- Email: [email protected]
Characterization of autoantibodies to vasoactive intestinal peptide in asthma
Journal of Neuroimmunology, 23 (1989) 133-142
133
Elsevier JNI 00779
Characterization of autoantibodies to vasoactive intestinal peptide in asthma Sudhir Paul 1, Sami I. Said 3, Austin B. T h o m p s o n 2, D e a n n a J. Voile 1, Devendra K. Agrawal Hussein F o d a 3 and Santiago de la R o c h a 5
4,
Departments of 1 Pharmacology and 2 Medicine, University of Nebraska Medical Center, Omaha, NE 68105, U.S.A., 3 Department of Medicine, University of lllinois College of Medicine at Chicago, Chicago, IL 60680, U.S.A., 4 Department of Medicine, Creighton University, Omaha, NE 68178, U.S.A., and 5 Department of Pediatrics, Oklahoma University Health Sciences Center, Oklahoma City, OK 73190, U.S.A.
(Received7 July 1988) (Revised, received17 October and 13 December1988) (Accepted 13 December 1988)
Key words: Vasoactiveintestinal peptide autoantibody; Asthma; Antibody affinity; Vasoactiveintestinal peptide receptor
Summary Vasoactive intestinal peptide (VIP) is a potent relaxant of the airway smooth muscle. In this study, VIP-binding autoantibodies were observed in the plasma of 18% asthma patients and 16% healthy subjects. Immunoprecipitation studies and chromatography on DEAE-cellulose and immobilized protein G indicated that the plasma VIP-binding activity was largely due to IgG antibodies. Saturation analysis of VIP binding by the plasmas suggested the presence of one or two classes of autoantibodies, distinguished by their apparent equilibrium affinity constants (Ka). The autoantibodies from asthma patients exhibited a larger VIP-binding affinity compared to those from healthy subjects ( K a 7.8 × 10 9 M -1 and 0.13 × 10 9 M - t , respectively; P < 0.005). The antibodies were specific for VIP, judged by their poor reaction with peptides bearing partial sequence homology with VIP (peptide histidine isoleucine, growth hormone releasing factor and secretin). IgG prepared from the plasma of an antibody-positive asthma patient inhibited the saturable binding of 125I-VIP by receptors in guinea pig lung membranes (by 39-59%; P < 0.001). These observations are consistent with a role for the VIP autoantibodies in the airway hyperresponsiveness of asthma.
Introduction The neuropeptide vasoactive intestinal peptide (VIP) is the likely physiological mediator of nonAddress for correspondence:Dr. Sudhir Paul, Department of Pharmacology, University of Nebraska Medical Center, 42nd and Dewey Avenue, Omaha, NE 68105, U.S.A. Supported by NHBLI grants 35506 and 40348.
adrenergic, non-cholinergic relaxation of the airway smooth muscle (Said, 1984, 1987). VIP is present in nerves supplying the smooth muscle (Laitinen et al., 1985), it is a potent relaxant of tracheal strips in vitro and is released by electrical stimulation of these strips (Matsuzaki et al., 1980), and specific receptors for VIP are present on the smooth muscle cells (Carstairs and Barnes, 1987). Asthma is characterized by a hyperresponsiveness
0165-5728/89/$03.50 © 1989 ElsevierSciencePublishers B.V. (BiomedicalDivision)
134
of the airways to a variety of stimuli. It has been suggested that airway hyperresponsiveness may arise from an imbalance in autonomic mechanisms regulating smooth muscle contraction and relaxation (Szentivanyi, 1968; Boushey et al., 1980). Thus, it is possible that a disturbance in VIP-mediated relaxation of the airways is a factor in the pathogenesis of asthma. Previously, we have shown the presence of autoantibodies directed against VIP in some healthy human subjects (Paul et al., 1985). VIP autoantibodies can potentially bring about a functional deficiency of VIP. This report describes circulating IgG antibodies directed against VIP in a subpopulation of asthma patients. These antibodies exhibited specific, high affinity binding of VIP and decreased the binding of VIP by lung receptors.
Materials and methods
Human subjects Asthma patients (n = 74) and non-asthmatic healthy volunteers (n = 98) were of either sex with mean ( + S E M ) ages 32.5+3.4 and 34.7+3.6 years, respectively. The asthmatics were drawn from university clinic populations. Their diagnosis was suggested by a history consistent with airways obstruction that was variable over time and responded clinically to bronchodilators. All of the asthmatic subjects had wheezing noted during auscultation either prior to or at the time of phlebotomy. Spirometry performed at diagnosis confirmed the presence of airway obstruction. At the time of obtaining blood samples for antibody analysis, obstruction was present in 69 patients and absent in five patients with well-controlled asthma. Selected spirometric values for 13 asthmatic subjects who were found positive for VIP antibodies are provided in Table 1. Twelve of these subjects showed increased FEV 1 values (by 12% to 62%) following administration of a fladrenergic agonist; this data was not available for one of the subjects included in Table 1, but this individual was clinically responsive to fl-adrenergic agonist therapy. All subjects had a history of asthma for at least 2 years. Seventy-two of the 74 asthmatics were being actively treated for their asthma, nine were receiving steroids by inhalation,
TABLE 1 S P I R O M E T R I C V A L U E S F O R SELECTED A S T H M A T I C S POSITIVE F O R P L A S M A VIP A N T I B O D I E S Subject code
FEV1 a
FVC b
FEVI / FVC
FEV1 ' postagonist c
A-1 A-2 A-3 A-4 A-5 A-6 A-7 A-8 A-9
2.80 1.27 2.87 0.92 1.95 0.63 2.10 2.52 0.75
4.32 2.12 5.16 1.37 3.52 1.53 2.62 2.85 1.53
65 60 56 67 55 41 80 88 49
3.54 (26) 1.70 (26) 3.31 (15) 1.23 (27) CR d 1.02 (62) 2.42 (15) 3.46 (37) 1.02 (36)
A-10
1.97 (54)
3.70 (84)
53
2.21 (12)
A-11 A-12 A-13
2.20 (57) 2.38 (59) 2.70 (63)
2.98 (65) 3.00 (62) 4.40 (83)
74 79 61
2.87 (30) 3.50 (47) 3.68 (36)
(62) (43) (63) (34) (66) (25) (69) (70) (17)
(77) (60) (95) (43) (89) (53) (74) (67) (30)
a Forced expiratory volume (liters) in 1 s. Values in parentheses are percent of predicted FEV 1. b Forced vital capacity (liters). Values in parentheses are percent of predicted FVC. c Values in parentheses represent percent increase over FEV 1 before fl-adrenergic agonist administration. a Clinically responsive to fl-adrenergic agonists. FEV 1 values not available.
but none were receiving orally administered steroids. The 98 non-asthmatic subjects denied any history suggestive of airways obstruction or cardiovascular, gastrointestinal, reproductive and nervous system abnormalities.
Detection and affinity of VIP-binding antibodies Blood (20 ml) was drawn from an antecubital vein in 4.5 mM EDTA (sodium salt), 50 ~M phenylmethylsulfonyl fluoride (PMSF), 5 ~tM pepstatin A and 100 units aprotinin/ml. 125I-VIP was prepared by the chloramine-T method and purified by reverse-phase high-performance liquid chromatography (HPLC) to a specific activity of 2 Ci//~mol (Paul et al., 1984). Plasma (100 #1), prepared by centrifugation of the blood (2000 × g, 15 min), was incubated in duplicate with lzsI-VIP (50 /~1; final concentration 40-60 pM) for 24 h (4 ° C) in the absence and presence of unlabeled VIP (50/~1; final concentration 2 /xM) in 7.5 mM sodium phosphate, pH 7.4, 0.64% sodium chloride,
135 0.5% bovine serum albumin, 25 m M EDTA, 0.005% bacitracin, 0.005% protamine sulfate, 0.073% sodium azide and 0.025% Tween-20. For initial screening of VIP binding activity, separation of bound and free VIP was by precipitation with polyethylene glycol (1 ml, final concentration 10% w / v ) , centrifugation (2000 × g , 4 ° C, 30 min) and aspiration of the supernatants. The pellets were counted for radioactivity (Beckm a n Model 5500 g a m m a spectrometer; 70% efficiency). Plasma samples at concentrations sufficient to give 2000-4000 cpm polyethylene glycol (PEG)-precipitable binding (undiluted to 10-fold diluted) were then assayed for anti-IgG and antiIgM precipitable binding. For this purpose, the assay mixtures were incubated (30 min, 4 ° C) with specific goat anti-human I g G serum or goat anti-human IgM serum (600 /~1, in the assay diluent containing 4% PEG) (Accurate Chem. Corp., Westbury, NY, U.S.A.). The anti-IgG and antiIgM dilutions employed for determination of binding by undiluted plasma were 1 : 2 and 1:4, respectively. Correspondingly larger dilutions of the antisera were used for the diluted plasma samples. Saturable binding was the difference between binding in the absence and presence of unlabeled VIP. Plasma samples with % B / T values (where B and T were cpm saturable binding and total 125I-VIP, respectively) greater than 5% (approximately 1500 cpm) were considered positive for the VIP-binding antibody. Saturable binding by two plasma samples (A-3 and NA-6 in Table 2) was measured after incubation with 125I-VIP for 0.5, 2, 6, 20, 24 and 48 h. In both cases, steady-state binding was achieved after incubation for 20 h. The affinity of the antibodies was determined under steady-state binding conditions, by measuring the displacement of a25I-VIP binding by increasing concentrations of unlabeled VIP (30 p M to 2 /zM), followed by analysis of the data with the programs EBDA and L I G A N D run on an IBM P S / 2 computer (McPherson, 1985). The protease inhibitors and reagents used for the binding assay were from Sigma (St. Louis, MO, U.S.A.) except for aprotinin, which was from M o b a y (New York, NY, U.S.A.). Purified porcine VIP was from Professor V. Mutt, Karolinska Institutet, Stockholm, Sweden, and porcine peptide histidine isoleucine (PHI), rat growth hormone releasing factor
TABLE 2 VIP BINDING CHARACTERISTICS OF PLASMA FROM ASTHMATIC AND NON-ASTHMATIC SUBJECTS Subject
% Saturable 125I-VIPbinding b
code a
PEG 10.3 11.9 50.2 45.6 67.2 19.3 10.7 20.0 15.4 13.0 32.3 5.2 8.3 25.2 12.6 13.6 12.2 36.7 7.1 28.2 24.1 9.2 47.5 21.0 9.9 13.3 37.5 13.1 7.2
A-1 A-2 A-3 A-4 A-5 A-6 A-7 A-8 A-9 A-10 A-11 A-12 A-13 NA-1 NA-2 NA-3 NA-4 NA-5 NA-6 NA-7 NA-8 NA-9 NA-10 NA-11 NA-12 NA-13 NA-14 NA-15 NA-16
Anti-IgG 79.8 90.1 77.2 91.3 107.9 96.6 88.1 75.5 98.3 100.2 53.5 43.5 85.5 73.3 87.5 100.8 84.6 98.9 89.0 85.5 83.5 61.9 46.7 56.5 96.4 57.5 55.2 99.2 86.8
KaC 28.6 1.9 3.1 1.4 6.6 12.5 0.3
0.300 0.061 0.083 0.008 0.016 0.205 0.355 0.022
a Subject type is denoted by 'A' (asthmatic) and 'NA' (nonasthmatic). b Values for PEG represent percent of total available xzSI-VIP and those for anti-IgG represent percent of the PEG-precipitable binding. c Affinity constants (× 109 M-1).
( G R F ) and porcine secretin from Peninsula, Belmont, CA, U.S.A.
DEAE-cellulose, protein G-Sepharose and size exclusion chromatography Plasma dialysed (24 h, 4 ° C ) against 0.05 M Tris-HC1, p H 8, was chromatographed on a DEAE-cellulose column (DE-52; 37 × 1 cm) (Whatman, Clifton, N J, U.S.A.) with a linear gradient from 0.05 M Tris-HC1, p H 8, to 0.3 M Tris-HC1, p H 4, over 35 min at a flow rate of 1
136
ml/min (Paul et al., 1985). Affinity purification of the plasma IgG fraction on a protein G-Sepharose column (8.5 × 0.7 cm) was according to the protocol supplied by the manufacturer (Pharmacia). Following application of the plasma, the column was washed with 50 mM Tris-HC1, pH 7.3, the retained IgG was eluted with 100 mM glycine, pH 2.7, and the eluate fractions were brought to pH 7.8 with 1 M Tris-HC1, pH 9. Size exclusion chromatography of VIP antibodies was on a Proteinpak 300 sw column (Waters Associates, Milford, MA, U.S.A.) using an ISCO HPLC apparatus (Lincoln, NE, U.S.A.). The colunm buffer was 50 mM Tris-HC1, pH 7.3 (flow rate: 0.5 ml/min). Dextran blue, ferritin, aldolase, bovine serum albumin, ovalbumin and ribonuclease (Pharmacia) were used for column calibration (PaUl et al., 1985). Purified human IgG was from Capell, Malvern, PA, U.S.A.
125I-VIP receptor binding Membranes from guinea pig lung were prepared by homogenization in 10 mM Tris-HC1, pH 7.2 (4 ° C) containing 0.25 M sucrose and protease inhibitors (Paul et al., 1988). Plasma IgG prepared by DEAE-cellulose chromatography was concentrated to 24 mg protein/ml using a YM-10 ultrafilter (10 kDa cutoff; Amicon, Danvers, MA, U.S.A.) and incubated with an equal volume of a2SI-VIP (240 pM) for 24 h at 4 ° C in the receptor assay diluent (100 mM Tris-HC1, pH 7.2, containing 5 mM MgC12 and 1% bovine serum albumin (BSA)). Increasing concentrations of the 125I-VIP were then mixed with lung membranes (50 /~g protein), the volume made up to 200/~1 with assay diluent and incubated for 30 rain at 4 ° C. Membrane-bound 125I-VIP was separated by centrifugation (Paul et al., 1988). Non-specific binding was determined by incubation in the presence of VIP (1/~M). Time-course experiments (results not shown) indicated that steady-state saturable binding was achieved under these conditions.
Results
125I-VIP binding by plasma Plasma samples from asthma patients and non-asthmatic healthy subjects were initially
screened for saturable binding of 125I-VIP precipitated by PEG. The binding activity was present in 18% of the asthma patients (n = 74) and 16% of healthy non-asthmatic subjects (n = 98) (P > 0.05, Fisher's exact test). The mean azsI-VIP binding values (%B/T) in the asthmatic and non-asthmatic subgroups positive for the VIPbinding factor (Table 2) were 23.8% _+ 5.3% (SEM) and 19.9% _+ 3.1%, respectively. Of the total binding activity observed in the 13 asthma and 16 non-asthmatic subjects using PEG as precipitant, 83.7% + 5.1% (SEM) and 79.0% + 4.5% was precipitated by a specific goat anti-human IgG serum, respectively (Table 2). The anti-IgG serum precipitated > 73% of the PEG-precipitable binding in most subjects. In a few subjects (two asthmatics and five non-asthmatics), the anti-IgG precipitated smaller proportions (43-62%) of the binding. Goat anti-human IgM antibodies did not precipitate the VIP-binding activity in plasma from the seven asthmatic and nine non-asthmatic subjects tested (codes A-l, A-2, A-3, A-7, A-8, A-11, A-12, NA-1, NA-3, NA-4, NA-5, NA-7, NA-8, NA-10, NA-13 and NA-15). Systematic investigation of the presence of other antibody classes with VIP-binding activity and the subclass distribution of the VIP-binding IgG antibodies was not done in this study because of unavailability of sufficient plasma volumes. The saturable binding of VIP was proportional to the concentration of the plasma for each of four samples tested (codes NA-1, NA-4, A-4 and A-5). The data for one of these plasma samples is shown in Fig. 1. The values for %B/T and non-saturable binding decreased progressively from 57.2% to 2.5% and 4.6% to 0.8%, respectively, by dilution of this plasma from 2- to 128-fold.
Chromatographic characterization of VIP antibodies The 125I-VIP binding activity in plasma from an asthma patient (code A-5) was found to cochromatograph with purified human IgG on a DEAE-cellulose column (Fig. 2), in agreement with findings on the binding factor present in healthy subjects (Paul et al., 1985). Of the starting VIPbinding activity present in plasma, 89% was recovered in the IgG fraction. This VIP-binding fraction exhibited a retention time of 14.5-15.0 min on a high-performance size exclusion column
137 20
$
\
I
o
16
pH2.r
x
\
13..
12"
i 0
\
8
IX.
I
2
x
d
Z :D O 133
o
"\
m
1
0.4
e~e.._. 0 t
I
I
I
2
4
8
16
I
I
64
128
I
32
PLASMA DILUTION, FOLD
Fig. l. Binding of ]251-VIP as a function of increasing dilution of plasma from an asthma patient (code A-5). (*) and (A) represent saturable binding and non-saturable binding precipitated by PEG, respectively. Total 125I-VIP in each tube was 31 100 cpm. Anti-IgG precipitated 107.9% of the PEG-precipitable binding by this plasma. Values plotted are means of three replicates. The coefficient of variation for each of the points was <5%.
(Proteinpak 300 sw), a value close to that of authentic IgG (14.8 min) (not shown). The VIPbinding activity in a plasma from another asthma subject (No. A-3) was retained by immobilized protein G, an agent that binds the Fc portion of IgG, and was then released by treatment at low
leg
1
m'
2.01
q"
t.el
x
.
3.
O m =E o.
2-
0.11t
I-
0.44
Q Z
0.8
:E in o
4 0
4.
~
1.e
1.2'
Z -j
O4
o
1
~'.0.
0 J .~ 0
,1)
2'o
so
EFFLUENT V O L U M E , ml
Fig. 2. DEAE-ceUulose chromatography of the VIP-binding activity in plasma (code A-5). The column effluent was screened for absorbance at 280 nm (A2s0) and saturable i25I-VIP binding precipitable by PEG was measured in 25/~1 aliquots of the colUmn fractions. The elution position of authentic IgG is shown by the arrow.
0
;
;
;o
,'5
~o
2"6
~o
E F F L U E N T VOLUME, ml
Fig. 3. Affinity purification of the VIP-binding antibodies on protein G-Sepharose. Plasma (1 ml) from an asthmatic subject (code A-3) was applied to the column, the column washed until /128o in the effluent had returned to baseline, and retained material was eluted with a pH 2.7 buffer (shown by the arrow). Saturable ]25I-VIP was measured in duplicate in 25 #1 aliquots of the column fractions using goat anti-human IgG antibodies to precipitate antibody-bound VIP.
pH (Fig. 3). This IgG fraction contained 96% of the binding activity originally present in the plasma (0.18 pmol 125I-VIP/ml).
Antibody affinity The binding affinity of VIP antibodies is likely to be an important factor determining the biological consequences of their presence. We determined the binding affinities of the antibodies present in seven asthma patients and eight non-asthmatic subjects. Displacement of a25I-VIP binding with increasing concentrations of unlabeled VIP was assayed (Fig. 4) using PEG to separate antibodybound VIP. The a25I-VIP binding in these experiments largely reflected binding of the peptide by IgG antibodies. This conclusion was based on the observations that: (i) 88.9% _+ 2.5% of the PEGprecipitable binding of the plasma samples assayed for binding affinity (n = 15) was precipitable by anti-human IgG antibodies (Table 2), and (ii) the VIP binding activity was present exclusively in IgG fractions prepared from two of these plasma samples by chromatography on DEAE-cellulose, size exclusion columns and protein GSepharose. Values for apparent K a (affinity constant) and binding capacity were obtained by computer-assisted analysis of the displacement
138 100.
T
100
o z
.
60
~"
o.2 0-
~ 60-
m W ._1
-~
6
B, Dmoles/rnl
o z z
80-
6O-
\
1~ 40
\\\
o
'x,\\ ix \,
%
VIP, - Log M
O.
Fig. 4. Saturable binding of 12SI-VIP by VIP antibodies in plasma from an asthma patient (code A-4) versus concentration of unlabeled VIE The Scatchard plot (inset) obtained using the computer program LIGAND (Elsevier-Biosoft, Cambridge, U.K.) was curvilinear, suggesting two antibody populations with K a 10.1 and 0.1 nM -1 and binding capacity 0.6 and 18.3 pmol VIP/ml plasma, respectively.
data (McPherson, 1985). Scatchard plots of the d a t a showed a single linear c o m p o n e n t (seven plasma samples) or were resolved into two linear c o m p o n e n t s (eight plasma samples), suggesting the presence of one and two classes of antibodies, respectively. The mean K a values for the high affinity autoantibodies of asthmatic (n = 7) and
T
11
0
10 v Z t-u3 Z 0 C)
I
9.
8.
I |
2 t.L
<
7. NONASTHMATICS
__~ 40. (D k~ CL
ASTHMATICS
Fig. 5. VIP-binding affinity of the antibodies present in asthma patients and healthy non-asthmatic subjects, computed from Scatchard plots such as the one in Fig. 4. Each point represents a different subject. The mean K a values for the antibodies present in the two groups of subjects are shown by broken lines.
-
11
-
-
10 PEPTIDE,
9 -log
-
I
8 M
Fig. 6. 125I-VIPbinding by antibodies in unfractionated plasma (e; code A-5) and its IgG fraction (o): displacement by unlabded VIP. Values, means of duplicates, are percent of binding by the plasma (4827 cpm) or IgG prepared by DEAEcellulose chromatography (3290 cpm) in the absence of unlabeled VIP.
non-asthmatic subjects (n = 8) were 7.8 × 109 M -1 and 0.13 X 109 M - I ( P < 0.005, t-test) (Fig. 5), and their V I P binding capacity values (in p m o l V I P / m l plasma) were 0.58 +__0.05 and 55.7 4- 39.2, respectively. The mean K a values for the low affinity a n t i b o d y of asthmatic (n = 5) and nonasthmatic subjects (n = 3) were similar (0.04 x 109 M -1 and 0.08 x 109 M -1, respectively; P > 0.05). The VIP-binding capacities (mean) for the low affinity antibodies of asthmatic and non-asthmatic subjects were 24.7 and 152.6 p m o l / m l , respectively. The a p p a r e n t affinity of these antibodies measured in this study m a y be influenced by contaminating factors present in plasma, including peptide hydrolases (although the plasma was collected in a mixture of peptide hydrolase inhibitors; see Materials and Methods) and peptides cross-reactive with V I P antibodies. To investigate the contribution of such factors, we c o m p a r e d the ability of V I P to displace a2SI-VIP binding by unfractionated plasma and its purified I g G fraction (Fig. 6). The two displacement curves were nearly superimposable and K a values calculated for the b i n d i n g b y plasma and purified I g G were similar (6.6 x 109 M -1 and 7.1 X 109 M -1, respectively; P > 0.05). Furthermore, the propor-
139
tions of 125I-VIP precipitated with 10% trichloroacetic acid after incubation of the peptide with plasma (n = 3) or assay diluent under conditions employed for antibody affinity measurements were comparable (73% and 91%, respectively). These observations suggest that peptide hydrolases and other agents present in plasma were not significant factors in the observed antibody-VIP interactions.
25I
o x
• 20.
12. O
c5 o m a.
15
•
10 •
_ s /J'" u3 t"4
0
Antibody specificity PHI, GRF and secretin, peptides partially identical to VIP in their amino acid sequence, were employed to examine the specificity of the antibodies. These peptides (1 #M) did not significantly displace the 125I-VIP binding by plasma from six asthma patients and four non-asthmatic subjects (Table 3). The plasma antibodies in one asthmatic and one non-asthmatic subject showed partial reactivity with PHI, GRF and secretin (21.9% to 33.4%). The poor reaction of the antibodies with these PHI, GRF and secretin suggests their high level of specificity for VIP.
SPECIFICITY O F P L A S M A VIP A N T I B O D I E S
A-3 (1 : 12) A-4 (1 : 6) A-5 (1 : 5) A-6 (1:4) A-8 (1:3) A-11 (1 : 3) N A - 4 (1 : 10) NA-5 (1 : 10) NA-7 (1 : 10) NA-8 (1 : 3)
I
[
I
I
40
60
80
Ioo
TOTAL 125 I-VIP,
CPM x 1 0 - 5
Fig. 7. Saturable binding of 125I-VIP by receptors in guinea pig lung membranes: effect of i m m u n e IgG from an asthmatic subject (code A-5) (A), n o n - i m m u n e IgG from a non-asthmatic subject ( • ) and assay diluent (0). IgG was purified by DEAEcellulose chromatography. 125I-VIP was preincubated with an equal volume of i m m u n e IgG (24 m g / m l , a concentration sufficient to bind 53.2% of the 125I-VIP), n o n - i m m u n e IgG or assay diluent. Receptor binding by the nSI-VIP was determined (see Materials and Methods). Total a m o u n t s of 1251VIP available for receptor binding are plotted on the x-axis. Values are m e a n s of three replicates each. The coefficient of variation for each of the points was < 7.5%.
Effect of the antibodies on VIP-receptor binding
TABLE 3
Subject code a
I
2O
% 125I-VIP displaced b VIP
PHI
GRF
Secretin
100 (2 589) 100 (5 066) 100 (4582) 100 (1350) 100(4173) 100 (3 686) 100 (2600) 100 (4512) 100 (1 855) 100 (3 567)
25.0 c - 3.9 2.7 4.1 -1.8 1.2 8.3 3.1 - 7.0 31.6 c
33.4 c 1.8 0.5 8.8 -1.2 6.7 3.3 2.9 12.1 33.0 c
21.9 1.5 5.1 11.9 -0.9 - 2.2 1.2 3.4 9.0 8.9
a Subject type is denoted by ' A ' (asthmatic) and ' N A ' (nonasthmatic). Values in parentheses are the dilutions at which the plasmas were assayed. b 125I.VI P binding was assayed in absence and presence of the peptides (1 /xM). The values, means of duplicates, are percent of displacement observed with VIP, with the c p m radioactivity displaced by VIP in parentheses. Non-specific binding ranged from 1563 to 3862, and the total available radioactivity was - 30000 cpm. c p < 0.05 versus binding in the absence of unlabeled peptide.
We have previously identified and characterized specific receptors for VIP in guinea pig lung membranes (Paul and Said, 1987; Paul et al., 1988). In the present study, the influence of the human antibodies on VIP binding by these receptors was investigated. 125I-VIP was preincubated with the receptor assay diluent, immune IgG purified from plasma of an asthma patient (code A-5) or non-immune IgG from a healthy subject. Saturable binding of increasing concentrations of the IESI-VIP by receptors in guinea pig lung membranes was then measured. The preincubation with immune IgG resulted in significantly reduced receptor binding at each of the uS I-VIP concentrations tested (by 39% to 59%; P<0.001 versus non-immune IgG) (Fig. 7). The nSI-VIP receptor binding curves after preincubation with non-immune IgG or receptor assay diluent were nearly indistinguishable. The free and antibody-bound VIP concentrations available to the receptors after preincubation with immune IgG were not measured. It is difficult, therefore, to obtain a reliable
140 estimate of receptor affinity in the presence of anti-VIP in this experiment. Scatchard analysis of receptor binding observed in the presence of nonimmune IgG suggested a binding site with Ka 5.3 X 10 9 M - 1 and binding capacity 0.9 pmol VIP/mg protein, consistent with previously observed values for the high affinity VIP receptor in guinea pig lung membranes (Paul and Said, 1987; Paul, 1988).
Discussion
This study demonstrates the presence of VIPbinding activity precipitated by polyethylene glycol (PEG) in the plasma of 29 plasma samples from, asthma patients and healthy non-asthmatic subjects. Immunoprecipitation and chromatographic analyses suggested that the VIP binding Was attributable largely to IgG antibodies. The VIP-binding IgG antibodies were detected by measuring their binding to porcine 125I-VIP. Human and porcine VIP are structurally identical (Itoh et al., 1983). Thus the porcine VIP-reactive antibodies found in asthma patients can be considered to represent autoantibodies. Previously, Bloom et al. (1979) observed that diabetics positive for plasma VIP antibodies had been treated with insulin preparations contaminated with VIP, suggesting that the formation of the antibodies was related to the VIP contaminant. In contrast, the autoantibodies observed in the present study were naturally occurring. The antigenic stimulus leading to formation of these autoantibodies cannot be identified with certainty. Candidate stimuli include exposure to viral determinants similar in sequence to VIP (e.g., peptide-T, an epitope found on the human immunodeficiency virus (Ruff et al., 1987)) and dietary ingestion of avian or fish VIP known to be structurally different from human VIP (Nilsson, 1975; Dimaline and Thorndyke, 1986). Another potential stimulus for VIP autoantibody formation is muscular exercise. We have observed that VIP autoantibodies are present more frequently in healthy subjects who habitually perform muscular exercise compared to those who do not (Paul and Said, 1988). Muscular exercise resuits in increased plasma VIP immunoreactivity (Paul et al., 1987a; Woie et al., 1987), an effect
shown to be more pronounced in subjects with exercise-induced asthma compared to non-asthmatic subjects (Hvidsten et al., 1986). Investigation of the molecular form(s) and antigenicity of VIP released during exercise may reveal more about the relationship between exercise and VIP autoantibody formation in healthy subjects. Irrespective of the type of antigenic stimulation leading to VIP autoantibody formation, these antibodies may possess important effects. To evaluate the potential functional role of the antibodies, we measured their apparent binding affinity (Ka). Since unfractionated plasma samples were used for K a measurements, peptide hydrolases and cross-reactive peptides were potential factors in the observed interaction of VIP with the antibodies. Peptide hydrolase inhibitors (EDTA, PMSF, aprotinin and pepstatin A) were added to the plasma samples to minimize degradation of VIP. The IgG purified from one of the plasma samples exhibited a K a value (7.1 × 10 9 M - 1 ) close to that observed for unfractionated plasma (6.6 × 10 9 M-1), suggesting a lack of interfering factors. The mean apparent g a for the antibodies in asthma patients was considerably larger (by 60-fold) than in healthy non-asthmatic subjects. The range of K a values observed for the high affinity autoantibodies of asthma patients (0.25-26.2 × 10 9 M - 1 ) was similar to that reported for VIP receptors present in the lung and other tissues (Rosselin, 1986; Carstairs and Barnes, 1987; Paul and Said, 1987). Moreover, preincubation with a purified antibody fraction from an asthmatic subject inhibited the saturable binding of VIP by guinea pig lung membranes, suggesting that the antibody was probably directed against an epitope involved in VIP receptor binding. These considerations suggest that VIP autoantibodies found in asthmatics could neutralize VIP receptor binding and the ensuing VIP-medi~ted biological effects. The mean concentration of VIP in plasma from asthma patients, measured by radioimmunoassay (Paul et al., 1987b), was 1.33 pmol/ml (n = 16). Computed on the basis of mean antibody affinity and binding capacity values (Rodbard, 1981), 82% and 22% of the plasma VIP was likely to be present bound to the high and low affinity antibodies, respectively. However, as a neurotransmitter, the smooth muscle relaxant action of VIP
141 in the airways would be determined mainly by the local, as opposed to the circulating levels of the peptide. Immunoglobulins are present in respiratory secretions, with the IgG levels approximating those in the plasma (Reynolds, 1987). IgG is thought to arrive in the airways mainly by transudation from the blood compartment, although secretion from lymphocytes present within the airways is an additional contributory factor (Hance et al., 1988). Airway inflammation, a common finding in asthmatic subjects, is likely to promote passage of antibodies from blood to airway fluids. Thus VIP autoantibodies observed in the blood of asthma patients in the present study may reach the microenvironment of airway smooth muscle. Here the antibodies could interfere with airway relaxation by VIP released from nerve terminals supplying the smooth muscle. Bronchial hyperresponsiveness in asthma may arise from an imbalance in the mechanisms that constrict and relax the airways (Szentivanyi, 1968; Boushey et al., 1980). The observations reported in this study are consistent with a role for VIP autoantibodies in the pathogenesis of asthma. Asthma has been linked to many different neuronal and immunological mediators known to govern the activity of airway smooth muscle (Nadel and Barnes, 1984; Nadel et al., 1987), and the pathophysiology of airway hyperresponsiveness in different individuals may be different. Our observation that the VIP antibodies were only present in some of the asthma patients is in line with this possibility. We were unable to find a clear difference between antibody-positive and antibody-negative asthmatic subjects with respect to age, sex, number of years the patients had had asthma, presence of obstruction judged by pulmonary function tests and type of inciting agent resulting in asthma episodes. However, we analysed only 13 antibody-positive subjects, and data from a larger number of patients may be more revealing. In addition to their well-documented role in disease (Rose and Mackay, 1985), autoantibodies may mediate physiologic functions (Kay et al., 1982; Khandsari et al., 1983). The significance of the low affinity VIP autoantibodies present in healthy subjects requires further study. Considerations relevant to the role of VIP autoantibodies in healthy subjects include: (i) the antibodies may
predate the appearance of some functional abnormality, (ii) because they possess a low VIPbinding affinity (mean K~ 0.13 × 109 M - I ) , the antibodies are likely to interfere with VIP-mediated effects only at concentrations greatly in excess of tissue VIP receptor concentrations, (iii) because their affinity is greater than the 1/Km value (inverse Michaelis constant) reported previously for VIP-degrading enzyme(s) (105 M -1) (Keltz et al., 1980), the antibodies could effectively inhibit the proteolytic inactivation'of VIIa and thus increase the availability of the peptide at tissue receptors. In conclusion, this study suggests that humoral autoimmunity is a potential modulatory factor in the biological actions of VIP in asthmatic and healthy humans. VIP possesses a broad range of actions in different organ systems. Further study of the pathophysiological significance of VIP autoantibodies is clearly warranted. Factors that are likely to influence the biological consequences of the VIP autoantibodies include their binding affinity; epitope specificity, site of accumulation and ability to activate inflammatory cells via Fc receptor binding. References Bloom, S.R., Barnes, A.J., Adrian, T.E. and Polak, J.M. (1979) Autoimmunity in diabetics induced by hormonal contaminants of insulin. Lancet i, 14-17. Boushey, H.A., Holtzman, M.J., Sheller, J.R. and Nadel, J.A. (1980) Bronchial hyperreactivity. Am. Rev. Respir. Dis. 121, 389-413. Carstairs, J.R. and Barnes, P.J. (1987) Visualizationof vasoactive intestinal peptide receptors in human and guinea pig lung. J. Pharmacol. Exp. Ther. 239, 249-255. Dimaline, R. and Thorndyke, M.C. (1986) Purification and characterization of VIP from two species of dogfish. Peptides 7 (Suppl. 1), 21-26. Hance, A.J., Saltini, C. and Crystal, R.J. (1988) Does de novo immunoglobulin synthesis occur on the epithelial surface of the human lower respiratory tract? Am. Rev. Respir. Dis. 137, 17-24. Hvidsten, D., Jenssen, T.G., Bone, R. and Burhol, P.G. (1986) Plasma gastrointestinal regulatory peptides in exercise induced asthma. Eur. J. Respir. Dis. 68, 326-331. Itoh, N., Obata, K.-I., Yanaihara, N. and Okamoto, H. (1983) Human preprovasoactiveintestinal polypeptide contains a novel PHI-27-1ikepeptide, PM-27. Nature 304, 547-549. Keltz, T.N., Straus, E. and Yalow, R.S. (1980) Degradation of vasoactive intestinal polypeptide by tissue homogenates. Biochem. Biophys. Res. Commun. 92, 669-674.
142 Kay, M.M,B., Sorenson, K., Wong, P. and Bolton, P. (1982) Antigenicity, storage and aging: physiologic autoantibodies to cell membrane and serum proteins and the senescent cell. Mol. Cell. Biochem. 49, 65-85. Khandsari, N. and Fudenberg, H.H. (1983) Immune elimination of aging platelets by autologous monocytes: role of membrane specific autoantibody? Eur. J. Immunol. 13, 990-994. Laitinen, A., Partanen, M., Hervonen, A., Peto-Huikko, M. and Laitinen, L.A. (1985) VIP like immunoreactive nerves in human respiratory tract. Light and electron microscopic study. Histochemistry 82, 313-319. Matsuzaki, Y., Hamsaki, Y. and Said, S.I. (1980) Vasoactive intestinal peptide: a possible transmitter of nonadrenergic relation of guinea pig airways. Science 210, 1252-1253. McPherson, G.A. (1985) Analysis of radioligand binding experiments. A collection of computer programs for the IBM PC. J. Pharmacol. Methods 14, 213-228. Nadel, J.A. and Barnes, P.J. (1984) Autonomic regulation of the airways. Annu. Rev. Med. 35, 451-467. Nadel, J.A., Pauweis, E. and Snashall, P.D. (Eds.) (1987) Bronchial Hyperresponsiveness, Blackwell, Oxford. Nilsson, A. (1975) Structure of the vasoactive intestinal peptide from chicken intestine. The amino acid sequence. FEBS Lett. 60, 322-326. Paul, S. (1988) Decreased selectivity of vasoactive intestinal peptide receptors by GTP. Biochem. Pharmacol. 38, 699-702. Paul, S. and Said, S.I. (1987) Characterization of receptors for vasoactive intestinal peptide from the lung. J. Biol. Chem. 262, 158-162. Paul, S. and Said, S.I. (1988) Human autoantibody to vasoactive intestinal peptide: increased incidence in muscular exercise. Life SCI. 43, 1079-1084. Paul, S., Wood, K. and Said, S.I. (1984) Purification of [125I]vasoactive intestinal peptide by reverse-phase HPLC. Peptides 5, 1085-1087.
Paul, S., Erian, P. and Said, S.I. (1985) Autoantibody to vasoactive intestinal peptide in human circulation. Biochem. Biophys. Res. Commun. 130, 479-485. Paul, S., Chou, J., Beckham, S., Liu, L.-W., Kubota, E., Dominguez, N. and Said, S.I. (1987a) Elevated levels of atrial natriuretic peptide and vasoactive intestinal peptide in exercising man. Clin. Res. 35, 112A (abstract). Paul, S., Chou, J. and Kubota, E. (1987b) High-affinity peptide-histidine-isoleucine preferring receptors in rat liver. Life SCI. 41, 2373-2380. Paul, S., Chou, J. and Said, S.I. (1988) Regulatory aspects of the VIP receptor in lung. Ann. N.Y. Acad. Sci. 527, 282-295. Reynolds, H.Y. (1987) Identification and role of immunoglobulins in respiratory secretions. Eur. J. Respir. Dis. 71 (Suppl. 153), 103-116. Rodbard, D. (1981) Radioimmunoassay data processing. In: Syllabus, Tenth Training Course on Hormonal Assay Techniques, The Endocrine Society, Bethesda, MD, p. 25. Rose, N.R. and Mackay, I.R. (Eds.) (1985) The Autoimmune Diseases, Academic Press, London. Rosselin, G. (1986) The receptors for the VIP family peptides (VIP, secretin, GRF, PHI, PHM, GIP, glueagon and oxyntomodulin). Specificities and identity. Peptides 7 (Suppl. 1), 89-100. Ruff, M.R., Martin, B.M., Ginns, E.I., Farrar, W.L. and Pert, C.B. (1987) CD4 receptor binding peptides that block HIV infectivity cause human monocyte chemotaxis. FEBS Lett. 211, 17-22. Said, S.I. (1984) Vasoactive intestinal peptide (VIP). Current status. Peptides 5, 143-150. Said, S.I. (1987) Influence of neuropeptides on airway smooth muscle. Am. Rev. Respir. Dis. 136, $52-$58. Szentivanyi, A. (1968) The fl-adrenergic theory of the atopic abnormality in bronchial asthma. J. Allergy 42, 203-232. Woie, L., Kaada, B. and Opstaad, P.K. (1987) Increase in plasma VIP in muscular exercise. Gen. Pharmacol. 17, 321-326.
133
Elsevier JNI 00779
Characterization of autoantibodies to vasoactive intestinal peptide in asthma Sudhir Paul 1, Sami I. Said 3, Austin B. T h o m p s o n 2, D e a n n a J. Voile 1, Devendra K. Agrawal Hussein F o d a 3 and Santiago de la R o c h a 5
4,
Departments of 1 Pharmacology and 2 Medicine, University of Nebraska Medical Center, Omaha, NE 68105, U.S.A., 3 Department of Medicine, University of lllinois College of Medicine at Chicago, Chicago, IL 60680, U.S.A., 4 Department of Medicine, Creighton University, Omaha, NE 68178, U.S.A., and 5 Department of Pediatrics, Oklahoma University Health Sciences Center, Oklahoma City, OK 73190, U.S.A.
(Received7 July 1988) (Revised, received17 October and 13 December1988) (Accepted 13 December 1988)
Key words: Vasoactiveintestinal peptide autoantibody; Asthma; Antibody affinity; Vasoactiveintestinal peptide receptor
Summary Vasoactive intestinal peptide (VIP) is a potent relaxant of the airway smooth muscle. In this study, VIP-binding autoantibodies were observed in the plasma of 18% asthma patients and 16% healthy subjects. Immunoprecipitation studies and chromatography on DEAE-cellulose and immobilized protein G indicated that the plasma VIP-binding activity was largely due to IgG antibodies. Saturation analysis of VIP binding by the plasmas suggested the presence of one or two classes of autoantibodies, distinguished by their apparent equilibrium affinity constants (Ka). The autoantibodies from asthma patients exhibited a larger VIP-binding affinity compared to those from healthy subjects ( K a 7.8 × 10 9 M -1 and 0.13 × 10 9 M - t , respectively; P < 0.005). The antibodies were specific for VIP, judged by their poor reaction with peptides bearing partial sequence homology with VIP (peptide histidine isoleucine, growth hormone releasing factor and secretin). IgG prepared from the plasma of an antibody-positive asthma patient inhibited the saturable binding of 125I-VIP by receptors in guinea pig lung membranes (by 39-59%; P < 0.001). These observations are consistent with a role for the VIP autoantibodies in the airway hyperresponsiveness of asthma.
Introduction The neuropeptide vasoactive intestinal peptide (VIP) is the likely physiological mediator of nonAddress for correspondence:Dr. Sudhir Paul, Department of Pharmacology, University of Nebraska Medical Center, 42nd and Dewey Avenue, Omaha, NE 68105, U.S.A. Supported by NHBLI grants 35506 and 40348.
adrenergic, non-cholinergic relaxation of the airway smooth muscle (Said, 1984, 1987). VIP is present in nerves supplying the smooth muscle (Laitinen et al., 1985), it is a potent relaxant of tracheal strips in vitro and is released by electrical stimulation of these strips (Matsuzaki et al., 1980), and specific receptors for VIP are present on the smooth muscle cells (Carstairs and Barnes, 1987). Asthma is characterized by a hyperresponsiveness
0165-5728/89/$03.50 © 1989 ElsevierSciencePublishers B.V. (BiomedicalDivision)
134
of the airways to a variety of stimuli. It has been suggested that airway hyperresponsiveness may arise from an imbalance in autonomic mechanisms regulating smooth muscle contraction and relaxation (Szentivanyi, 1968; Boushey et al., 1980). Thus, it is possible that a disturbance in VIP-mediated relaxation of the airways is a factor in the pathogenesis of asthma. Previously, we have shown the presence of autoantibodies directed against VIP in some healthy human subjects (Paul et al., 1985). VIP autoantibodies can potentially bring about a functional deficiency of VIP. This report describes circulating IgG antibodies directed against VIP in a subpopulation of asthma patients. These antibodies exhibited specific, high affinity binding of VIP and decreased the binding of VIP by lung receptors.
Materials and methods
Human subjects Asthma patients (n = 74) and non-asthmatic healthy volunteers (n = 98) were of either sex with mean ( + S E M ) ages 32.5+3.4 and 34.7+3.6 years, respectively. The asthmatics were drawn from university clinic populations. Their diagnosis was suggested by a history consistent with airways obstruction that was variable over time and responded clinically to bronchodilators. All of the asthmatic subjects had wheezing noted during auscultation either prior to or at the time of phlebotomy. Spirometry performed at diagnosis confirmed the presence of airway obstruction. At the time of obtaining blood samples for antibody analysis, obstruction was present in 69 patients and absent in five patients with well-controlled asthma. Selected spirometric values for 13 asthmatic subjects who were found positive for VIP antibodies are provided in Table 1. Twelve of these subjects showed increased FEV 1 values (by 12% to 62%) following administration of a fladrenergic agonist; this data was not available for one of the subjects included in Table 1, but this individual was clinically responsive to fl-adrenergic agonist therapy. All subjects had a history of asthma for at least 2 years. Seventy-two of the 74 asthmatics were being actively treated for their asthma, nine were receiving steroids by inhalation,
TABLE 1 S P I R O M E T R I C V A L U E S F O R SELECTED A S T H M A T I C S POSITIVE F O R P L A S M A VIP A N T I B O D I E S Subject code
FEV1 a
FVC b
FEVI / FVC
FEV1 ' postagonist c
A-1 A-2 A-3 A-4 A-5 A-6 A-7 A-8 A-9
2.80 1.27 2.87 0.92 1.95 0.63 2.10 2.52 0.75
4.32 2.12 5.16 1.37 3.52 1.53 2.62 2.85 1.53
65 60 56 67 55 41 80 88 49
3.54 (26) 1.70 (26) 3.31 (15) 1.23 (27) CR d 1.02 (62) 2.42 (15) 3.46 (37) 1.02 (36)
A-10
1.97 (54)
3.70 (84)
53
2.21 (12)
A-11 A-12 A-13
2.20 (57) 2.38 (59) 2.70 (63)
2.98 (65) 3.00 (62) 4.40 (83)
74 79 61
2.87 (30) 3.50 (47) 3.68 (36)
(62) (43) (63) (34) (66) (25) (69) (70) (17)
(77) (60) (95) (43) (89) (53) (74) (67) (30)
a Forced expiratory volume (liters) in 1 s. Values in parentheses are percent of predicted FEV 1. b Forced vital capacity (liters). Values in parentheses are percent of predicted FVC. c Values in parentheses represent percent increase over FEV 1 before fl-adrenergic agonist administration. a Clinically responsive to fl-adrenergic agonists. FEV 1 values not available.
but none were receiving orally administered steroids. The 98 non-asthmatic subjects denied any history suggestive of airways obstruction or cardiovascular, gastrointestinal, reproductive and nervous system abnormalities.
Detection and affinity of VIP-binding antibodies Blood (20 ml) was drawn from an antecubital vein in 4.5 mM EDTA (sodium salt), 50 ~M phenylmethylsulfonyl fluoride (PMSF), 5 ~tM pepstatin A and 100 units aprotinin/ml. 125I-VIP was prepared by the chloramine-T method and purified by reverse-phase high-performance liquid chromatography (HPLC) to a specific activity of 2 Ci//~mol (Paul et al., 1984). Plasma (100 #1), prepared by centrifugation of the blood (2000 × g, 15 min), was incubated in duplicate with lzsI-VIP (50 /~1; final concentration 40-60 pM) for 24 h (4 ° C) in the absence and presence of unlabeled VIP (50/~1; final concentration 2 /xM) in 7.5 mM sodium phosphate, pH 7.4, 0.64% sodium chloride,
135 0.5% bovine serum albumin, 25 m M EDTA, 0.005% bacitracin, 0.005% protamine sulfate, 0.073% sodium azide and 0.025% Tween-20. For initial screening of VIP binding activity, separation of bound and free VIP was by precipitation with polyethylene glycol (1 ml, final concentration 10% w / v ) , centrifugation (2000 × g , 4 ° C, 30 min) and aspiration of the supernatants. The pellets were counted for radioactivity (Beckm a n Model 5500 g a m m a spectrometer; 70% efficiency). Plasma samples at concentrations sufficient to give 2000-4000 cpm polyethylene glycol (PEG)-precipitable binding (undiluted to 10-fold diluted) were then assayed for anti-IgG and antiIgM precipitable binding. For this purpose, the assay mixtures were incubated (30 min, 4 ° C) with specific goat anti-human I g G serum or goat anti-human IgM serum (600 /~1, in the assay diluent containing 4% PEG) (Accurate Chem. Corp., Westbury, NY, U.S.A.). The anti-IgG and antiIgM dilutions employed for determination of binding by undiluted plasma were 1 : 2 and 1:4, respectively. Correspondingly larger dilutions of the antisera were used for the diluted plasma samples. Saturable binding was the difference between binding in the absence and presence of unlabeled VIP. Plasma samples with % B / T values (where B and T were cpm saturable binding and total 125I-VIP, respectively) greater than 5% (approximately 1500 cpm) were considered positive for the VIP-binding antibody. Saturable binding by two plasma samples (A-3 and NA-6 in Table 2) was measured after incubation with 125I-VIP for 0.5, 2, 6, 20, 24 and 48 h. In both cases, steady-state binding was achieved after incubation for 20 h. The affinity of the antibodies was determined under steady-state binding conditions, by measuring the displacement of a25I-VIP binding by increasing concentrations of unlabeled VIP (30 p M to 2 /zM), followed by analysis of the data with the programs EBDA and L I G A N D run on an IBM P S / 2 computer (McPherson, 1985). The protease inhibitors and reagents used for the binding assay were from Sigma (St. Louis, MO, U.S.A.) except for aprotinin, which was from M o b a y (New York, NY, U.S.A.). Purified porcine VIP was from Professor V. Mutt, Karolinska Institutet, Stockholm, Sweden, and porcine peptide histidine isoleucine (PHI), rat growth hormone releasing factor
TABLE 2 VIP BINDING CHARACTERISTICS OF PLASMA FROM ASTHMATIC AND NON-ASTHMATIC SUBJECTS Subject
% Saturable 125I-VIPbinding b
code a
PEG 10.3 11.9 50.2 45.6 67.2 19.3 10.7 20.0 15.4 13.0 32.3 5.2 8.3 25.2 12.6 13.6 12.2 36.7 7.1 28.2 24.1 9.2 47.5 21.0 9.9 13.3 37.5 13.1 7.2
A-1 A-2 A-3 A-4 A-5 A-6 A-7 A-8 A-9 A-10 A-11 A-12 A-13 NA-1 NA-2 NA-3 NA-4 NA-5 NA-6 NA-7 NA-8 NA-9 NA-10 NA-11 NA-12 NA-13 NA-14 NA-15 NA-16
Anti-IgG 79.8 90.1 77.2 91.3 107.9 96.6 88.1 75.5 98.3 100.2 53.5 43.5 85.5 73.3 87.5 100.8 84.6 98.9 89.0 85.5 83.5 61.9 46.7 56.5 96.4 57.5 55.2 99.2 86.8
KaC 28.6 1.9 3.1 1.4 6.6 12.5 0.3
0.300 0.061 0.083 0.008 0.016 0.205 0.355 0.022
a Subject type is denoted by 'A' (asthmatic) and 'NA' (nonasthmatic). b Values for PEG represent percent of total available xzSI-VIP and those for anti-IgG represent percent of the PEG-precipitable binding. c Affinity constants (× 109 M-1).
( G R F ) and porcine secretin from Peninsula, Belmont, CA, U.S.A.
DEAE-cellulose, protein G-Sepharose and size exclusion chromatography Plasma dialysed (24 h, 4 ° C ) against 0.05 M Tris-HC1, p H 8, was chromatographed on a DEAE-cellulose column (DE-52; 37 × 1 cm) (Whatman, Clifton, N J, U.S.A.) with a linear gradient from 0.05 M Tris-HC1, p H 8, to 0.3 M Tris-HC1, p H 4, over 35 min at a flow rate of 1
136
ml/min (Paul et al., 1985). Affinity purification of the plasma IgG fraction on a protein G-Sepharose column (8.5 × 0.7 cm) was according to the protocol supplied by the manufacturer (Pharmacia). Following application of the plasma, the column was washed with 50 mM Tris-HC1, pH 7.3, the retained IgG was eluted with 100 mM glycine, pH 2.7, and the eluate fractions were brought to pH 7.8 with 1 M Tris-HC1, pH 9. Size exclusion chromatography of VIP antibodies was on a Proteinpak 300 sw column (Waters Associates, Milford, MA, U.S.A.) using an ISCO HPLC apparatus (Lincoln, NE, U.S.A.). The colunm buffer was 50 mM Tris-HC1, pH 7.3 (flow rate: 0.5 ml/min). Dextran blue, ferritin, aldolase, bovine serum albumin, ovalbumin and ribonuclease (Pharmacia) were used for column calibration (PaUl et al., 1985). Purified human IgG was from Capell, Malvern, PA, U.S.A.
125I-VIP receptor binding Membranes from guinea pig lung were prepared by homogenization in 10 mM Tris-HC1, pH 7.2 (4 ° C) containing 0.25 M sucrose and protease inhibitors (Paul et al., 1988). Plasma IgG prepared by DEAE-cellulose chromatography was concentrated to 24 mg protein/ml using a YM-10 ultrafilter (10 kDa cutoff; Amicon, Danvers, MA, U.S.A.) and incubated with an equal volume of a2SI-VIP (240 pM) for 24 h at 4 ° C in the receptor assay diluent (100 mM Tris-HC1, pH 7.2, containing 5 mM MgC12 and 1% bovine serum albumin (BSA)). Increasing concentrations of the 125I-VIP were then mixed with lung membranes (50 /~g protein), the volume made up to 200/~1 with assay diluent and incubated for 30 rain at 4 ° C. Membrane-bound 125I-VIP was separated by centrifugation (Paul et al., 1988). Non-specific binding was determined by incubation in the presence of VIP (1/~M). Time-course experiments (results not shown) indicated that steady-state saturable binding was achieved under these conditions.
Results
125I-VIP binding by plasma Plasma samples from asthma patients and non-asthmatic healthy subjects were initially
screened for saturable binding of 125I-VIP precipitated by PEG. The binding activity was present in 18% of the asthma patients (n = 74) and 16% of healthy non-asthmatic subjects (n = 98) (P > 0.05, Fisher's exact test). The mean azsI-VIP binding values (%B/T) in the asthmatic and non-asthmatic subgroups positive for the VIPbinding factor (Table 2) were 23.8% _+ 5.3% (SEM) and 19.9% _+ 3.1%, respectively. Of the total binding activity observed in the 13 asthma and 16 non-asthmatic subjects using PEG as precipitant, 83.7% + 5.1% (SEM) and 79.0% + 4.5% was precipitated by a specific goat anti-human IgG serum, respectively (Table 2). The anti-IgG serum precipitated > 73% of the PEG-precipitable binding in most subjects. In a few subjects (two asthmatics and five non-asthmatics), the anti-IgG precipitated smaller proportions (43-62%) of the binding. Goat anti-human IgM antibodies did not precipitate the VIP-binding activity in plasma from the seven asthmatic and nine non-asthmatic subjects tested (codes A-l, A-2, A-3, A-7, A-8, A-11, A-12, NA-1, NA-3, NA-4, NA-5, NA-7, NA-8, NA-10, NA-13 and NA-15). Systematic investigation of the presence of other antibody classes with VIP-binding activity and the subclass distribution of the VIP-binding IgG antibodies was not done in this study because of unavailability of sufficient plasma volumes. The saturable binding of VIP was proportional to the concentration of the plasma for each of four samples tested (codes NA-1, NA-4, A-4 and A-5). The data for one of these plasma samples is shown in Fig. 1. The values for %B/T and non-saturable binding decreased progressively from 57.2% to 2.5% and 4.6% to 0.8%, respectively, by dilution of this plasma from 2- to 128-fold.
Chromatographic characterization of VIP antibodies The 125I-VIP binding activity in plasma from an asthma patient (code A-5) was found to cochromatograph with purified human IgG on a DEAE-cellulose column (Fig. 2), in agreement with findings on the binding factor present in healthy subjects (Paul et al., 1985). Of the starting VIPbinding activity present in plasma, 89% was recovered in the IgG fraction. This VIP-binding fraction exhibited a retention time of 14.5-15.0 min on a high-performance size exclusion column
137 20
$
\
I
o
16
pH2.r
x
\
13..
12"
i 0
\
8
IX.
I
2
x
d
Z :D O 133
o
"\
m
1
0.4
e~e.._. 0 t
I
I
I
2
4
8
16
I
I
64
128
I
32
PLASMA DILUTION, FOLD
Fig. l. Binding of ]251-VIP as a function of increasing dilution of plasma from an asthma patient (code A-5). (*) and (A) represent saturable binding and non-saturable binding precipitated by PEG, respectively. Total 125I-VIP in each tube was 31 100 cpm. Anti-IgG precipitated 107.9% of the PEG-precipitable binding by this plasma. Values plotted are means of three replicates. The coefficient of variation for each of the points was <5%.
(Proteinpak 300 sw), a value close to that of authentic IgG (14.8 min) (not shown). The VIPbinding activity in a plasma from another asthma subject (No. A-3) was retained by immobilized protein G, an agent that binds the Fc portion of IgG, and was then released by treatment at low
leg
1
m'
2.01
q"
t.el
x
.
3.
O m =E o.
2-
0.11t
I-
0.44
Q Z
0.8
:E in o
4 0
4.
~
1.e
1.2'
Z -j
O4
o
1
~'.0.
0 J .~ 0
,1)
2'o
so
EFFLUENT V O L U M E , ml
Fig. 2. DEAE-ceUulose chromatography of the VIP-binding activity in plasma (code A-5). The column effluent was screened for absorbance at 280 nm (A2s0) and saturable i25I-VIP binding precipitable by PEG was measured in 25/~1 aliquots of the colUmn fractions. The elution position of authentic IgG is shown by the arrow.
0
;
;
;o
,'5
~o
2"6
~o
E F F L U E N T VOLUME, ml
Fig. 3. Affinity purification of the VIP-binding antibodies on protein G-Sepharose. Plasma (1 ml) from an asthmatic subject (code A-3) was applied to the column, the column washed until /128o in the effluent had returned to baseline, and retained material was eluted with a pH 2.7 buffer (shown by the arrow). Saturable ]25I-VIP was measured in duplicate in 25 #1 aliquots of the column fractions using goat anti-human IgG antibodies to precipitate antibody-bound VIP.
pH (Fig. 3). This IgG fraction contained 96% of the binding activity originally present in the plasma (0.18 pmol 125I-VIP/ml).
Antibody affinity The binding affinity of VIP antibodies is likely to be an important factor determining the biological consequences of their presence. We determined the binding affinities of the antibodies present in seven asthma patients and eight non-asthmatic subjects. Displacement of a25I-VIP binding with increasing concentrations of unlabeled VIP was assayed (Fig. 4) using PEG to separate antibodybound VIP. The a25I-VIP binding in these experiments largely reflected binding of the peptide by IgG antibodies. This conclusion was based on the observations that: (i) 88.9% _+ 2.5% of the PEGprecipitable binding of the plasma samples assayed for binding affinity (n = 15) was precipitable by anti-human IgG antibodies (Table 2), and (ii) the VIP binding activity was present exclusively in IgG fractions prepared from two of these plasma samples by chromatography on DEAE-cellulose, size exclusion columns and protein GSepharose. Values for apparent K a (affinity constant) and binding capacity were obtained by computer-assisted analysis of the displacement
138 100.
T
100
o z
.
60
~"
o.2 0-
~ 60-
m W ._1
-~
6
B, Dmoles/rnl
o z z
80-
6O-
\
1~ 40
\\\
o
'x,\\ ix \,
%
VIP, - Log M
O.
Fig. 4. Saturable binding of 12SI-VIP by VIP antibodies in plasma from an asthma patient (code A-4) versus concentration of unlabeled VIE The Scatchard plot (inset) obtained using the computer program LIGAND (Elsevier-Biosoft, Cambridge, U.K.) was curvilinear, suggesting two antibody populations with K a 10.1 and 0.1 nM -1 and binding capacity 0.6 and 18.3 pmol VIP/ml plasma, respectively.
data (McPherson, 1985). Scatchard plots of the d a t a showed a single linear c o m p o n e n t (seven plasma samples) or were resolved into two linear c o m p o n e n t s (eight plasma samples), suggesting the presence of one and two classes of antibodies, respectively. The mean K a values for the high affinity autoantibodies of asthmatic (n = 7) and
T
11
0
10 v Z t-u3 Z 0 C)
I
9.
8.
I |
2 t.L
<
7. NONASTHMATICS
__~ 40. (D k~ CL
ASTHMATICS
Fig. 5. VIP-binding affinity of the antibodies present in asthma patients and healthy non-asthmatic subjects, computed from Scatchard plots such as the one in Fig. 4. Each point represents a different subject. The mean K a values for the antibodies present in the two groups of subjects are shown by broken lines.
-
11
-
-
10 PEPTIDE,
9 -log
-
I
8 M
Fig. 6. 125I-VIPbinding by antibodies in unfractionated plasma (e; code A-5) and its IgG fraction (o): displacement by unlabded VIP. Values, means of duplicates, are percent of binding by the plasma (4827 cpm) or IgG prepared by DEAEcellulose chromatography (3290 cpm) in the absence of unlabeled VIP.
non-asthmatic subjects (n = 8) were 7.8 × 109 M -1 and 0.13 X 109 M - I ( P < 0.005, t-test) (Fig. 5), and their V I P binding capacity values (in p m o l V I P / m l plasma) were 0.58 +__0.05 and 55.7 4- 39.2, respectively. The mean K a values for the low affinity a n t i b o d y of asthmatic (n = 5) and nonasthmatic subjects (n = 3) were similar (0.04 x 109 M -1 and 0.08 x 109 M -1, respectively; P > 0.05). The VIP-binding capacities (mean) for the low affinity antibodies of asthmatic and non-asthmatic subjects were 24.7 and 152.6 p m o l / m l , respectively. The a p p a r e n t affinity of these antibodies measured in this study m a y be influenced by contaminating factors present in plasma, including peptide hydrolases (although the plasma was collected in a mixture of peptide hydrolase inhibitors; see Materials and Methods) and peptides cross-reactive with V I P antibodies. To investigate the contribution of such factors, we c o m p a r e d the ability of V I P to displace a2SI-VIP binding by unfractionated plasma and its purified I g G fraction (Fig. 6). The two displacement curves were nearly superimposable and K a values calculated for the b i n d i n g b y plasma and purified I g G were similar (6.6 x 109 M -1 and 7.1 X 109 M -1, respectively; P > 0.05). Furthermore, the propor-
139
tions of 125I-VIP precipitated with 10% trichloroacetic acid after incubation of the peptide with plasma (n = 3) or assay diluent under conditions employed for antibody affinity measurements were comparable (73% and 91%, respectively). These observations suggest that peptide hydrolases and other agents present in plasma were not significant factors in the observed antibody-VIP interactions.
25I
o x
• 20.
12. O
c5 o m a.
15
•
10 •
_ s /J'" u3 t"4
0
Antibody specificity PHI, GRF and secretin, peptides partially identical to VIP in their amino acid sequence, were employed to examine the specificity of the antibodies. These peptides (1 #M) did not significantly displace the 125I-VIP binding by plasma from six asthma patients and four non-asthmatic subjects (Table 3). The plasma antibodies in one asthmatic and one non-asthmatic subject showed partial reactivity with PHI, GRF and secretin (21.9% to 33.4%). The poor reaction of the antibodies with these PHI, GRF and secretin suggests their high level of specificity for VIP.
SPECIFICITY O F P L A S M A VIP A N T I B O D I E S
A-3 (1 : 12) A-4 (1 : 6) A-5 (1 : 5) A-6 (1:4) A-8 (1:3) A-11 (1 : 3) N A - 4 (1 : 10) NA-5 (1 : 10) NA-7 (1 : 10) NA-8 (1 : 3)
I
[
I
I
40
60
80
Ioo
TOTAL 125 I-VIP,
CPM x 1 0 - 5
Fig. 7. Saturable binding of 125I-VIP by receptors in guinea pig lung membranes: effect of i m m u n e IgG from an asthmatic subject (code A-5) (A), n o n - i m m u n e IgG from a non-asthmatic subject ( • ) and assay diluent (0). IgG was purified by DEAEcellulose chromatography. 125I-VIP was preincubated with an equal volume of i m m u n e IgG (24 m g / m l , a concentration sufficient to bind 53.2% of the 125I-VIP), n o n - i m m u n e IgG or assay diluent. Receptor binding by the nSI-VIP was determined (see Materials and Methods). Total a m o u n t s of 1251VIP available for receptor binding are plotted on the x-axis. Values are m e a n s of three replicates each. The coefficient of variation for each of the points was < 7.5%.
Effect of the antibodies on VIP-receptor binding
TABLE 3
Subject code a
I
2O
% 125I-VIP displaced b VIP
PHI
GRF
Secretin
100 (2 589) 100 (5 066) 100 (4582) 100 (1350) 100(4173) 100 (3 686) 100 (2600) 100 (4512) 100 (1 855) 100 (3 567)
25.0 c - 3.9 2.7 4.1 -1.8 1.2 8.3 3.1 - 7.0 31.6 c
33.4 c 1.8 0.5 8.8 -1.2 6.7 3.3 2.9 12.1 33.0 c
21.9 1.5 5.1 11.9 -0.9 - 2.2 1.2 3.4 9.0 8.9
a Subject type is denoted by ' A ' (asthmatic) and ' N A ' (nonasthmatic). Values in parentheses are the dilutions at which the plasmas were assayed. b 125I.VI P binding was assayed in absence and presence of the peptides (1 /xM). The values, means of duplicates, are percent of displacement observed with VIP, with the c p m radioactivity displaced by VIP in parentheses. Non-specific binding ranged from 1563 to 3862, and the total available radioactivity was - 30000 cpm. c p < 0.05 versus binding in the absence of unlabeled peptide.
We have previously identified and characterized specific receptors for VIP in guinea pig lung membranes (Paul and Said, 1987; Paul et al., 1988). In the present study, the influence of the human antibodies on VIP binding by these receptors was investigated. 125I-VIP was preincubated with the receptor assay diluent, immune IgG purified from plasma of an asthma patient (code A-5) or non-immune IgG from a healthy subject. Saturable binding of increasing concentrations of the IESI-VIP by receptors in guinea pig lung membranes was then measured. The preincubation with immune IgG resulted in significantly reduced receptor binding at each of the uS I-VIP concentrations tested (by 39% to 59%; P<0.001 versus non-immune IgG) (Fig. 7). The nSI-VIP receptor binding curves after preincubation with non-immune IgG or receptor assay diluent were nearly indistinguishable. The free and antibody-bound VIP concentrations available to the receptors after preincubation with immune IgG were not measured. It is difficult, therefore, to obtain a reliable
140 estimate of receptor affinity in the presence of anti-VIP in this experiment. Scatchard analysis of receptor binding observed in the presence of nonimmune IgG suggested a binding site with Ka 5.3 X 10 9 M - 1 and binding capacity 0.9 pmol VIP/mg protein, consistent with previously observed values for the high affinity VIP receptor in guinea pig lung membranes (Paul and Said, 1987; Paul, 1988).
Discussion
This study demonstrates the presence of VIPbinding activity precipitated by polyethylene glycol (PEG) in the plasma of 29 plasma samples from, asthma patients and healthy non-asthmatic subjects. Immunoprecipitation and chromatographic analyses suggested that the VIP binding Was attributable largely to IgG antibodies. The VIP-binding IgG antibodies were detected by measuring their binding to porcine 125I-VIP. Human and porcine VIP are structurally identical (Itoh et al., 1983). Thus the porcine VIP-reactive antibodies found in asthma patients can be considered to represent autoantibodies. Previously, Bloom et al. (1979) observed that diabetics positive for plasma VIP antibodies had been treated with insulin preparations contaminated with VIP, suggesting that the formation of the antibodies was related to the VIP contaminant. In contrast, the autoantibodies observed in the present study were naturally occurring. The antigenic stimulus leading to formation of these autoantibodies cannot be identified with certainty. Candidate stimuli include exposure to viral determinants similar in sequence to VIP (e.g., peptide-T, an epitope found on the human immunodeficiency virus (Ruff et al., 1987)) and dietary ingestion of avian or fish VIP known to be structurally different from human VIP (Nilsson, 1975; Dimaline and Thorndyke, 1986). Another potential stimulus for VIP autoantibody formation is muscular exercise. We have observed that VIP autoantibodies are present more frequently in healthy subjects who habitually perform muscular exercise compared to those who do not (Paul and Said, 1988). Muscular exercise resuits in increased plasma VIP immunoreactivity (Paul et al., 1987a; Woie et al., 1987), an effect
shown to be more pronounced in subjects with exercise-induced asthma compared to non-asthmatic subjects (Hvidsten et al., 1986). Investigation of the molecular form(s) and antigenicity of VIP released during exercise may reveal more about the relationship between exercise and VIP autoantibody formation in healthy subjects. Irrespective of the type of antigenic stimulation leading to VIP autoantibody formation, these antibodies may possess important effects. To evaluate the potential functional role of the antibodies, we measured their apparent binding affinity (Ka). Since unfractionated plasma samples were used for K a measurements, peptide hydrolases and cross-reactive peptides were potential factors in the observed interaction of VIP with the antibodies. Peptide hydrolase inhibitors (EDTA, PMSF, aprotinin and pepstatin A) were added to the plasma samples to minimize degradation of VIP. The IgG purified from one of the plasma samples exhibited a K a value (7.1 × 10 9 M - 1 ) close to that observed for unfractionated plasma (6.6 × 10 9 M-1), suggesting a lack of interfering factors. The mean apparent g a for the antibodies in asthma patients was considerably larger (by 60-fold) than in healthy non-asthmatic subjects. The range of K a values observed for the high affinity autoantibodies of asthma patients (0.25-26.2 × 10 9 M - 1 ) was similar to that reported for VIP receptors present in the lung and other tissues (Rosselin, 1986; Carstairs and Barnes, 1987; Paul and Said, 1987). Moreover, preincubation with a purified antibody fraction from an asthmatic subject inhibited the saturable binding of VIP by guinea pig lung membranes, suggesting that the antibody was probably directed against an epitope involved in VIP receptor binding. These considerations suggest that VIP autoantibodies found in asthmatics could neutralize VIP receptor binding and the ensuing VIP-medi~ted biological effects. The mean concentration of VIP in plasma from asthma patients, measured by radioimmunoassay (Paul et al., 1987b), was 1.33 pmol/ml (n = 16). Computed on the basis of mean antibody affinity and binding capacity values (Rodbard, 1981), 82% and 22% of the plasma VIP was likely to be present bound to the high and low affinity antibodies, respectively. However, as a neurotransmitter, the smooth muscle relaxant action of VIP
141 in the airways would be determined mainly by the local, as opposed to the circulating levels of the peptide. Immunoglobulins are present in respiratory secretions, with the IgG levels approximating those in the plasma (Reynolds, 1987). IgG is thought to arrive in the airways mainly by transudation from the blood compartment, although secretion from lymphocytes present within the airways is an additional contributory factor (Hance et al., 1988). Airway inflammation, a common finding in asthmatic subjects, is likely to promote passage of antibodies from blood to airway fluids. Thus VIP autoantibodies observed in the blood of asthma patients in the present study may reach the microenvironment of airway smooth muscle. Here the antibodies could interfere with airway relaxation by VIP released from nerve terminals supplying the smooth muscle. Bronchial hyperresponsiveness in asthma may arise from an imbalance in the mechanisms that constrict and relax the airways (Szentivanyi, 1968; Boushey et al., 1980). The observations reported in this study are consistent with a role for VIP autoantibodies in the pathogenesis of asthma. Asthma has been linked to many different neuronal and immunological mediators known to govern the activity of airway smooth muscle (Nadel and Barnes, 1984; Nadel et al., 1987), and the pathophysiology of airway hyperresponsiveness in different individuals may be different. Our observation that the VIP antibodies were only present in some of the asthma patients is in line with this possibility. We were unable to find a clear difference between antibody-positive and antibody-negative asthmatic subjects with respect to age, sex, number of years the patients had had asthma, presence of obstruction judged by pulmonary function tests and type of inciting agent resulting in asthma episodes. However, we analysed only 13 antibody-positive subjects, and data from a larger number of patients may be more revealing. In addition to their well-documented role in disease (Rose and Mackay, 1985), autoantibodies may mediate physiologic functions (Kay et al., 1982; Khandsari et al., 1983). The significance of the low affinity VIP autoantibodies present in healthy subjects requires further study. Considerations relevant to the role of VIP autoantibodies in healthy subjects include: (i) the antibodies may
predate the appearance of some functional abnormality, (ii) because they possess a low VIPbinding affinity (mean K~ 0.13 × 109 M - I ) , the antibodies are likely to interfere with VIP-mediated effects only at concentrations greatly in excess of tissue VIP receptor concentrations, (iii) because their affinity is greater than the 1/Km value (inverse Michaelis constant) reported previously for VIP-degrading enzyme(s) (105 M -1) (Keltz et al., 1980), the antibodies could effectively inhibit the proteolytic inactivation'of VIIa and thus increase the availability of the peptide at tissue receptors. In conclusion, this study suggests that humoral autoimmunity is a potential modulatory factor in the biological actions of VIP in asthmatic and healthy humans. VIP possesses a broad range of actions in different organ systems. Further study of the pathophysiological significance of VIP autoantibodies is clearly warranted. Factors that are likely to influence the biological consequences of the VIP autoantibodies include their binding affinity; epitope specificity, site of accumulation and ability to activate inflammatory cells via Fc receptor binding. References Bloom, S.R., Barnes, A.J., Adrian, T.E. and Polak, J.M. (1979) Autoimmunity in diabetics induced by hormonal contaminants of insulin. Lancet i, 14-17. Boushey, H.A., Holtzman, M.J., Sheller, J.R. and Nadel, J.A. (1980) Bronchial hyperreactivity. Am. Rev. Respir. Dis. 121, 389-413. Carstairs, J.R. and Barnes, P.J. (1987) Visualizationof vasoactive intestinal peptide receptors in human and guinea pig lung. J. Pharmacol. Exp. Ther. 239, 249-255. Dimaline, R. and Thorndyke, M.C. (1986) Purification and characterization of VIP from two species of dogfish. Peptides 7 (Suppl. 1), 21-26. Hance, A.J., Saltini, C. and Crystal, R.J. (1988) Does de novo immunoglobulin synthesis occur on the epithelial surface of the human lower respiratory tract? Am. Rev. Respir. Dis. 137, 17-24. Hvidsten, D., Jenssen, T.G., Bone, R. and Burhol, P.G. (1986) Plasma gastrointestinal regulatory peptides in exercise induced asthma. Eur. J. Respir. Dis. 68, 326-331. Itoh, N., Obata, K.-I., Yanaihara, N. and Okamoto, H. (1983) Human preprovasoactiveintestinal polypeptide contains a novel PHI-27-1ikepeptide, PM-27. Nature 304, 547-549. Keltz, T.N., Straus, E. and Yalow, R.S. (1980) Degradation of vasoactive intestinal polypeptide by tissue homogenates. Biochem. Biophys. Res. Commun. 92, 669-674.
142 Kay, M.M,B., Sorenson, K., Wong, P. and Bolton, P. (1982) Antigenicity, storage and aging: physiologic autoantibodies to cell membrane and serum proteins and the senescent cell. Mol. Cell. Biochem. 49, 65-85. Khandsari, N. and Fudenberg, H.H. (1983) Immune elimination of aging platelets by autologous monocytes: role of membrane specific autoantibody? Eur. J. Immunol. 13, 990-994. Laitinen, A., Partanen, M., Hervonen, A., Peto-Huikko, M. and Laitinen, L.A. (1985) VIP like immunoreactive nerves in human respiratory tract. Light and electron microscopic study. Histochemistry 82, 313-319. Matsuzaki, Y., Hamsaki, Y. and Said, S.I. (1980) Vasoactive intestinal peptide: a possible transmitter of nonadrenergic relation of guinea pig airways. Science 210, 1252-1253. McPherson, G.A. (1985) Analysis of radioligand binding experiments. A collection of computer programs for the IBM PC. J. Pharmacol. Methods 14, 213-228. Nadel, J.A. and Barnes, P.J. (1984) Autonomic regulation of the airways. Annu. Rev. Med. 35, 451-467. Nadel, J.A., Pauweis, E. and Snashall, P.D. (Eds.) (1987) Bronchial Hyperresponsiveness, Blackwell, Oxford. Nilsson, A. (1975) Structure of the vasoactive intestinal peptide from chicken intestine. The amino acid sequence. FEBS Lett. 60, 322-326. Paul, S. (1988) Decreased selectivity of vasoactive intestinal peptide receptors by GTP. Biochem. Pharmacol. 38, 699-702. Paul, S. and Said, S.I. (1987) Characterization of receptors for vasoactive intestinal peptide from the lung. J. Biol. Chem. 262, 158-162. Paul, S. and Said, S.I. (1988) Human autoantibody to vasoactive intestinal peptide: increased incidence in muscular exercise. Life SCI. 43, 1079-1084. Paul, S., Wood, K. and Said, S.I. (1984) Purification of [125I]vasoactive intestinal peptide by reverse-phase HPLC. Peptides 5, 1085-1087.
Paul, S., Erian, P. and Said, S.I. (1985) Autoantibody to vasoactive intestinal peptide in human circulation. Biochem. Biophys. Res. Commun. 130, 479-485. Paul, S., Chou, J., Beckham, S., Liu, L.-W., Kubota, E., Dominguez, N. and Said, S.I. (1987a) Elevated levels of atrial natriuretic peptide and vasoactive intestinal peptide in exercising man. Clin. Res. 35, 112A (abstract). Paul, S., Chou, J. and Kubota, E. (1987b) High-affinity peptide-histidine-isoleucine preferring receptors in rat liver. Life SCI. 41, 2373-2380. Paul, S., Chou, J. and Said, S.I. (1988) Regulatory aspects of the VIP receptor in lung. Ann. N.Y. Acad. Sci. 527, 282-295. Reynolds, H.Y. (1987) Identification and role of immunoglobulins in respiratory secretions. Eur. J. Respir. Dis. 71 (Suppl. 153), 103-116. Rodbard, D. (1981) Radioimmunoassay data processing. In: Syllabus, Tenth Training Course on Hormonal Assay Techniques, The Endocrine Society, Bethesda, MD, p. 25. Rose, N.R. and Mackay, I.R. (Eds.) (1985) The Autoimmune Diseases, Academic Press, London. Rosselin, G. (1986) The receptors for the VIP family peptides (VIP, secretin, GRF, PHI, PHM, GIP, glueagon and oxyntomodulin). Specificities and identity. Peptides 7 (Suppl. 1), 89-100. Ruff, M.R., Martin, B.M., Ginns, E.I., Farrar, W.L. and Pert, C.B. (1987) CD4 receptor binding peptides that block HIV infectivity cause human monocyte chemotaxis. FEBS Lett. 211, 17-22. Said, S.I. (1984) Vasoactive intestinal peptide (VIP). Current status. Peptides 5, 143-150. Said, S.I. (1987) Influence of neuropeptides on airway smooth muscle. Am. Rev. Respir. Dis. 136, $52-$58. Szentivanyi, A. (1968) The fl-adrenergic theory of the atopic abnormality in bronchial asthma. J. Allergy 42, 203-232. Woie, L., Kaada, B. and Opstaad, P.K. (1987) Increase in plasma VIP in muscular exercise. Gen. Pharmacol. 17, 321-326.