- Email: [email protected]
The accessory cell function of murine Peyer's patches
CELLULAR
IMMUNOLOGY
92, 41-52 (1985)
The Accessory
Cell Function of Murine Peyer’s Patches’
WALTER G. BARR,* STEPHENJ. CHALLAcoMBE,t ALEX YEM,$ AND THOMAS B. TOMASI~~ *Department of Rheumatology, Loyola University, Chicago, Illinois 60626; tDepartment of Oral Immunology, Guys Hospital, London SE1 9RT, England; and #Department of Cell Biology, University of New Mexico Cancer Center, Albuquerque, New Mexico 87131 Received August 6, 1984; accepted November 30, 1984 The cellular composition and certain functional characteristics of murine Peyer’s patches (PP) were examined and compared with other lymphoid tissues. The composition of PP resembled most closely that of the spleen with the exception of a significant decrease in the number of adherent and phagocytic cells. Very few cells with dendritic morphology could he identified in Peyer’s patches. Whole PP (and the nonadherent population) were capable of presenting antigen ovalbumin, human gammaglobulin, and purified protein derivative in a T proliferative assay to sensitized lymph node cells and to an antigen-specific T-cell clone. The antigen-presenting cell in both the spleen and PP was concentrated in the low-density population which floated on 1.080 bovine plasma albumin. However, equal numbers of whole and PP tloaters were deficient in their capacity to present antigen compared with similar populations from spleen. Moreover, in PP the antigen-presenting cell appeared in the nonadherent rather than the adherent population as found with other lymphoid tissues. Similar results were obtained with (B6A)F,, CBA, A.TFR-1 and BI0.S (12R) mice, suggesting that the inability of adherent cells from PP to present antigen effectively was not genetically determined. Whole and nonadherent PP contained cells capable of stimulating an allogeneic MLR, although again they were generally inferior to those of the spleen when comparable numbers of cells were employed. The adherent population of PP did not elicit an MLR. However, whole PP contained accessory cells needed for mitogen-induced proliferation since passage over nylonwool columns resulted in a nonadherent fraction which did not respond to concanavalin A or phytohemagglutinin and the additionof adherentperitonealexudatecellsrestored the lectin
response. Thedifferences notedin theaccessory cellfunctionin PPandotherlymphoidtissues suggest the possibility that quantitative or qualitative differences in the function of these cells may explain some of the previously observed characteristics of PP, such as the inability to detect a primary antibody response in this tissue. The possibility that the development of gutassociated suppressor cells and their migration to peripheral tissues may he involved in the systemic tolerance that follows oral immunization and that these may be related to numerical 8 198s Academic and/or functional differences in macrophages or accessory cellsis discussed. Resa Inc.
INTRODUCTION
Peyer’s patches (PP) are subepithelial lymphoid follicles found throughout the mammalian small intestine where they comprise a major component of the gut ’ This work was supported in part by USPHS Grant AM-3 1448. ‘To whomcorrespondence should be sent at: Cancer Center, The University of New Mexico, 900 Camino de Salud NE, Albuquerque, N. Mex. 87131. 41
42
BARR ET AL.
associated lymphoid tissue (GALT). Specialized epithelial cells (M cells) overlie their surface and are primarily responsible for the transport of luminal antigens to the underlying lymphoid cells. Antigens deposited within the gut lumen have been shown to elicit IgA antibodies at sites distal from the original application such as in the colostral, lacrimal, and salivary secretions of animal and man (1-3) and the salivery secretions in the rhesus monkey (4). It appears that antigen-sensitized cells, largely precommitted to IgA production, arise within the PP and then migrate to the widely separated mucosal surfaces that collectively make up the secretory immune system. Paradoxically, oral and systemic immunization with soluble antigens do not appear to lead to the production of antibody-secreting cells or an effective T-cellproliferative response in the PP themselves (5-7) although irregular responses have been reported in the rabbit (8) and with particulate antigens such as killed Streptococcus mutans or their cell walls which stimulate a T-cell response in PP of the mouse (7). Kagnoff and Campbell reported (9) the failure of PP cells to support the induction of a primary humoral response to sheep red blood cells (SRBC) or the induction of cytotoxicity to allogeneic cells. Both of these functions were restored by the addition of adherent peritoneal exudate cells (APEC) or 2-mercaptoethanol (2-ME) to the cultures. This suggested that the immunological shortcomings of PP cells might exist secondary to a decrease in the number or functional capacity of its accessory cells. This study was undertaken to further explore parameters of accessory cell function within the Peyer’s patch. MATERIALS
AND
METHODS
Animals. Six- to twelve-week-old CBA/J and B6A mice were obtained from the Jackson Laboratories, Bar Harbor, Maine. BlO.S5 (12R) mice and A.TFR-1 mice were bred in our own animal facilities and were kindly provided by Dr. Chella David. Media. Calcium- and magnesium-free Hanks’ balanced salt solution (HBSS) containing 25 mM Hepes buffer (GIBCO, Grand Island, N.Y.) and 10% fetal calf serum (FCS) was used as the standard preparation medium prior to final culture medium as described below. In culture RPM1 1640 (GIBCO) supplemented with 25 mM Hepes, 100 U/ml of penicillin, 100 pg/ml of streptomycin, 2 mM glutamine, 400 &ml of garamycin, 30 clg/ml of fungizone, and 2-mercaptoethanol (5 X lo-’ M) was employed. Heat-inactivated (56”C, 30 min) human serum (2.5% v/v concentration) was used to supplement cultures involving presentation to sensitized lymph node cells (SLNC); mixed lymphocyte reactions (MLR) were carried out in 5% FCS and presentation of antigens to T cells in 10% FCS. Preparation of cell suspensions.Spleen, mesenteric lymph nodes (MLN), PP, and inguinal and para-aortic lymph nodes (PLN) were removed from normal animals and placed in HBSS- 10 ( 10% serum). The tissues were gently crushed using a Teflon tissue homogenizer. The cell suspensions were passed through a nylon mesh to remove clumps and debris. The single-cell suspensions were collected and washed. Cytocentrifuge preparation of PP cells were analyzed for Ig-containing plasma cells by the fluorescent antibody technique and preparations containing >l% plasma cells were considered contaminated with lamina propria and discarded. Resident peritoneal exudate cells (PEC) were obtained by washing the peritoneal cavity with
ACCESSORY CELLS OF PEYER’S PATCHES
43
HBSS. Cell suspensions were adjusted to 10 X lo6 cells/ml before placing in petri dishes. Sensitized lymph node cells for the proliferative assays were obtained from the inguinal and para-aortic lymph nodes of mice given 100 pg of ovalbumin (Sigma Chemical Co., St. Louis, MO.) in H37Ra adjuvant (Difco, Detroit, Mich.) subcutaneously into the tail 8 days previously. Adherent cells. Aliquots (2.5 ml) of a cell suspension in HBSS-2.5 (2.5% FCS) were placed in 5-cm diameter plastic tissue culture petri dishes and the dishes were incubated for 60 min at 37°C. The dishes were washed with warm HBSS-2.5 and then 5 ml of HBSS-2.5 was added to each dish and incubation continued for an additional 60-l 20 min. The dishes were washed with 5 ml of warm HBSS-20 and then 5 ml of phosphate-buffered saline (PBS) containing 0.02% (w/v) tetrasodium EDTA was added to remove adherent cells. After 15 min on ice the dishes were washed three times with 10 ml of HBSS-2.5; washes were centrifuged at 4°C and the cell pellets resuspended in cold HBSS2.5. After two additional washings the cells were resuspended in the appropriate medium for testing. In experiments to determine the number of adherent cells, the cells recovered after EDTA treatment were counted and expressed as a percentage of the number applied to the plates. In an attempt to enrich for accessory and MLR-stimulating cells, suspensions of cells from spleen or PP were allowed to adhere to plastic tissue culture plates for 2 hr at 37°C; nonadherent cells were removed by gentle pipetting with cold medium. The plates were then covered with medium (RPM1 1640 supplemented with 25 mM Hepes, 100 U/ml penicillin, 100 &ml of streptomycin, 2 mM glutamine, 400 pg/ml of garamycin, and 30 pg/ml of fungizone with 5% FCS and 2-mercaptoethanol (5 X 10e5 M)) and returned to the 37°C CO2 incubator overnight. During this period many of the adherent cells either spontaneously eluted from the plastic surface or were easily dislodged by pipetting with cold medium. These cells were washed and the number of viable cells was determined by trypan blue exclusion. Presentation of antigen to SLNC. Method I. Cell preparations were incubated in triplicate in flat-bottomed tissue culture plates (Falcon 3042, Falcon, Oxnard, Calif.) at six different concentrations calculated to give 0.1 to 3.2 X 10’ adherent cells per well. After incubation for 60 min at 37°C nonadherent cells were removed by washing the cells three times in cold RPMI-2.5 (2.5% FCS). For pulsing 50 pg of ovalbumin in 200 ~1 of medium was added to each well and the plates were incubated for a further 2 hr or, in some experiments, 24 hr. After washing three times with warm medium, 4 X 10’ SLNC were added to each well and the plates incubated for 4 days with no further addition of antigen. Tritiated thymidine (1 &i) was added to each well 16 hr prior to harvesting the cultures. As controls, SLNC were added to duplicate cultures of nonpulsed adherent cells or pulsed adherent cells were incubated alone. For comparison the proliferation of SLNC was determined by incubation of 4 X 10’ SLNC with 50 pg and 100 Fg of ovalbumin in 200 ~1 of media. The presentation experiments were performed three times and the means were calculated. To test for the effects of H-2 or non-H-2 genes on antigen presentation, experiments were repeated in A.TFR-1 mice, which differ from CBA/J in background but have the same I region determinants (H-2k) and in B 1O.S (12R) mice (H-2”) which differ from CBA/J at both H-2 and non-H-2 loci. In additional experiments human gammaglobulin (HGG) (Sigma) or purified protein derivative (PPD) (Mycobacterium tuberculosis, Connaught Laboratories, Toronto, Canada) were used as antigens.
44
BARR ET AL.
Method ZZ. In a separate series of experiments, cells were allowed to adhere to 60-cm plastic petri dishes as outlined above. The cells were pulsed with ovalbumin (1 mg/ml) for 2 hr and then thoroughly washed in RPMI-2.5. Adherent cells were removed with 0.02% EDTA as described and added to 4 X lo5 SLNC in concentrations ranging from 0.1 X lo5 to 1.6 X lo5 (viable cells) in a total volume of 200 ~1 and incubated for 4 days at 37°C. Results in both methods were expressed as the difference in counts between pulsed adherent cells plus SLNC, and nonpulsed adherent cells plus SLNC. Enrichment for low-density populations. Populations of spleen and PP cells were enriched for low-density cells by equilibrium density centrifugation on dense bovine plasma albumin (BPA) solutions. Single-cell suspensions were spun at 10,OOOg in an SW39 swinging bucket rotor for 20 min at 4°C. Two cell fractions (pellicle and pellet) were then harvested with Pasteur pipets and washed in RPM1 prior to suspension in the final medium for counting. BPA solutions were prepared from bovine plasma albumin powder (fraction V, Reheis Chemical Co., Phoenix, Ariz.). Routinely 100 g of BPA were mixed with 186 ml of Ca’+- and Mg*+-free PBS, 20 ml 1 N NaOH, and 65 ml H20. The pH and density were adjusted if necessary to give a final solution with a pH 7.35-7.45 and a density of 1.080-l .082. The solution was subjected to millipore filtration and kept at 4°C prior to use. Morphologic examination for dendritic cells. Examination by phase-contrast microscopy was performed on specimens fixed for 5 min at room temperature in 2.5% glutaraldehyde buffered with PBS, pH 7.40, following adherence to glass coverslips for 2 hr at 37°C. Nonadherent cells were bound to coverslips previously coated with poly(L-lysine) (PLL; Type VII, Sigma; 25 &ml in PBS). The cells were allowed to adhere for 20 min, fixed with glutaraldehyde, and examined for the morphological feature of dendritic cells as described by Steinman and Cohn (10). Preparation of T cells. (a) Cloned T cells specific for poly(~-G1u6’, L-Ala3’, L-Tyr’O) (GAT) were kindly donated by Dr. Masao Kimoto and Dr. C. G. Fathman and were prepared as previously described (11). (b) Alloreactive long-term T-cell cultures were the result of long-term repetitive MLR with initially unfractionated responder cell populations. Responder cells from PP or spleen (3-6 X lo6 cells) were maintained in 20 ml of RPM1 supplemented with 25 mM Hepes, 100 U/ml penicillin, 100 &ml of streptomycin, 2 m&Z glutamine, 400 &ml of garamycin, and 30 pg/ml of fungizone with 5% FCS and 2-mercaptoethanol (5 X 10e5 M) in 75-cm tissue culture flasks (Coming Glass Works, Coming, N.Y.). Every lo- 14 days these cells were pulsed with 50-60 X lo6 irradiated (3300 R) spleen cells. T cells recognizing (B6A)Fi determinants were derived from CBA Peyer’s patch cells stimulated with irradiated (B6A)Fi spleen cells. T-cell lines recognizing CBA determinants were derived from B 10 spleen cells stimulated by irradiated CBA spleen cells. Mixed lymphocyte reactions. Alloreactive T cells ( 105) served as responder cells and variable numbers of irradiated (3300 R) spleen or Peyer’s patch cells as stimulators. Cells were cultured in round-bottom tissue culture plates (Linbro, Flow Laboratories, Hamden, Conn.) and stimulation was measured by uptake of tritiated thymidine in a 2-day proliferative assay. Primary MLR were performed as described by Murgita and Tomasi (12). Presentation of GAT to cloned T cells. Cloned GAT-reactive T cells ( 104) were mixed with 50 pg of GAT and variable numbers of irradiated cells from the spleen
ACCESSORY CELLS OF PEYER’S PATCHES
45
or PP of (B6A)Fl in a total volume of 200 ~1. Cells were cultured in flat-bottom tissue culture plates (Falcon 3042) for 3 days and stimulation was measured by uptake of tritiated thymidine. Accessory cell function for mitogen responses. Concanavalin A (Con A)- and phytohemagglutinin (PHA)-induced proliferative reactions were carried out as previously described ( 12). Accessory cells were depleted by passage over a nylonwool column. Syngeneic adherent peritoneal exudate cells (5%) were added in an attempt to restore the response. RESULTS Antigen presentation to sensitized lymph node cells. Adherent spleen cells (AX) pulsed with antigen (Method I-Material and Methods) were most efficient in presenting antigen to SLNC. Significant stimulation of SLNC was found with 2.5% ASC and the proliferative response increased in proportion to the number of cells added to a maximum at 40% ASC (Fig. 1). Addition of pulsed APEC also resulted in increased stimulation of SLNC to a maximum at lo-20% APEC; greater concentrations caused a progressive decrease in the response. Presentation of antigen by adherent lymph node cells (ALNC) appeared to be less effective than with ASC or APEC, and significant counts were not found until 10% adherent cells were added. At concentrations of 40% or greater, pulsed adherent mesenteric lymph node cells (AMLNC) showed significantly greater counts than ALNC (P < 0.05) although the gross cellular characteristics of mesenteric and peripheral lymph nodes are very similar (Table 1). Data very similar to those shown in Fig. 1 with ovalbumin (OVA) as antigen were also obtained with HGG and PPD. As shown in Fig. 1, pulsed
3 hour
adherent
cells
1~o14x1051
FIG. I. Presentation of ovalbumin to sensitized lymph node cells (SLNC) by adherent cells from Peyer’s patches (APPC, l ), spleen (ASC, A), mesenteric lymph nodes (AMLNC, A), peripheral lymph nodes (ALNC, 0), and peritoneal exudate (APEC, 0). Adherent cells were pulsed with 200 JL~ovalbumin for 2 hr at 37’C in plates and washed, SLNC were added at 4 X 10’ per well and cuhured for 4 days. n = SLNC control, 50 ccgovalbumin added to 4 X lo5 SLNC. Means of three experiments. Control values (nonpuked adherent cells plus SLNC) have been subtracted.
46
BARR ET AL. TABLE 1 Cellular Characteristics of Peyer’s Patches and Mesenteric Node Cells from CBA/J Mice Compared with Other Lymphoid Tissues Peyer’s patches
B cells (anti-kappa positive) T cells (anti-Thy 1.2 positive) Phagocytic cells Adherent cells Esterase positive cells
49 31 5.2 3.8 13.9
+4 f6 f 1.7 31 1.8* + 1.4**
Spleen 44 ?I 41 +9 11.1 + 2.4 13.1 f 2.1 12.0 k 2.6
Peritoneal exudate 22 17 51.8 63.1 48.6
+ + f f +
2 2 10.5 17 6.8
Lymph nodes 17 68 5.7 6.4 4.5
*5 +9 + 1.8 + 1.9 + 1.4
Mesenteric lymph nodes 16 70 5.6 7.6 4.6
+3 +5 f 1.9 + 1.5 + 0.8
Note. Mean percentage + SD of between 5 and 10 experiments for each determination. Tissues from normal I-week-old mice. Percentage B and T cells by fluorescent staining with anti-kappa or anti-Thy 1.2 antisera. Percentage phagocytic cells by latex ( 1.1 pm) ingestion. Adherent cells = percentage adhered and recovered from petri dishes after 2 hr (see Methods and Materials). * P < 0.02 compared with lymph nodes or spleen. ** P < 0.01 compared with lymph nodes.
adherent Peyer’s patch cells (APPC) did not elicit stimulation of SLNC at any of the concentrations tested between 2.5 and 80% of the number of SLNC. Lengthening the period of antigen pulsing to 24 hr did not improve presentation, Antigen-pulsed APPC were also unable to present HGG or PPD to SLNC. Ten percent APPC added to APEC or ASC did not significantly affect their ability to present antigen, suggesting that suppressor cells were not responsible for the deficient presentation by APPC. When adherent cells were pulsed in petri dishes, removed with EDTA, and transferred to plates (Method II-Material and Methods) in contrast to being allowed to adhere to plates, pulsed, washed and SLNC added, similar results were obtained with the exception that at the highest concentration of APPC examined (40%) minimal presentation of ovalbumin was detected. However, this was significantly less with APPC than with AMLNC or ASC as shown in Fig. 2. It was noted that control values (nonpulsed adherent cells + SLNC) were higher for all cell types examined than with Method I, and increased in proportion to the number of adherent cells added. Viabilities of the adherent cells for each of the cell preparations were similar. The lack of presentation by APPC did not appear to be restricted to CBA/J mice since identical results were found with mice of the A.TFR-1 and BIO.S (12R) strains. This suggested that the failure of APPC to present antigen was probably not restricted by H-2 or non-H-2 loci. Antigen presentation to cloned T cells. Both unfractionated PP and spleen cells from (B6A)Fi mice were shown to be capable of presenting antigen to cloned GATreactive T cells. Density centrifugation on BPA (p = 1.080) separates spleen and PP cells into two fractions: a low-density population (“floaters”) and a high density fraction (“sinkers”). The antigen-presenting cell in the spleen was significantly enriched in the floaters (containing 5-15% of total cell population) as shown in Fig. 3. Similar findings were found with PP cells (see Fig. 4) where low-density cells made up lo-30% of the total. Repeated experiments showed PP floaters to be less
ACCESSORY CELLS OF PEYER’S PATCHES
3 hours
adherent I cells
47
cells % of 4 x lo5 translered 1
FIG. 2. Presentation of ovalbumin to SLNC by adherent cells from Peyer’s patches (APPC, l ), mesenteric lymph nodes (AMLNC, A) and spleen (ASC, Cl). Adherent cells were pulsed with 1 mg/ml ovalbumin for 2 hr at 37°C in petri dishes, removed with 0.02% EDTA, washed, and added to SLNC. Control values (nonpulsed adherent cells plus SLNC) have been subtracted. Means of four experiments.
capable of presenting antigen to the cloned cells than spleen cells at the same cell concentrations. MLR stimulation by Peyers patch cells. Whole Peyer’s patch cells provide good stimulation in the allogeneic MLR. However, as shown in Table 2, the adherent cell population from PP, unlike those from the spleen, did not act as effective stimulators. The number of adherent PP cells employed in Table 2 gave optimal stimulation. The stimulating capacity of the nonadherent population was similar to that of the whole PP. In other experiments, long-term T-cell cultures enriched for alloreactive cells by repetitive MLR were used as responder cells. Stimulation of lo5 of these T cells by as few as 2 X lo4 irradiated spleen or PP cells provided an excellent response when 2-day MLR were assessed. Separation of spleen and PP cells into high- and low-density cells resulted in enrichment for the stimulating cell
Whole
Floaters
Sinkers
FIG. 3. Antigen presentation by irradiated (3300 R) syngeneic spleen cells (0.25 X 106) to (B6A)Fr cloned GAT-specific T cells (10” cells). GAT (50 &ml) was added and [‘Hlthymidine incorporation was measured after 3 days of culture. CPM on whole spleen, 13,031 + 813; floaters, 25,631 + 1682; sinkers, 1003 rt 273. Low-density (floaters) and highdensity (sinkers) cells separated on BPA (see Methods and Materials).
48
BARR ET AL.
Whole
FlC&0nl
Sinkers
FIG. 4. Antigen presentation by irradiated (3300 R) syngeneic Peyer’s patch cells (1 X lo6 cells) to (B6A)Fr cloned GAT-specific T cells (IO4 cells). GAT (200 &ml) was added and [3H]thymidine incorporation was measured after 2 days of culture. CPM on whole Peyer’s patches, 2309 ? 155; tloatem, 3,321 + 82; sinkers, 1,517 + 109. Floaters and sinkers separated on BPA (see Methods and Materials).
in the low-density fraction (Fig. 5). In repeated experiments PP floaters provided good stimulation though consistently less than spleen cells (Fig. 5). PP and spleen cells differed when low-density cells were separated into adherent and nonadherent fractions. In spleen cell populations the MLR-stimulating cell was present primarily in the adherent fraction while in the PP it was recovered only in the nonadherent population (Fig. 6). Experiments using T cells derived from BlO mice recognizing CBA determinants yielded similar results to those outlined above. Accessory cell function of Peyers patch cells in mitogen responses. In view of the well known requirement for accessory cells in Con A- and PHA-induced proliferation we examined PP for accessory cell function in these responses. As shown in Table 3, Con A and PHA induced reasonable proliferative responses which were nearly abrogated by passage over a nylon-wool column. Addition of irradiated (3300 R) APEC (5%) to the nylon-nonadherent fraction from PP restored their reactivity. Dendritic cells in Peyer’s patches. Phase contrast examination of low-density adherent cells (LODAC) from PP revealed significant differences when compared to the spleen. Spleen LODAC preparations were consistently characterized by 40% or more cells with typical morphological features of dendritic cells (DC) as described by Steinman and Cohn (10). PP LODAC populations infrequently contained small TABLE 2 Mixed Lymphocyte Reaction
Source of stimulator
Responder (CBA LNC)
MLR
Whole spleen Whole PP *ASC *APPC
536 605 1021 444
139,590 18,180 41,419 951
Stimulator (BALBfc) 1110 116 1586 73
SI 85 25 16 1.8
Note. cpm = Counts per minute of [‘Hlthymidine incorporated after 4day culture in 2.5% human AB serum. Results with 5 X 10e5 M 2-ME added were similar. ASC = adherent spleen cells. APPC = adherent Peyer’s patch (PP) cells. Number of cells per well: CBA LNC, 5 X 10’; BALB/c whole spleen, 1 X 106; whole PP, 1 X lob, ASC, 1 X 10s; and APPC, 1 X 10’. * Adherence carried out for 2 hours at 37°C in plastic petri dishes.
ACCESSORY CELLS OF PEYER’S PATCHES Psyrr’a
WhOI. Cbal.”
49
Patch
skew,
FIG. 5. Mixed lymphocyte reactions (MLR) comparing irradiated (3300 R) stimulators from CBA spleen and Peyer’s patch (PP) populations (LOS/well) separated on BPA gradients. Responding cells ( 105/ well) were obtained from long-term T-cell cultures derived from Bl0.A spleen and PP cells repetitively stimulated with CBA spleen cells in vitro. Whole spleen SI, 41; spleen floaters SI, 158; spleen sinkers SI, 26. Whole PP SI, 49; PP floaters SI, 113; PP sinkers SI, 15.
numbers (~5%) of cells resembling splenic DC. Other large, spread out mononuclear cells of indeterminate lineage were also seen but the surface characteristics and functional significance of these cells are unknown. The nonadherent low-density cells from PP were attached to glass coverslips previously treated with poly(L-lysine). However, these cells did not demonstrate the typical morphology exhibited by splenic DC after adherence to PLL-coated coverslips. Furthermore, observation of the nonadherent population in culture for up to 5 days did not reveal development of the cytologic features similar to those of splenic dendritic cells. We find no evidence that enzymatic dissociation with dispase releases cells with dendritic morphology. DISCUSSION Adherent cells from Peyer’s patches are deficient in their ability to present soluble antigens (OVA, HGG, and PPD) to sensitized lymph node cells when compared to adherent cells from peripheral lymph nodes, spleen, or peritoneal exudate. The number of adherent cells in Peyer’s patches is significantly less than in spleen or lymph nodes (Table 1) although we cannot be sure that this is not attributable to technical problems such as the manner in which the patches are collected. Perhaps macrophages are concentrated close to the intestinal lumen. In this regard we have not been able to improve yields of adherent cells by using collagenase (Type II) or
*mur.nt
Na.dkmu
Aummt
Nnum.r.nt
FIG.6. Mixed lymphocyte reactions (MLR) with irradiated (3300 R) low-density spleen and Peyer’s patch (PP) cells comparing their adherent and nonadherent fractions. Responders were long-term T-cell cultures ( 10s/well). Stimulators Born the spleen (2 X 10’) were from (B6A)Fi mice and responding T cells were CBA in origin. PP stimulators (IS X 10’) were from CBA mice with T-cell responders from BIO.A. Spleen adherent SI, 228; spleen nonadherent SI, 65. PP adherent SI, 6; PP nonadherent SI, 174.
50
BARR ET AL. TABLE 3 Effect of Removal of Adherent Cells (AC) from Peyer’s Patches (PP) on the Proliferative Response to Concanavalin A (Con A) and Phytohemagglutinin (PHA) AC-depleted Whole Peyer’s patch cells
Medium Con A (4 pg./well) PHA (2 &vell)
2,199 95,841 37,616
f 1 IO* f 5693 2 4167
ACdepleted
PP
185 f 41 2416 f 326 2053 + 876
* Mean cpm f SE of triplicate cultures. PEC = normal were incubated at 4 X 10’ cells per well and [‘Hlthymidine
PP with addition
1%
2.5%
572 f 53 15,993 f 6587 12,155 k 2148
816 f 88 115,274 f 3947 22,870 + 7019
syngeneic peritoneal exudate incorporation was measued
of PEC 5% 1,384 105,380 41,412
k 196 f 4058 f 219
cells added (W of 4 X 10’). Cells after 3d of culture.
trypsin plus collagenase, nor do we find significant differences with the manner in which the patches are obtained (deep cuts versus more superficial) unless the lamina propria is included. Contamination with lamina propria is a key issue in all studies on PP and can be ascertained by determining of the number of Ig-containing plasma cells on cytocentrifuge preparations. Normally, PP lack plasma cells (except for scattered Ig-containing cells localized to the dome) and preparations containing > 1% should in our view be considered contaminated and discarded. Defective antigen presentation cannot be attributed to cell numbers alone since in the antigen presentation experiments equivalent numbers of adherent cells were used from each source. It also seems unlikely that APPC contain a population of suppressor cells. The addition of APPC to SLNC did not suppress the response and mixing APPC with APEC or ASC did not alter their ability to present antigen. It has not been excluded that the inability of APPC to present antigen may apply only to certain antigens. This possibility is raised by the observation that the T-cell proliferative response of PP cells following intragastric administration of soluble (tolerogenic) protein antigens such as OVA is negligible whereas significant stimulation occurs after administration of a particulate antigen such as S. mutans (7). Work is in progress attempting to elucidate this point. We have found that flotation of whole spleen cells on BPA (p = 1.080) produces a heterogeneous low-density population (5-15% of the total spleen cells) which includes macrophages, B and T cells, and dendritic cells as first reported by Steinman and Cohn (10). This population is enriched in antigen-presenting cells. Flotation of whole Peyer’s patch cells produces a low-density fraction which represents lo-30% of the total population and is also enriched in antigen-presenting capabilities. However, repeated experiments have shown that equivalent numbers of low-density cells from the Peyer’s patches are less efficient in their ability to present antigen to cloned T cells than those derived from spleen. Similarly, whole PP are deficient compared to whole spleen in antigen presentation in this system. In both the spleen and PP the MLR-stimulating cell was found predominantly within the low-density population. However, while adherent floaters from the spleen are potent stimulators of the MLR, this activity is found exclusively in the nonadherent low-density population of PP cells. In this regard rat lymph node dendritic cells are found in the low-density nonadherent fraction ( 13). The low-density nonadherent population from the PP does not, however, contain cells with the typical morphological characteristics of dendritic cells when examined in suspension or after adherence to
ACCESSORY CELLS OF PEYER’S PATCHES
51
poly(L-lysine)-treated glass coverslips; nor do the cells assume those characteristics when kept in cultures for up to 5 days. We have not been able to identify dendritic cells in cell populations obtained with dispase but have shown accessory cell function (for mitogens) in such preparations as reported by others (14). It is possible that dendritic cells are present in low but functionally significant numbers within the PP and/or have different morphological characteristics than those in the spleen, perhaps related to technical factors involved in obtaining cell preparations from these sources. Alternatively, a different type of cell or cells in PP may be responsible for antigen presentation. B cells have been reported (15) to have antigen-presenting capacity and we are currently investigating whether PP B cells may present certain antigens. Recently we (Yem and Tomasi, unpublished data) have obtained actively presenting preparations using the fluorescent activated cell sorter (Becton-Dickinson, FACS III) which was gated to sort and collect large cells and contained >85% surface Ig-positive cells. These preparations, enriched for B blasts, are currently being analyzed and sorted for macrophages (assaying with MAC-l, -2, and -3 reagents) and examined for dendritic cells by cytotoxicity using an antidendritic cell-specific antibody (anti-DC- 1, courtesy of Dr. S. Steinman). Our previous studies (16) have defined the expression of Ia antigens on PP cells and determined that a large proportion of these cells (70-85%) expressed both I-A and I-E including most PP B cells and 20% of T cells. We could not however, determine whether the small number of phagocytic cells present in PP preparations possess Ia antigens. Because of the importance of Ia in antigen presentation further studies of PP are underway involving Ia expression and the affect of anti-Ia on isolated actively presenting populations. The relative lack of antigen presentation by APPC is in agreement with a previous report (9) that PP lack adherent accessory cells needed for a primary in vitro antibody response and were deficient in antibody production after oral or parenteral immunization (5,6). However, the accessory cells necessary for antigen presentation, as well as for the Con A and PHA response, appear to be present in whole and nonadherent PP and when removed they can be replaced by irradiated (3300 R) APEC. In analogy with other systems it has been suggested (17) that the deficient antigen presentation and presumably the failure to generate a large number of proliferating T helper cells could be related to the apparent ease with which suppressor cells are developed. This would have implications pertinent to a number of central issues in mucosal immunity. For example, the generation of suppressor cells for IgG and IgM and their dissemination to peripheral lymphoid tissues (18, 19-22) occurs concomitantly with the development of antigen-specific systemic tolerance and a (local) secretory IgA response (7). This is associated with the appearance of IgA-specific helper cells in PP and MLN (22). Speculatively, antigen presentation in the gut may occur by unique mechanisms which generate T helper cells for IgA and T suppressor cells for IgG and IgM (21). The possibility that gut macrophages have some degree of antigen and/or isotype specificity in their cell interactions as previously suggested (23) has not been excluded. REFERENCES I. We&-Canington, Proc. Natl.
P., Roux, M. E., McWilliams,
Acad. Sci. 75, 2928,
1978.
M., Phillips-Quagliata,
J. M, and Lamm, M. E.,
52
BARR
ET
AL.
2. Mestecky, J., McGhee, J. R., Arnold, R. R., Michalek, S. M., Prince, S. J., and Babb, J. L., .I. Clin. Znvesf.61, 731, 1978. 3. Michalek, S. M., McGhee, J. R., and Babb, J. L., Znfect.Zmmun. 19, 217, 1978. 4. Challacombe, S. J., and Lehner, T., Arch. Oral Biol. 24, 917, 1980. 5. Bienenstcck, J., and Dolezel, J., J. Zmmunol. 106, 938, 1971. 6. Henry, C., Faulk, W. P., Kuhn, Z., Yoffey, J. M., and Fudenberg, H. H., J. Exp. Med. 131, 1200, 1970. 7. Challacombe, S., and Tomasi, T. B., J. Exp. Med. 152, 1459, 1980. 8. Clancy, R., and Pucci, A., AJEBAK 56, 337, 1978. 9. Kagnoff, M. F., and Campbell, S., J. Exp. Med. 139, 398, 1974. 10. Steinman, R. M., and Cohn, Z. A., J. Exp. Med. 137, 1142, 1973. 11. Kimoto, M., and Fathman, C. G., J. Exp. Med. 152, 759, 1980. 12. Mu&a, R. A., and Tomasi, T. B., J. Exp. Med. 141,440, 1975. 13. Klinkert, W. E., LaBadie, J. H., O’Brien, J. P., Beyer, C. F., and Bowers, W. E., Proc. Nat/. Acad. Sci. 77, 5414, 1980. 14. Spalding, D. M., Koopman, W. J., and McGhee, J. R., Ann. N.Y. Acad. Sci. 409, 880, 1983. 15. Chestnut, R. W., and Grey, H. M., J. Zmmunol. 126, 1075, 1981. 16. Krco, C. J., Challacombe, S. J., Lafuse, W. P., David, C. S., and Tomasi, T. B., Cell. Zmmunol. 57, 420, 1981. 17. Pierres, M., and Germain, R. N., J. Immunol. 121, 1306, 1978. 18. Mattingly, J. A., and Waksman, B. H., J. Zmmunol. 121, 1878, 1978. 19. Husband, A. J., and Gowans, J. L., .I Exp. Med. 148, 1146, 1978. 20. Hanson, D. G., Vas, N. M., Maia, Z. C. S., Hombrook, M. M., Lynd, J. M., and Roy, C. A., Arch. Allergy Appl. Zmmunol. 55, 526, 1977. 21. Asherson, G. L., Zembala, M., Perera, M. A. C., Mayhew, B., and Thomas, W. R., Cell. Zmmunol. 33, 145, 1977. 22. Richman, L. K., Graeff, A. S., Yarchan, R., and Strober, W., J. Zmmunol. 126, 2079, 1981. 23. Tomasi, T. B., Transplantation 29, 353, 1980.
IMMUNOLOGY
92, 41-52 (1985)
The Accessory
Cell Function of Murine Peyer’s Patches’
WALTER G. BARR,* STEPHENJ. CHALLAcoMBE,t ALEX YEM,$ AND THOMAS B. TOMASI~~ *Department of Rheumatology, Loyola University, Chicago, Illinois 60626; tDepartment of Oral Immunology, Guys Hospital, London SE1 9RT, England; and #Department of Cell Biology, University of New Mexico Cancer Center, Albuquerque, New Mexico 87131 Received August 6, 1984; accepted November 30, 1984 The cellular composition and certain functional characteristics of murine Peyer’s patches (PP) were examined and compared with other lymphoid tissues. The composition of PP resembled most closely that of the spleen with the exception of a significant decrease in the number of adherent and phagocytic cells. Very few cells with dendritic morphology could he identified in Peyer’s patches. Whole PP (and the nonadherent population) were capable of presenting antigen ovalbumin, human gammaglobulin, and purified protein derivative in a T proliferative assay to sensitized lymph node cells and to an antigen-specific T-cell clone. The antigen-presenting cell in both the spleen and PP was concentrated in the low-density population which floated on 1.080 bovine plasma albumin. However, equal numbers of whole and PP tloaters were deficient in their capacity to present antigen compared with similar populations from spleen. Moreover, in PP the antigen-presenting cell appeared in the nonadherent rather than the adherent population as found with other lymphoid tissues. Similar results were obtained with (B6A)F,, CBA, A.TFR-1 and BI0.S (12R) mice, suggesting that the inability of adherent cells from PP to present antigen effectively was not genetically determined. Whole and nonadherent PP contained cells capable of stimulating an allogeneic MLR, although again they were generally inferior to those of the spleen when comparable numbers of cells were employed. The adherent population of PP did not elicit an MLR. However, whole PP contained accessory cells needed for mitogen-induced proliferation since passage over nylonwool columns resulted in a nonadherent fraction which did not respond to concanavalin A or phytohemagglutinin and the additionof adherentperitonealexudatecellsrestored the lectin
response. Thedifferences notedin theaccessory cellfunctionin PPandotherlymphoidtissues suggest the possibility that quantitative or qualitative differences in the function of these cells may explain some of the previously observed characteristics of PP, such as the inability to detect a primary antibody response in this tissue. The possibility that the development of gutassociated suppressor cells and their migration to peripheral tissues may he involved in the systemic tolerance that follows oral immunization and that these may be related to numerical 8 198s Academic and/or functional differences in macrophages or accessory cellsis discussed. Resa Inc.
INTRODUCTION
Peyer’s patches (PP) are subepithelial lymphoid follicles found throughout the mammalian small intestine where they comprise a major component of the gut ’ This work was supported in part by USPHS Grant AM-3 1448. ‘To whomcorrespondence should be sent at: Cancer Center, The University of New Mexico, 900 Camino de Salud NE, Albuquerque, N. Mex. 87131. 41
42
BARR ET AL.
associated lymphoid tissue (GALT). Specialized epithelial cells (M cells) overlie their surface and are primarily responsible for the transport of luminal antigens to the underlying lymphoid cells. Antigens deposited within the gut lumen have been shown to elicit IgA antibodies at sites distal from the original application such as in the colostral, lacrimal, and salivary secretions of animal and man (1-3) and the salivery secretions in the rhesus monkey (4). It appears that antigen-sensitized cells, largely precommitted to IgA production, arise within the PP and then migrate to the widely separated mucosal surfaces that collectively make up the secretory immune system. Paradoxically, oral and systemic immunization with soluble antigens do not appear to lead to the production of antibody-secreting cells or an effective T-cellproliferative response in the PP themselves (5-7) although irregular responses have been reported in the rabbit (8) and with particulate antigens such as killed Streptococcus mutans or their cell walls which stimulate a T-cell response in PP of the mouse (7). Kagnoff and Campbell reported (9) the failure of PP cells to support the induction of a primary humoral response to sheep red blood cells (SRBC) or the induction of cytotoxicity to allogeneic cells. Both of these functions were restored by the addition of adherent peritoneal exudate cells (APEC) or 2-mercaptoethanol (2-ME) to the cultures. This suggested that the immunological shortcomings of PP cells might exist secondary to a decrease in the number or functional capacity of its accessory cells. This study was undertaken to further explore parameters of accessory cell function within the Peyer’s patch. MATERIALS
AND
METHODS
Animals. Six- to twelve-week-old CBA/J and B6A mice were obtained from the Jackson Laboratories, Bar Harbor, Maine. BlO.S5 (12R) mice and A.TFR-1 mice were bred in our own animal facilities and were kindly provided by Dr. Chella David. Media. Calcium- and magnesium-free Hanks’ balanced salt solution (HBSS) containing 25 mM Hepes buffer (GIBCO, Grand Island, N.Y.) and 10% fetal calf serum (FCS) was used as the standard preparation medium prior to final culture medium as described below. In culture RPM1 1640 (GIBCO) supplemented with 25 mM Hepes, 100 U/ml of penicillin, 100 pg/ml of streptomycin, 2 mM glutamine, 400 &ml of garamycin, 30 clg/ml of fungizone, and 2-mercaptoethanol (5 X lo-’ M) was employed. Heat-inactivated (56”C, 30 min) human serum (2.5% v/v concentration) was used to supplement cultures involving presentation to sensitized lymph node cells (SLNC); mixed lymphocyte reactions (MLR) were carried out in 5% FCS and presentation of antigens to T cells in 10% FCS. Preparation of cell suspensions.Spleen, mesenteric lymph nodes (MLN), PP, and inguinal and para-aortic lymph nodes (PLN) were removed from normal animals and placed in HBSS- 10 ( 10% serum). The tissues were gently crushed using a Teflon tissue homogenizer. The cell suspensions were passed through a nylon mesh to remove clumps and debris. The single-cell suspensions were collected and washed. Cytocentrifuge preparation of PP cells were analyzed for Ig-containing plasma cells by the fluorescent antibody technique and preparations containing >l% plasma cells were considered contaminated with lamina propria and discarded. Resident peritoneal exudate cells (PEC) were obtained by washing the peritoneal cavity with
ACCESSORY CELLS OF PEYER’S PATCHES
43
HBSS. Cell suspensions were adjusted to 10 X lo6 cells/ml before placing in petri dishes. Sensitized lymph node cells for the proliferative assays were obtained from the inguinal and para-aortic lymph nodes of mice given 100 pg of ovalbumin (Sigma Chemical Co., St. Louis, MO.) in H37Ra adjuvant (Difco, Detroit, Mich.) subcutaneously into the tail 8 days previously. Adherent cells. Aliquots (2.5 ml) of a cell suspension in HBSS-2.5 (2.5% FCS) were placed in 5-cm diameter plastic tissue culture petri dishes and the dishes were incubated for 60 min at 37°C. The dishes were washed with warm HBSS-2.5 and then 5 ml of HBSS-2.5 was added to each dish and incubation continued for an additional 60-l 20 min. The dishes were washed with 5 ml of warm HBSS-20 and then 5 ml of phosphate-buffered saline (PBS) containing 0.02% (w/v) tetrasodium EDTA was added to remove adherent cells. After 15 min on ice the dishes were washed three times with 10 ml of HBSS-2.5; washes were centrifuged at 4°C and the cell pellets resuspended in cold HBSS2.5. After two additional washings the cells were resuspended in the appropriate medium for testing. In experiments to determine the number of adherent cells, the cells recovered after EDTA treatment were counted and expressed as a percentage of the number applied to the plates. In an attempt to enrich for accessory and MLR-stimulating cells, suspensions of cells from spleen or PP were allowed to adhere to plastic tissue culture plates for 2 hr at 37°C; nonadherent cells were removed by gentle pipetting with cold medium. The plates were then covered with medium (RPM1 1640 supplemented with 25 mM Hepes, 100 U/ml penicillin, 100 &ml of streptomycin, 2 mM glutamine, 400 pg/ml of garamycin, and 30 pg/ml of fungizone with 5% FCS and 2-mercaptoethanol (5 X 10e5 M)) and returned to the 37°C CO2 incubator overnight. During this period many of the adherent cells either spontaneously eluted from the plastic surface or were easily dislodged by pipetting with cold medium. These cells were washed and the number of viable cells was determined by trypan blue exclusion. Presentation of antigen to SLNC. Method I. Cell preparations were incubated in triplicate in flat-bottomed tissue culture plates (Falcon 3042, Falcon, Oxnard, Calif.) at six different concentrations calculated to give 0.1 to 3.2 X 10’ adherent cells per well. After incubation for 60 min at 37°C nonadherent cells were removed by washing the cells three times in cold RPMI-2.5 (2.5% FCS). For pulsing 50 pg of ovalbumin in 200 ~1 of medium was added to each well and the plates were incubated for a further 2 hr or, in some experiments, 24 hr. After washing three times with warm medium, 4 X 10’ SLNC were added to each well and the plates incubated for 4 days with no further addition of antigen. Tritiated thymidine (1 &i) was added to each well 16 hr prior to harvesting the cultures. As controls, SLNC were added to duplicate cultures of nonpulsed adherent cells or pulsed adherent cells were incubated alone. For comparison the proliferation of SLNC was determined by incubation of 4 X 10’ SLNC with 50 pg and 100 Fg of ovalbumin in 200 ~1 of media. The presentation experiments were performed three times and the means were calculated. To test for the effects of H-2 or non-H-2 genes on antigen presentation, experiments were repeated in A.TFR-1 mice, which differ from CBA/J in background but have the same I region determinants (H-2k) and in B 1O.S (12R) mice (H-2”) which differ from CBA/J at both H-2 and non-H-2 loci. In additional experiments human gammaglobulin (HGG) (Sigma) or purified protein derivative (PPD) (Mycobacterium tuberculosis, Connaught Laboratories, Toronto, Canada) were used as antigens.
44
BARR ET AL.
Method ZZ. In a separate series of experiments, cells were allowed to adhere to 60-cm plastic petri dishes as outlined above. The cells were pulsed with ovalbumin (1 mg/ml) for 2 hr and then thoroughly washed in RPMI-2.5. Adherent cells were removed with 0.02% EDTA as described and added to 4 X lo5 SLNC in concentrations ranging from 0.1 X lo5 to 1.6 X lo5 (viable cells) in a total volume of 200 ~1 and incubated for 4 days at 37°C. Results in both methods were expressed as the difference in counts between pulsed adherent cells plus SLNC, and nonpulsed adherent cells plus SLNC. Enrichment for low-density populations. Populations of spleen and PP cells were enriched for low-density cells by equilibrium density centrifugation on dense bovine plasma albumin (BPA) solutions. Single-cell suspensions were spun at 10,OOOg in an SW39 swinging bucket rotor for 20 min at 4°C. Two cell fractions (pellicle and pellet) were then harvested with Pasteur pipets and washed in RPM1 prior to suspension in the final medium for counting. BPA solutions were prepared from bovine plasma albumin powder (fraction V, Reheis Chemical Co., Phoenix, Ariz.). Routinely 100 g of BPA were mixed with 186 ml of Ca’+- and Mg*+-free PBS, 20 ml 1 N NaOH, and 65 ml H20. The pH and density were adjusted if necessary to give a final solution with a pH 7.35-7.45 and a density of 1.080-l .082. The solution was subjected to millipore filtration and kept at 4°C prior to use. Morphologic examination for dendritic cells. Examination by phase-contrast microscopy was performed on specimens fixed for 5 min at room temperature in 2.5% glutaraldehyde buffered with PBS, pH 7.40, following adherence to glass coverslips for 2 hr at 37°C. Nonadherent cells were bound to coverslips previously coated with poly(L-lysine) (PLL; Type VII, Sigma; 25 &ml in PBS). The cells were allowed to adhere for 20 min, fixed with glutaraldehyde, and examined for the morphological feature of dendritic cells as described by Steinman and Cohn (10). Preparation of T cells. (a) Cloned T cells specific for poly(~-G1u6’, L-Ala3’, L-Tyr’O) (GAT) were kindly donated by Dr. Masao Kimoto and Dr. C. G. Fathman and were prepared as previously described (11). (b) Alloreactive long-term T-cell cultures were the result of long-term repetitive MLR with initially unfractionated responder cell populations. Responder cells from PP or spleen (3-6 X lo6 cells) were maintained in 20 ml of RPM1 supplemented with 25 mM Hepes, 100 U/ml penicillin, 100 &ml of streptomycin, 2 m&Z glutamine, 400 &ml of garamycin, and 30 pg/ml of fungizone with 5% FCS and 2-mercaptoethanol (5 X 10e5 M) in 75-cm tissue culture flasks (Coming Glass Works, Coming, N.Y.). Every lo- 14 days these cells were pulsed with 50-60 X lo6 irradiated (3300 R) spleen cells. T cells recognizing (B6A)Fi determinants were derived from CBA Peyer’s patch cells stimulated with irradiated (B6A)Fi spleen cells. T-cell lines recognizing CBA determinants were derived from B 10 spleen cells stimulated by irradiated CBA spleen cells. Mixed lymphocyte reactions. Alloreactive T cells ( 105) served as responder cells and variable numbers of irradiated (3300 R) spleen or Peyer’s patch cells as stimulators. Cells were cultured in round-bottom tissue culture plates (Linbro, Flow Laboratories, Hamden, Conn.) and stimulation was measured by uptake of tritiated thymidine in a 2-day proliferative assay. Primary MLR were performed as described by Murgita and Tomasi (12). Presentation of GAT to cloned T cells. Cloned GAT-reactive T cells ( 104) were mixed with 50 pg of GAT and variable numbers of irradiated cells from the spleen
ACCESSORY CELLS OF PEYER’S PATCHES
45
or PP of (B6A)Fl in a total volume of 200 ~1. Cells were cultured in flat-bottom tissue culture plates (Falcon 3042) for 3 days and stimulation was measured by uptake of tritiated thymidine. Accessory cell function for mitogen responses. Concanavalin A (Con A)- and phytohemagglutinin (PHA)-induced proliferative reactions were carried out as previously described ( 12). Accessory cells were depleted by passage over a nylonwool column. Syngeneic adherent peritoneal exudate cells (5%) were added in an attempt to restore the response. RESULTS Antigen presentation to sensitized lymph node cells. Adherent spleen cells (AX) pulsed with antigen (Method I-Material and Methods) were most efficient in presenting antigen to SLNC. Significant stimulation of SLNC was found with 2.5% ASC and the proliferative response increased in proportion to the number of cells added to a maximum at 40% ASC (Fig. 1). Addition of pulsed APEC also resulted in increased stimulation of SLNC to a maximum at lo-20% APEC; greater concentrations caused a progressive decrease in the response. Presentation of antigen by adherent lymph node cells (ALNC) appeared to be less effective than with ASC or APEC, and significant counts were not found until 10% adherent cells were added. At concentrations of 40% or greater, pulsed adherent mesenteric lymph node cells (AMLNC) showed significantly greater counts than ALNC (P < 0.05) although the gross cellular characteristics of mesenteric and peripheral lymph nodes are very similar (Table 1). Data very similar to those shown in Fig. 1 with ovalbumin (OVA) as antigen were also obtained with HGG and PPD. As shown in Fig. 1, pulsed
3 hour
adherent
cells
1~o14x1051
FIG. I. Presentation of ovalbumin to sensitized lymph node cells (SLNC) by adherent cells from Peyer’s patches (APPC, l ), spleen (ASC, A), mesenteric lymph nodes (AMLNC, A), peripheral lymph nodes (ALNC, 0), and peritoneal exudate (APEC, 0). Adherent cells were pulsed with 200 JL~ovalbumin for 2 hr at 37’C in plates and washed, SLNC were added at 4 X 10’ per well and cuhured for 4 days. n = SLNC control, 50 ccgovalbumin added to 4 X lo5 SLNC. Means of three experiments. Control values (nonpuked adherent cells plus SLNC) have been subtracted.
46
BARR ET AL. TABLE 1 Cellular Characteristics of Peyer’s Patches and Mesenteric Node Cells from CBA/J Mice Compared with Other Lymphoid Tissues Peyer’s patches
B cells (anti-kappa positive) T cells (anti-Thy 1.2 positive) Phagocytic cells Adherent cells Esterase positive cells
49 31 5.2 3.8 13.9
+4 f6 f 1.7 31 1.8* + 1.4**
Spleen 44 ?I 41 +9 11.1 + 2.4 13.1 f 2.1 12.0 k 2.6
Peritoneal exudate 22 17 51.8 63.1 48.6
+ + f f +
2 2 10.5 17 6.8
Lymph nodes 17 68 5.7 6.4 4.5
*5 +9 + 1.8 + 1.9 + 1.4
Mesenteric lymph nodes 16 70 5.6 7.6 4.6
+3 +5 f 1.9 + 1.5 + 0.8
Note. Mean percentage + SD of between 5 and 10 experiments for each determination. Tissues from normal I-week-old mice. Percentage B and T cells by fluorescent staining with anti-kappa or anti-Thy 1.2 antisera. Percentage phagocytic cells by latex ( 1.1 pm) ingestion. Adherent cells = percentage adhered and recovered from petri dishes after 2 hr (see Methods and Materials). * P < 0.02 compared with lymph nodes or spleen. ** P < 0.01 compared with lymph nodes.
adherent Peyer’s patch cells (APPC) did not elicit stimulation of SLNC at any of the concentrations tested between 2.5 and 80% of the number of SLNC. Lengthening the period of antigen pulsing to 24 hr did not improve presentation, Antigen-pulsed APPC were also unable to present HGG or PPD to SLNC. Ten percent APPC added to APEC or ASC did not significantly affect their ability to present antigen, suggesting that suppressor cells were not responsible for the deficient presentation by APPC. When adherent cells were pulsed in petri dishes, removed with EDTA, and transferred to plates (Method II-Material and Methods) in contrast to being allowed to adhere to plates, pulsed, washed and SLNC added, similar results were obtained with the exception that at the highest concentration of APPC examined (40%) minimal presentation of ovalbumin was detected. However, this was significantly less with APPC than with AMLNC or ASC as shown in Fig. 2. It was noted that control values (nonpulsed adherent cells + SLNC) were higher for all cell types examined than with Method I, and increased in proportion to the number of adherent cells added. Viabilities of the adherent cells for each of the cell preparations were similar. The lack of presentation by APPC did not appear to be restricted to CBA/J mice since identical results were found with mice of the A.TFR-1 and BIO.S (12R) strains. This suggested that the failure of APPC to present antigen was probably not restricted by H-2 or non-H-2 loci. Antigen presentation to cloned T cells. Both unfractionated PP and spleen cells from (B6A)Fi mice were shown to be capable of presenting antigen to cloned GATreactive T cells. Density centrifugation on BPA (p = 1.080) separates spleen and PP cells into two fractions: a low-density population (“floaters”) and a high density fraction (“sinkers”). The antigen-presenting cell in the spleen was significantly enriched in the floaters (containing 5-15% of total cell population) as shown in Fig. 3. Similar findings were found with PP cells (see Fig. 4) where low-density cells made up lo-30% of the total. Repeated experiments showed PP floaters to be less
ACCESSORY CELLS OF PEYER’S PATCHES
3 hours
adherent I cells
47
cells % of 4 x lo5 translered 1
FIG. 2. Presentation of ovalbumin to SLNC by adherent cells from Peyer’s patches (APPC, l ), mesenteric lymph nodes (AMLNC, A) and spleen (ASC, Cl). Adherent cells were pulsed with 1 mg/ml ovalbumin for 2 hr at 37°C in petri dishes, removed with 0.02% EDTA, washed, and added to SLNC. Control values (nonpulsed adherent cells plus SLNC) have been subtracted. Means of four experiments.
capable of presenting antigen to the cloned cells than spleen cells at the same cell concentrations. MLR stimulation by Peyers patch cells. Whole Peyer’s patch cells provide good stimulation in the allogeneic MLR. However, as shown in Table 2, the adherent cell population from PP, unlike those from the spleen, did not act as effective stimulators. The number of adherent PP cells employed in Table 2 gave optimal stimulation. The stimulating capacity of the nonadherent population was similar to that of the whole PP. In other experiments, long-term T-cell cultures enriched for alloreactive cells by repetitive MLR were used as responder cells. Stimulation of lo5 of these T cells by as few as 2 X lo4 irradiated spleen or PP cells provided an excellent response when 2-day MLR were assessed. Separation of spleen and PP cells into high- and low-density cells resulted in enrichment for the stimulating cell
Whole
Floaters
Sinkers
FIG. 3. Antigen presentation by irradiated (3300 R) syngeneic spleen cells (0.25 X 106) to (B6A)Fr cloned GAT-specific T cells (10” cells). GAT (50 &ml) was added and [‘Hlthymidine incorporation was measured after 3 days of culture. CPM on whole spleen, 13,031 + 813; floaters, 25,631 + 1682; sinkers, 1003 rt 273. Low-density (floaters) and highdensity (sinkers) cells separated on BPA (see Methods and Materials).
48
BARR ET AL.
Whole
FlC&0nl
Sinkers
FIG. 4. Antigen presentation by irradiated (3300 R) syngeneic Peyer’s patch cells (1 X lo6 cells) to (B6A)Fr cloned GAT-specific T cells (IO4 cells). GAT (200 &ml) was added and [3H]thymidine incorporation was measured after 2 days of culture. CPM on whole Peyer’s patches, 2309 ? 155; tloatem, 3,321 + 82; sinkers, 1,517 + 109. Floaters and sinkers separated on BPA (see Methods and Materials).
in the low-density fraction (Fig. 5). In repeated experiments PP floaters provided good stimulation though consistently less than spleen cells (Fig. 5). PP and spleen cells differed when low-density cells were separated into adherent and nonadherent fractions. In spleen cell populations the MLR-stimulating cell was present primarily in the adherent fraction while in the PP it was recovered only in the nonadherent population (Fig. 6). Experiments using T cells derived from BlO mice recognizing CBA determinants yielded similar results to those outlined above. Accessory cell function of Peyers patch cells in mitogen responses. In view of the well known requirement for accessory cells in Con A- and PHA-induced proliferation we examined PP for accessory cell function in these responses. As shown in Table 3, Con A and PHA induced reasonable proliferative responses which were nearly abrogated by passage over a nylon-wool column. Addition of irradiated (3300 R) APEC (5%) to the nylon-nonadherent fraction from PP restored their reactivity. Dendritic cells in Peyer’s patches. Phase contrast examination of low-density adherent cells (LODAC) from PP revealed significant differences when compared to the spleen. Spleen LODAC preparations were consistently characterized by 40% or more cells with typical morphological features of dendritic cells (DC) as described by Steinman and Cohn (10). PP LODAC populations infrequently contained small TABLE 2 Mixed Lymphocyte Reaction
Source of stimulator
Responder (CBA LNC)
MLR
Whole spleen Whole PP *ASC *APPC
536 605 1021 444
139,590 18,180 41,419 951
Stimulator (BALBfc) 1110 116 1586 73
SI 85 25 16 1.8
Note. cpm = Counts per minute of [‘Hlthymidine incorporated after 4day culture in 2.5% human AB serum. Results with 5 X 10e5 M 2-ME added were similar. ASC = adherent spleen cells. APPC = adherent Peyer’s patch (PP) cells. Number of cells per well: CBA LNC, 5 X 10’; BALB/c whole spleen, 1 X 106; whole PP, 1 X lob, ASC, 1 X 10s; and APPC, 1 X 10’. * Adherence carried out for 2 hours at 37°C in plastic petri dishes.
ACCESSORY CELLS OF PEYER’S PATCHES Psyrr’a
WhOI. Cbal.”
49
Patch
skew,
FIG. 5. Mixed lymphocyte reactions (MLR) comparing irradiated (3300 R) stimulators from CBA spleen and Peyer’s patch (PP) populations (LOS/well) separated on BPA gradients. Responding cells ( 105/ well) were obtained from long-term T-cell cultures derived from Bl0.A spleen and PP cells repetitively stimulated with CBA spleen cells in vitro. Whole spleen SI, 41; spleen floaters SI, 158; spleen sinkers SI, 26. Whole PP SI, 49; PP floaters SI, 113; PP sinkers SI, 15.
numbers (~5%) of cells resembling splenic DC. Other large, spread out mononuclear cells of indeterminate lineage were also seen but the surface characteristics and functional significance of these cells are unknown. The nonadherent low-density cells from PP were attached to glass coverslips previously treated with poly(L-lysine). However, these cells did not demonstrate the typical morphology exhibited by splenic DC after adherence to PLL-coated coverslips. Furthermore, observation of the nonadherent population in culture for up to 5 days did not reveal development of the cytologic features similar to those of splenic dendritic cells. We find no evidence that enzymatic dissociation with dispase releases cells with dendritic morphology. DISCUSSION Adherent cells from Peyer’s patches are deficient in their ability to present soluble antigens (OVA, HGG, and PPD) to sensitized lymph node cells when compared to adherent cells from peripheral lymph nodes, spleen, or peritoneal exudate. The number of adherent cells in Peyer’s patches is significantly less than in spleen or lymph nodes (Table 1) although we cannot be sure that this is not attributable to technical problems such as the manner in which the patches are collected. Perhaps macrophages are concentrated close to the intestinal lumen. In this regard we have not been able to improve yields of adherent cells by using collagenase (Type II) or
*mur.nt
Na.dkmu
Aummt
Nnum.r.nt
FIG.6. Mixed lymphocyte reactions (MLR) with irradiated (3300 R) low-density spleen and Peyer’s patch (PP) cells comparing their adherent and nonadherent fractions. Responders were long-term T-cell cultures ( 10s/well). Stimulators Born the spleen (2 X 10’) were from (B6A)Fi mice and responding T cells were CBA in origin. PP stimulators (IS X 10’) were from CBA mice with T-cell responders from BIO.A. Spleen adherent SI, 228; spleen nonadherent SI, 65. PP adherent SI, 6; PP nonadherent SI, 174.
50
BARR ET AL. TABLE 3 Effect of Removal of Adherent Cells (AC) from Peyer’s Patches (PP) on the Proliferative Response to Concanavalin A (Con A) and Phytohemagglutinin (PHA) AC-depleted Whole Peyer’s patch cells
Medium Con A (4 pg./well) PHA (2 &vell)
2,199 95,841 37,616
f 1 IO* f 5693 2 4167
ACdepleted
PP
185 f 41 2416 f 326 2053 + 876
* Mean cpm f SE of triplicate cultures. PEC = normal were incubated at 4 X 10’ cells per well and [‘Hlthymidine
PP with addition
1%
2.5%
572 f 53 15,993 f 6587 12,155 k 2148
816 f 88 115,274 f 3947 22,870 + 7019
syngeneic peritoneal exudate incorporation was measued
of PEC 5% 1,384 105,380 41,412
k 196 f 4058 f 219
cells added (W of 4 X 10’). Cells after 3d of culture.
trypsin plus collagenase, nor do we find significant differences with the manner in which the patches are obtained (deep cuts versus more superficial) unless the lamina propria is included. Contamination with lamina propria is a key issue in all studies on PP and can be ascertained by determining of the number of Ig-containing plasma cells on cytocentrifuge preparations. Normally, PP lack plasma cells (except for scattered Ig-containing cells localized to the dome) and preparations containing > 1% should in our view be considered contaminated and discarded. Defective antigen presentation cannot be attributed to cell numbers alone since in the antigen presentation experiments equivalent numbers of adherent cells were used from each source. It also seems unlikely that APPC contain a population of suppressor cells. The addition of APPC to SLNC did not suppress the response and mixing APPC with APEC or ASC did not alter their ability to present antigen. It has not been excluded that the inability of APPC to present antigen may apply only to certain antigens. This possibility is raised by the observation that the T-cell proliferative response of PP cells following intragastric administration of soluble (tolerogenic) protein antigens such as OVA is negligible whereas significant stimulation occurs after administration of a particulate antigen such as S. mutans (7). Work is in progress attempting to elucidate this point. We have found that flotation of whole spleen cells on BPA (p = 1.080) produces a heterogeneous low-density population (5-15% of the total spleen cells) which includes macrophages, B and T cells, and dendritic cells as first reported by Steinman and Cohn (10). This population is enriched in antigen-presenting cells. Flotation of whole Peyer’s patch cells produces a low-density fraction which represents lo-30% of the total population and is also enriched in antigen-presenting capabilities. However, repeated experiments have shown that equivalent numbers of low-density cells from the Peyer’s patches are less efficient in their ability to present antigen to cloned T cells than those derived from spleen. Similarly, whole PP are deficient compared to whole spleen in antigen presentation in this system. In both the spleen and PP the MLR-stimulating cell was found predominantly within the low-density population. However, while adherent floaters from the spleen are potent stimulators of the MLR, this activity is found exclusively in the nonadherent low-density population of PP cells. In this regard rat lymph node dendritic cells are found in the low-density nonadherent fraction ( 13). The low-density nonadherent population from the PP does not, however, contain cells with the typical morphological characteristics of dendritic cells when examined in suspension or after adherence to
ACCESSORY CELLS OF PEYER’S PATCHES
51
poly(L-lysine)-treated glass coverslips; nor do the cells assume those characteristics when kept in cultures for up to 5 days. We have not been able to identify dendritic cells in cell populations obtained with dispase but have shown accessory cell function (for mitogens) in such preparations as reported by others (14). It is possible that dendritic cells are present in low but functionally significant numbers within the PP and/or have different morphological characteristics than those in the spleen, perhaps related to technical factors involved in obtaining cell preparations from these sources. Alternatively, a different type of cell or cells in PP may be responsible for antigen presentation. B cells have been reported (15) to have antigen-presenting capacity and we are currently investigating whether PP B cells may present certain antigens. Recently we (Yem and Tomasi, unpublished data) have obtained actively presenting preparations using the fluorescent activated cell sorter (Becton-Dickinson, FACS III) which was gated to sort and collect large cells and contained >85% surface Ig-positive cells. These preparations, enriched for B blasts, are currently being analyzed and sorted for macrophages (assaying with MAC-l, -2, and -3 reagents) and examined for dendritic cells by cytotoxicity using an antidendritic cell-specific antibody (anti-DC- 1, courtesy of Dr. S. Steinman). Our previous studies (16) have defined the expression of Ia antigens on PP cells and determined that a large proportion of these cells (70-85%) expressed both I-A and I-E including most PP B cells and 20% of T cells. We could not however, determine whether the small number of phagocytic cells present in PP preparations possess Ia antigens. Because of the importance of Ia in antigen presentation further studies of PP are underway involving Ia expression and the affect of anti-Ia on isolated actively presenting populations. The relative lack of antigen presentation by APPC is in agreement with a previous report (9) that PP lack adherent accessory cells needed for a primary in vitro antibody response and were deficient in antibody production after oral or parenteral immunization (5,6). However, the accessory cells necessary for antigen presentation, as well as for the Con A and PHA response, appear to be present in whole and nonadherent PP and when removed they can be replaced by irradiated (3300 R) APEC. In analogy with other systems it has been suggested (17) that the deficient antigen presentation and presumably the failure to generate a large number of proliferating T helper cells could be related to the apparent ease with which suppressor cells are developed. This would have implications pertinent to a number of central issues in mucosal immunity. For example, the generation of suppressor cells for IgG and IgM and their dissemination to peripheral lymphoid tissues (18, 19-22) occurs concomitantly with the development of antigen-specific systemic tolerance and a (local) secretory IgA response (7). This is associated with the appearance of IgA-specific helper cells in PP and MLN (22). Speculatively, antigen presentation in the gut may occur by unique mechanisms which generate T helper cells for IgA and T suppressor cells for IgG and IgM (21). The possibility that gut macrophages have some degree of antigen and/or isotype specificity in their cell interactions as previously suggested (23) has not been excluded. REFERENCES I. We&-Canington, Proc. Natl.
P., Roux, M. E., McWilliams,
Acad. Sci. 75, 2928,
1978.
M., Phillips-Quagliata,
J. M, and Lamm, M. E.,
52
BARR
ET
AL.
2. Mestecky, J., McGhee, J. R., Arnold, R. R., Michalek, S. M., Prince, S. J., and Babb, J. L., .I. Clin. Znvesf.61, 731, 1978. 3. Michalek, S. M., McGhee, J. R., and Babb, J. L., Znfect.Zmmun. 19, 217, 1978. 4. Challacombe, S. J., and Lehner, T., Arch. Oral Biol. 24, 917, 1980. 5. Bienenstcck, J., and Dolezel, J., J. Zmmunol. 106, 938, 1971. 6. Henry, C., Faulk, W. P., Kuhn, Z., Yoffey, J. M., and Fudenberg, H. H., J. Exp. Med. 131, 1200, 1970. 7. Challacombe, S., and Tomasi, T. B., J. Exp. Med. 152, 1459, 1980. 8. Clancy, R., and Pucci, A., AJEBAK 56, 337, 1978. 9. Kagnoff, M. F., and Campbell, S., J. Exp. Med. 139, 398, 1974. 10. Steinman, R. M., and Cohn, Z. A., J. Exp. Med. 137, 1142, 1973. 11. Kimoto, M., and Fathman, C. G., J. Exp. Med. 152, 759, 1980. 12. Mu&a, R. A., and Tomasi, T. B., J. Exp. Med. 141,440, 1975. 13. Klinkert, W. E., LaBadie, J. H., O’Brien, J. P., Beyer, C. F., and Bowers, W. E., Proc. Nat/. Acad. Sci. 77, 5414, 1980. 14. Spalding, D. M., Koopman, W. J., and McGhee, J. R., Ann. N.Y. Acad. Sci. 409, 880, 1983. 15. Chestnut, R. W., and Grey, H. M., J. Zmmunol. 126, 1075, 1981. 16. Krco, C. J., Challacombe, S. J., Lafuse, W. P., David, C. S., and Tomasi, T. B., Cell. Zmmunol. 57, 420, 1981. 17. Pierres, M., and Germain, R. N., J. Immunol. 121, 1306, 1978. 18. Mattingly, J. A., and Waksman, B. H., J. Zmmunol. 121, 1878, 1978. 19. Husband, A. J., and Gowans, J. L., .I Exp. Med. 148, 1146, 1978. 20. Hanson, D. G., Vas, N. M., Maia, Z. C. S., Hombrook, M. M., Lynd, J. M., and Roy, C. A., Arch. Allergy Appl. Zmmunol. 55, 526, 1977. 21. Asherson, G. L., Zembala, M., Perera, M. A. C., Mayhew, B., and Thomas, W. R., Cell. Zmmunol. 33, 145, 1977. 22. Richman, L. K., Graeff, A. S., Yarchan, R., and Strober, W., J. Zmmunol. 126, 2079, 1981. 23. Tomasi, T. B., Transplantation 29, 353, 1980.