations), it may be hypothesized that it is the cellular arm lymphocytes may serve at birth a s a temporary surveil- of ADCC to be positively affected by pregnancy. In conclusion, these results demonstrate that natural lance mechanism which uses lytic the cellular machinery of neonates to whom antibacterial specificity is conferred antibacterial mechanisms are present and efficient at by maternal antibodies. And these anti-LPS maternalIgG birth, and these may be important in protecting neonates 7% Antibacterial Activlty

Experimental Group

TABLE X Antibacterial activity ofFACS separatedT lymphocytes before and after treatment with purijied human IgA against S . typhi Purified I g A 0.6 pglrnl

Effector Cells

PBMCb

CD4' 49 CBMCb CD4' CD3' a

-

55

+3 7 -

+

7% Antibacterial Activity 0.35" 1.5

-86 37

0.75

50

25

8

21

-6 1 -27

-35 -25

3.112.5

18

-64 0

E/T ratio. Non-adherent PBMC from a normal adult separated atFACS according to theCD4 marker Non-adherent CBMC separated atFACS according to CD3 and CD4 markers. p 5 0 . 0 5 vs CD4+ PBMC without purified 1gA.

6.2

100

24 51

32 62

32 57

39 6gd

-44 6

5

26 35

32 37

9

41 51

2518

: v i

NATURALANTIBACTERIAL

ACTIVITY IN CORD BLOOD

1974. Lymphocyte subpopulations in the blood of newborn infants. Clin. Exp. Immunol. 18:469. 10. Baley, J. E.. a n d B. 2.Schacter. 1985. Mechanisms of diminished natural killer cell activityinpregnant women andneonates. J. Immunol. 134:3042. 11. Wetson, W. L.. B. S . Carson, R. M. Barkin, G . E. Slater, R. D. Dustin, and S . K. Hecht. 1977. Monocyte-macrophage function in the newborn. A m . J. Dis. Child. 131:1241. 0 4 8 12 16 20 12. Hoffman, A.A.,A.R. Hayward, J. T. Kumick, E, C. Defreitas, J. McGregor, a n d R. J. Harbeck. 1981. Presentation of antigen by 100 IgG human newborn monocytes to maternal tetanus toxoid-specific Tcell blasts. J. Clin. Immunol. 1:217. 13. Nencioni, L.,L.Villa. D. Boraschi. B. Berti, andA. Tagliabue. 1983. Natural and antibody-dependent cell-mediated activity against Salmonella typhimuriumby peripheral and intestinallymphoid cells in mice. J. Immunol. 130:903. 14. Tagliabue, A.. L. Nencioni, L. Villa, D. F. Keren, G . H. Lowell, and D. Boraschi. 1983. Antibody-dependent cell-mediated antibacterial IgA. Nature activity of intestinal lymphocytes with secretory 306: 184. 15. Tagliabue, A..D. Boraschi, L. Villa, D. F. Keren, G . H. Lowell, R Rappuoli, and L. Nencioni L. 1984. IgA-dependent cell-mediated activity against enteropathogenic bacteria: distribution, specificity 100 and characterization of the effectorcells. J. Immunol. 133:988. 0 20 40 60 80 16. Tagliabue. A., L. Villa, D. Boraschi, G . Peri, V. De Gori, and L. Salmonella Nencioni. 1985. Natural antibacterial activity against 8 IgM typhi by human T4' lymphocytes armed with IgA antibodies. J. lmmunol. 135:4178. 17. Tagliabue. A.. L. Nencioni. A. Caffarena. L. Villa, D. Boraschi. G . Cazzola. a n d S. Cavalieri. 1985.Cellular immunity against Salmonella typhi after live oral vaccine. Clin. Exp. Immunol. 62:242. 18. Tagliabue, A., L. Villa, M. T. De Magistris, M. Romano, S. Silvestri. D. Boraschi, a n d L. Nencioni.1986. IgA-driven T-cell mediated antibacterial immunity in man after live oral Ty 21a vaccine. J. MOTHER (pg/mll Immunol. 137: 1504. 19. Colotta, F., G. Peri, A. Villa, and A. Mantovani. 1984. Rapid killing Figure 2. Specific antibodies anti-S. typhi LPS in newborn cord blood of actinomycin D-treated tumor cells by human mononuclear cells. and in peripheralblood from corresponding mothers. Correlation coeffi1. Effectors belong to the monocyte-macrophagelineage. J. Immunol. d e n t s were R = 0.656 forIgA, R = 0.979 forIgG, and R= -0.342 for IgM. 132:936. 20. Julius, M. H.. F. Simpson, and L. A. Herzenberg. 1973. A rapid method for the isolation of functional thymus-derived murine lymfrom infections. In previous in vitro studies, NA activity phocytes. Eur. J. Immunol. 3:645. against S . typhi was increased after an oral vaccination 21. Lundgren, G.,C.F. Zukoski. and G . Moller G . 1968. Differential effects of human granulocytes and lymphocytes on human fibrocapable of inducing protection against typhoid fever ( 17, blasts f n vitro. Clin.Exp. Immunol. 3r817. 18). it has been thought for a long time to passively 22. Corte. G.. M. C. Mingari, A. Moretta. G . Damiani. L. Moretta, and A. Bargellesi.1982.HumanT cell subpopulations defined by a immunize the fetus against infection by actively immumonoclonal antibody. J. Immunol. 128:16. nizing the mother through vaccination (31).Inasmuch as 23.Malavasi, F., G . Bellone, L. Matera, C. Milanese, E. Ferrero. A. we have recently observed that an oral live vaccine highly Funero. S . De Maria, F. Caligaris-Cappio, a n d G . Camussi. 1985. Murine monoclonal antibody as probe for the phenotype, functional effective against typhoid (32) was able to augment NA and molecular analysis of discrete peripheralblood lymphocyte popactivity of adult volunteers (18),it is likely that measureulations exerting natural killer activity in vitro. Hum. Immunol. 14:87. ment of NA activity at the cord blood level after vaccinaE.. a n d P. Perlmann. 1972. Enzyme-linked immunosorbent tion of the mothers may become a useful tool to verify 24. Engvall, assay. 111. Quantitation of specific antibodiesby enzyme labelled antitransfer of the immunity to neonates. immunoglobulin in antigen coated tubes.J. Immunol. 103: 129. 25. Finney, D. J. 1971. Parallel line assays. In Stattstical Methods in Biological A s s a y , 2nd e d . Griffin, London, p. 99. REFERENCES 26. Cohen. M. C.. J. K. Gutonski, and S. Cohen. 1986. Inflammatory lymphokines in hypersensitivity reactions.In The Reticuloendothe1. Van Furth. R.. H. R. E. Schuit, and W. Humans. 1965. The immulial S y s t e m : A Comprehensive Treatise. Vol. 9.S . M. Phillips and nological development of the human fetus.J. Exp. Med. 122:1173. M. R. Escobar, eds. Plenum Press.New York, p. 209-232. 2. Miirtensson. L., a n d H. H. Fudenberg.1965. Gm genesand G27, Suga, T., M. Endoh. H. Sakai. M. Miura. Y.Tomino. and Y.Nomoto. globulin synthesis in the human fetus. J. Immunol. 94:514. 1985. T cell subsetsinhumanperipheral blood. J. Immunol. 3. Gitlin. L. D., a n d A. Biasucci. 1969. Development of g, A, @'lc/lA, 134: 1327. C' 1 esterase inhibitor, ceruloplasmin, transferrin, hemopexin. hap- 28. Kawanishi. H.. E. Saltzman, and W. Strober. 1983. Mechanisms toglobin. fibrinogen, plasminogen, a,-antitrypsin, orosomucoid, @regulating IgA-class specific immunoglobulin production in murine lipoprotein. @',-macroglobulinand prealbumin in the human concepgut-associated lymphoid tissues. I. T cells derived from Peyer's tus. J. Clin. Invest.48: 1433. patches which switch sIgM B cells to slgA B cells in vitro. J. Exp. 4. Andersson, U., A. G. Bird, S. Britton, and R. Palacios. 1981. HuMed. 157:433. moral and cellular immunity in humans studied at thecell level from 29. Kiyono. H., M. D. Cooper, J. F. Kearney, L. M. Mosteller. S. M. birth to two years of age. Immunol. Rev. 57:5. Michalek. W. J. Kwpman, and J. R. McGhee. 1984. lsotype speci5. Alford, R. H., B. B. Cartwright, andS . H. V. Sell. 1976. Ontogeny of ficity of helper T cell clones. Peyer's patch Th cells preferentially invftro human cell-mediatedimmunity:age-relatedvariationin collaborate with mature IgA B cells for IgA responses. J. Exp. Med. infantile lymphocyte transformation. Infect. Immun. 13:1170. 159:718. 30. Amino, N., 0.Tanizawa, K. Miyai, F. Tanaka, C. Hayashi, M. 6. Stites, D. P.. M. C. Carr, and H.H. Fudenberg. 1974. Ontogeny of of serum immunoglobKawashima. a n d K.Ichihara. 1978. Changes cellular immunity inthe human fetus.Development of responses to ulins lgG, IgA. IgM, and 1gE duringpregnancy. Obstet. Gynecol. phytohemagglutinin and to allogenic cells. Cell. Immunol. 11:257. 52:415. 7. Rayfield. L. S., L. Brent, and C. H. Rodeck. 1980. Development of cell-mediated lympholysis in human fetal blood lymphocytes. C l f n . 31. Amstey, M. S., R.A. Insel. Lnd M. E. Pichichero. 1984. Neonatal passive immunization by maternalvaccination. Obstet. Gynecol. Exp. Imrnunol. 42:561. 63: 105. 8. Granberg. C.. T. Hirvonen, a n d P. Toivanen. 1979. Cell-mediated 32. Wahden. M. H.. C. Sene, Y.Cerisier, S. Sallam, andR. Germanier. lympholysis by human maternal and neonatal lymphocytes: moth1982. A controlled field trial of live Salmonella typhi strain Ty2la er's reactivity against neonatal cells and vice versa. J. Immunol. oralvaccineagainst typhoid: three year results. J. Infect.Dis. 123:2563. 145:292. 9. Campbell, A. C., C. Waller, J. Wood, A. Aynsley-Green. a n d V. Yu.

-1 /I

1