- Email: [email protected]
Oral feeding of renal tubular antigen abrogates interstitial nephritis and renal failure in Brown Norway rats
Kidney International, Vol. 52 (1997), pp.
725—732
Oral feeding of renal tubular antigen abrogates interstitial nephritis and renal failure in Brown Norway rats KEVIN PHAM, WILLIAM E. SMOYER, D. ClAY ARCHER, F1NcIs GuBM, and CAROLYN J. KELLY Division of Nephrology-Hypertension, Department of Medicine, University of California, San Diego, and the Veterans Affairs Medical Center, San Diego, California; and the Department of Pediatrics, University of Michigan, Ann Arbor, Michigan, USA
Oral feeding of renal tubular antigen abrogates interstitial nephritis and renal failure in Brown Norway rats. We have examined whether oral feeding of antigen can regulate the expression of autoimmune interstitial nephritis induced by antigen-in-adjuvant (RTA'CFA) immunization of Brown Norway rats. Male rats were divided into six experimental groups:
anergy, depending the characteristics of the antigen and feeding protocol [reviewed in 3]. Our laboratory has had a longstanding interest in developing modalities of targeted immunosuppression in autoimmune renal
disease. One model system we utilize is that of anti-tubular basement membrane (TBM) disease with interstitial nephritis. This disease is induced in susceptible strains of guinea pigs, rats, and mice by a single immunization of a renal tubular antigen Renal histology, inulin clearance, DTH responses to RTA, and IgG (RTA) emulsified in complete Freund's adjuvant (CFA) [9—li]. antibody responses to RTA were monitored as endpoints of the study. Our The resultant humoral and cell-mediated immune response to results demonstrated that Group III and IV animals had significantly less RTA results in deposition of specific IgG along the tubular severe renal injury, as assessed by inulin clearance and extent of renal cortical involvement by mononuclear cells. Group II and IV animals had basement membrane and mononuclear cell infiltration of renal interstitium. This mononuclear cell infiltrate includes both CD4 suppressed DTH responses, and only Group IV animals had significantly Group I, RTAJCFA immunization alone; Groups II, III, and IV were
pretreated with 1 mg (Group II), 5 mg (Group III), and 25 mg (Group IV) of oral tubular antigen every other day for ten days, followed by RTAICFA immunization; Group V was pretreated with a control antigen, followed by RTAJCFA immunization; and Group VI was immunized with CFA alone.
depressed antigen-specific IgG serum titers. Group III animals had neither suppressed DTH responses or IgG titers. We conclude that oral administration of tubular antigen can modulate the intensity of interstitial nephritis produced by immunization, but that the regulatory mechanism is not dependent (at all doses of fed antigen) on suppressed DTH reactivity to RTA or suppressed antigen-specific IgG.
The term oral tolerance refers to the induction of systemic hyporesponsiveness to an antigen following oral administration of this antigen. Although the phenomenon was initially described in
the first half of this century [1, 2] there has been a recent resurgence of interest in it because of its potential as a therapeutic
and CD8 T cells, as well as macrophages, natural killer cells, and B cells [121. The relative importance of these various immune cell subsets, as well as antigen-specific lgG, in mediating renal injury
varies among species [13]. For example, the passive transfer of anti-TBM antibodies can induce disease in guinea pigs [14, 15], to a lesser degree in rats [16], and not at all in mice. Susceptibility to disease depends, at the minimum, on expression of discrete MHC
genes [11], expression of antigen in the kidney [17], and the development of a susceptible T cell phenotype following immunization [18—20]. The animals develop progressive renal failure. Previous work in this model has documented that this nephritogenie immune response can be modulated by immunosuppressive drugs [21, 22] dietary manipulation [231, PGEI [24], and through the induction of regulatory T cells [9, 25—28]. Regulatory T cells can be induced through the presentation of antigen in a tolerogenie manner, that is, emulsified with IFA rather than CFA [9], or
modality for autoimmune disease [3]. Oral feeding of proteins that are target antigens of autoimmune T and B lymphocytes can protect the host against subsequent immunization with those antigens to induce disease. Amelioration of organ pathology by coupled to the surface of syngeneic splenocytes and injected this approach has been demonstrated in experimental allergic intravenously [27]. In the studies reported here, we examined encephalomyelitis [4, 5], myasthenia gravis [6], uveoretinitis [7], whether oral administration of renal tubular antigen can alter the and collagen arthritis [81. The mechanisms by which oral feeding development of interstitial nephritis. We find that oral feeding of of protein antigens alter the nature of the subsequent immune RTA can abrogate the severity of the induced interstitial ncphriresponse to antigenic challenge are complex. Recent work has tis, and that this effect is demonstrable at doses of fed antigen at demonstrated that the phenomenon of oral tolerance can be which anti-1BM lgG titers and DTH reactivity arc not altered. attributed to active suppression by T lymphocytes or by clonal METHODS Key words: oral feeding in TIN, interstitial nephritis, antigens, renal failure, autoimmune disease.
Animals
Received for publication August 16, 1996 and in revised form February 27, 1997 Accepted for publication April 7, 1997
from Charles River Laboratories (Boston, MA, USA) and housed in the animal facility at the San Diego Veterans Affairs Medical
© 1997 by the International Society of Nephrology
Center. All care of animals was performed under guidelines
Male Brown Norway rats weighing 200 to 25(1 g were purchased
725
726
Pham et al: Oral tolerance in interstitial nephritis
approved by the American Association for the Accreditation of challenge by an independent observer blinded to the identity of Laboratory Animal Care. the experimental groups, using a spring-loaded engineer's micrometer. The change in footpad thickness was recorded for each Immunization of animals animal and the result expressed as the mean for each experimenRabbit renal tubular basement membranes were isolated by a tal group SEM. differential sieving technique, sonicated, lyophilized, and stored at —20°C [9]. Lyophilized renal tubular antigen was emulsified in Assessment of renal function At 17 to 18 days following initial immunization, all rats undercomplete Freund's adjuvant (RTAJCFA). Antigen concentrations for both RTA and PPD (M.Th. H37Ra; Difco Laboratories, went an inulin clearance as a measure of glomerular filtration Detroit, MI, USA) were 4 mg/mI. After the induction of anesthe- rate, using established techniques. Briefly, following anesthesia sia with Brevital (7 mg/100 g body wt., i.p.; Eli Lilly & Co., with mactin (10 mgIlOO g body wt, i.p.; BYK, Konstanz, GermaIndianapolis, IN, USA), animals were immunized subcutaneously ny), animals were tracheally intubated, and catheters were ininto four sites in the axillary and groin regions with a total of 0.75 serted in the jugular vein, femoral artery, and urinary bladder. ml of RTA/CFA. One group of animals received 0.75 ml of CFA The femoral artery catheter was used for periodic blood sampling alone (4 mg/mI PPD) as a control for the immunization proce- and monitoring mean arterial pressure with a transducer (model dure. In these studies no animal received footpad immunization P23DB; Statham Instruments, Gould Divsion, Hato Rey, Puerto Rico) and recorded on a chart recorder. All rats received an with RTA/CFA or CFA. infusion of NaC1-NaHCO3 throughout the experiment containing Feeding protocol (3H) inulin at a rate of 2.2 ml/hr. After 60 minutes of equilibraFor the three groups of rats that received RTA or control tion, plasma and urine samples were collected for three consecuantigens (ovalbumin or hemoglobin; Sigma Chemical Co., St. tive 20-minute periods from femoral artery and urinary bladder, Louis, MO, USA) in the food, the antigen was mixed thoroughly respectively. Aliquots of plasma and urine samples were counted in powdered rat chow prior to feeding. Animals were housed in in a scintillation counter (Packard Instruments, Downer's Grove, individual cages and fed 14 g of standard rat chow daily for a IL, USA). Glomerular filtration rate was calculated from the period of ten days. (This quantity of powdered rat chow was radioactivity counts of the plasma and urine aliquots and the selected because it was completely consumed daily). Antigen in volume of urine collected per 20 minute period. designated amounts was mixed with the food on alternate days. The two groups of rats that did not receive RTA or control ELISA for anti-RTA antibodies ELISA microtiter plates were coated overnight at 4°C with antigens in the food were also housed individually and given 14 g of food daily for ten days. After the 10-day regimen of food intake, collagenase solubilized RTA at 40 jig/mI. Following washing with all rats were fed normal rat chow and water ad libitum and housed dH2O, the plates were blocked for two to three hours at room temperature with PBS containing 4% BSA, 0.05% Tween 20. at three rats per cage. Following extensive washing with Tris-buffered saline with 0.05%
Experimental groups Tween 20, pH 7.6, multiple dilutions of test sera from individual Animals were divided into six groups. Group I (N = 6) was animals were placed in wells, and incubated at 37°C for 2.5 hours. immunized with RTA/CFA two days following completion of Following extensive washing, the appropriate secondary reagent
feeding regimen. This group served as a positive control for [goat anti-rat IgG (H+L), goat anti-rat IgGI or goat anti-rat disease severity. Group II (N = 4) was fed with I mg of RTA in powdered rat chow on alternate days as described above (total of 5 mg fed antigen). Group III (N = 8) was fed with 5 mg of RTA
IgG2a diluted in PBS with 4% BSA, 0.05% Tween 20, all alkaline phosphatase conjugated] was added and incubated for 2.5 hours at 37°C. Following washing, the plates were developed with substrate
hemoglobin as control antigens. All animals in groups II to V were immunized with RTA/CFA two days following completion of the feeding regimen. Group VI (N = 4) was immunized with CFA two
in carbonate buffer, pH 9.6, for the appropriate time period (45 mm for total IgG, 5 hr for IgG subtype ELISA). The O.D. at 405 nm was measured in a microplate reader. The best linear curve fit for each group (as determined by the linear least-squares regression method) was calculated, and the serum titer for each group was calculated as the serum dilution at which the test sera O.D. 405 nm was equal to the O.D. 405 for CFA serum.
in powdered rat chow on alternate days (total of 25 mg fed antigen). Group IV (N = 4) was fed with 25 mg of RTA in powdered rat chow on alternate days (total of 125 mg fed antigen). Group V (N = 5) was fed with either ovalbumin or
days following completion of feeding regimen. This group was used as a control for normal renal function. Previous work from our laboratory has shown that at one and three weeks following CFA immunization, glomerular filtration rate is indistinguishable from that of animals never immunized (C. Kelly and F. Gabbai, unpublished data). Delayed-type hypersensitivity testing At 13 to 15 days following subcutaneous immunization with RTA/CFA, all rats received a challenge dose of antigen, consisting of 75 jd of I mg/ml collagenase solubilized RTA in PBS in one
Assessment of histologic renal disease Animals were sacrificed following the clearance measurements (at day 18 following immunization). The left kidney of each rat was harvested and immersion fixed in 10% buffered formalin. The
extent of interstitial involvement was determined by a single observer from 4 mm, paraffin-embedded kidney sections stained with hematoxylin and eosin. The specimens were coded to prevent observer bias, and graded at a single session by an observer who additionally was unaware of the results of the clearance measurements. Cellular infiltration and tubular damage leading to cortical
hind footpad. The other hind footpad was injected with PBS. interstitial destruction was graded on a scale of 0 to 4: 0 = no Footpad thickness was measured before and 24 hours after lesions; 1.0 or greater = 10 to 25% cortical involvement by
727
Phain et at: Oral tolerance in interstitial nephritis
5
2.5
4
2
Cl)
ci)
a)
C)
1.5
(ci—
0 C
*
B C
1
0.5
1
0
0 II
Ill
IV
V
VI
Experimental groups
Fig. 1. Oral feeding of RTA can diminish the severity of interstitial nephritis. Experimental groups (designated by Roman numerals on the
horizontal axis) of Brown Norway rats were fed with varying total doses of RTA prior to immunization to produce interstitial nephritis. The feeding
II
Ill
IV
V
VI
Experimental groups Fig. 2. Oral feeding of antigen leads to improved renal function. Inulin
clearances were performed on the same experimental groups outlined in the Figure 1 legend and the Methods section. *P < 0.05 compared with Group I.
protocols are outlined in the Methods section. Experimental group designations are: I, RTAICFA immunized with no oral antigen fed; II, fed with I mg RTA every other day for ten days prior to immunization with RTA/CFA; HI, fed with 5 mg RTA every other day for ten days prior to immunization with RTAJCFA; IV, fed with 25 mg RTA every other day for ten days prior to immunization with RTAICFA; V, fed with control Ag prior to immunization with RTAJCFA; VI, no antigen feeding, immunized with CFA alone. Group VI animals have no detectable disease. p < 0.05 compared with Group I.
inflammatory cells; 2.0 or greater 25 to 50% cortical involvement by inflammatory cells; 3.0 or greater = 50 to 75% cortical involvement by inflammatory cells; 4.0 = 75% or greater cortical involvement by inflammatory cells. 4.0 was the maximum score assigned.
Statistical analysis Results are presented as mean
differences were statistically significant. CFA immunized animals
(Group VI) do not have detectable disease. Oral feeding of tubular antigen leads to preserved GFR following RTAJCFA immunization
To have a precise quantitative endpoint (GFR) by which to compare experimental groups, we performed inulin clearances on all animals at 17 to 18 days following RTAICFA immunization. These data are shown in Figure 2. The mean inulin clearance for the CFA immunized rats (Group VI) was 1.9 mi/mm, establishing
the control values for male BN rats of this weight. RTA/CFA immunized animals (Group I) had a markedly diminished GFR at
approximately 0.5 mI/mm. Oral antigen feeding had a dose dependent effect on GFR. Although the mean GFR for Group II was not significantly different than Group I, the mean GFR for
SEM. ANOVA with post hoc Groups III and IV, the 5x and 25 X oral antigen groups was tests was used for comparison among groups. The level of significantly different from Group I and, in fact, not significantly
statistical significance was defined as a value of P < 0.05. RESULTS
Oral feeding of tubular antigen blocks the development of interstitial nephritis following RTAICFA immunization
Figure 1 shows the effect of oral feeding of RTA on the histologic severity of the interstitial injury. The grading scale for interstitial injury extends from 0 to 4 and represents the extent of
different than that of the normal controls (Group VI). Control antigen feeding (Group V) did not significantly improve GFR. Close correlation between the GFR and histology score in Groups I-V Figure 3 depicts a plot of mean inulin clearance versus mean
histologic score for each experimental group immunized with RTA/CFA. There is a close inverse correlation between GFR and
cortical involvement by the mononuclear cell infiltrate. The extent of cortical interstitial inflammation (linear correlation positive control animals (Group I), those fed Ix oral antigen coefficient, r = 0.972). (Group II), and those fed the control Ags (Group V), all had mean histologic scores in the 2.75 to 3.0 range, signifying approximately 50% cortical involvement. Animals fed either 5x (Group
Antigen-feeding resulted in a dose-dependent inhibition of DTH responses to RTA
III) or 25x (Group IV) oral antigen had histologic scores of
The DTH response to RTA has correlated well in previous
approximately 1, signifying 10% cortical involvement. These latter
studies with the presence of nephritogenic lymphocytes and active
728
Pham et al: Oral tolerance in interstitial nephritis
3.5
60
3
50
• Group II Group V
w 2.5
00
cc
' 1.5
U 30
40
o?
IH Group IV
0.5
0
20
10
0.5
1.5 lnulin clearance
1
2
mi/mm
Fig. 3. A close correlation exists between inulin clearance and histologic score in the experimental groups. The mean inulin clearances from each experimental group (except Group VI, where the histologic score is zero) were plotted on the horizontal axis versus the mean histologic scores on the vertical axis. The best linear curve fit for these data (as determined by the linear least-squares regression method) was determined using CACricket Graph III" software (Computer Associates, International). The equation for this line is y = —1.70x + 4.21, r = 0.972.
*
0 II
Ill
I\/
\'I
Experimental groups Fig. 4. "Low" and "high" dose oral antigen feeding inhibit DTH reactiv-
ity to RTA. DTH testing for reactivity to RTA was performed on all experimental groups several days prior to performance of the inulin clearances. Footpads were read on unanaesthetized animals using an engineer's micrometer in cage-blind fashion. Results are depicted as the
change in footpad thickness measurement from that prior to RTA injection to that measured 24 hours following RTA challenge. p < 0.01 compared with group I.
interstitial nephritis [9, 20, 29]. Figure 4 illustrates the DTH responses to collagenase solubilized RTA at 14 days after immunization in all six groups of animals. The groups fed with repeated
biologic significance of this slightly depressed anti-TBM IgG titer is uncertain.
doses of 1 mg (Group II) or 25 mg (Group IV) of RTA
Since other groups have reported a preferential inhibition of demonstrated marked suppression of DTH responses to RTA ThI CD4 cells over Th2 responses at low doses of oral antigen, compared to Group I animals. Group Ill, fed with repeated doses we additionally examined the titers of the IgGi and IgG2a of 5 mg of RTA, had a DTH response not significantly different subtypes in the csTBM antibody responses in these animals. IgGI than Group I. The group fed with irrelevant antigens also responses typically correlate with a Th2 response, and IgG2a displayed no suppression of DTH to RTA, as expected. The responses correlate with a Thi response. We found no significant change in footpad thickness in response to PBS injection alone differences in the ratio of these subtypes compared with control ranged from a mean of 1.0 to 5.9 in the different groups. (The data from the PBS injected footpad are not depicted on Fig. 4 for the sake of clarity.) These experiments demonstrate that oral feeding of this antigen preparation is capable of modulating one marker of the cell-mediated immune response to RTA.
Antigen feeding does not affect the anti-TBM antibody response
We also examined the titers of aTBM IgG antibodies in the various experimental groups (Fig. 5). The mean O.D. readings versus serum dilution for the different experimental groups are depicted in Figure SA. The titers derived from these readings for each experimental group are compared in Figure SB. Although the mean O.D. readings at some serum dilutions were depressed for both Groups II and V, only group IV had a significantly
animals (data not shown). DISCUSSION
Our studies clearly demonstrate that oral feeding of RTA to Brown Norway rats prior to immunization with RTAJCFA can block the subsequent development of interstitial nephritis and renal insufficiency. The effect of antigen feeding is dose depen-
dent. Although certain doses of fed antigen had significant suppressive effects on DTH responses to RTA, the DTH effect did not correlate well with improvement in histologic disease or GFR. When one compares the two experimental groups in which there was improvement in GFR and histologic disease (Groups III and
IV), it is evident that neither inhibition of DTH reactivity to the immunogen, nor inhibition of the IgG antibody titer is required to
depressed total lgG aTBM Ab serum titer. Even with this slightly ameliorate this disease process within the time frame of our depressed log10 titer, the kidneys of Group IV animals still had 4 studies. positive linear TBM staining for IgG by direct immunofluoresThe term "oral tolerance" is used to describe the alterations in cence and were indistinguishable from the Group I animals by immune responsiveness seen following feeding of various antiintensity of immunofluorescence (data not shown). Therefore, the gens. While this widely used phrase evokes the protective nature
Pham et al: Oral tolerance in interstitial nephritis
A
729
require oral feeding. Similarly protective effects have been shown with aerosolized antigen [301 and antigen applied to the nasal mucosa [311. In these instances, the antigen processing by either bronchial or nasal mucosa associated lymphoid tissue presumably
0.6 0.5 -
induces the same type of protective response as does the gutassociated lymphoid tissue.
Our study is the first to demonstrate abrogation of renal inflammation and preservation of renal function by the oral
0.4 E
administration of a renal target antigen. Two previous studies in models of immune-mediated renal disease have supported the
0 0.3 IC)
notion that presentation of antigen via the gut mucosa can
0
modulate the expression of renal disease. Oral feeding of HSA can decrease the incidence of immune complex disease in the kidney despite continued presence of high levels of circulating immune complexes [32]. The decreased incidence of immune
0.2 -
complex disease was attributed to a possible change in the affinity
0.1 0
of the immune complexes. In a model of IgA nephropathy I
I
3
3.7
I
4 Log10 serum dilution
4.7
5
development of oral tolerance, thereby supporting a paradigm whereby defective oral tolerance facilitates the development of this form of glomerulonephritis [33]. The mechanisms underlying how oral feeding of antigen abrogates autoimmune responses have been evaluated in several other models of organ-specific autoimmune disease. Experimental allergic encephalomyelitis (EAE) has been the most extensively
4.75
studied model system [3]. Studies have been conducted in both the
B
rat and mouse models of this disease. In rat EAE, clinical and histologic manifestations of disease can be suppressed by low doses of fed myelin basic protein [4, 51. In this model, animals fed with MBP have a CD8 T cell population demonstrable in their spleen and mesenterie lymph nodes that can adoptively transfer protection against EAE [34, 35]. The antigenic epitopes of MBP recognized by these regulatory T cells are distinct from those
a)
C
0)
induced by oral feeding of bovine gamma globulin, the concomitant administration of cyclophosphamide or estradiol increased Ag-specific IgG responses, glomerular IgG immune deposits and the development of hematuria. Both drugs can interfere with the
4.5
0 -J
recognized by encephalitogenic T cells [36]. In the murine model of EAE, antigen feeding induces Th2-like CD4 regulatory T cells
4.25
that produce TGF-f3 and IL-4 and functionally suppress EAE induced with either MBP or proteolipid protein [37]. These studies suggest that inactivation, rather than deletion, of effector T cells is a generic operative mechanism in oral tolerance. This notion is further supported by recent studies utilizing a system of
4 I
II
Ill
V
Experimental groups Fig. 5. Anti-RTA IgC titers are suppressed only in Group IV animals. (.4)
The mean O.D. readings flr each experimental group at the indicated serum dilutions. Only positive error defiections (1 su) are shown. Icons for Groups Ito VI are as follows: () Group Group U; (0) Group III; (0) Group IV; () Group V; (+) Group VI. The best linear curve fit for each group (as determined by the linear leastsquares regression method)
I; ()
was calculated, and the serum titer taken as the serum dilution at which
the test sera 0D405 nm was equal to the 00405 for CFA serum. (B) Values for the linear curve fit. Only the Group TV data are significantly different. Where error bars are not depicted, they were insignificant in
size. P 0.05.
of the subsequent immune response to challenge immunizations with the same antigen (in the ease of autoimmune disease), it is not, strictly speaking, tolerance, nor does this protective effect
ovalbumin specific TCR transgenic lymphocytes transplanted into BALB/c recipients which were subsequently fed and challenged with ovalbumin [38]. in this system, the proliferative noriresponsiveness seen in fed animals could be normalized for the number of antigen specific cells. This calculation demonstrated a marked reduction in proliferation per lymphocyte expressing the ovalbumm specific TCR in antigen fed animals, supporting a tolerance mechanism of anergy (inactivation) rather than deletion [38]. It is
important to point out this kind of proliferative nonresponsiveness could be secondary to the induction of Th2-like cytokines in the TCR transgenic T cells. Previous experimental work in both the rat and murinc models of
18,
730
P/sam
et al: Oral tolerance in interstitial nephritis
24, 25, 29], which abrogate disease have consistently also abro-
TGF-p as an immunosuppressive cytokine would be fully consis-
gated Ag-specific DTH responses. The relationship between DTH-reactivity and nephritogenicity is more than a correlative presumption of pathogenicity. T cells purified from diseased kidneys mediate DTH to the tubular Ag, and renal-eluted DTH reactive cells can infiltrate the renal cortex following adoptive
tent with our previous observations regarding TGF-/3 in this model system [45—47]. This finding has important practical impli-
cations, as it implies that it is not necessary to purify a target
antigen to homogeneity, or even identify a target antigen, in order to induce this protective effect. Our own studies confirm this, as we used a renal tubular antigen preparation which consists of 15 to 20 distinct protein bands by SDS-PAGE [48]. Previous studies
subcapsular transfer into naive hosts [20, 39]. Moreover, tubular Ag-reactive T cell clones that can mediate DTH are also nephri. togenic following adoptive transfer [40, 41]. These findings have have demonstrated that rodents mount a significant immune
strongly supported the notion that DTH-reactive T cells are
response to a single component of RTA, the 3M—I protein,
important effector cells in experimental interstitial nephritis. The uncoupling of DTH reactivity to tubular Ag and protection from interstitial nephritis (as displayed by Group III in this study) was surprising to us. It is possible that DTH-reactivity to the Ag is an
following immunization with RTAJCFA [48]. While the remaining
the critical effector cell, but that antigen feeding protocols result
pressed in physical proximity to the target antigen could abrogate
proteins in the RTA preparation are innocuous from the stand-
point of generating a nephritogenic immune response, it is
possible that they generate protective responses when presented epiphenomenon and unrelated to the immunopathogenesis of via the gut mucosa, and thereby contribute to the induced disease. For the reasons outlined above, we do not favor this tolerance. Another implication of immunosuppression by Aginterpretation. Another possibility is that the DTH-reactive cell is nonspecific cytokines is that feeding with a single antigen ex-
in impaired homing to the kidney, perhaps through cytokine immune responses directed to the target antigen, by local produc-
induced alterations in integrins or selectin ligands. An alternative tion of immunosuppressive cytokines [491. This raises the possiinterpretation is that the oral feeding regimen results in the bility that feeding an "irrelevant" antigen, but one expressed in inactivation of a different T cell population, not DTH reactive, the tubules or interstitium, may stimulate a host protective which is required for interstitial injury. This could be a CD8 CTL response. population, or a functionally distinct CD4 subpopulation. The Our studies demonstrated equivalent protection by the antigen last two scenarios described above (impaired homing vs. a non- doses in Groups III and IV. These studies have not established the DTH-reactive effector cell) are not mutually exclusive. The mech- minimum total dose or timing necessary to induce oral tolerance. anism underlying the preserved DTH response in group III has We believe that the dose administered to Group II, 5 mg total, is not yet been elucidated. In preliminary studies, we have found on the border of what is effective. The large standard error that the mechanisms underlying suppression of DTH in Group II displayed in this group in both Figures 1 and 2 is the result of and IV animals are different. This finding suggests that Group III some animals displaying protection and other animals displaying animals received an antigen dose orally which did not activate none. The correlation between histology and inulin clearance
either suppressive mechanism. We were not surprised by the within Group II is strong, that is, those animals with minimal discordance between Ag-specific IgG titers, histologic renal dis- histologic abnormalities have the higher inulin clearances. We ease, and functional deterioration. Previous studies have demon- have not seen any protection with doses of this RTA antigen strated marked suppression of renal disease in this model, without preparation below this dose (data not shown). It is also possible alterations in circulating Ag-specific IgG or kidney bound IgG [9]. There was a striking correlation between histologic scores and
that the gastrointestinal absorption of the antigen is variable. The quantity of antigen used may be minimized by coupling to cholera toxin [50]. An important second-order study will be to determine the efficacy of oral Ag feeding after induction of disease. Preliminary data from several clinical trials suggested that Ag
GFR in our studies. We performed both measurements as they provided a complementary analysis. The inclusion of the inulin clearance contributed a "hard" endpoint to the analysis. Given the well-recognized inverse correlation between degree of interstitial feeding may have beneficial effects clinically in autoimmune damage and glomerular filtration rate in human renal disease disease, with minimal toxicity or side effects [51, 521, although [42-44], it is not surprising that an experimental model of more recent observations in non-human primates partially temper interstitial inflammation should result in impaired GFR. The this early enthusiasm for the therapeutic potential of immune close correlation between function and histology in the current deviation to the Th2 phenotype [53]. The appeal of this as a study is reassuring support for the use of the latter as an important potential therapy for human immune-mediated renal disease is endpoint. Despite this strong correlation, we still believe that both that the knowledge of a precise target antigen is probably not measurements are important. In the context of immunomodula- required. It is additionally possible that in complex lesions, such as tion, it is quite possible that an immunosuppressive modality that the interstitial inflammatory disease accompanying heavy proteinaltered the types of cytokines expressed by Ag-reactive cells might uric states, inhibition of an antigen-specific secondary immune not change the degree of mononuclear cell infiltration but would response targeted to the interstitium might block progression of improve renal function. In contrast, other immunosuppressive disease. Further work on oral tolerance in experimental renal agents that result in the virtual elimination of Ag-reactive lym- disease may provide a foundation for a novel approach to phocytes from the renal cortex might depress GFR, through either treatment of immune-mediated renal disease. functional or toxic mechanisms. The finding that oral administration of antigen can prime T ACKNOWLEDGMENTS cells which produce cytokines with immunosuppressive properties This work was presented in part at the 1995 Annual Meeting of the (that is, IL-4, IL-b, and TGF-13) suggested that the final mediator American Society of Nephrology and published in abstract form JAm Soc of immunosuppression was Ag-nonspecific. TGF-p in particular Nephrol 1995). The studies were supported by grants from the NIH
has been implicated as a mediator [36, 37]. A potential role for (DK42155 and DK45346) and the Office of Research and Development,
Pham et al: Oral tolerance in interstitial nephritis
Medical Research Service, Department of Veterans Affairs. Kevin Pham
received support from the Student Affairs Office of UCSD Medical School. We thank Martin Kagnoff, M.D. for helpful discussions and Ser Khang for assistance with clearance measurements.
731
19. MANN R, KELLY CJ, HINES WH, CLAYMAN M, BLANCh-lARD N, SUN
MJ, NEILSON EG: Effector T cell differentiation in experimental interstitial nephritis I. The development and modulation of effector lymphocyte maturation by 1-J regulatory T cells. J Immunol 136: 4200—4208, 1987 20. NEILSON EG, MCCAFFERTY E, MANN R, MICHAUD L, CLAYMAN M:
Reprint requests to Carolyn J. Kelly, M.D., UCSD and VAMC 111-H, 3350 La Jolla Village Drive, San Diego, California 92161, USA. E-mail: [email protected]
Murine interstitial nephritis. III. The selection of phenotypic (Lyt and L3T4) and idiotypic (RE-Id) T cell preferences by genes in Jgh-1 and H-2K characterize the cell-mediated potential for disease expression:
Susceptible mice provide a unique effector T cell repertoire in REFERENCES 1. WELLS H: Studies on the chemistry of anaphylaxis. III. Experiments
with isolated proteins, especially those of hen's egg. J Infect Dis 9:147—151, 1911
response to tubular antigen. J Immunol 134:2375—2382, 1985 21. AGus D, MANN R, CLAYMAN M, KELLY C, MICHAUD L, COHN D, NEILSON EG: The effects of daily cyclophosphamide administration on
the development and extent of primary experimental interstitial ncphritis in rats. Kidney mt 29:635—640, 1986
2. ChASE MW: Inhibition of experimental drug allergy by prior feeding of the sensitizing agent. Proc Soc Exp Biol Med 61:257—259, 1946
22. SHIH W, HINES W, NEII.soN EG: Effects of cyclosporin A on the development of immune-mediated interstitial nephritis. Kidney mt
3. WEINER HL, FRIEDMAN A, MILLER A, KHOURY Si, AL-SABBAGH A, SANTOS L, SAYEGII M, NUSSENBLAYT RB, TREN1HAM DE, HAFLER
33:1113—1118, 1988 23. AGES D, MANN R, COhN D, MICHAUD L, KELIX C, CLAYMAN M,
DA: Oral tolerance: Immunologic mechanisms and treatment of animal and human organ-specific autoimmune diseases by oral administration of autoantigens. Annu Rev Immunol 12:809—837, 1994 4. BITAR DM, WHITACRE CC: Suppression of experimental autoimmune
encephalomyelitis by the oral administration of myelin basic protein. Cell Immunol 112:364—370, 1988 5. HIGGINS PJ, WEINER 1-IL: Suppression of experimental autoimmune encephalomyelitis by oral administration of myelin basic protein and its fragments. J Immunol 140:440—445, 1988 6. WANG Z-Y, QIAO L, LINK H: Suppression of experimental autoimmune myasthenia gravis by oral administration of acetylcholine receptor. J Neuroimmunol 44:209—214, 1993 7. NUSSENBLATF RB, CASPI RR, MAHDI R, CHAN CC, ROBERGE F, LIDER 0, WEINER HL: Inhibition of S-antigen induced experimental
autoimmunc uveorctinitis by oral induction of tolerance with Santigen. J Immunol 144:1689—1695, 1990 8. NAGLER-ANDERSON C, BouEI LA, ROBINSON ME, SISKIND GW, THORBECKE FJ: Suppression of type II collagen-induced arthritis by
NEILSON EG: Inhibitory role of dietary protein restriction on the development and expression of immune-mediated antitubular basement membrane-induced tubulointerstitial nephritis in rats. J Clin Invest 76:930—936, 1985 24. KELLY CJ, ZURIER RB, KRAKAUER KA, BLANCHARD N, NEILSON EG:
Prostaglandin El inhibits effector T cell induction and tissue damage in experimental murine interstitial nephritis. J Clin Invest 79:782—789, 1987 25. KELLY CJ, CLAYMAN MD, HINES WH, NEILSON EG: Therapeutic immune regulation in experimental interstitial nephritis with suppressor T cells and their soluble factors. Ciba Found Symp 129:73—87, 1987 26. KELLY Ci, NEILSON EG: Contrasuppression in autoimmunity. Abnormal contrasuppression facilitates expression of nephritogenic effector T cells and interstitial nephritis in kdkd mice.JExp Med 165:107—123, 1987 27. NEILSON EG, MCCAFFERTY F, MANN R, MICHAUD L, CLAYMAN M:
intragastrie administration of soluble type II collagen. Proc NatlAcad
Tubular antigen derivatized cells induce a disease protective, antigenspecific, and idiotype-specific suppressor T cell network restricted by IJ and Igh-V in mice with experimental interstitial nephritis. J Exp
Sci USA 83:7443—7446, 1986 9. KELLY Ci, CLAYMAN MD, NEII,soN EG: Immunoregulation in exper-
Med 162:215—230, 1985 28. NEILSON EG, KELLY J, CLAYMAN MD, HINES WH, HAVERTY T, SUN
imental interstitial nephritis: Immunization with renal tubular antigen
Mi, BI,ANCHARD N: Murine interstitial nephritis. VII. Suppression of
in incomplete Freund's adjuvant induces major histocompatibility restricted 0X8+ suppressor T cells which are antigen specific and inhibit the expression of disease. J Immunol 136:903—907, 1986 10. NEII.SON EG, PHILLIPS SM: Cell-mediated immunity in interstitial
nephritis. 1. T lymphocyte systems in nephritis guinea pigs: The natural history and diversity of the immune response. J Immunol 123:2373—2380, 1979
11. NEILSON EG, PI-IILLIPS SM: Murinc interstitial nephritis I. Analysis of
disease susceptibility and its relationship to pleiomorphic gene products defining both immune response genes and restrictive requirement for cytotoxic T cells at H-2K. J Exp Med 155:1075—1085, 1982 12. MAMPASO FM, WILSON CB: Characterization of inflammatory cells in
renal injury after treatment with soluble suppressor factor TsFI. J Immunol 139:1518—1524, 1987 29. KELLY CJ, SILVERS W, NEILSON EG: Tolerance to parenchymal self:
The regulatory role of an RTI-restricted, OX8+ suppressor T cell for autologous renal tubular antigen in experimental interstitial nephritis. JExp Med 162:1892—1903, 1985 30. MCMENAMIN C, OlIVER J, GIRN B, HOLT BJ, KEES UR, THOMAS WR,
HOLT PG: Regulation of T-cell sensitization at epithelial surfaces in the respiratory tract: Suppression of IgE responses to inhaled antigens
by CD3 Tcra--lymphocytes (putative y/b T cells). Immunology 74:234-239, 1991 31. DICK AD, CHENG YF, MCKINNON A, LIVERSIDGE J, FORRESTER JV:
457, 1983 13. KELLY CJ, ROTH DA, MEYERS CM: Immune recognition and re-
Nasal administration of retinal antigens suppresses the inflammatory response in experimental allergic uveoretinitis. A preliminary report of intranasal induction of tolerance with retinal antigens. Br J Oph-
sponse to the renal interstitium. Kidney mt 31:518—530, 1991 14. STI:BLAY RW, RUDoFSKY U: Transfer of experimental autoimmune renal cortical tubular and interstitial disease in guinea pigs by serum. Science 180:966—968, 1973
32. DEVEY ME, BLEASDALE K: Antigen feeding modifies the course of antigen-induced immune complex disease. C/in Exp Immunol 56:637— 644, 1984
15. VAN ZWIETEN M, BHAN A, MCCLUSKEY R, COLLINS A: Studies on the
33. GESUALDO L, LAMM ME, EMANCIPATOR SN: Defective oral tolerance
pathogenesis of experimental anti-tubular basement membrane ne-
promotes nephritogenesis in experimental IgA nephropathy induced by oral immunization. J Immunol 145:3684—3691, 1990
autoimmune tubulointcrstitial nephritis in rats. Kidney mt 23:448—
phritis in the guinea pig. Am J Pathol 83:53 1—546, 1976 16. BANNISTER K, WII.SON C: Transfer of tubulointerstitial nephritis in the
thahnol 77:171—175, 1993
34. MII.LER A, LIDER 0, ROBERTS AB, SPRON MB, WEINER HL: Suppres-
Brown Norway rat with anti-tubular basement membrane antibody: Quantitation and kinetics of binding and effect of deconiplcmentation.
sor T cells generated by oral tolerization to myelin basic protein
J immunol 135:3911—3917, 1985
transforming growth factor beta after antigen-specific triggering.
17. NEILSON EG, GASSER DL, MCCAFFLRTY E, ZAKHEIM B, PhIlLIPS SM:
Polymorphism of genes involved in anti-tubular basement membrane disease in rats. Immunogenetics 17:555—565, 1983
18. KELLY CJ, MOK H, NEIISON EG: The selection of effector T cell phenotype by contrasuppression modulates susceptibility to autoimmune injury. J Immunol 141:3022—3028, 1988
suppress both in vitro and in vivo immune responses by the release of PNAS USA 89:421—425, 1992 35. LIDER 0, SANTOS LMB, LEE CSY, HIGGINS PJ, WEINER IlL: Suppres-
sion of experimental autoimmune enccphalomyelitis by oral administration of myelin basic protein. II. Suppression of disease and in vitro immune responses is mediated by antigen-specific CD8+ T lymphocytes. J Immunol 142:748—752, 1989
732
Pham et a!: Oral tolerance in interstitial nephritis
36. MILLER A, AL-SABBAGH A, SANTOS LM, DAS MP, WEINER HL:
Epitopes of myelin basic protein that trigger TGF-beta release after
oral tolerization are distinct from encephalitogenic epitopes and
sion of a nephritogenic murine T cell clone Kidney In! 46:1295—1301, 1994
37. CHeN Y, KUCHROO VK, INOBE J, HAFLER DA, WEINER HL: Regula-
46. MEYERS CM, KELLY CJ: Inhibition of murine nephritogenic effector T cells by a clone-specific suppressor factor. J Clin Invest 94:2093—2104, 1994 47. BAILEY NC, FRISHBERG Y, KELLY CJ: Loss of high affinity TGF-131
tory T cell clones induced by oral tolerance: Suppression of autoim-
binding to a nephritogenic T cell results in absence of growth
mune encephalomyelitis. Science 265:1237—1240, 1994
inhibition by TGF-j31 and augmented nephritogenicity. J Immunol
mediate epitope-driven bystander suppression. J Immunol 151:7307— 7315, 1993
38. VAN HOUTEN N, BLAKE SF: Direct measurement of anergy of antigen-specific T cells following oral tolerance induction. J Immunol 157:1337—1341, 1996
156:3009—3016, 1996 48. CLAYMAN MD, MARTINEZ-HERNANDEZ A, MICHAUD L, ALPER R, MANN R, KEFALIDES NA, NEILSON EG: Isolation and characterization
39. KELLY CJ, KORNGOLD R, MANN R, CLAYMAN MD, HAVERTY TP,
of the nephritogenic antigen producing anti-tubular basement mem-
NEILSON EG: Spontaneous nephritis in kdkd mice. II. Characterization of a tubular antigen-specific, H-2K restricted Lyt-2 effector T
49. AL-SABBAGH A, MILLER A, SANTOS LM, WEINER HL: Antigen-driven
cell that mediates destructive tubulointerstitial injury. J Immunol 136:526—531, 1986
40. MEYERS CM, KELLY CJ: Effector mechanisms in organ-specific auto-
immunity. 1. Characterization of a CD8+ T cell line that mediates murine interstitial nephritis. J Clin Invest 88:408—416, 1991 41. MEYERS CM, KELLY Ci: Adoptive transfer of an antigen-specific DTH-reactive T cell clone induces murine interstitial nephritis. (abstract) JAm Soc Nephrol 2:553, 1991 42. SCHAINUCK LI, STRIKER GE, CUTLER RE, BENDITT EP: Structural-
functional correlations in renal disease. Hum Pathol 1:631—641, 1970 43. BOHLE A, MACKENSEN-HAEN S, VON GISE H: Significance of tuhuloin-
erstitial changes in the renal cortex for excretory function and concentration ability of the kidney: A morphometric contribution. Am J Nephrol 7:421—433, 1987 44. BOI-ILE A, MACKENSEN-HAEN S, VON GISE H, GRUND KE, WEHRMANN
M, BATZ C, BOGENSCHUTZ 0, SCHMITF H, NAGY J, MULLER C, MULLER G: The consequences of tubulo-interstitial changes for renal function in glomerulopathies. Path Res Pract 186:135—144, 1990 45. MEYERS CM, KELLY CJ: Immunoregulation and TGF-/31. Suppres-
brane disease. J Exp Med 161:290—305, 1985
tissue-specific suppression following oral tolerance: Orally administered myelin basic protein suppresses proteolipid protein-induced experimental autoimmune encephalomyelitis in the SJL mouse. Eur J Immunol 24:2104—2 109, 1994 50. SUN JB, HOLMOREN J, CZERKINSKY C: Cholera toxin B subunit: An efficient transmucosal carrier-delivery system for induction of periph-
eral immunological tolerance. Proc Nat! Acad Sci USA 91:10795— 10799, 1994
51. WEINER HL, MACKIN GA, MATSUI M, ORAV Ei, KUOURY Si, DAwsoN DM, HAFLER DA: Double-blind pilot trial of oral tolerization with myelin antigens in multiple sclerosis. Science 259:1321—1324, 1993 52. TRENTHAM DE, DYNESIUS-TRENTHAM RA, ORAV EJ, COMBIrCHI D,
LORENZO C, SEWELL KL, HAFLER DA, WEINER HL: Effects of oral
administration of collagen on rheumatoid arthritis. Science 261:1727— 1730, 1993 53. GENAIN C, ABEL K, BELMAR N, VILLINGER F, ROSENBERG D, LINING-
TON C, RAINE C, HAUSER S: Late complications of immune deviation therapy in a nonhuman primate. Science 274:2054—2057, 1996
725—732
Oral feeding of renal tubular antigen abrogates interstitial nephritis and renal failure in Brown Norway rats KEVIN PHAM, WILLIAM E. SMOYER, D. ClAY ARCHER, F1NcIs GuBM, and CAROLYN J. KELLY Division of Nephrology-Hypertension, Department of Medicine, University of California, San Diego, and the Veterans Affairs Medical Center, San Diego, California; and the Department of Pediatrics, University of Michigan, Ann Arbor, Michigan, USA
Oral feeding of renal tubular antigen abrogates interstitial nephritis and renal failure in Brown Norway rats. We have examined whether oral feeding of antigen can regulate the expression of autoimmune interstitial nephritis induced by antigen-in-adjuvant (RTA'CFA) immunization of Brown Norway rats. Male rats were divided into six experimental groups:
anergy, depending the characteristics of the antigen and feeding protocol [reviewed in 3]. Our laboratory has had a longstanding interest in developing modalities of targeted immunosuppression in autoimmune renal
disease. One model system we utilize is that of anti-tubular basement membrane (TBM) disease with interstitial nephritis. This disease is induced in susceptible strains of guinea pigs, rats, and mice by a single immunization of a renal tubular antigen Renal histology, inulin clearance, DTH responses to RTA, and IgG (RTA) emulsified in complete Freund's adjuvant (CFA) [9—li]. antibody responses to RTA were monitored as endpoints of the study. Our The resultant humoral and cell-mediated immune response to results demonstrated that Group III and IV animals had significantly less RTA results in deposition of specific IgG along the tubular severe renal injury, as assessed by inulin clearance and extent of renal cortical involvement by mononuclear cells. Group II and IV animals had basement membrane and mononuclear cell infiltration of renal interstitium. This mononuclear cell infiltrate includes both CD4 suppressed DTH responses, and only Group IV animals had significantly Group I, RTAJCFA immunization alone; Groups II, III, and IV were
pretreated with 1 mg (Group II), 5 mg (Group III), and 25 mg (Group IV) of oral tubular antigen every other day for ten days, followed by RTAICFA immunization; Group V was pretreated with a control antigen, followed by RTAJCFA immunization; and Group VI was immunized with CFA alone.
depressed antigen-specific IgG serum titers. Group III animals had neither suppressed DTH responses or IgG titers. We conclude that oral administration of tubular antigen can modulate the intensity of interstitial nephritis produced by immunization, but that the regulatory mechanism is not dependent (at all doses of fed antigen) on suppressed DTH reactivity to RTA or suppressed antigen-specific IgG.
The term oral tolerance refers to the induction of systemic hyporesponsiveness to an antigen following oral administration of this antigen. Although the phenomenon was initially described in
the first half of this century [1, 2] there has been a recent resurgence of interest in it because of its potential as a therapeutic
and CD8 T cells, as well as macrophages, natural killer cells, and B cells [121. The relative importance of these various immune cell subsets, as well as antigen-specific lgG, in mediating renal injury
varies among species [13]. For example, the passive transfer of anti-TBM antibodies can induce disease in guinea pigs [14, 15], to a lesser degree in rats [16], and not at all in mice. Susceptibility to disease depends, at the minimum, on expression of discrete MHC
genes [11], expression of antigen in the kidney [17], and the development of a susceptible T cell phenotype following immunization [18—20]. The animals develop progressive renal failure. Previous work in this model has documented that this nephritogenie immune response can be modulated by immunosuppressive drugs [21, 22] dietary manipulation [231, PGEI [24], and through the induction of regulatory T cells [9, 25—28]. Regulatory T cells can be induced through the presentation of antigen in a tolerogenie manner, that is, emulsified with IFA rather than CFA [9], or
modality for autoimmune disease [3]. Oral feeding of proteins that are target antigens of autoimmune T and B lymphocytes can protect the host against subsequent immunization with those antigens to induce disease. Amelioration of organ pathology by coupled to the surface of syngeneic splenocytes and injected this approach has been demonstrated in experimental allergic intravenously [27]. In the studies reported here, we examined encephalomyelitis [4, 5], myasthenia gravis [6], uveoretinitis [7], whether oral administration of renal tubular antigen can alter the and collagen arthritis [81. The mechanisms by which oral feeding development of interstitial nephritis. We find that oral feeding of of protein antigens alter the nature of the subsequent immune RTA can abrogate the severity of the induced interstitial ncphriresponse to antigenic challenge are complex. Recent work has tis, and that this effect is demonstrable at doses of fed antigen at demonstrated that the phenomenon of oral tolerance can be which anti-1BM lgG titers and DTH reactivity arc not altered. attributed to active suppression by T lymphocytes or by clonal METHODS Key words: oral feeding in TIN, interstitial nephritis, antigens, renal failure, autoimmune disease.
Animals
Received for publication August 16, 1996 and in revised form February 27, 1997 Accepted for publication April 7, 1997
from Charles River Laboratories (Boston, MA, USA) and housed in the animal facility at the San Diego Veterans Affairs Medical
© 1997 by the International Society of Nephrology
Center. All care of animals was performed under guidelines
Male Brown Norway rats weighing 200 to 25(1 g were purchased
725
726
Pham et al: Oral tolerance in interstitial nephritis
approved by the American Association for the Accreditation of challenge by an independent observer blinded to the identity of Laboratory Animal Care. the experimental groups, using a spring-loaded engineer's micrometer. The change in footpad thickness was recorded for each Immunization of animals animal and the result expressed as the mean for each experimenRabbit renal tubular basement membranes were isolated by a tal group SEM. differential sieving technique, sonicated, lyophilized, and stored at —20°C [9]. Lyophilized renal tubular antigen was emulsified in Assessment of renal function At 17 to 18 days following initial immunization, all rats undercomplete Freund's adjuvant (RTAJCFA). Antigen concentrations for both RTA and PPD (M.Th. H37Ra; Difco Laboratories, went an inulin clearance as a measure of glomerular filtration Detroit, MI, USA) were 4 mg/mI. After the induction of anesthe- rate, using established techniques. Briefly, following anesthesia sia with Brevital (7 mg/100 g body wt., i.p.; Eli Lilly & Co., with mactin (10 mgIlOO g body wt, i.p.; BYK, Konstanz, GermaIndianapolis, IN, USA), animals were immunized subcutaneously ny), animals were tracheally intubated, and catheters were ininto four sites in the axillary and groin regions with a total of 0.75 serted in the jugular vein, femoral artery, and urinary bladder. ml of RTA/CFA. One group of animals received 0.75 ml of CFA The femoral artery catheter was used for periodic blood sampling alone (4 mg/mI PPD) as a control for the immunization proce- and monitoring mean arterial pressure with a transducer (model dure. In these studies no animal received footpad immunization P23DB; Statham Instruments, Gould Divsion, Hato Rey, Puerto Rico) and recorded on a chart recorder. All rats received an with RTA/CFA or CFA. infusion of NaC1-NaHCO3 throughout the experiment containing Feeding protocol (3H) inulin at a rate of 2.2 ml/hr. After 60 minutes of equilibraFor the three groups of rats that received RTA or control tion, plasma and urine samples were collected for three consecuantigens (ovalbumin or hemoglobin; Sigma Chemical Co., St. tive 20-minute periods from femoral artery and urinary bladder, Louis, MO, USA) in the food, the antigen was mixed thoroughly respectively. Aliquots of plasma and urine samples were counted in powdered rat chow prior to feeding. Animals were housed in in a scintillation counter (Packard Instruments, Downer's Grove, individual cages and fed 14 g of standard rat chow daily for a IL, USA). Glomerular filtration rate was calculated from the period of ten days. (This quantity of powdered rat chow was radioactivity counts of the plasma and urine aliquots and the selected because it was completely consumed daily). Antigen in volume of urine collected per 20 minute period. designated amounts was mixed with the food on alternate days. The two groups of rats that did not receive RTA or control ELISA for anti-RTA antibodies ELISA microtiter plates were coated overnight at 4°C with antigens in the food were also housed individually and given 14 g of food daily for ten days. After the 10-day regimen of food intake, collagenase solubilized RTA at 40 jig/mI. Following washing with all rats were fed normal rat chow and water ad libitum and housed dH2O, the plates were blocked for two to three hours at room temperature with PBS containing 4% BSA, 0.05% Tween 20. at three rats per cage. Following extensive washing with Tris-buffered saline with 0.05%
Experimental groups Tween 20, pH 7.6, multiple dilutions of test sera from individual Animals were divided into six groups. Group I (N = 6) was animals were placed in wells, and incubated at 37°C for 2.5 hours. immunized with RTA/CFA two days following completion of Following extensive washing, the appropriate secondary reagent
feeding regimen. This group served as a positive control for [goat anti-rat IgG (H+L), goat anti-rat IgGI or goat anti-rat disease severity. Group II (N = 4) was fed with I mg of RTA in powdered rat chow on alternate days as described above (total of 5 mg fed antigen). Group III (N = 8) was fed with 5 mg of RTA
IgG2a diluted in PBS with 4% BSA, 0.05% Tween 20, all alkaline phosphatase conjugated] was added and incubated for 2.5 hours at 37°C. Following washing, the plates were developed with substrate
hemoglobin as control antigens. All animals in groups II to V were immunized with RTA/CFA two days following completion of the feeding regimen. Group VI (N = 4) was immunized with CFA two
in carbonate buffer, pH 9.6, for the appropriate time period (45 mm for total IgG, 5 hr for IgG subtype ELISA). The O.D. at 405 nm was measured in a microplate reader. The best linear curve fit for each group (as determined by the linear least-squares regression method) was calculated, and the serum titer for each group was calculated as the serum dilution at which the test sera O.D. 405 nm was equal to the O.D. 405 for CFA serum.
in powdered rat chow on alternate days (total of 25 mg fed antigen). Group IV (N = 4) was fed with 25 mg of RTA in powdered rat chow on alternate days (total of 125 mg fed antigen). Group V (N = 5) was fed with either ovalbumin or
days following completion of feeding regimen. This group was used as a control for normal renal function. Previous work from our laboratory has shown that at one and three weeks following CFA immunization, glomerular filtration rate is indistinguishable from that of animals never immunized (C. Kelly and F. Gabbai, unpublished data). Delayed-type hypersensitivity testing At 13 to 15 days following subcutaneous immunization with RTA/CFA, all rats received a challenge dose of antigen, consisting of 75 jd of I mg/ml collagenase solubilized RTA in PBS in one
Assessment of histologic renal disease Animals were sacrificed following the clearance measurements (at day 18 following immunization). The left kidney of each rat was harvested and immersion fixed in 10% buffered formalin. The
extent of interstitial involvement was determined by a single observer from 4 mm, paraffin-embedded kidney sections stained with hematoxylin and eosin. The specimens were coded to prevent observer bias, and graded at a single session by an observer who additionally was unaware of the results of the clearance measurements. Cellular infiltration and tubular damage leading to cortical
hind footpad. The other hind footpad was injected with PBS. interstitial destruction was graded on a scale of 0 to 4: 0 = no Footpad thickness was measured before and 24 hours after lesions; 1.0 or greater = 10 to 25% cortical involvement by
727
Phain et at: Oral tolerance in interstitial nephritis
5
2.5
4
2
Cl)
ci)
a)
C)
1.5
(ci—
0 C
*
B C
1
0.5
1
0
0 II
Ill
IV
V
VI
Experimental groups
Fig. 1. Oral feeding of RTA can diminish the severity of interstitial nephritis. Experimental groups (designated by Roman numerals on the
horizontal axis) of Brown Norway rats were fed with varying total doses of RTA prior to immunization to produce interstitial nephritis. The feeding
II
Ill
IV
V
VI
Experimental groups Fig. 2. Oral feeding of antigen leads to improved renal function. Inulin
clearances were performed on the same experimental groups outlined in the Figure 1 legend and the Methods section. *P < 0.05 compared with Group I.
protocols are outlined in the Methods section. Experimental group designations are: I, RTAICFA immunized with no oral antigen fed; II, fed with I mg RTA every other day for ten days prior to immunization with RTA/CFA; HI, fed with 5 mg RTA every other day for ten days prior to immunization with RTAJCFA; IV, fed with 25 mg RTA every other day for ten days prior to immunization with RTAICFA; V, fed with control Ag prior to immunization with RTAJCFA; VI, no antigen feeding, immunized with CFA alone. Group VI animals have no detectable disease. p < 0.05 compared with Group I.
inflammatory cells; 2.0 or greater 25 to 50% cortical involvement by inflammatory cells; 3.0 or greater = 50 to 75% cortical involvement by inflammatory cells; 4.0 = 75% or greater cortical involvement by inflammatory cells. 4.0 was the maximum score assigned.
Statistical analysis Results are presented as mean
differences were statistically significant. CFA immunized animals
(Group VI) do not have detectable disease. Oral feeding of tubular antigen leads to preserved GFR following RTAJCFA immunization
To have a precise quantitative endpoint (GFR) by which to compare experimental groups, we performed inulin clearances on all animals at 17 to 18 days following RTAICFA immunization. These data are shown in Figure 2. The mean inulin clearance for the CFA immunized rats (Group VI) was 1.9 mi/mm, establishing
the control values for male BN rats of this weight. RTA/CFA immunized animals (Group I) had a markedly diminished GFR at
approximately 0.5 mI/mm. Oral antigen feeding had a dose dependent effect on GFR. Although the mean GFR for Group II was not significantly different than Group I, the mean GFR for
SEM. ANOVA with post hoc Groups III and IV, the 5x and 25 X oral antigen groups was tests was used for comparison among groups. The level of significantly different from Group I and, in fact, not significantly
statistical significance was defined as a value of P < 0.05. RESULTS
Oral feeding of tubular antigen blocks the development of interstitial nephritis following RTAICFA immunization
Figure 1 shows the effect of oral feeding of RTA on the histologic severity of the interstitial injury. The grading scale for interstitial injury extends from 0 to 4 and represents the extent of
different than that of the normal controls (Group VI). Control antigen feeding (Group V) did not significantly improve GFR. Close correlation between the GFR and histology score in Groups I-V Figure 3 depicts a plot of mean inulin clearance versus mean
histologic score for each experimental group immunized with RTA/CFA. There is a close inverse correlation between GFR and
cortical involvement by the mononuclear cell infiltrate. The extent of cortical interstitial inflammation (linear correlation positive control animals (Group I), those fed Ix oral antigen coefficient, r = 0.972). (Group II), and those fed the control Ags (Group V), all had mean histologic scores in the 2.75 to 3.0 range, signifying approximately 50% cortical involvement. Animals fed either 5x (Group
Antigen-feeding resulted in a dose-dependent inhibition of DTH responses to RTA
III) or 25x (Group IV) oral antigen had histologic scores of
The DTH response to RTA has correlated well in previous
approximately 1, signifying 10% cortical involvement. These latter
studies with the presence of nephritogenic lymphocytes and active
728
Pham et al: Oral tolerance in interstitial nephritis
3.5
60
3
50
• Group II Group V
w 2.5
00
cc
' 1.5
U 30
40
o?
IH Group IV
0.5
0
20
10
0.5
1.5 lnulin clearance
1
2
mi/mm
Fig. 3. A close correlation exists between inulin clearance and histologic score in the experimental groups. The mean inulin clearances from each experimental group (except Group VI, where the histologic score is zero) were plotted on the horizontal axis versus the mean histologic scores on the vertical axis. The best linear curve fit for these data (as determined by the linear least-squares regression method) was determined using CACricket Graph III" software (Computer Associates, International). The equation for this line is y = —1.70x + 4.21, r = 0.972.
*
0 II
Ill
I\/
\'I
Experimental groups Fig. 4. "Low" and "high" dose oral antigen feeding inhibit DTH reactiv-
ity to RTA. DTH testing for reactivity to RTA was performed on all experimental groups several days prior to performance of the inulin clearances. Footpads were read on unanaesthetized animals using an engineer's micrometer in cage-blind fashion. Results are depicted as the
change in footpad thickness measurement from that prior to RTA injection to that measured 24 hours following RTA challenge. p < 0.01 compared with group I.
interstitial nephritis [9, 20, 29]. Figure 4 illustrates the DTH responses to collagenase solubilized RTA at 14 days after immunization in all six groups of animals. The groups fed with repeated
biologic significance of this slightly depressed anti-TBM IgG titer is uncertain.
doses of 1 mg (Group II) or 25 mg (Group IV) of RTA
Since other groups have reported a preferential inhibition of demonstrated marked suppression of DTH responses to RTA ThI CD4 cells over Th2 responses at low doses of oral antigen, compared to Group I animals. Group Ill, fed with repeated doses we additionally examined the titers of the IgGi and IgG2a of 5 mg of RTA, had a DTH response not significantly different subtypes in the csTBM antibody responses in these animals. IgGI than Group I. The group fed with irrelevant antigens also responses typically correlate with a Th2 response, and IgG2a displayed no suppression of DTH to RTA, as expected. The responses correlate with a Thi response. We found no significant change in footpad thickness in response to PBS injection alone differences in the ratio of these subtypes compared with control ranged from a mean of 1.0 to 5.9 in the different groups. (The data from the PBS injected footpad are not depicted on Fig. 4 for the sake of clarity.) These experiments demonstrate that oral feeding of this antigen preparation is capable of modulating one marker of the cell-mediated immune response to RTA.
Antigen feeding does not affect the anti-TBM antibody response
We also examined the titers of aTBM IgG antibodies in the various experimental groups (Fig. 5). The mean O.D. readings versus serum dilution for the different experimental groups are depicted in Figure SA. The titers derived from these readings for each experimental group are compared in Figure SB. Although the mean O.D. readings at some serum dilutions were depressed for both Groups II and V, only group IV had a significantly
animals (data not shown). DISCUSSION
Our studies clearly demonstrate that oral feeding of RTA to Brown Norway rats prior to immunization with RTAJCFA can block the subsequent development of interstitial nephritis and renal insufficiency. The effect of antigen feeding is dose depen-
dent. Although certain doses of fed antigen had significant suppressive effects on DTH responses to RTA, the DTH effect did not correlate well with improvement in histologic disease or GFR. When one compares the two experimental groups in which there was improvement in GFR and histologic disease (Groups III and
IV), it is evident that neither inhibition of DTH reactivity to the immunogen, nor inhibition of the IgG antibody titer is required to
depressed total lgG aTBM Ab serum titer. Even with this slightly ameliorate this disease process within the time frame of our depressed log10 titer, the kidneys of Group IV animals still had 4 studies. positive linear TBM staining for IgG by direct immunofluoresThe term "oral tolerance" is used to describe the alterations in cence and were indistinguishable from the Group I animals by immune responsiveness seen following feeding of various antiintensity of immunofluorescence (data not shown). Therefore, the gens. While this widely used phrase evokes the protective nature
Pham et al: Oral tolerance in interstitial nephritis
A
729
require oral feeding. Similarly protective effects have been shown with aerosolized antigen [301 and antigen applied to the nasal mucosa [311. In these instances, the antigen processing by either bronchial or nasal mucosa associated lymphoid tissue presumably
0.6 0.5 -
induces the same type of protective response as does the gutassociated lymphoid tissue.
Our study is the first to demonstrate abrogation of renal inflammation and preservation of renal function by the oral
0.4 E
administration of a renal target antigen. Two previous studies in models of immune-mediated renal disease have supported the
0 0.3 IC)
notion that presentation of antigen via the gut mucosa can
0
modulate the expression of renal disease. Oral feeding of HSA can decrease the incidence of immune complex disease in the kidney despite continued presence of high levels of circulating immune complexes [32]. The decreased incidence of immune
0.2 -
complex disease was attributed to a possible change in the affinity
0.1 0
of the immune complexes. In a model of IgA nephropathy I
I
3
3.7
I
4 Log10 serum dilution
4.7
5
development of oral tolerance, thereby supporting a paradigm whereby defective oral tolerance facilitates the development of this form of glomerulonephritis [33]. The mechanisms underlying how oral feeding of antigen abrogates autoimmune responses have been evaluated in several other models of organ-specific autoimmune disease. Experimental allergic encephalomyelitis (EAE) has been the most extensively
4.75
studied model system [3]. Studies have been conducted in both the
B
rat and mouse models of this disease. In rat EAE, clinical and histologic manifestations of disease can be suppressed by low doses of fed myelin basic protein [4, 51. In this model, animals fed with MBP have a CD8 T cell population demonstrable in their spleen and mesenterie lymph nodes that can adoptively transfer protection against EAE [34, 35]. The antigenic epitopes of MBP recognized by these regulatory T cells are distinct from those
a)
C
0)
induced by oral feeding of bovine gamma globulin, the concomitant administration of cyclophosphamide or estradiol increased Ag-specific IgG responses, glomerular IgG immune deposits and the development of hematuria. Both drugs can interfere with the
4.5
0 -J
recognized by encephalitogenic T cells [36]. In the murine model of EAE, antigen feeding induces Th2-like CD4 regulatory T cells
4.25
that produce TGF-f3 and IL-4 and functionally suppress EAE induced with either MBP or proteolipid protein [37]. These studies suggest that inactivation, rather than deletion, of effector T cells is a generic operative mechanism in oral tolerance. This notion is further supported by recent studies utilizing a system of
4 I
II
Ill
V
Experimental groups Fig. 5. Anti-RTA IgC titers are suppressed only in Group IV animals. (.4)
The mean O.D. readings flr each experimental group at the indicated serum dilutions. Only positive error defiections (1 su) are shown. Icons for Groups Ito VI are as follows: () Group Group U; (0) Group III; (0) Group IV; () Group V; (+) Group VI. The best linear curve fit for each group (as determined by the linear leastsquares regression method)
I; ()
was calculated, and the serum titer taken as the serum dilution at which
the test sera 0D405 nm was equal to the 00405 for CFA serum. (B) Values for the linear curve fit. Only the Group TV data are significantly different. Where error bars are not depicted, they were insignificant in
size. P 0.05.
of the subsequent immune response to challenge immunizations with the same antigen (in the ease of autoimmune disease), it is not, strictly speaking, tolerance, nor does this protective effect
ovalbumin specific TCR transgenic lymphocytes transplanted into BALB/c recipients which were subsequently fed and challenged with ovalbumin [38]. in this system, the proliferative noriresponsiveness seen in fed animals could be normalized for the number of antigen specific cells. This calculation demonstrated a marked reduction in proliferation per lymphocyte expressing the ovalbumm specific TCR in antigen fed animals, supporting a tolerance mechanism of anergy (inactivation) rather than deletion [38]. It is
important to point out this kind of proliferative nonresponsiveness could be secondary to the induction of Th2-like cytokines in the TCR transgenic T cells. Previous experimental work in both the rat and murinc models of
18,
730
P/sam
et al: Oral tolerance in interstitial nephritis
24, 25, 29], which abrogate disease have consistently also abro-
TGF-p as an immunosuppressive cytokine would be fully consis-
gated Ag-specific DTH responses. The relationship between DTH-reactivity and nephritogenicity is more than a correlative presumption of pathogenicity. T cells purified from diseased kidneys mediate DTH to the tubular Ag, and renal-eluted DTH reactive cells can infiltrate the renal cortex following adoptive
tent with our previous observations regarding TGF-/3 in this model system [45—47]. This finding has important practical impli-
cations, as it implies that it is not necessary to purify a target
antigen to homogeneity, or even identify a target antigen, in order to induce this protective effect. Our own studies confirm this, as we used a renal tubular antigen preparation which consists of 15 to 20 distinct protein bands by SDS-PAGE [48]. Previous studies
subcapsular transfer into naive hosts [20, 39]. Moreover, tubular Ag-reactive T cell clones that can mediate DTH are also nephri. togenic following adoptive transfer [40, 41]. These findings have have demonstrated that rodents mount a significant immune
strongly supported the notion that DTH-reactive T cells are
response to a single component of RTA, the 3M—I protein,
important effector cells in experimental interstitial nephritis. The uncoupling of DTH reactivity to tubular Ag and protection from interstitial nephritis (as displayed by Group III in this study) was surprising to us. It is possible that DTH-reactivity to the Ag is an
following immunization with RTAJCFA [48]. While the remaining
the critical effector cell, but that antigen feeding protocols result
pressed in physical proximity to the target antigen could abrogate
proteins in the RTA preparation are innocuous from the stand-
point of generating a nephritogenic immune response, it is
possible that they generate protective responses when presented epiphenomenon and unrelated to the immunopathogenesis of via the gut mucosa, and thereby contribute to the induced disease. For the reasons outlined above, we do not favor this tolerance. Another implication of immunosuppression by Aginterpretation. Another possibility is that the DTH-reactive cell is nonspecific cytokines is that feeding with a single antigen ex-
in impaired homing to the kidney, perhaps through cytokine immune responses directed to the target antigen, by local produc-
induced alterations in integrins or selectin ligands. An alternative tion of immunosuppressive cytokines [491. This raises the possiinterpretation is that the oral feeding regimen results in the bility that feeding an "irrelevant" antigen, but one expressed in inactivation of a different T cell population, not DTH reactive, the tubules or interstitium, may stimulate a host protective which is required for interstitial injury. This could be a CD8 CTL response. population, or a functionally distinct CD4 subpopulation. The Our studies demonstrated equivalent protection by the antigen last two scenarios described above (impaired homing vs. a non- doses in Groups III and IV. These studies have not established the DTH-reactive effector cell) are not mutually exclusive. The mech- minimum total dose or timing necessary to induce oral tolerance. anism underlying the preserved DTH response in group III has We believe that the dose administered to Group II, 5 mg total, is not yet been elucidated. In preliminary studies, we have found on the border of what is effective. The large standard error that the mechanisms underlying suppression of DTH in Group II displayed in this group in both Figures 1 and 2 is the result of and IV animals are different. This finding suggests that Group III some animals displaying protection and other animals displaying animals received an antigen dose orally which did not activate none. The correlation between histology and inulin clearance
either suppressive mechanism. We were not surprised by the within Group II is strong, that is, those animals with minimal discordance between Ag-specific IgG titers, histologic renal dis- histologic abnormalities have the higher inulin clearances. We ease, and functional deterioration. Previous studies have demon- have not seen any protection with doses of this RTA antigen strated marked suppression of renal disease in this model, without preparation below this dose (data not shown). It is also possible alterations in circulating Ag-specific IgG or kidney bound IgG [9]. There was a striking correlation between histologic scores and
that the gastrointestinal absorption of the antigen is variable. The quantity of antigen used may be minimized by coupling to cholera toxin [50]. An important second-order study will be to determine the efficacy of oral Ag feeding after induction of disease. Preliminary data from several clinical trials suggested that Ag
GFR in our studies. We performed both measurements as they provided a complementary analysis. The inclusion of the inulin clearance contributed a "hard" endpoint to the analysis. Given the well-recognized inverse correlation between degree of interstitial feeding may have beneficial effects clinically in autoimmune damage and glomerular filtration rate in human renal disease disease, with minimal toxicity or side effects [51, 521, although [42-44], it is not surprising that an experimental model of more recent observations in non-human primates partially temper interstitial inflammation should result in impaired GFR. The this early enthusiasm for the therapeutic potential of immune close correlation between function and histology in the current deviation to the Th2 phenotype [53]. The appeal of this as a study is reassuring support for the use of the latter as an important potential therapy for human immune-mediated renal disease is endpoint. Despite this strong correlation, we still believe that both that the knowledge of a precise target antigen is probably not measurements are important. In the context of immunomodula- required. It is additionally possible that in complex lesions, such as tion, it is quite possible that an immunosuppressive modality that the interstitial inflammatory disease accompanying heavy proteinaltered the types of cytokines expressed by Ag-reactive cells might uric states, inhibition of an antigen-specific secondary immune not change the degree of mononuclear cell infiltration but would response targeted to the interstitium might block progression of improve renal function. In contrast, other immunosuppressive disease. Further work on oral tolerance in experimental renal agents that result in the virtual elimination of Ag-reactive lym- disease may provide a foundation for a novel approach to phocytes from the renal cortex might depress GFR, through either treatment of immune-mediated renal disease. functional or toxic mechanisms. The finding that oral administration of antigen can prime T ACKNOWLEDGMENTS cells which produce cytokines with immunosuppressive properties This work was presented in part at the 1995 Annual Meeting of the (that is, IL-4, IL-b, and TGF-13) suggested that the final mediator American Society of Nephrology and published in abstract form JAm Soc of immunosuppression was Ag-nonspecific. TGF-p in particular Nephrol 1995). The studies were supported by grants from the NIH
has been implicated as a mediator [36, 37]. A potential role for (DK42155 and DK45346) and the Office of Research and Development,
Pham et al: Oral tolerance in interstitial nephritis
Medical Research Service, Department of Veterans Affairs. Kevin Pham
received support from the Student Affairs Office of UCSD Medical School. We thank Martin Kagnoff, M.D. for helpful discussions and Ser Khang for assistance with clearance measurements.
731
19. MANN R, KELLY CJ, HINES WH, CLAYMAN M, BLANCh-lARD N, SUN
MJ, NEILSON EG: Effector T cell differentiation in experimental interstitial nephritis I. The development and modulation of effector lymphocyte maturation by 1-J regulatory T cells. J Immunol 136: 4200—4208, 1987 20. NEILSON EG, MCCAFFERTY E, MANN R, MICHAUD L, CLAYMAN M:
Reprint requests to Carolyn J. Kelly, M.D., UCSD and VAMC 111-H, 3350 La Jolla Village Drive, San Diego, California 92161, USA. E-mail: [email protected]
Murine interstitial nephritis. III. The selection of phenotypic (Lyt and L3T4) and idiotypic (RE-Id) T cell preferences by genes in Jgh-1 and H-2K characterize the cell-mediated potential for disease expression:
Susceptible mice provide a unique effector T cell repertoire in REFERENCES 1. WELLS H: Studies on the chemistry of anaphylaxis. III. Experiments
with isolated proteins, especially those of hen's egg. J Infect Dis 9:147—151, 1911
response to tubular antigen. J Immunol 134:2375—2382, 1985 21. AGus D, MANN R, CLAYMAN M, KELLY C, MICHAUD L, COHN D, NEILSON EG: The effects of daily cyclophosphamide administration on
the development and extent of primary experimental interstitial ncphritis in rats. Kidney mt 29:635—640, 1986
2. ChASE MW: Inhibition of experimental drug allergy by prior feeding of the sensitizing agent. Proc Soc Exp Biol Med 61:257—259, 1946
22. SHIH W, HINES W, NEII.soN EG: Effects of cyclosporin A on the development of immune-mediated interstitial nephritis. Kidney mt
3. WEINER HL, FRIEDMAN A, MILLER A, KHOURY Si, AL-SABBAGH A, SANTOS L, SAYEGII M, NUSSENBLAYT RB, TREN1HAM DE, HAFLER
33:1113—1118, 1988 23. AGES D, MANN R, COhN D, MICHAUD L, KELIX C, CLAYMAN M,
DA: Oral tolerance: Immunologic mechanisms and treatment of animal and human organ-specific autoimmune diseases by oral administration of autoantigens. Annu Rev Immunol 12:809—837, 1994 4. BITAR DM, WHITACRE CC: Suppression of experimental autoimmune
encephalomyelitis by the oral administration of myelin basic protein. Cell Immunol 112:364—370, 1988 5. HIGGINS PJ, WEINER 1-IL: Suppression of experimental autoimmune encephalomyelitis by oral administration of myelin basic protein and its fragments. J Immunol 140:440—445, 1988 6. WANG Z-Y, QIAO L, LINK H: Suppression of experimental autoimmune myasthenia gravis by oral administration of acetylcholine receptor. J Neuroimmunol 44:209—214, 1993 7. NUSSENBLATF RB, CASPI RR, MAHDI R, CHAN CC, ROBERGE F, LIDER 0, WEINER HL: Inhibition of S-antigen induced experimental
autoimmunc uveorctinitis by oral induction of tolerance with Santigen. J Immunol 144:1689—1695, 1990 8. NAGLER-ANDERSON C, BouEI LA, ROBINSON ME, SISKIND GW, THORBECKE FJ: Suppression of type II collagen-induced arthritis by
NEILSON EG: Inhibitory role of dietary protein restriction on the development and expression of immune-mediated antitubular basement membrane-induced tubulointerstitial nephritis in rats. J Clin Invest 76:930—936, 1985 24. KELLY CJ, ZURIER RB, KRAKAUER KA, BLANCHARD N, NEILSON EG:
Prostaglandin El inhibits effector T cell induction and tissue damage in experimental murine interstitial nephritis. J Clin Invest 79:782—789, 1987 25. KELLY CJ, CLAYMAN MD, HINES WH, NEILSON EG: Therapeutic immune regulation in experimental interstitial nephritis with suppressor T cells and their soluble factors. Ciba Found Symp 129:73—87, 1987 26. KELLY Ci, NEILSON EG: Contrasuppression in autoimmunity. Abnormal contrasuppression facilitates expression of nephritogenic effector T cells and interstitial nephritis in kdkd mice.JExp Med 165:107—123, 1987 27. NEILSON EG, MCCAFFERTY F, MANN R, MICHAUD L, CLAYMAN M:
intragastrie administration of soluble type II collagen. Proc NatlAcad
Tubular antigen derivatized cells induce a disease protective, antigenspecific, and idiotype-specific suppressor T cell network restricted by IJ and Igh-V in mice with experimental interstitial nephritis. J Exp
Sci USA 83:7443—7446, 1986 9. KELLY Ci, CLAYMAN MD, NEII,soN EG: Immunoregulation in exper-
Med 162:215—230, 1985 28. NEILSON EG, KELLY J, CLAYMAN MD, HINES WH, HAVERTY T, SUN
imental interstitial nephritis: Immunization with renal tubular antigen
Mi, BI,ANCHARD N: Murine interstitial nephritis. VII. Suppression of
in incomplete Freund's adjuvant induces major histocompatibility restricted 0X8+ suppressor T cells which are antigen specific and inhibit the expression of disease. J Immunol 136:903—907, 1986 10. NEII.SON EG, PHILLIPS SM: Cell-mediated immunity in interstitial
nephritis. 1. T lymphocyte systems in nephritis guinea pigs: The natural history and diversity of the immune response. J Immunol 123:2373—2380, 1979
11. NEILSON EG, PI-IILLIPS SM: Murinc interstitial nephritis I. Analysis of
disease susceptibility and its relationship to pleiomorphic gene products defining both immune response genes and restrictive requirement for cytotoxic T cells at H-2K. J Exp Med 155:1075—1085, 1982 12. MAMPASO FM, WILSON CB: Characterization of inflammatory cells in
renal injury after treatment with soluble suppressor factor TsFI. J Immunol 139:1518—1524, 1987 29. KELLY CJ, SILVERS W, NEILSON EG: Tolerance to parenchymal self:
The regulatory role of an RTI-restricted, OX8+ suppressor T cell for autologous renal tubular antigen in experimental interstitial nephritis. JExp Med 162:1892—1903, 1985 30. MCMENAMIN C, OlIVER J, GIRN B, HOLT BJ, KEES UR, THOMAS WR,
HOLT PG: Regulation of T-cell sensitization at epithelial surfaces in the respiratory tract: Suppression of IgE responses to inhaled antigens
by CD3 Tcra--lymphocytes (putative y/b T cells). Immunology 74:234-239, 1991 31. DICK AD, CHENG YF, MCKINNON A, LIVERSIDGE J, FORRESTER JV:
457, 1983 13. KELLY CJ, ROTH DA, MEYERS CM: Immune recognition and re-
Nasal administration of retinal antigens suppresses the inflammatory response in experimental allergic uveoretinitis. A preliminary report of intranasal induction of tolerance with retinal antigens. Br J Oph-
sponse to the renal interstitium. Kidney mt 31:518—530, 1991 14. STI:BLAY RW, RUDoFSKY U: Transfer of experimental autoimmune renal cortical tubular and interstitial disease in guinea pigs by serum. Science 180:966—968, 1973
32. DEVEY ME, BLEASDALE K: Antigen feeding modifies the course of antigen-induced immune complex disease. C/in Exp Immunol 56:637— 644, 1984
15. VAN ZWIETEN M, BHAN A, MCCLUSKEY R, COLLINS A: Studies on the
33. GESUALDO L, LAMM ME, EMANCIPATOR SN: Defective oral tolerance
pathogenesis of experimental anti-tubular basement membrane ne-
promotes nephritogenesis in experimental IgA nephropathy induced by oral immunization. J Immunol 145:3684—3691, 1990
autoimmune tubulointcrstitial nephritis in rats. Kidney mt 23:448—
phritis in the guinea pig. Am J Pathol 83:53 1—546, 1976 16. BANNISTER K, WII.SON C: Transfer of tubulointerstitial nephritis in the
thahnol 77:171—175, 1993
34. MII.LER A, LIDER 0, ROBERTS AB, SPRON MB, WEINER HL: Suppres-
Brown Norway rat with anti-tubular basement membrane antibody: Quantitation and kinetics of binding and effect of deconiplcmentation.
sor T cells generated by oral tolerization to myelin basic protein
J immunol 135:3911—3917, 1985
transforming growth factor beta after antigen-specific triggering.
17. NEILSON EG, GASSER DL, MCCAFFLRTY E, ZAKHEIM B, PhIlLIPS SM:
Polymorphism of genes involved in anti-tubular basement membrane disease in rats. Immunogenetics 17:555—565, 1983
18. KELLY CJ, MOK H, NEIISON EG: The selection of effector T cell phenotype by contrasuppression modulates susceptibility to autoimmune injury. J Immunol 141:3022—3028, 1988
suppress both in vitro and in vivo immune responses by the release of PNAS USA 89:421—425, 1992 35. LIDER 0, SANTOS LMB, LEE CSY, HIGGINS PJ, WEINER IlL: Suppres-
sion of experimental autoimmune enccphalomyelitis by oral administration of myelin basic protein. II. Suppression of disease and in vitro immune responses is mediated by antigen-specific CD8+ T lymphocytes. J Immunol 142:748—752, 1989
732
Pham et a!: Oral tolerance in interstitial nephritis
36. MILLER A, AL-SABBAGH A, SANTOS LM, DAS MP, WEINER HL:
Epitopes of myelin basic protein that trigger TGF-beta release after
oral tolerization are distinct from encephalitogenic epitopes and
sion of a nephritogenic murine T cell clone Kidney In! 46:1295—1301, 1994
37. CHeN Y, KUCHROO VK, INOBE J, HAFLER DA, WEINER HL: Regula-
46. MEYERS CM, KELLY CJ: Inhibition of murine nephritogenic effector T cells by a clone-specific suppressor factor. J Clin Invest 94:2093—2104, 1994 47. BAILEY NC, FRISHBERG Y, KELLY CJ: Loss of high affinity TGF-131
tory T cell clones induced by oral tolerance: Suppression of autoim-
binding to a nephritogenic T cell results in absence of growth
mune encephalomyelitis. Science 265:1237—1240, 1994
inhibition by TGF-j31 and augmented nephritogenicity. J Immunol
mediate epitope-driven bystander suppression. J Immunol 151:7307— 7315, 1993
38. VAN HOUTEN N, BLAKE SF: Direct measurement of anergy of antigen-specific T cells following oral tolerance induction. J Immunol 157:1337—1341, 1996
156:3009—3016, 1996 48. CLAYMAN MD, MARTINEZ-HERNANDEZ A, MICHAUD L, ALPER R, MANN R, KEFALIDES NA, NEILSON EG: Isolation and characterization
39. KELLY CJ, KORNGOLD R, MANN R, CLAYMAN MD, HAVERTY TP,
of the nephritogenic antigen producing anti-tubular basement mem-
NEILSON EG: Spontaneous nephritis in kdkd mice. II. Characterization of a tubular antigen-specific, H-2K restricted Lyt-2 effector T
49. AL-SABBAGH A, MILLER A, SANTOS LM, WEINER HL: Antigen-driven
cell that mediates destructive tubulointerstitial injury. J Immunol 136:526—531, 1986
40. MEYERS CM, KELLY CJ: Effector mechanisms in organ-specific auto-
immunity. 1. Characterization of a CD8+ T cell line that mediates murine interstitial nephritis. J Clin Invest 88:408—416, 1991 41. MEYERS CM, KELLY Ci: Adoptive transfer of an antigen-specific DTH-reactive T cell clone induces murine interstitial nephritis. (abstract) JAm Soc Nephrol 2:553, 1991 42. SCHAINUCK LI, STRIKER GE, CUTLER RE, BENDITT EP: Structural-
functional correlations in renal disease. Hum Pathol 1:631—641, 1970 43. BOHLE A, MACKENSEN-HAEN S, VON GISE H: Significance of tuhuloin-
erstitial changes in the renal cortex for excretory function and concentration ability of the kidney: A morphometric contribution. Am J Nephrol 7:421—433, 1987 44. BOI-ILE A, MACKENSEN-HAEN S, VON GISE H, GRUND KE, WEHRMANN
M, BATZ C, BOGENSCHUTZ 0, SCHMITF H, NAGY J, MULLER C, MULLER G: The consequences of tubulo-interstitial changes for renal function in glomerulopathies. Path Res Pract 186:135—144, 1990 45. MEYERS CM, KELLY CJ: Immunoregulation and TGF-/31. Suppres-
brane disease. J Exp Med 161:290—305, 1985
tissue-specific suppression following oral tolerance: Orally administered myelin basic protein suppresses proteolipid protein-induced experimental autoimmune encephalomyelitis in the SJL mouse. Eur J Immunol 24:2104—2 109, 1994 50. SUN JB, HOLMOREN J, CZERKINSKY C: Cholera toxin B subunit: An efficient transmucosal carrier-delivery system for induction of periph-
eral immunological tolerance. Proc Nat! Acad Sci USA 91:10795— 10799, 1994
51. WEINER HL, MACKIN GA, MATSUI M, ORAV Ei, KUOURY Si, DAwsoN DM, HAFLER DA: Double-blind pilot trial of oral tolerization with myelin antigens in multiple sclerosis. Science 259:1321—1324, 1993 52. TRENTHAM DE, DYNESIUS-TRENTHAM RA, ORAV EJ, COMBIrCHI D,
LORENZO C, SEWELL KL, HAFLER DA, WEINER HL: Effects of oral
administration of collagen on rheumatoid arthritis. Science 261:1727— 1730, 1993 53. GENAIN C, ABEL K, BELMAR N, VILLINGER F, ROSENBERG D, LINING-
TON C, RAINE C, HAUSER S: Late complications of immune deviation therapy in a nonhuman primate. Science 274:2054—2057, 1996