Current Drug Targets - Inflammation & Allergy, 2002, 1, 45-52 45

Oral Tolerance in the Treatment of Rheumatoid Arthritis

ric A. Toussirot*

Department of Rheumatology, University Hospital Jean Minjoz, Bd A. Fleming, F-25030

Besanon Cdex France

Abstract: Oral tolerance (OT) consists of the oral administration of antigens (Ag) that
could alter the response of the immune system. This is a form of peripheral immune
tolerance in which mature lymphocytes in the peripheral lymphoid tissues are rendered
non fonctional or hyporesponsive by prior oral administration of Ag. It was first
described in 1911 in animal models of anaphylaxis. This therapeutic approach requires
the orally administration of Ag and the active participation of the gut-associated
lymphoid tissue (GALT), a tissue comprising Peyers patches, intraepitelial cells and
villi containing epithelials cells which is a well organized immune network.

The mechanisms by which OT is mediated included deletion or anergy and active cellular suppression. The
primary factor determining which form of tolerance will be developed after oral administration of Ag is the Ag
dosage. Thus, it is thought that low doses of Ag induce the generation of active suppression, via regulatory T
cells in the GALT, which then migrate to the systemic immune system. These regulatory T cells produce down-
regulatory cytokines such as IL4, IL10 and TGF, a Th2 / Th3 cytokine pattern. Conversely, high dose of Ag
favors anergy or clonal deletion. The phenomenon in which regulatory cells, as generated by oral tolerization,
are primed in an Ag specific manner, but act in the respective microenvironment in a non-Ag specific manner is
called bystander suppression. This phenomenon is of particular interest and explained the use of OT in T cell
mediated autoimmune diseases such as rheumatoid arthritis (RA), multiple sclerosis (MS) and type I diabetes,
some diseases in which the autoAg remains unknown or where there are reactivities to multiple autoAgs.

There were several studies demonstrating the effectiveness of orally administered Ag in different animal
models of autoimmune diseases, such as experimental allergic encephalomyelitis, collagen induced arthritis,
diabetes, but also uveitis, myastenia gravis and transplantation. These mouse or rat models of autoimmune
diseases gave the rationale for the therapeutic use of OT in human and this therapeutic approach has been tried
in MS and RA, using oral myelin or oral collagen, respectively. In RA, 4 trials of oral type II collagen (CII) in
RA have been published. Taken together, these studies suggested that oral CII in RA gave a trend toward
clinical improvement, with significance in only 2 studies. Bacterial extract from Escherichia coli containing
heat shock proteins has been tried in oral treatment for RA. Two placebo controlled trials and 2 comparative
studies gave favorable results for this bacterial extract with no or mild adverse events.

Altough oral/mucosal tolerance has given successful results in animal models of autoimmune diseases, the
enthusiasm for this therapeutic approach in human diseases must be tempered. The discrepancies between the
effectiveness of OT in animal models and the resullts in human trials raise some questions, the identification
of the subgroup of patients who might respond to this treatment and the source (or nature) of the administered
Ag (homologous versus heterologous), for instance.

Key- words: Oral tolerance- Immunomodulation- Rheumatoid arthritis- Collagen induced arthritis.

INTRODUCTION self determinants. This self immune tolerance includes a

Oral tolerance (OT) refers to the oral administration of
protein antigen (Ag) which induces a state of non or
hyporesponsiveness specific for the fed Ag. It is a method to
induce orally peripheral immune tolerance [1-4]. The
immune system must be capable to maintain tolerance to the
hosts own Ags in order to avoid immune response against
*Address correspondence to this author at the Department of
Rheumatology, University Hospital Jean Minjoz, Bd A. Fleming, F-25030
Besanon Cdex France; Phone: 33-3 81-66-82-41; Fax: 33-3 81-66-86-
86; E-mail: eric.toussirot@ufc-chu.univ-fcomte.fr
central tolerance, which is acquired in the thymus by
presentation of foreign Ags, and a peripheral tolerance which
occurs outside the thymus. The mechanisms explaining such
a tolerance state include central clonal deletion, clonal
anergy, peripheral clonal deletion and suppression of self
immune responses [3].
Inflammatory autoimmune diseases can be regarded as a
failure of the immune system to maintain tolerance to certain
self determinants. The treatment of these diseases is difficult
since the etiology is generally unknown and the immune
response is directed to a wide range of self Ags in the target

1568-010X/02 $35.00+.00 2002 Bentham Science Publishers Ltd.

46 Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1
tissue. For instance, rheumatoid arthritis (RA) is a chronic
inflammatory disease characterized by joint inflammation
with progressive joint destruction. Autoreactive T cells to
certain articular components such as cartilage Ags have been
described, but the nature of the auto-Ag(s) is still unknown.
Thus, OT has been proposed for treating some autoimmune
diseases, giving promising results in animal models and a
rationale for its utilisation in human autoimmune diseases
[1,5,6] .
The phenomenon of OT was first described in 1911 by
Wells, who reported that guinea pigs were protected from
systemic anaphylaxis by previous feeding of hens egg
proteins [7]. A few years later, Chase reported the inhibition
of contact hypersensitivity response to haptens by first
feeding the hapten [8]. Subsequently, OT was investigated
using oral administration of ovalbumin or red blood cells
and then, has been used for the treatment of autoimmune
diseases in animal models or in human diseases [1,9].
By definition, OT represents a form of immune tolerance
obtained by the oral administration of Ag to the peripheral
immune system [1,2,5,10-12]. Oral tolerance is a form of
mucosal acquired immune tolerance which also included
nasal and bronchial tolerance [6,13]. There are two main
mechanisms for explaining this oral Ag-driven tolerance:
clonal deletion and anergy, and active suppression [1,2,5].
Oral tolerance has been used to treat RA patients [12,14].
Some experimental data in animal models of RA and results
of therapeutic trials involving RA patients were available.
This review focuses on the mechanisms, the cellular
participants and the regulatory events of OT, and also the
animal models and the therapeutic trials in RA based on this
form of immunotherapy.
THE GUT EPITHELIAL BARRIER AND THE GUT
ASSOCIATED LYMPHOID TISSUE SYSTEM
Orally administered antigens first encounter the gut
associated lymphoid tissue (GALT), a well developed
immune system in the gut [5,15,16]. GALT is devoted for
immune recognization and protection of the host from
ingested pathogens but also for the prevention of immune
response to ingested proteins. A breakdown in this tolerance
to food antigens or to components of the gut flora or foods
will result in food-allergic diseases or inflammatory bowel
diseases, respectively. The cells that comprise the GALT
include epithelial cells, intraepithelial lymphocytes,
lymphocytes in the lamina propria of the gut wall, and
lymphocytes in specialized lymphoid follicles or Peyer
patches. Dentritic cells and macrophages are also found in
the GALT and act as antigen presenting cells (APC). Peyer
patches are organized follicles containing T cells of the CD4
lineage, but also CD8+ cells, B lymphocytes, macrophages
and dentritic cells. Overlying the Peyer patches figure M
cells (or specialized microfold epithelial cells), which are
devoted to the transport of Ag and to their presentation by
APC in Peyer patches. In these lymphoid nodules, specific
immune responses are initiated. Additionally, Peyer patches
are an important source for IgA-producing B cells and the
generation of regulatory cells [1,15]. The regulatory cells
ric A. Toussirot
which play an important role in OT have been well
identified to be CD4+ T cells of the T helper (Th) 2-like
phenotype. Indeed, some experimental data have
demonstrated the generation of Th2 cells producing IL4 and
TGF in Peyer patches following oral administration of Ag
[9,17]. By contrast, intraepithelial lymphocytes and
lymphocytes through the lamina propria probably do not
play an important role in OT [15,16]. Once ingested, Ags
are generally degraded by proteolytic digestion. However,
some Ags are absorbed intact through the gut epithelium,
and then reach the mesenteric and then systemic circulation,
or the hepatic portal circulation [16].
MECHANISMS FOR ORAL TOLERANCE
It is believed that OT is mediated through two main
mechanisms: induction of active cellular suppression, and
clonal anergy or deletion. The main determining factor
influencing these two processes is the Ag dosage. Low dose
of Ag ingestion favors active cellular suppression while high
dose of Ag the alternative mechanism, ie deletion or anergy.
These two mechanisms of OT were suggested by
experimental autoimmune diseases and largely illustrated by
the works of Weiner et al [1,5,9]. Moreover, transfert of OT
could be achieved in the case of active cellular suppression
while the induction of immune ignorance (or anergy) or
clonal deletion are non transferable [5,18,19].
Active suppression is mediated in the GALT, precisely
in the Peyers patches, leading to the induction of regulatory
T cells which then migrate to the systemic immune system
and to the inflammatory area. One of the primary mechanism
of active cellular suppression is the secretion of
downregulating or suppressive cytokines such as TGF, IL4
and IL10 [2,5,17]. Following the oral administration of low
dose Ag, it was demonstrated in mice transgenic for an
OVA-specific T cell receptor, that initial Th1 response was
initiated before induction of tolerance. However, the GALT
system favors the induction of Th2 cells and cells that
secrete TGF. TGF plays an important role since its favors
the switch for IgA production. This cytokine is also
involved in the homing of cells to high endothelial veinules.
TGF is produced by both CD4+ and CD8+ GALT derived
T cells. TGF secreting CD4 cells have been identified as a
particular cellular subset and have been termed Th3 cells [5].
Indeed, these cells help for the IgA production and primarily
secrete TGF. Beside this cellular pattern, Th2 response is
generated in the gut, resulting in the production of cells
which secrete IL4 and IL10. More recently was described
another regulatory cellular subset, Tr1 cells or regulatory T
type 1 cells, which are also a mediator of OT [4]. CD4+ T
cells are considered to be the most important cellular subset
mediating OT. However, CD8 T cells may also mediate OT
in certain conditions. It was thus demonstrated that these
CD8 cells mediating OT bore the TCR. In the model of
OT following oral feeding of ovalbumin, depletion of these
CD8 cells diminished OT [18]. These cells are also
involved in nasal tolerance.
By contrast to the low dose Ag-driven tolerance, a
second mechanism of tolerance could be achieved using high
dose of Ag. This results in a systemic passage through the
Treatment of Rheumatoid Arthritis
gut barrier, and a direct entrance in the systemic circulation
as intact protein. This mechanism is explained by the
induction of unresponsiveness of Th1 cells via clonal
anergy. This is supported by findings that cells impaired to
respond to Ags could be restored by exogenous IL2
[5,18,19]. Finally, a mechanism of clonal deletion has been
proposed in OT, but seems rarely involved. This mechanism
supposes a long lasting phenomenon of tolerance. Apoptotic
events involving the fas system are though to explain such
clonal deletion. More recently, the p55 TNF receptor has
been implicated in this mediated apoptosis-OT [4,18].
Beside these classical mechanisms for explaining OT
induction, another concept recently emerged and corresponds
to the cognate interactions between T cells and APC. Indeed,
the production of inhibitory cytokines or the involvement of
costimulatory molecules on APC explains the suppressive
effects of regulatory or anergic cells. However, the effects of
regulatory cells may also be mediated via a cognate
interaction. A tolerized T cell may modify APC such that
they can induce tolerance in subsequently encountered naive
T cells. A cognate interaction supposes the delivery of a
tolerogenic signal by an APC to the next T cell encountered.
A possible role for the receptor ligand pair Notch-Serrate has
been suspected in such a system [4,19].
An important property of the OT phenomenon that have
practical implications is the by-stander suppression effect
[5,13,20]. In fact, in OT, tolerized T cells secrete
suppressive cytokines in an Ag specific fashion, after feeding
the Ag, but the release of these cytokines in the
inflammatory environment may also suppress immune
response to an unrelated but co-localized Ag. This property
is important in specific therapy for inflammatory
autoimmune diseases in which the autoAg is generally
unknown or in which multiple Ags are involved. A direct
application of by-stander suppression is that for an
inflammatory autoimmune disease, it is not necessary to
know the auto-Ag, but one could induce OT by feeding an
Ag that is capable to induce regulatory cells, which then
migrate to the target tissue and thus suppress the
inflammatory response [5,13].
FACTORS INFLUENCING ORAL TOLERANCE AND
ITS REGULATION
Different mechanisms are involved in OT regulation. As
seen above, the Ag dosage seems the main factor. However,
the nature of Ag, some host factors (genetic background, gut
flora) and also the Ag uptake and processing are important
factors influencing the phenomenon of OT [2,15]. In this
sense, different cells have been proposed to play a role for
APC: enterocytes, conventional APC such as dentritic cells
in Peyer patches and B lymphocytes are supposed to be
tolerogenic APC. Aberrant presentation of Ag by these cells
may influence the occurrence of OT [15,16].
Oral tolerance may be regarded as the inhibition of Th1
response in the peripheral immune system either by high
dose of antigen that induce anergy /or deletion or by gut-
induced Th2 or Th3 regulatory cells. Thus, mechanisms that
favor a Th1 response would abrogate OT and conversely. In
Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1 47
this sense, large doses of IFN abrogate OT and the same is
observed with IL-12. Indeed, this cytokine is involved in
TGF secretion and T cell apopotosis. Conversely, IL4
administration enhances OT in low dose OT in animal
model of experimental allergic encephalomyelitis (EAE). In
the same way, lipopolysaccharide also favors OT to myelin
basic protein (MBP). Certain potent mucosal adjuvants have
been studied such as cholera toxin. The B subunit of cholera
toxin seems to enhance OT. Finally, antibodies to the
chemokine monocyte chemotactic protein 1 (MCP-1)
abrogate OT [5,15,16,18].
ORAL TOLERANCE IN EXPERIMENTAL MODELS
OF AUTOIMMUNE DISEASE
Autoimmune diseases can be regarded as a failure of the
immune system to maintain tolerance to self Ags. Since the
auto-Ag(s) is (are) generally unknown in most autoimmune
diseases, the challenges treatment is to re-establish tolerance
in a previously sensitized host. In this sense, the aim of
mucosal tolerance is to treat and prevent autoimmune
diseases via the recognization of self Ags. A number of
studies has demonstrated that orally administered Ags or
peptides from self Ags can suppress or ameliorate some
experimental models of autoimmunity by inducing a shift in
a Th1 to Th2/Th3 immune response or inducing regulatory
cells or anergy [1,5,6,13].
There are several models of experimental autoimmune
diseases that have been treated by OT: EAE as a model of
multiple sclerosis (MS), collagen-induced arthritis (CIA) or
adjuvant- induced arthritis (AA) as models of RA,
autoimmune uveitis, myasthenia gravis, diabetes,
transplantation etc. [1,9,15]
The most studied models were certainly EAE and animal
models for arthritis [21]. Oral tolerance in EAE was largely
studied by Weiner et al [5,9]. These authors well
demonstrated that orally administered MBP, one of the auto-
Ag in MS, or its fragments in Lewis rats can prevent
animals from developing EAE [22]. Mice can be protected
from relapses of EAE following oral administration of MBP
or proteolipid, another potential autoantigen in MS [5].
The effects of oral administration of collagen, a plausible
auto-Ag for RA, has been evaluated in various models of
arthritis, including CIA, AA, pristane-induced arthritis.
These diseases are induced in rats and susceptible strains of
mice, giving clinical and histopathological changes similar
to those observed in RA [5,11,15]. The Ag used was bovine
or human collagen. Orally administered type II collagen
(CII) can suppress CIA and also AA [23-28] . In the same
way, peptides from type CII can also modulate CIA.
Collagen-induced arthritis with chicken CII can be prevented
by feeding human collagen prior to challenge. Adjuvant-
induced arthritis can be suppressed by transfert of T cells
from collagen-fed animals. Of note is that oral
administration of mycobacterial 65kD heat shock protein
could prevent or attenuate AA and CIA [29,30]. Both
mechanisms of OT, clonal anergy and active cellular
suppression, have been suggested as possible mechanisms
mediating OT in these animal models. Indeed, clonal anergy
48 Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1
has been demonstrated in Lewis rats fed ovalbumin (OVA)
and later challenged with OVA. In the CIA model, results
supported a Th2-like response in collagen fed animals
[9,15]. Another Ag used in animal models of arthritis was
the human cartilage gp39 glycoprotein. Nasal administration
of gp39 to Balb/c mice induced a diminished
hypersensitivity response [31].
THERAPEUTIC TRIALS OF ORAL TOLERANCE IN
HUMANS
Rheumatoid arthritis is the most common inflammatory
joint disease in which the synovial membrane becames
inflamed, with the release of inflammatory cytokines causing
damage to joint components, cartilage and bone. There are
some evidences suggesting that autoreactive T cells
participate in joint inflammation. Th1 cells secreting TNF
and IFN are found in joints of RA patients and thus may
play an important role in the persistence of inflammation in
RA synovium. Thus, therapeutic approach targeting Th1 cell
subset is of particular interest in RA management. Indeed,
animal models of RA have showed that functional activity,
differenciation of Th1 cells can be inhibited by IL4, IL10
and/or TGF, resulting in improvement of the inflammatory
process. Consequently, OT which aimed to induce a shift in
the Th1 / Th2 balance seems promising and appropriate to
RA treatment [14,32].
Furthermore, in RA pathophysiology, the nature of the
recognized auto-Ag responsible for the induction of the
disease remains unknown. However, there is a list of
potential auto-Ag, the most popular being CII. Type II
collagen is the most abundant protein of cartilage. In RA
patients, B and T cell reactivity to CII has been
demonstrated. Antibodies to CII have been detected in early
and late stages of RA and the prevalence of such antibodies
is known to range from 20 to 30%. Further, the possible
induction of animal model of arthritis with CII (i.e. CIA)
gives another argument for the implication of CII in RA
pathophysiology [14]. Minor collagens such as type IX and
XI collagens have also been suspected as potential auto-Ag
in RA. Indeed, in RA, antibody response to a variety of
epitopes on native and denatured type IX and XI collagens,
as well as CII, has been reported. Additionally, other
cartilage components could play a contributory role in
autoimmune pathogenesis of RA: human cartilage gp39,
proteoglycan aggrecan and link protein [14]. In fact, humoral
or cellular reactivity to these different proteins has been
evidenced in patients with RA.
ORAL COLLAGEN IN THE TREATMENT
RHEUMATOID ARTHRITIS
Based on these different aspects of RA pathophysiology
and the mechanisms of action of OT, CII has been used in
the treatment of RA patients.
Firstly, a open trial enrolling 10 patients and using
chicken CII was performed. Patients had recalcitrant disease
and were taken off their disease modifying anti-rheumatic
ric A. Toussirot
(DMARD) and immunosuppressive drugs. 100 g/ day of
solubilized CII were given for one month followed by a
higher dose of 500 g daily for the next 2 months. Six of
10 patients appeared to be improved while the others
worsened the disease due to the discontinuation of their
DMARD. However, no adverse events were noted [33].
Based on the results of this open phase I study, a
placebo-controlled study was then conducted by the same
group. The same dosage regimen and escalation was given
but patients were recruited in excluding those with severe
and recalcitrant disease. In fact, since the open trial results
have suggested that oral CII did not benefit RA patients
with extremely severe disease, such patients might not
respond to this treatment intervention. As it was the case in
the pilot study, DMARDs were stopped and the treatment
period lasted 3 months. Sixty patients with a mean disease
duration of 10 years participated to the study and a
significant improvement of clinical indexes of disease
activity (swollen and painful joint scores) was observed in
the CII treated group. However, no change in erythrocyte
sedimentation rate (ESR) was noted as well as changes in
rheumatoid factor titers and antibodies to CII. As it was
observed in the open study, no side effects were recorded
[33] (Table 1).
This study was then completed by a multicenter, double
blind, placebo-controlled trial involving a large number of
patients (N=274). In this study, 4 different dosages of oral
chicken CII were used: 20, 100, 500 and 2500 g/ day for a
six month period. Patients had a disease duration ranging
from 10 to 13 years. The analyzed criteria for response were
the Paulus criteria, the American College of Rheumatology
(ACR) criteria for improvement, and the requirement for an
improvement of 30% in both the tender and swollen joint
scores. A total of 228 patients completed the 6 month
treatment period, with a majority of dropouts due to a lack
of efficacy. Statistiscal analysis revealed that only the 20 g/
day CII group showed a significant improvement in the
percentage of patients who met the Paulus criteria. However,
numeric trends in favor of the 20 g/day CII group were
seen in the other analyzed measures. Using the Paulus
criteria, an inverse dose response relationship to collagen
feeding was observed. Of interest, the presence of serum
antibodies to CII at baseline was associated with the
likelihood of responding to treatment [34].
Another placebo-controlled, double blind, randomized
study using bovine CII has been published. This study was
performed in Germany, and the CII dosages were 1 or 10 mg
daily. Duration of the trial was 3 months, enrolling 90
patients with early disease (disease duration between 10 and
20 months). The evaluated criteria for response was the ACR
OF
20 (improvement in both tender and joint scores 20% and an
additional 20% improvement in at least 3 of the following
parameters: patient assessment of disease activity, physician
assessment of disease activity, pain assessment, patient self-
assessed disability and acute phase reactant, ESR or CRP).
Most patients had not received prior DMARD therapy. The
results showed no difference between the treated groups and
the placebo group, although a higher prevalence of
responders was observed in CII-treated groups, but without
significance. Collagen II was well tolerated with an equally
Treatment of Rheumatoid Arthritis Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1 49

Table 1. Clinical Trials Using Oral Type II Collagen in Rheumatoid Arthritis

Author Number of Fed antigen Daily dosage and Disease duration Criteria of Results
(Reference) patients duration of of rheumatoid efficacy
administration arthritis

Trentham [33] 59: 28 CII, 31 chicken CII 0.1 mg/kg one 10 years swollen and significant
placebo month, 0.5 mg/kg 2 tender joint scores improvement in
months joint scores

Sieper [35] 90: 30 CII 1 mg, bovine CII 1 mg or 10 mg < 3 years ACR 20% no significant
30 CII 10 mg, 30 during 12 weeks (between 10 and difference
placebo 20 months) between placebo
and CII groups

Barnett [34] 274: 217 CII, 57 chicken CII 20, 100, 500 or 10-13 years Paulus criteria, significant
placebo 2500 g during 6 ACR criteria for improvement in
months improvement, the 20 g CII
tender and group using Paulus
swollen joint criteria
counts

Mc Known [36] 190: 95 CII, 95 chicken CII with 0.1 mg/kg one 12-14 years ACR 20% no significant
placebo ongoing DMARD month, 0.5 mg/kg 5 difference
months between CII and
placebo group

Barnett [37] open trial in 10 chicken CII 0.1 mg one month, 4.3 years swollen and reduction in joint
JRA patients 0.5 mg 2 months tender joint scores scores in 8/10
cases
CII: type II collagen- ACR: American College of Rheumatology- DMARD: disease modifying antirheumatic drug- JRA: juvenile rheumatoid arthritis.

distributed side-effect rate between treatment groups and
placebo group [35].
An additional study using also bovine CII examined
whether oral CII could be an additive therapy to an ongoing
DMARD. In this study, 190 RA patients received 0.1 mg
CII / day for one month and then 0.5 mg CII/ day for the
next 5 months or a placebo. The mean disease duration of
the disease was 13 months and patients continued to take
their DMARD (methotrexate, hydroxychloroquine, gold,
sulfasalasine, azathioprine, D-penicillamine or prednisone).
The analyzed parameter for response was the ACR 20
definition of improvement in RA. No difference was
observed in the rate of response between the treated group
and the placebo group. On the contrary of previous studies
using oral CII, a high withdrawal rate was observed, either
in the placebo and in the treated group (27.4 and 26.3%,
respectively).These dropouts were due to lack of efficacy
[36].
Finally, an open pilot trial using chicken CII during 12
weeks was performed in 10 cases of chronic juvenile arthritis
(with polyarticular disease onset in 3 cases, pauciarticular in
3 cases and systemic in 4 cases). Reduction in both swollen
and tender joints was observed in 8 cases/ 10 without
adverse event [37] (Table 1).
BACTERIAL EXTRACT IN THE TREATMENT OF
RHEUMATOID ARTHRITIS
Another potential autoAg has been used in RA treatment.
Indeed, heat shock proteins (hsp) have been implicated in the
process of joint inflammation of RA. Heat shock proteins
are highly conserved proteins which fonctioned as molecular
chaperones for protecting cells during stress conditions.
They also play different roles in the immune system.
Members of the hsp60 are known to be immunogen. Both
humoral and cellular autoreactivities have been demonstrated
in RA patients, and hsp 60 has been found expressed in
synovium from patients with chronic juvenile arthritis
[38,39]. In addition, protection against the disease may be
achieved by immunization with mycobacterial hsp60 in AA
animals [40]. Furthermore, the Escherichia coli (E. coli) hsp
dnaJ expresses the rheumatoid shared epitope or QKRAA
motif, and molecular mimicry has been demonstrated
between this bacterial protein and HLA DRB1*0401, giving
another argument for the implication of hsp in the RA
pathogenesis [41]. Finally, the shared epitope has the
properties to bind the 70 kD hsp from E. coli [42].
A lyophilized bacterial extract, obtained by fractionating
the hydrolysate of several E. coli strains has been tested in
RA trials. In fact, this compound contains hsp60, 65 and 70
kD and has proved to have immunomodulatory properties
[43-45]. It can reduce the severity of AA and the joint
swelling in the avridine-induced polyarthritis [46,47]. Thus,
this bacterial protein extract (or OM-89) has been used in the
treatment of RA patients, giving favorable results in 2
placebo controlled trials. These studies enrolled 83 and 72
patients in a 6 months period treatment. Both studies
showed a superior effect of the bacterial extract on joint
swelling score, pain, or patient global evaluation [48,49]
(Table 2). Two other studies documented a similar clinical
efficacy between this treatment and classical DMARD
(penicillamine and oral gold) [50,51].
50 Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1 ric A. Toussirot

Table 2. Clinical Trials Using Oral Administration of E. coli Bacterial Extracts in the Treatment of Rheumatoid Arthritis

Reference Number of Dosage of fed Mean disease duration Criteria of efficacy Results
patients antigen and duration (years)
of administration

Brackertz [48] 103: 50 E. coli, 53 24 mg during 6 E. coli group: 7.9 tender and swollen joint significant improvement in
placebo months counts, morning the assessed variables in
Placebo: 7.6 stiffness, ESR, pain the E. coli group

Hauzer [49] 72: 40 E. coli, 32 24 mg during 60 E. coli group: 9 tender and swollen joint significant improvement
placebo months counts, ESR, pain favoring the E. coli group
Placebo: 6.3

Verstraeten [50] 34: 17 E. coli, 17 D 24 mg during 12 E. coli group: 5.1 Ritchie index, tender E. coli equivalent to D
Penicillamine months and swollen joint Penicillamine
D Penicillamine: 6.5 counts, grip strength,
ESR, pain

Vischer [51] 145: 75 E. coli, 70 24 mg during 6 E. coli group: 8.3 Ritchie index, tender E. coli equivalent to
Auranofin months and swollen joint Auranofin
Auranofin: 8.5 counts, grip strength,
ESR, pain

E. Coli: Escherichia coli- ESR: erythrocyte sedimentation rate.

DISCUSSION using animal CII, human CII or CII-derived peptides will be

A total of 4 controlled and 2 open studies investigating
the efficacy and tolerance of oral CII in RA are available in
the current literature. Most have been performed with
chicken CII and 2 with bovine CII. A variable daily dose has
been used, ranging from 0.1 mg to 10 mg, and the period
treatment lasted between 3 and 6 months. In general, the
treatment was well tolerated inducing minor or mild side
effects, with a frequency that did not differ with the placebo
group. Only 2 studies gave evidences of a statistical
improvement in the CII group, only on clinical parameters
(swollen and tender joint score, and Paulus criteria for
improvement) [33,34]. In the responder patients, the
treatment effect was observed at one month of therapy and
can be maintained for 2 months after cessation of oral CII
administration. No statistical changes in the levels of
biological indexes of disease activity (ESR and CRP) were
observed during these studies. Further, the duration of the
trials did not allow to analyse the effects of oral CII on
radiological progression of erosions. Finally, it did not seem
that oral CII have an additional effect to ongoing DMARDs.
A number of comments could be adressed regarding the
results of oral therapy using CII in RA. There are some
discrepancies in the results of these trials. This could be
explained by:
- a different dosage regimen.
- the difference in the CII species that has been used
(chicken versus bovine).
- the duration of the disease may also have an influence
on the clinical response.
Indeed, it could be speculated that the exact dose and
also the nature of the fed Ag are inadequate. Perhaps, the
route of the Ag may also be of importance. Thus, instead of
more effective as well as another route such as the nasal
route [32]. In the CII trials, both early RA and RA patients
with long disease duration have been recruited with a much
evident efficacy in patients with long stage disease. For
explaining their negative results in early RA, Sieper and coll
hypothesized that CII therapy in RA will only benefit
patients with established and erosive disease, in which
cartilage is damaged, thus liberating Ags such as collagen.
Such antigenic peptides from collagen could generate
autoreactive T cells and in this case, oral/mucosal tolerance
could be effective. Conversely, in the case of non erosive
disease, collagen is not a target of autoreactive T cells and
therefore, immune down regulation using this Ag will be
inoperant [35].
In addition, the existing RA treatments may certainly
influence the efficacy of oral CII. Still, NSAIDs may
damage gut mucosa and this could modify the passage of
immunogenic particules through the gut. In the same way,
changes in the gut flora could also influence the induction
and effects of OT. In addition, gut damage can alter
digestion of the Ag which is an essential stage for OT
induction [15,16].
Another problem is that CII is a highly insoluble protein
and was dissolved in acetic acid solution This solublization
procedure requires a cold temperatue to preserve the native
triple helical structure of collagen. In the different trials,
collagen solution was added to cold orange juice in order to
mask the acidic taste. Preparation of orally administered CII
is thus an important phase of OT.
It would also be of interest to identify responder RA
patients to oral CII treatment. In this sense, Barnett et al
have found a correlation between the presence of anti-CII
antibodies at baseline with the likelihood of responding to
treatment [34]. It has been suggested that antibodies to CII
correlated with rapidly progressive disease in early RA.
Treatment of Rheumatoid Arthritis Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1 51

Thus, testing the humoral and cellular reactivity to CII prior MBP = Myelin basic protein
to the start of treatment may help for distinguishing
responders from non responders. MS = Multiple sclerosis

The security of such treatment modality also needs to be CIA = Collagen induced arthritis
considered. In fact , a bacterial, viral or even prion
contamination could be present with the use of animal AA = Adjuvant arthritis
extracts. For this reason, bovine CII does not represent the
ideal Ag for treating human diseases. CII = Type II collagen

Oral tolerance in RA has mainly been tested using CII. DMARD = Disease modifying anti rheumatic drug
However, the other cartilage auto-Ags have not been used
and thus, human trials with oral type IX or XI or oral ESR = Erythrocyte sedimentation rate
administration of gp39 would be of interest.
ACR = American College of Rheumatology
Orally administration of bacterial extract from E. coli in
hsp = Heat shock protein
RA patients has given clinical promising results in placebo
and comparative studies. However, the mechanisms and the
exact protein responsible for the therapeutic effect in RA
need precisions. REFERENCES

[1] Weiner, H.L., Friedman, A., Miller, A. et al. Ann. Rev.

Immunol., 1994, 12, 809-37.
CONCLUSION
[2] Strobel S. Mucosal immunity and the gut epithelium.
Animal experimental models have given the rationale for Interactions in health and disease. Dyn. Nutr. Res., Basel,
the use of OT in human autoimmune diseases. Oral tolerance Karger, 1995.
appears as an attractive therapeutic principle which is both
cheap and easy to administer for application in human [3] Ilan, Y. Can. J. Gastroenterol., 1999, 13 , 829- 35.
diseases. Rheumatoid arthritis is one of the autoimmune [4] Smith, K.M., Eaton, A.D., Finlayson, L.M. et al. Am. J.
diseases that has benefited from this form of Respir. Crit. Care Med., 2000, 162, 175-8.
immunotherapy, using mainly CII. Although enthusiastic
results have been recorded in RA trials, these studies showed [5] Weiner, H.L. Immunol. Today, 1997, 18 , 335-43.
only a trend toward clinical improvement with the use of
CII. In addition, only patients with mild disease participated [6] Xiao, B.G., Link, H. Clin. Immunol. Immunopath., 1997,
to these different trials. Some questions remain: the adequate 85 , 119-128.
dosage and route of the Ag, its nature and species (protein [7] Wells, H. J. Infect. Dis., 1911, 9, 147-51.
versus peptide, chicken versus bovine or human) and the
duration of the treatment. Thus, if CII has been well [8] Chase, M.W. Proc. Soc. Exp. Biol. Med., 1946, 61 , 257-9.
tolerated in the clinical trials, we dont know about the
effects with a long term therapy and the risk of flare or [9] Faria, A.M., Weiner, H.L. Advances Immunol., 1999, 73 ,
exacerbation of the disease [52]. Another unanswered 153- 264.
question is the influence of such treatment on the [10] Weiner, H.L. J. Clin. Invest., 2000, 106, 935-7.
progression of erosions. Thus, the exact place of OT using
CII or another auto-Ag in RA management requires further [11] Meyer, O. Rev. Rhum. (Ed. Fr.), 2000, 67 , 593-603.
studies.
[12] Cleland, L.G., Mayrhoffer, G. Br. J. Rheumatol., 1997, 36 ,
1139-41.
LIST OF ABBREVIATIONS [13] Krause, I., Blank, M., Shoenfeld, Y. Isr. Med. Assoc. J.,
1999, 1, 45-9.
OT = Oral tolerance
[14] Trentham, D.E. Rheum. Dis. Clin. North. Am., 1998, 24 ,
Ag = Antigen 525- 36.

RA = Rheumatoid arthritis [15] Strobel, S., Mowat, A.M. Immunol. Today, 1998, 19 , 173-
81.
GALT = Gut asociated lymphoid tissue
[16] Garside, P., Mowat, A.M. Semin. Immunol., 2001, 13 , 177-
APC = Antigen presenting cell 85.

[17] Yoshino, S. J. Immunol., 1998, 160, 3067-71.

Th = T helper
[18] Krause, I., Blank, M, Shoenfeld, Y. Crit. Rev. Immunol.,
EAE = Experimental allergic encephalomyelitis 2000, 20 , 1-16.
52 Current Drug Targets - Inflammation & Allergy, 2002, Vol. 1, No. 1 ric A. Toussirot

[19] Wardrop, R.M., Whitacre, C.C. Inflamm. Res., 1999, 48 , [36] McKown, K.M., Carbone, L.D., Kaplan, S.B. et al.
106-19. Arthritis Rheum., 1999, 42 , 1204-1208.

[20] Miller, A., Lider, O., Weiner, H.L. J. Exp. Med., 1991, 174, [37] Barnett, M.L., Combitchi, B., Trentham, D.E. Arthritis
791-8. Rheum., 1996, 39 , 623-8.

[21] Meyer, A.L., Benson, J.M., Gienapp, I.E. et al. J. Immunol., [38] Vischer, T.L., Van Eden, W. Ann. Rheum. Dis., 1994, 53 ,
1996, 157, 4230-8. 708-10.

[22] Brod, S.A., Al Sabbagh, A., Sobel, R.A. et al. Ann. Neurol., [39] Bonnin, D., Albani, S. Biotherapy, 1998, 10 , 213-21.
1991, 29 , 615-22.
[40] Van Eden, W. Infect. Dis. Obstet. Gynecol., 1999, 7, 49-
[23] Thomson, H.S.G., Staines, N.A. Clin. Exp. Immunol., 1985, 54.
64 , 581-6.
[41] Albani, S., Tuckwell, J.E., Esparza, L.E. et al. J. Clin.
[24] Nagler, A.C., Bober, L.A., Robinson, M. et al. Proc. Natl. Invest., 1992, 89 , 327-31.
Acad. Sci. USA, 1986, 83 , 7443-6.
[42] Auger, I., Escola, J.M., Gorvel, J.P. et al. Nature Med.,
[25] Thompson, S.J., Harper, N., Bevan, D.J. et al. 1996, 2, 306-10.
Autoimmunity, 1993, 16 , 189-99.
[43] Polla, B., Baladi, S., Fuller, K. et al. Experentia, 1995, 51 ,
[26] Zhang, Z.J., Lee, C.SY., Lider, O. et al. J. Immunol., 1990, 775- 9.
145, 2489-93.
[44] Farine, J.C., Meredith, M. Advances in anti-rheumatic
[27] Khare, S.D., Krco, C.J., Griffiths, M.M. et al. J. Immunol., therapy, CRC Press, Boca Raton, Florida, 1996.
1995, 155, 3653-9.
[45] Peters, D.H., Goa, K.L. Clin. Immunother., 1994, 2, 65-77.
[28] Thompson, S.J., Thompson, H.S., Harper, N. et al.
Immunology, 1993, 79 , 152-7. [46] Willis, D., Moore, A.R., Gowland, G. et al. Br. J.
Rheumatol., 1995, 34 , 525-8.
[29] Billingham, M.E.J., Carney, S., Butler, R. et al. J. Exp.
Med., 1990, 171, 339-44. [47] Willis, D., Moore, A.R., Gowland, G. et al. Br. J.
Rheumatol., 1995, 34 , 1135-8.
[30] Jorgensen, C., Gedon, E, Jaquet, C. et al. J. Rheumatol.,
1998, 25 , 763-7. [48] Brackertz, D., Vischer, T.L. J. Rheumatol., 1989, 16 , 19-
23.
[31] 31-Joosten, L.A.B., Coenen-De Roo, C.J.J., Helsen,
M.M.A. et al. Arthritis Rheum., 1999 (Suppl), 42 , S120. [49] Hauzer, J.P., Appelboom, T. Rheumatol. Int., 1989, 9, 71-
6.
[32] Kalden, J.R., Sieper, J. Arthritis Rheum., 1998, 41 , 191- 4.
[50] Verstraeten, A., Sigeghem, A., Dequecker, J. Scand. J.
[33] Trentham, D.E., Dynesius-Trentham, R.A., Orav, E.J. et al. Rheumatol., 1990, 19 , 422-31.
Science, 1993, 261, 1727-30.
[51] Vischer, T.L. Ann. Rheum. Dis., 1988, 47 , 582-7.
[34] Barnett, M.L., Kremer, J.M., St Clair, E.W. et al. Arthritis
Rheum., 1998, 41 , 290-7. [52] Blanas, E., Heath, W.R. Intern. Rev. Immunol., 1999, 18 ,
217-228.
[35] Sieper, J., Kary, S., Srensen, H. et al. Arthritis Rheum.,
1996, 39 , 41-51.