Phenotypic characterization and functional activity of human milk macrophages

Immunoh~gr Letu'r.~. 8 (1984) 249 256.

Elsexicr lmlct 512

PHENOTYPIC

CHARACTERIZATION HUMAN

MILK

AND

FUNCTIONAL

ACTIVITY

OF

MACROPHAGES

F. L E Y V A - C O B I A N and J. C L E M E N T E IhT~artnu'Ht ~!/ Immun~hd,,.l'. ('entr~ "Ram~Di ~ Cafizr'. ('arrctera de C~dtnetlar h'm. 9. I00. 28034 .Aladrid. Spaitl

(Received 14 Ma~ 1984) (Modified ~ersion receked 27 Jul', 1984) (Accepted 31)July 1984)

I. Summary A large population (about 80%) of the cells obtained from colostrum and early human milk were considered to be macrophages by the following criteria: nonspecific esterase stain, adherence, phagocytosis and IgG-Fc receptor expression. The majority of freshly isolated human milk macrophages (H M Md,) stain for the monocyte antigen O K M 1. Another monocyte antigen, 61D3, was expressed only by 30% of HMM~b. Class !1 antigens were expressed by H M M~b. About 85% of the cells were 1)R-positive whereas 50% were DS-positive as assessed with a panel of monoclonal antibodies directed against class I1 antigens. Monocyte and class II antigens were gradually lost during in vitro culture. H M Md~ can support proliferative response to antigens and mitogens when cocultured with autologous peripheral 1 cells. The proliferative response was significantly reduced when monoclonal antibodies to DR or DS were added to the assay. These results indicate that H M M~b have the phenotype and functional characteristics of antigen presenting cells.

presentation to T lymphocytes. Monocytes, culture derived macrophages (M~b), alveolar Md~, Kiipffer cells, Langerhans cells, peritoneal Md,, veiled and dendritic cells from lymphoid organs and microglial cells have been described both as bearing la-like antigens and as being able to act as antigen-presenting cells (APC) [1,2]. About 80% of the cells obtained from colostrum and early human milk are cells of the M PS [3]. Although this percentage represents a large Md, population, there is very little information about the characteristics of these cells and their functions are only partly understood. Surprisingly, only one report has been published recently dealing with H MM~, as A P C [4]. We have previously communicated the existence of a large la-like positive subpopulation in M~b from human lung, breast milk and the peritoneal cavity [5,6]. In the present paper we have characterized both function and phenotype of HMM~b.

3. Materials and Methods 3. I. ( ' o l l e c t i o n ~ f m i l k a n d b l o o d s a m p l e s

2. Introduction Cells from the mononuclear phagocyte system (MPS) are the principal antigen-presenting cells although other related cells are also capable of antigen Key u'ord~: antigen presenting cells I)R antigens DS antigens

macrophages milk cells 0165 2478 84. $3.00 © ElsevierScience Publishers B.V.

Colostrum and milk samples were obtained with a sterile breast pump after cleaning the nipple with an alcohol swab from a group of women which included primigravidae and multigravidae from the first to fifth day postpartum. Samples were collected in sterile conical tubes containing antibiotics and processed within I h. Women with clinical evidence of mastitis and those who had received estrogens to suppress lactation were excluded. 249

Peripheral blood was drawn, in sterile Vacutaincr tubes containing heparin, from women who showed positive candidin skin tests.

3.2. Preparation o.f cell su.wenshms The volume of the colostrum or milk samples was measured and centrifuged at 357 × g for 20 rain at 4 °C. The middle and fat layers were then removed and the cell pellet washed once in Hanks" Balanced Salt Solution (HBSS) without Ca 2" and Mg 2" before resuspending in RPMI-1640. The H M M O suspension (16 ml at a time) was gently layered onto 9 ml volumes of Ficoll- Hypaque (FH) (density 1.007) before centrifugation at 562 × g for 35 min at 20 °C. l h e interface was then removed and washed twice with HBSS. Finally, the cells were counted and resuspended in RPMI-1640 supplemented with 10% fetal calf serum, penicillin (100 U .' ml), streptomycin (100~g ml),nystatin(100 U ml) a n d 2 m M g l u t a m inc. M ononuclear peripheral cells (M PC) were separated by Ficoll- Hypaque gradient centrifugation. T cell-enriched populations were obtained using nylon wool columns. Brielly, the M P C were incubated on the nylon-wool columns (Fenwal Laboratories, Deerfield, 11,, U.S.A.) for45 min at 37 °C and the resulting adherent cell-depleted T cells (T) were collected by dropwise elution. In other experiments, M PC were cultured in Petri dishes for 2 h at 37 °C in an atmosphere containing 5% CO 2, and the non-

adherent M P C collected for further use. Identification of H M M~b and peripheral blood monocytes (PBMo) was accomplished using a nonspecific esterase stain [7]. The recovered cells were also assayed for viability by trypan blue exclusion. Wright's stain and supravital staining with acridine orange were used for differential counting. ()nix' samples containing 8()c/,~ of HMM~h were used in our assays. 3.3. Monoclonal antibodies for cell surlace pheno-

t) pe The monoclonal antibodies (MoAb) used in our assays are listed in Table I. For control purposes, the following M o A b of irrelevant specificity ~.ere used: MOPC21, MOPCI95, M O P I 0 4 E and U P C I 0 mouse myeloma proteins obtained from Litton Bionetics Inc. (Kensington, MI), U.S.A.). They were of the same class and subclass and were in the same form as the M o A b being tested. Phenotypic analysis of H M MO using these M o A b was performed by indirect immunofluorescence as was described previously [6]. Briefly, I × 106 H M Md, in H BSS containing 0.005% sodium azide were incubated for 30 rain at 4 °C with the appropriate MoAb. After washing twice, the cells were incubated for 45 min with lluorescein-conjugated F(ab')2 fragment of goat antiserum to mouse igG (Cappel Lab., Cochranville, PA, U.S.A.). Cells were washed three times, mounted and examined.

lable I Monoclonal antibodies assayed M onoclonal antibod}

Specificity

('lass

('RI

III.A-A,B,C (broad polymorphic) DR (monomorphic) DR (monomorphic) DR tmonomorphic) I)R (monomorphic) DR (monomorphic) I)R (nearly monomorphic) DR and SB DS (D(') antigen on most monocytes monocytes peripheral l lymphc~ytcs (l.eu-7) NK K cells

IgGl IgG~ lg(il IgG?t, lgG?. IgG2h IgG:,, IgG! IgG lg(i2~, IgGt IgCi2,, IgM

59.5 61.4 7.2 1.243 El)U-1 Q5.13 DA6.231 r[.122 OKMI 61D3 OKT3 IINK-I

250

Reference

8

9 10 II 12 13 14 15 16 17 18

3.4. Other /unctional assay.~ H M M4) adherence was evaluated as described [6,19]. Briefly, H M M ¢ were allowed to adhere to polystyrene I'etri dishes and incubated for 2 h at 37 °C in an atmosphere containing 59,: CO,. The nonadherent H M M ¢ were rinsed offand both total and differential counts performed. The Fc-lgG (EA) receptors were detected by standard rosetting techniques using rabbit lgG-coated sheep erythrocytes [6,19]. Candida alhicans phagocytosis was assayed by a standard method reported previously [6.19]. Briefly, preopsonized yeasts were mixed with I × 105 HMMcb in HBSS and incubated in Lab--Tek tissue culture chambers for 30 rain at 37 °C in humid atmosphere with 5% CO,. The percentage of phagocytic H M Mob was calculated by optic microscopy. 3.5. A ssa.v ~?/ lymphocyte reactivit i" Assays to determine the proliferative reactivity of cells to mitogens (concanavalin A. Pharmacia Fine Chemicals, Uppsala, Sweden) (Con A) and antigens (('am~ida alhicans antigen, Hollister Slier, Spokane, WA, U.S.A.) (('and) were performed by standard techniques. ! n each case, two concentrations of Con A or Cand were used. Lymphocyte reactivity was measured by tritiated thymidine ([~H]TdR) (Amersham International, Amersham, U.K.)incorporation into I)NA during the final 18 h of a 3-5 day culture with Con A or Cand, respectively. Each experiment was performed in triplicate and the results expressed as counts per minute (cpm). 3.6. Statistics Standard error (SE) was used as an estimate of variance and one-tailed Student's t-test for paired samples was used to compare means.

of undistinguishable cells. More than 80ci (88.1 = 7.5) of cells were M~ as judged by a combination of nonspecific esterase staining, adherence ability, yeast phagocytosis and l-c-lgG receptor procedures. Viability was above 909i. The H M Mob isolated from individual donors were cultured for up to 30 days. Reactivity with O K M I and 61 I)3 was studied with freshly isolated H M M ¢ after different culture periods. Figure l represents the percentages of H M M¢b expressing monocvte antigens. The O K M I antigen was expressed by ahnost all the HMMcf. However, 61 I)3 was expressed on a smaller proportion. After 24 days in culture. 61 I)3 antigens were almost lost. whereas about 30el H M Mg, remained OK M l positive. Figure 2 shows the expression of class II antigens on HMM4,. The H M M ¢ routinely showed variation in la-like antigen expression ranging from 70 to 889/- positive cells by indirect immunotluorescence. The lack of staining of a minor H M M4~ subset occurred with all M o A b directed to the I)R locus. With a M o A b to I)S locus (IU22), the negative subset was considerably higher. M o A b used as negalive controls were tested in parallel through all the experiments and did not show much positive staining (less than I~). The staining was also judged specific since it was not detected when the cell suspensions were reacted with FITC conjugated F(ab')~ goat anti-mouse IgG or IgM bv direct immunofluo-

o,.,

+

÷

'-I-' 4. Results

Over 100 colostral or milk samples from 86 women have been processed in our laboratory. M4~ percentages before FH separation in 12 selected samples used in the present study ranged from 51 to 74~. After FH separation, these samples contained less than 10~ o f T lymphocytes (as detected using OKT3), less than 5% of polymorphonuclear leukocytes (as judged with Wright's stain) and around 2c¥

clone number

days

in

Culture

0

I

!

e

20

40

60

Positive

stained

i

80 cells,

100

:':

Fig. I. Expression of OK M I and 611)3 monocytc antigens on human milk macrophagcs. Percentage t~l positi'~e stained cells was determined in t'reshl) isolated macmphages and alter the 24th da~. of in '.itro cuhurc. Results are expressed as the mean • gl:. l h e number of experiments is indicated in the bars.

251

sg.s

Is

61.4

la

7.2

3

L243

:

15

05113

19

oA6.2at

I"

.

,

•

• number

i

Is

ou-t

clone

gcn was gradually lost and bv the 30th day only a few cells retained the ability to express the antigen. When I)R antigens wcrc detected in old cultures they could be observed on all cell types, including the multinuclcated giant cells (Fig. 3C, I)). In other experiments, correlation between I)R antigen expression and phagocytosis by H M Md, was carried out. The DR-positive H M MO consisted of a large subset of phagocytic cells and another nonphagocytic subpopulation. These subsets were also present in the I)R negative HMM~b subpopulation (Fig. 5). l-inally. HMM4b did not stain for l,eu-7, a marker of NK cells, and virtually all HMM4b expressed class 1 antigens as detected with CR I MoAb. Some assays wcrc designed to study H M M O i cell interaction in the proliferative response to the mitogen Con A or to ('and antigen. Proliferative responses of the cultures to Con A and Cand wcrc assayed by measuring the incorporation of [tH]'TdR into I)NA. Only four individual experiments arc illustrated in detail (Table 2). Purified populations of T cells and H M M46 failed to proliferate in response to Con A and (-'and. Addition of 10~;i7 I I M M O to 2.5 × 10s T cells resulted in a proliferative response comparable to that of unfractionated M PC. Also, when HMM4b were added to M P C cultures, the proliferative response remained unchanged, suggesting the absence of a possible suppressive effect. Vary-

I i

I

I

i

f

0

20

40

60

80

Positive

stained

cells,

I 100 ",/.

fig. 2. Detection of class 11 antigens on tluil/an milk t u t t i o phages. Resulls are expressed as indicated ifl trig 1.

rescence. l h e membrane of positive It M MO subsets typically presented a linear fluorescent pattern (Fig. 3A, BJ although a patchy pattern was also seen. l-:resh H M M O were cultured for 30 days in vitro. The kinetics of EI)U-[ antigen expression on H M M05 is shown in Fig. 4. During the first days of culture, a large population of cells continued to express the EDU-I antigen. After this period, the anti-

l a ble 2 Antigen- a n d mitogen-induced I cell prolifcrati(m supported b', h u m a n milk m a c r o p h a g e s ('ells added

Reacti', it.,, to (cpm - I11 ~

MoAb added

...... I xperiment 2

Experinlcnt I ('and .

M PC T H M MG5 M I'(" - H M M~D 1 • HMMcb I - HMM~b I . HMMO l • ItMM4J

.

EI)t-I I)A6.231 TU22

.

.

.

2(I.0 8.2 3.2 259 19.3 4.4 92 16.2

.

C o n ..\ .

.

.

42.6 9.0 1.0 66.3 514.2 27.9 10.7 13.8

.

Cand .

.

5.3 1.14 0.4 7.5 6.3

t).8 NI) NI)

.

Experiment 3

[:'on A .

.

.

190 4.9 11.3 28.4 10.7 5.5 NI) NI)

.

('and .

.

.

200 6.2 18 24.0 17.5 5.3 6.2 NI)

Experiment 4

Con A .

.

.

37.6 10.6 1.2 45.9 21.0 13.3 N[) NI)

.

('and

(.'(~n ,\

14.5 14.6 6.2 I 1.2 11.6 5.9 5.3 67

5(I. .7 17.8 4.0 289 25 `7' 7.f, 111.14 119

.

M o n o n u c l e a r peripheral cells ( M P(.') or adherent cell-depleted I cells (T) were reconstituted with 10c; milk m a c r o p h a g e s ( It M MO). t-i', c ,ul of an a p p r o p r i a t e d dilution ot the m o n o c l o n a l a n t i b o d y ( M o A b ) to be assayed were a d d e d as indicated. Proliferati',e reacti', it,.' of the cultures to Candida antigen ( ( ' a n d ) or c o n c a n a , , a h n A (('t)n A) ',,,as assayed b), measuring the incorporation o l [ . q t i'ldR into I)NA. Figures represent the mean of triplicate cultures.

252

r,.

O Fig. 3. ['xpression of the Q5.13 antigen on h u m a n milk macrophages in ,,itro. A and (." shov, immunotluorescence. B and 1.). phase contrast micro~,copy. Magnification, × 40(). A and B shov, 2 h-old macrophages. (" and 1) shov, I~ day-old macrophages. Autofluorcscence is seen in some cells. Ilcterogcneity in the intensit.', ot fluorescence is obserxed l h e I)R-positi'.c cells are indicated by arrox,.s. 100

ing Con A and Cand concentrations did not affect the response (results not shown). The effects of monoclonal antibodics against class II antigens on T cell proliferation in response to Con A or Cand were assayed in cultures of T cells plus 10~ H M M ~ and are also shown in Tablc 2. In general, all three M o A b proved capable of inhibiting lymphocyte activation either by Con A or Cand although the most pronounced inhibitory effect was obtained with EDU-I.

~z

,~

80

4~ U

~

6o

C a

o.

20

•

I

I

I

i

0

7

14

21

28

Time

in c u l t u r e , d a y s

Fig. 4. Kinetics of Et)U-I antigen expression o n cultured human macrophages. Results are expressed as the mean - St! of 4 experiments.

5. Discussion Relatively little information is available concerning HMMa5 as APC. Only one report showed that 253

fig. 5. Relation of 3east phagocytosis to the presence of I)R antigens on short cultured human milk macrophages. Preopsonized ('amfida ~ere incubated v,ith macrophages for 30 rain. followed by fixation in It; paraformaldehyde and then stained with the 59.5 MoAb and examined under lluorcscencc(A) and phase contrast microscopy(B). I~hagocytic(a) and mmphagocytic(b) l)R-negati~c subsets and phagocytic I)R-positi~c (c) subsets arc obserxed (~" 400).

H M M ~ could present tetanus toxoid or P P D antigen to T cell blasts and M PC [4]. I h e present report indicates that H M M O bear the phenotype of M~b from other anatomical sites and function as accessory cells in mitogen and antigen induced T cell proliferation and their cooperation is class II antigen dependent. H u m a n milk contains up to 80% t t M M O which can ingest Camlida yeast, adhere to plastic, contain nonspecific cstcrase and bear Fc-lgG receptors like other human M~b [3,5.6.191. l h c changes in surface phcnotype of HMMcb during in vitro culture are of interest. Whereas 85(,: of freshly isolated H M M05 expressed OK M I, only 30ci were positive with 61 D3. Previous reports showed that human alveolar M05 and skin I,angerhans cells did not stain with 6[ I)3. l'his marker appears to bc specific for monocytes alone [16]. tlowe~.er, it was present in a small subpopulation of HM M~5. These contradictory results could in part be due to contaminant monocytes in H M MO populations. In culture, 61 I)3 positive cells (10C.?) seemed to belong to a population of small cells in these cultures. The marker is progressively lost as the cells mature or differentiate in vitro. Also, the disappearance of the surface antigen recognized by the 611)3 is faster than the disappearance of the marker recognized by the ( ) K M I . These observations suggest that not all the antibodies used can recognize the same antigenic determinants and these markers should be potentially 254

important tools for studying M~b differentiation in vitro. The present study shows that H M Mob express class !1 antigens. DR antigens localized by indirect immunofluorescence, were found to be in the 80% range with all the monoclonal antibodies assayed. Similar percentages were reported previously with these and other M o A b directed to class I[ antigens with M~b from other anatomical sites [5,6,19 21 ]. DS is a D-region second locus coding for class I1 molecules other than I)R and SB [22]. TU22 seems to be specific for M B I, the serologically defined specificity, carried for I)S molecules [14,22]. In our study, TU22 recognized a minor subpopulation of H M M~b. "Ibis observation suggests that M B can be expressed differentially on I)R-positive H M MO. Serological data have been presented recently suggesting that MB and M I can be expressed independently on class It positive cells [22]. Class II antigens also exhibited differential expression on abnormal B lymphocytes. Most B cells from patients with chronic lymphocytic leukaemia, had low M B positive cells while an anti-DR M o A b stained most cells [22]. It has also been suggested that an ordered sequence of expression of class II antigens exists in normal B cell maturation [22] and there is no reason to suppose that the same circumstances could not occur with the MPS. Previous studies have demonstrated that DR antigens expressed by freshly seeded human P B M o

and M(h are progressively lost as the cells mature or differentiate in vitro both on glass and on a collagen matrix [5,21]. This phenomenon also occurs with H M M ~ cultured on glass. It is not clear from these studies and from the present report whether the differentiation of M05 in vitro is influenced by the culture substrate or other time-dependent processes. There is evidence to show that antigen presentation is restricted to la and DR antigen positive P B M o or M05 in humans and mice [I,2]. However, conflicting reports have been published about human alveolar M05 as antigen presenting cells. Thus, according to some authors [23], alveolar M~b have poor accessory cell function whereas others [24] consider that alveolar M ~ are adequate APC. H M M ~ behave as well as P B M o in supporting proliferative responses to Con A and Cand. It has been documented that, in most cases, milk lymphocytes are completely unresponsive to ('andida albicans although blood lymphocytes from the same patients respond to the antigen [25]. Also, milk lymphocytes are hyporesponsive to P H A or Con A [25]. Our data seem to indicate that this dichotomy of reactivity does not result from suppressive HMMq5 because these cells could support Con A and Cand stimulation when cocultured with autologous peripheral T cells. Other studies provided functional data to support the existence of a human determinant, HI,A-DS, which is important in antigen presentation and stimulation in the autologous mixed lymphocyte reaction [26]. These authors used the Rb03 antiserum, which is a rabbit serum directed against DS. Our preliminary results with TU22 M o A b are consistent with these conclusions. In our system the presence of optimal dilutions of TU22 was capable of abrogating the proliferative responses to Con A and Cand. More definitive studies to directly determine the role of DS molecules in antigen presentation have to be carried out in the near future. Our findings and those of others [4], are consistent with the fact that these cells may play an important role in the immunological development of the ne~'born.

Acknowledgements This work was partly supported by a grant from Fundaci6n H. Koch (Madrid, Spain). The authors wish to thank the patients and the nurseD' staff of the Department of Obstetrics (Hospital " l a Paz", Madrid) for their collaboration in obtained blood and milk samples; l)r. D..laraquemada (The London Hospital Medical College, London) for the gift of DA6.231, TU22 and 61 D3 monoclonal antibodies; Dr. S. Fcrrone (Columbia University, New York) for the gift of CRI and Q5., 13 monoclonal antibodies. We also wish to acknowledge the comments of Dr. M. Hernfindez (Universidad de Alcalb, de Henares, Madrid) and the valuable criticism of Dr. I. M. Outschoorn (Universidad Aut6noma de Madrid).

References I1 ] Ferrone, S. and Da'.id, C. (1982) la Antigens, ,,ol. I. ( ' R . C . Press. Florida. [2] Unanuc. 1-. R. (1981)Ad~. lmmunol. 31, I 136. 131 Pitt..I. (1979) Pediatrics 64. 745 749. [4] Mori, M. and ilavv, ard. A. R. (1982)Clin. lmmunol. Immunopathol. 23, 94 99. 151 l.ey,.a-Cobi/m, F., GabricI-Maliniak, 1.. Mampaso. F. and Clerici, N. (1982) lmmunobiolog.', 163. 120 12 I. 161 Clerici. N, Reboiras, S., I'ierro. C. and I,ey'.a-Cobi/m. I-. (1984) Clin. Exp. Immunol., m press. [7] Iucker. S. B., Pierre. R. V. and Jordon, R. E. (1977)3. Immunol. Methods 14. 267 269. [8] Pellegrino, M. A., Ng, A. K.. I;',usso. C. and Ferrone. S. (1982) Transplantation 34. 18 23. [9] Hanscn, J. A., Martin, P. ,1. and No'~.inski. R. C. (1980) Immunogenetics 10, 247 260. [101 I,ampson, L.A. andl.cvy, R.(1980) J. Immunol. 125. 293 299. [111 Vilclla. R., Yagt~e, .I. and Vives, ,I. (1983) Ituman lmmunol. 6, 53 62. [12] Quaranta. V.. Pellegrino, M. A. and [:erronc, S.(1981)3. Immunol. 126. 548 552. 113] Steel. C. M.. ,,an Heyningen, V., Guy, K., ('ohcn, B. B. and l)eane, D. 1,. (1982) Immunology 47. 597 603. 114] Pawelcc. G. P., Sha'o., S., Ziegler, A.. Mtiller. C. and Wernet, P. (1982).I. Immunol. 129, 1070 1072. 115] Breard..1., Rcinherz. E. I,., Kung, P. C., Goldstein. (i. and Schlossman, S. F. (1980).I. lmmunol. 124. 1943 1948. [16] Nfifie,', G., t'golini. V.. ('apra, .I. t). and Stasny. P. (1982) Scan& .I. Immunol. 16, 515 523. [17] Kung, P. C.. (.ioldstcin. G., Reinherz. E. 1.. and Schlossman. S. F. (1979) Science 206. 347 349.

255

11~] Abo. l+and Balch,('. M.(19NI)J. lmmtmol. 127. 11)24 1029. [19] ('lcrici. N. (1983) I-studio mortologico +vhmcit>nal dcl macr6la[zo al'. ¢olar humano, lhcsis, t~ nix crsidad C¢>mplutcusc, Madrid. [20] N , e . A . K . Zhang. Y. H.,Russo, C. and I-crrone. S.(19811 ('ell Immunol. 66, 78 87. [21] Kaplan, G. a n d ( h t u d c r n a c k , ( i . 11982).l. I!xp. Mcd. 156. I101 1114. [22] Beckman. 1. (1984) h+rmlunt+l. 1 oda 3 5.29 32+

256

1231 1 oe'.'+s, (i. I1., Vml. W. C.. l)unn, M. M.. (;tt//ctt,t. I~.+ NtJtic/. G.. Stash+v, P. and I.ip~,comb, M. |:. (1984),1. Immunol. 132, 181 186. [241 laughtcr, A. H.. Martin, R. R. a n d l v . o m c 3 . . I ,1+11977),1+ l a b . ( ' l i n . Mcd. 89. 1326 1332. 1251 t~armcly. M..1.. Beer. A. E. ~tnd Billint.zh.ml. R. t-. 11976).I. [!xp. Mcd. 144. 358 371). [26] (ionv.a. 1. A., Picker. I. J., R~fll. H. V.. (io+vcrt. g. M., ."Gil'.cr..I. arid gtobt~, .l. 1). ( 1983] .I. lrnmun~l. 130. 706 71 I