Journal of Neuroimmunology, 39 (1992) 101-108

101

© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-5728/92/$05.00 JNI 02203

Murine mucosal T cells have VIP receptors functionally distinct from those on intestinal epithelial cells A r t h u r M. Blum, R a n j i t M a t h e w , G e o r g e A. Cook, A h m e d Metwali, R o b e r t F e l m a n and Joel V. W e i n s t o c k Dic'ision of Gastroenterology-Hepatology, Department of Internal Medicine, University of Iowa College of Medicine, Iowa City, IA, USA (Received 10 January 1992) (Revised, received 3 March 1992) (Accepted 3 March 1992)

Key words. Immunoregulation; Neuropeptide; Interleukin-5

Summary Reports suggest that vasoactive intestinal peptide (VIP) binds to lymphocytes and modulates immune responses. The intestines are richly innervated with VIP-producing nerves. Thus, VIP from nerves or other sources may participate in mucosal immunoregulation. To explore this hypothesis further, murine intestinal mucosa inflammatory cells were scrutinized for functional VIP receptors. An [125I]V1P competitive binding assay characterized VIP receptors. Unfractionated lamina propria inflammatory cells bound [125I]VIP specifically. This binding was abrogated by T cell depletion. The VIP receptor on lamina propria T cells was of a single class with a K d of 9.08 × 10 - 9 M. It bound PHI and other peptide analogs poorly. The intestinal epithelial cell had a high-affinity VIP receptor ( K d 4.17 × 10 -1° M) that bound one VIP analog with moderate affinity. Both VIP and ConA stimulated mucosal inflammatory cells to release interleukin-5 (IL-5). Mucosal inflammatory ceils depleted of T cells did not release IL-5 in response to VIP or ConA. It is concluded that: (1) some murine mucosal T lymphocytes have VIP receptors that may be distinct from those displayed on mucosal epithelial cells; (2) VIP affects mucosal T lymphocyte function.

Introduction Vasoactive intestinal peptide (VIP) is a neurotransmitter produced by the central and peripheral nervous systems, and by inflammatory cells (Said, 1986; Weinstock and Blum, 1990). In the intestine, VIP affects smooth muscle contraction

Correspondence to: J.V. Weinstock, GastroenterologyHepatology Division, 4607 JCP Department of Internal Medicine, University of Iowa College of Medicine, Iowa City, IA 52242, USA.

and stimulates epithelial cells to secrete water and electrolytes. Specific VIP receptors mediate these functions (Luis et al., 1988; Robberecht et al., 1990). Studies have implicated VIP as an immunoregulatory molecule (Weinstock, 1991). VIP receptors are present on human peripheral blood and mouse splenic lymphocytes, and on several lymphocyte cell lines (O'Dorisio, 1987). The aim of this study was to determine whether murine mucosal T cells express VIP receptors. It was found that mucosal T lymphocytes have VIP receptors that may be distinct from those dis-

102 played on intestinal epithelial ceils. Furthermore, VIP stimulates mucosal T cells to release IL-5, an interleukin potentially important for mucosal B cell and eosinophil development.

Materials and methods

Intestinal cell isolation The ileum and distal jejunum of normal 8week-old CBA mice (Cumberland Laboratories) were the source of epithelial cells and lamina propria inflammatory cells. About 25 cm of the distal small bowel was flushed free of feces with 10 ml RPMI (RPMI 1640 medium) injected through a 10-ml syringe. Next, macroscopically visible Peyer's patches were excised with a dissection scissors. The intestine was opened longitudinally and cut into 5-cm segments. The intestinal segments from about five mice were put into a 50-ml siliconized Erlenmeyer flask containing 30 ml of Ca2+/Mg2+ free HBSS and 0.5 mM EDTA. The segments were agitated in a shaker waterbath for 30 rain at 37°C. Then, the supernatant containing epithelial cells was poured off and saved. Next, the intestinal segments were washed three times with sequential 30-ml aliquots of CaZ+/Mg2+-free HBSS containing 0.5 mM EDTA. These supernatants were discarded. To remove residual epithelial cells, the tissue was processed twice more as described above. The gut pieces were then agitated in the siliconized Erlenmeyer flask for 30 rain at 37°C in 20 ml RPMI containing 5 m g / m l collagenase (Type IV, Sigma). The liberated lamina propria cells were washed through coarse gauze with 20 ml RPMI to remove debris and the cells were pelleted at 250 × g. The cells were then washed with three consecutive 10-ml aliquots of RPMI by repeat centrifugation and re-suspension. The epithelial and lamina propria cells were further enriched on a 3 0 % / 7 0 % discontinuous Percoll gradient. Percoll was diluted with PBS. 4 ml 30% Percoll was layered on top of 4 ml 70% Percoll in a 15-ml centrifuge tube. The cells (3-5 x 107), in 4 ml RPMI, were layered on top of the 30% Percoll. The gradients were centrifuged at 1000 × g for 20 min at room temperature. Epithelial cells collected at the R P M I / 3 0 % Percoll

interface, while lamina propria inflammatory cells migrated to the 3 0 % / 7 0 % Percoll interface. Both cell layers were collected and washed three times with RPMI. The approximate cell yields per mouse were 3 × 107 and 1-1.5 × 107 for epithelial and inflammatory cells, respectively. Viability of both cell populations exceeded 85% by eosin Y exclusion. The lamina propria cells were about 35% T lymphocytes and 20% B cells as determined with fluorescein-conjugated Thy-l.2 monoclonal antibody and fluorescein-conjugated F(ab')2 goat anti-mouse Ig, respectively (Accurate Chemical). They also contained about 20% polymorphonuclear cells. There were no apparent epithelial cells. The epithelial cell preparations contained < 3% T or B lymphocytes, and no polymorphonuclear ceils. In some experiments, T cells were eliminated by two consecutive treatments with anti-Thy 1.2 and complement (Accurate Chemical Corp., Westbury, NY). Cells were handled as suggested by the manufacturer. Fluorescein-conjugated Thy-l.2 antiserum documented efficacy of T cell lysis. Control cells were treated with normal mouse serum (NMS) prior to complement treatment. Spleens were dispersed to obtain single-cell preparations. Spleens were mechanically disrupted and pressed through a no. 100 stainless steel mesh. Splenic T ceils were isolated by gentle elution through a nylon wool column. The nylon (Travenol, Deerfield, IL) was boiled in four changes of distilled water, packed into a 10-ml syringe barrel to the 2.5-ml mark, and rinsed with RPMI. Spleen cells (5 x 107) in 1 ml RPMI containing 10% fetal calf serum were placed on the column and incubated for 45 min at 37°C. The cells were then gently eluted from the column with 5 ml RPMI. The yield was greater than 90% T cells as determined by fluorescein-conjugated Thy-l.2 monoclonal antibody.

Cell permeabilization Cells were permeabilized with digitonin (Sigma) to enhance specific binding (Brooker and Pedone, 1986). The digitonin was prepared by dissolving 6 g of digitonin in 60 ml of boiling water. The solution was cooled overnight at room temperature, filtered to remove precipitate and

103

lyophillized. Cells (2 X 106) were incubated in 10 ml RPMI containing 37.5/zg digitonin for 15 min at 4°C, washed twice in RPMI and resuspended in the appropriate solution for assay of radioligand binding.

VIP competitive binding assay The binding assay was performed in quadruplicate on suspensions of freshly isolated, permeabilized cells. The binding medium was RPMI containing 5% bovine serum albumin (Sigma) and 0.1% sodium azide. Cells (up to 5 x 10 6) in 200 /xl of binding medium containing [lzSI]VIP (10-u M) (New England Nuclear, Boston, MA) (sp. act. 2200 Ci/mmol) were incubated 30 min in polystyrene 12 × 75 mm round-bottom tubes at room temperature in the presence or absence of various concentrations of unlabelled, synthetic VIP (Sigma). We chose a 30-min incubation time because binding equilibrium was achieved after 20 min with all cell types and remained stable for at least 30 rain. In some experiments, unlabelled VIP was replaced with various hormone analogs that included PHI, [Ac-Tyr l, D-Phe2]-GRF (analog 1), VIP receptor binding inhibitor L-8-K (analog 2) and [P-4C1-D-Phe 6, LeulV]-vIP (analog 3) (all from Peninsula Laboratories, Belmont CA). The cells were then washed twice at 4°C with 400 /zl of the binding medium by centrifugation at 3 0 0 x g . Using a multichannel cell harvester (Brandel, Gaithersburg, MD), the cells were then removed from their test tubes, deposited onto glass fiber filters and washed with five 1.5-ml aliquots of distilled water. Cell-associated radioactivity was measured in a gamma counter. Specific binding was defined as the difference between cell-bound radioactivity in the p~esence or absence of 10 -5 M unlabelled VIP. Total binding comprised < 10% of the total counts added to each tube. Non-specific binding was usually less than 20% of the total binding. Binding data were analysed by computer using the LIGAND program.

bated for 24 h at 37°C in 5% C O 2. Some cultures contained ConA or VIP. An IL-5 ELISA measured IL-5 concentration in cell supernatants, previously stored at -70°C. Samples were assayed using two rat anti-IL-5 mAbs in a two-site sandwich ELISA (Schumacher et al., 1988). The anti-IL-5 mAbs were purified by ammonium precipitation from supernatants of antibody secreting hybridoma clones (TRFK 4 and 5, a generous gift of Dr. Robert Coffman, DNA research Institute, Palo Alto, CA). Microtiter plates were coated with 50/xl of 1/xg/ml coating antibody (TRFK 5) in PBS containing 0.5 /~l/ml Tween 20 (PBS-T), and incubated at 4°C overnight. Wells were then blocked by the addition of 150 /zl of 10% FCS in PBS with incubation at 37°C for 30 min. An IL-5-containing supernatant from ConA-activated spleen cells from schistosomiasis-infected mice was standardized against rIL-5. Sample and standard dilutions were made in RPMI containing 10% FCS in a separate 96-well flat-bottom microtiter plate, and 50 /xl volumes were transferred to the ELISA plates which had been washed three times in PBS-T. Samples were incubated in the assay plates for 1 h at 37°C. After washing three times in PBS-T, each well received 50 /xl of antibody (TRFK 4)-biotin conjugate at 0.5 /xg/ml in 1% BSA/PBS-T. Plates were incubated at room temperature for 1 h followed by washing three times in PBS-T. Streptavidin-horseradish peroxidase conjugate (75 ~1) (Zymed, San Francisco, CA) was added at 1 /zg/ml in 1% BSA/PBS-T and incubated at room temperature for 1 h. Plates were then washed six times in fresh PBS-T, and 100/xl of substrate (2.2'-azino(3-ethylbenzthiazoline sulfonic acid) (Zymed) at 1 mg/ml in 44 mM NazHPO4, 28 mM citric acid, and 0.003% H202) were added. The colored product was measured at a wavelength of 405 nm using a microplate reader.

Results

Measurement of IL-5 release Cells were suspended at 2 × 106 per ml in RPMI containing 10% fetal calf serum, 2 mM glutamine, 100 U / m l penicillin, 100/zg/ml streptomycin and 25 mM Hepes. The cells were incu-

Binding of VIP to intestinal cells Radiolabelled VIP specifically bound to intestinal epithelial cells and to inflammatory cells isolated from the lamina propria or spleen (Table

104 TABLE 1 [125I]VIP M A X I M U M SPECIFIC B I N D I N G Binding experiments were done as described in the Materials and methods section. Cells were incubated 30 rain with [12sl]VIP in the presence or absence of 10 -6 M unlabelled VIP. The maximum specific binding data are means + SD for five separate experiments. Non-specific binding never exceeded 20% of total binding. Each tube received about 12000

cpm total counts Specific binding ( c p m / 5 × 106 cells) Epithelial cells Lamina propria cells Splenic T cells

803 _+249 389+ 70 950_+216

1). Binding equilibrium was achieved after 20 min with all cell types and remained stable for at least 30 rain. With regard to the epithelial cells, unlabelied VIP displaced the radioligand with a K d of 4.17 x 10 -1° M. Analysis of the competitive binding data suggested the presence of a single class of high-affinity VIP receptor. VIP bound to lamina propria inflammatory cells with a K~ 9.08 X 10 - 9 M. The profile of this binding curve suggested the presence of a single class of intermediate affinity receptor (Fig. 1). Normal splenic T cells bound V1P with affinity similar to that of the intestinal inflammatory cells ( K d of 1.07 x 10 -8 M). Normal spleen ceils were obtained by mechanically disrupting spleens in the absence of E D T A or collagenase. Yet, treatment of spleen cells with E D T A a n d / o r collagenase did not alter the VIP binding constant. Similarly, the K d of epithelial cells, liberated by E D T A exposure, was not changed by collagenase treatment. Thus, the various cell isolation techniques used did not appear to account for the variance between the binding constants of the epithelial and lamina propria or splenic T cells.

antagonize VIP binding in other systems. Analog 1 and 2, used at concentrations up to 10 -6 M, did not displace radioligand. Analog 3 did weakly oppose radioligand binding. However, it showed appreciably higher affinity for the VIP receptors displayed on the epithelial cells than for those expressed on the inflammatory ceils (Fig. 2). Splenic T cells bound these peptide analogs in a fashion similar to that of lamina propria inflammatory cells (data not shown).

Identification of T cells as one VIP-b&ding, inflammatory cell subpopulation The dispersed lamina propria cells contained lymphocytes and other inflammatory cell types. Our studies using splenic T cells suggested that some mucosal T lymphocytes would have VIP receptors. Treatment of dispersed lamina propria cells with anti-Thy 1.2 decreased the T lymphocyte content from 31.5 to 6.5%. Depletion of T cells substantially reduced VIP-specific binding

~-~oC100 ° o

or,tz° ~.0 z_

=_ '>_

Next, we examined the capacity of peptide analogs, other than VIP, to displace radiolabeUed VIP from its receptor. PHI, a naturally occurring VIP analog, was a weak competitor for the VIP receptor displayed on either the epithelial cells or intestinal inflammatory cells ( K d 5 x 10 - 7 M). Also tested were synthetic analogs reported to

~

J

SOl

SO

iiiiiiii

y

//

MINA

PROPRIA

CELLS

40

~:

l 0

Specificity of VIP b&ding

L EPITHELIACEL L

1 !1

/ t /

I 10

1 9 VIP

I 8

I 7

I IS

(-logMJ

Fig. 1. Competitive inhibition of [12511VIP binding to (A) intestinal epithelial cells or (B) lamina propria inflammatory cells. Cells were incubated 30 min at room temperature with radiolabelled VIP in the absence or presence of various concentrations of unlabelled VIP. All values are the m e a n s of triplicate determinations from three separate experiments + SE.

105

B

A Ioo o c

g ~

ao

m ~

u.

6o

G.

/

L_

~

J

2o

m

¥ z J

1

I:z:.-~+-~

ANALOG I

i/

I 9

I 0

I 7

2 I 6

111

VIP (-log M}

I f0

/

I

a

I

?

I

$

VIP (-log MI

Fig. 2. Competitive inhibition of [t25I]VIP binding by various peptides. (A) Intestinal epithelial or (B) lamina propria lymphocytes were incubated 30 min at room temperature with radiolabelled VIP in the absence or presence of various concentrations of unlabelled peptides. Values are the means of quadruplicate determination from three separate experiments + SE.

20o

A

20O

B

A

loo

.~ lOO u')

t~

0

Cells

Cells + Con h (5 plJt/ml)

Cells + VIP (10"6M)

0

NMS+C

NM$+C Thy 1.2+C + VIP (10"SM)

~ 1.2+C + VIP (10"6M)

Fig. 3. IL-5 in the culture supernatants of mucosal inflammatory cells. (A) Inflammatory cells (2 × 106) were cultured in vitro in 200 /zl of medium for 24 h in the presence or absence of ConA or VIP. (B) In some experiments, inflammatory cells were first treated with either anti-Thy 1.2 or normal mouse serum (NMS) and complement (C) to lyse T cells, and then cultured as described above. An ELISA measured the IL-5 released into the culture supernatants. Data are means + SE of multiple determinations from three separate experiments.

106 TABLE 2

Discussion

T C E L L D E P L E T I O N A T T E N U A T E S VIP B I N D I N G T O L A M I N A P R O P R I A CELLS Anti-Thy 1.2+C ~ Tcells(%) b Specific binding ( c p m / 5 × 106 cells) c

6.5_+ 2.1

160

_+11

NMS + C 31.5_+ 7.8

517

+36

P