вход по аккаунту


Induction of arthritis in rats by aqueous suspensions of mycobacteria without the use of oil.

код для вставкиСкачать
We report for the first time the induction of
arthritis by an aqueous, rather than an oil, suspension
of killed tubercle bacilli. This was accomplished in the
highly susceptible dark Agouti strain of rats, by intraperitoneal injection during the healing phase of chemically induced peritonitis. The same procedure (injection
after the induction of peritonitis) augmented the incidence of arthritis produced by bovine type I1 collagen
and Freund’s complete adjuvant. Enhanced delivery of
antigen from the peritoneal cavity to regional lymph
nodes in the postinflammatory state was responsible for
this increase in the induction of arthritis.
The absorption of pigments, oils, normal or
neoplastic cells, tissue suspensions, proteins, and peptides from the peritoneal cavity into the draining
mediastinal lymph nodes of rats is increased during the
healing phase of chemically induced peritonitis (1-9).
Increased absorption of xenogeneic erythrocytes has
been shown to lead to an enhancement of hemagglutinin titers (4).Increased absorption of central nervous
system tissue facilitated the induction of experimental
allergic encephalomyelitis (EAE) in a resistant rat
and increased the severity of EAE or
strain (9,
produced a relapsing form of EAE in susceptible
strains (3,6). With the use of this method, EAE could
be produced by the pure encephalitogenic antigen,
myelin basic protein, or by certain of its constituent
From the Pathology Department, New York Medical College, Valhalla.
Supported by NIH grant NS-22261.
Seymour Levine, MD: Professor of Pathology; Arthur
Saltzman, BS: Research Associate.
Address reprint requests to Seymour Levine, MD, Pathology Department, New York Medical College, Valhalla, NY 10595.
Submitted for publication May 3 1, 1990; accepted in revised
form July 27, 1990.
Arthritis and Rheumatism, Val. 34, No. 1 (January 1991)
peptides in aqueous solution, without the need for
oil-based adjuvants (8,9). This latter finding suggested
the possibility that adjuvant arthritis could be induced
by injecting mycobacteria in an aqueous, rather than
an oil, suspension. This has not been reported previously, and, in fact, the lack of arthritogenicity of
aqueous suspensions of mycobacteria has been reported (10).
We report herein that mycobacteria in an aqueous suspension will induce arthritis if injected into the
peritoneal cavity of highly susceptible rats during the
postinflammatory healing phase of chemically induced
peritonitis. This procedure (injection during the healing phase of peritonitis) also enhances the effectiveness of conventional arthritogenic inocula, previously
considered ineffective in the peritoneal cavity (lCrl5).
Male and female Lewis rats were obtained from
Harlan-Sprague-Dawley (Indianapolis, IN) or bred in our
laboratory, and dark Agouti (DA) rats were obtained from
the Trudeau Institute (Saranac Lake, NY). The weight range
of the rats was 110-180 gm. They were maintained in hanging
metal cages with wire mesh floors, and were fed a diet of
Rodent Chow 5001 (Purina, St. Louis, MO) and tap water ad
libitum. They were not given any food the night before
intraperitoneal (IP) injections, in order to minimize accidental injection into the gastrointestinal tract. Chemically induced peritonitis was produced by IP injection of 5 ml of
sodium hypochlorite (NaOCI; household bleach diluted
1:200 in saline) per 100 grams of body weight. After injection, the rats were quickly rotated 3 times to ensure wide
distribution of the NaOCI. This procedure does not cause
any obvious pain or discomfort, but after 1 week, peritoneal
fibrosis causes shrinkage of the omentum, rounding of the
liver, adhesions, especially around the spleen, and occasional ascites.
One week after NaOCl injection, rats were injected
Table 1. Effect of chemically induced peritonitis on the induction of arthritis by intraperitoneal injection of aqueous or oil-based suspensions
of Mycobacterium tuberculosis, bovine type I1 collagen in adjuvant, or M tuberculosis in adjuvant, in dark Agouti (DA) and Lewis rats*
Incidence of
Arthritis severity score?
Weight of hindfoot (gm)$
Weight of lymph nodes
104 f 53
103 f 33
98 -+ 33
134 k 54
1.09 k 0.05
1.58 f 0.03
1.22 f 0.05
1.70 f 0.11
36 f 9
161 f 51
45 f 8
87 t 20
1.14 f 0.11
1.25 f 0.09
1.23 f 0.07
1.38 f 0.18
74 f 28
197 2 80
58 f 5
138 f 10
2.64 k 0.30
1.72 k 0.38
41 f 6
143 f 10
46 f 9
119 k 17
M tuberculosis in saline
No NaOCl
After NaOCl
Collagen in FIA
After NaOCl
No NaOCl
After NaOCl
M tuberculosis in oil
No NaOCl
After NaOCl
* 0.18
* ND
= not done; FIA = Freund’s incomplete adjuvant; FCA = Freund’s complete adjuvant.
t Values are the mean score on a 0 4 scale, where 0 = no swelling and 4 = most severe swelling.
i Values are the mean t SD weights of freshly isolated hindfeet (a measure of arthritis severity) and of all mediastinal lymph nodes (a measure
of absorption of the intraperitoneal inoculum).
intraperitoneally with 1 ml of a saline suspension of heatkilled Mycobacterium tuberculosis. The Jamaica 22 stain
and a strain of unknown source were used. The mycobacteria had been delipidated by successive extractions, during
a 2-week period, with 4 portions of ether:absolute ethanol
(1: 1, volumeivolume), 4 portions of chloroform, and 4 portions of ch1oroform:methanol (2:1, v/v), all at room temperature. The bacteria were centrifuged after each extraction
and air dried after the last treatment.
The dried preparation was ground and suspended in
saline at a concentration of 4 mg/ml. The delipidation reduced the hydrophobicity of the mycobacteria slightly, so
that a fairly fine suspension was obtained. Nevertheless,
vigorous agitation was necessary to prevent settling. Experiments involving the use of oil or adjuvants were performed
at different times than the experiments with aqueous inocula.
For the last experiment, all glassware was rinsed with ether
to avoid even trace contamination with oils, and gloves were
In the experiments with the more conventional oilbased inocula, rats were given an IP injection of 0.2 ml (1
mg) of bovine type I1 collagen, dissolved in 0.1M acetic acid
at a concentration of 10 mg/ml, and emulsified in an equal
volume of Freund’s incomplete adjuvant (FIA; mineral
oi1:Arlacel A emulsifying agent, 85: 15, v/v). Alternatively,
the rats were immunized with an IP injection of 0.05 ml of
Freund’s complete adjuvant (FCA; 2 mg/ml of killed M
tuberculosis suspended in a mixture of mineral oil and
Arlacel A emulsifying agent, 92:8, v/v), or with the same
dose of mycobacteria suspended in oil without the emulsifying agent.
Arthritis in each limb was scored daily, on a scale of
0-4,where 0 = no swelling and 4 = severe swelling. The
maximum scores observed for each limb were averaged.
Iiour weeks after inoculation, the rats were anesthetized and
bled. Freshly isolated hindfeet (cut off at the ankle joint) and
mediastinal lymph nodes were weighed (as measures of the
severity of arthritis and the extent of the absorption of the
inoculum, respectively), fixed in Bouin’s fluid, embedded in
paraffin, sectioned, and stained with hematoxylin and eosin.
Histologic evaluation of absorption of inoculum into lymph
nodes (granulomas, oil drops, suppurative foci) was done
using coded, randomized slides.
Six experiments were performed with aqueous
suspensions of mycobacteria. Some or all of the DA
rats in all 6 experiments developed arthritis after an
incubation period of 10-20 days, but only after pretreatment with NaOCl (Table 1). In the same experiments, only 1 of the Lewis rats developed arthritis,
also after pretreatment with NaOC1. There was no
difference in the induction of arthritis between the 2
strains of M tuberculosis that were used. In 2 experiments, the delipidation was done with ether:ethanol
only or with chloroform only, and the results were
similar to those obtained with the full sequence of the
3 solvents used for delipidation. The weakest response
(development of arthritis in only 1 of 8 DA rats) was
obtained in the experiment using the most settled
suspension, perhaps reflecting insufficient grinding in
saline. (Particle size has been shown to be important
even when bacilli are injected in oil [16].)
The induced arthritis usually started in a hindpaw, spread to the other extremities, and increased in
severity during the next few days. In a few instances,
the arthritis was as severe as that produced by conventional inocula of tubercle bacilli in oil. There were
no differences in the clinical appearance and evolution
or in the histologic features of the arthritis (Figures 1
and 2), compared with those seen in arthritis induced
using inocula in oil.
Figure 1. Histologic analysis of the synovium of a dark Agouti rat
with arthritis induced by intraperitoneal injection of an aqueous
mycobacterial suspension. after pretreatment with NaOCl (chemically induced peritonitis). The synovial membrane is infiltrated by
many mononuclear inflammatory cells. Articular cartilage and subchondral bone are seen along the right margin (hematoxylin and
eosin stained, original magnification x 200).
Figure 3. Histologic analysis of the omentum of a dark Agouti rat
that was inoculated with an aqueous mycobacterial suspension, but
was not pretreated with NaOCl and did not develop arthritis. There
are many sequestered clumps of acid-fast bacilli (round central area
of gray amorphous material), which are surrounded by radially
disposed epithelioid cells. Adjacent areas are edematous (hematoxylin and eosin stained, original magnification x 200).
Swelling or nodules on the tail and ears were
seen occasionally. In 2 of the rats whose data are
summarized in Table 1, the initial inoculation with
mycobacteria in saline was repeated for the next 3
days, without resulting in any increase in the severity
of arthritis. Some of the rats were given a quadruple
dose all at 1 time, which did not affect the incidence of
arthritis, but seemed to increase severity, albeit irregularly.
The results of autopsies of these rats were
revealing in a number of ways. The omentum of rats
that had not been pretreated with NaOCl and had not
developed arthritis was full of residual inoculum in the
form of clumps of tubercle bacilli, up to 100 pm in
diameter, with epithelioid cell granulomas, focal suppuration, giant cells, fibrosis, and edema (Figure 3 ) .
This was the expected result of the normal sequestering function of the omentum. Rats that had been
pretreated with NaOCl had the expected evidence of
healing peritonitis, with fibrosis of the liver and spleen
capsules, and omentum (Figure 4). Some of these
shrunken and fibrotic omenta did not contain any
clumps of tubercle bacilli and, when these clumps
were present, they were not as numerous as in the
omentum of the rats that were not pretreated with
The important consequences of these differences between the group of rats with chemically
induced peritonitis and the group without chemically
induced peritonitis were demonstrated by the findings
in the draining mediastinal lymph nodes. These consequences were not evident in the weights of the
lymph nodes, which had marked individual variability
but showed no overall difference between the 2
groups. However, microscopic analysis revealed more
granulomas in lymph nodes of NaOC1-pretreated rats
than in the non-pretreated group, and in 1 experiment,
small suppurative foci were seen only in some of the
NaOC1-pretreated rats. These lesions undoubtedly
Figure 2. Histologic analysis of the synovium of a dark Agouti rat
with arthritis induced by intraperitoneal injection of an aqueous
mycobacterial suspension, after pretreatment with NaOCl (chemically induced peritonitis). The lower synovial surface has ulcerated,
and fibrinous exudate protrudes into the lumen (hematoxylin and
eosin stained, original magnification x 100).
Figure 4. Histologic analysis of the omentum of a dark Agouti rat
that was inoculated with an aqueous mycobacterial suspension,
after pretreatment with NaOCl (chemically induced peritonitis). The
surface of the omentum is fibrotic, and there is no sequestration of
mycobacterial clumps. (In some other rats that were inoculated with
an aqueous mycobacterial suspension after NaOCl pretreatment,
there was some sequestration of mycobacterial clumps, but to a
much smaller degree than was seen in rats that were inoculated with
an aqueous mycobacterial suspension without NaOCl pretreatment.) (Hematoxylin and eosin stained, original magnification x
represent a reaction to the absorbed inoculum. Previous work has established a causal relationship between the reduced sequestering ability of the fibrotic
omentum and the enhanced absorption of inoculum
into the lymph nodes (1-9).
It was of interest to study the effect of NaOCl
pretreatment on arthritis produced by conventional
inocula, for comparison with that produced by aqueous suspensions of mycobacteria. Arthritis did not
develop in any of the 24 rats that were given 1P
injections of collagen in FIA but did not have prior
NaOC1-induced peritonitis, a finding that is consistent
with the results of other investigations (15). In contrast, most DA rats and many Lewis rats did develop
collagen-induced arthritis when immunized by IP injection after NaOCl treatment (Table 1). The draining
mediastinal lymph nodes were 2 4 times larger in
NaOCl-treated rats than in the non-pretreated animals.
This enlargement was caused by an increase in granulomas and oil drops related to absorption of the
FCA, despite its content of mycobacteria, was
not effective in the production of adjuvant arthritis
when injected intraperitoneally, a result that confirmed the findings of other investigators (10-14). Only
2 of 20 rats that were not pretreated with NaOCl
developed arthritis, and it was mild in both cases.
Pretreatment with NaOCl increased the incidence of
arthritis, but not to the degree noted with the use of
collagen (Table I).
In all the experiments described above, NaOCl
pretreatment was helpful or essential in the induction
of arthritis, and DA rats were more susceptible than
Lewis rats. However, IP injection of a suspension of
tubercle bacilli in oil was much more effective in
inducing arthritis than was FCA or collagen, as previously reported by others (17,18). It produced arthritis
in all DA and Lewis rats tested, independent of
pretreatment with NaOCl (Table 1).
Although NaOCl had only a moderate influence
on the arthritogenicity of FCA and no effect on the
arthritogenicity of tubercle bacilli suspended in oil, in
all cases it caused a striking increase in the size
(weight) of and number of histologic lesions in draining
lymph nodes, which was related to absorption of
inocula (oil drops, granulomas, necrotic foci). As an
additional control, NaOCl was injected intraperitoneally without any further treatment. In addition to
showing a healing peritonitis, these rats had slight
hyperplasia and enlargement of mediastinal lymph
nodes (16 mg and 19 mg, respectively, compared with
5 mg and 9 mg, respectively, for normal controls);
however, there were no granulomas, oil drops, suppurative foci, or necrotic foci. Therefore, none of the
changes described above and in Table 1 can be ascribed solely to the NaOCl.
We have induced what is usually called “adjuvant arthritis” or “adjuvant disease” without the use
of adjuvant, but with the mycobacterial component
contained therein. Perhaps a better designation would
be “mycobacterial arthritis” or “mycobacterial disease.” This arthritis seems to be analogous to Poncet’s
disease (tuberculous rheumatism not caused by tuberculous infection of the joints) (19), and it may be
related to the arthritis that occurs during therapy with
bacillus Calmette-Guerin (20).
The induction of arthritis by mycobacteria in
aqueous suspension emphasizes the value of the
postinflammatory state in enhancing absorption of
antigens into the lymph nodes. Indeed, the incidence
of arthritis was related to the antigenic load delivered
to regional lymph nodes for the collagen and FCA
inocula as well, a finding observed previously by
another investigator who used other routes of inoculation (21). This conclusion is also consistent with our
previous observations in other immunologic systems
(1-9). The apparent lack of influence of NaOCl pretreatment on arthritis induced by mycobacteria in oil,
despite enhanced absorption, may simply reflect the
strong arthritogenicity of that particular inoculum, so
that absorption was not a limiting factor.
Our previous work has shown that the fibrosis
and shrinkage of the omentum in the postinflammatory
state interferes with its normal function of sequestering the antigenic inoculum. More of the antigen is able
to reach the diaphragm, where it is absorbed into the
lymphatic system. However, this chain of events does
not exclude the possibility that pretreatment with
NaOCl may also have augmented the essential steps of
antigen processing and presentation by activated macrophages or by increased numbers of macrophages or
dendritic cells in the peritoneal cavity or draining
lymph nodes. Indeed, the slight hyperplasia of the
lymph nodes observed after treatment with only
NaOCl is consistent with this possibility.
We do not advocate the routine use of the IP
route of inoculation, but it does have certain advantages. Specifically, the procedure is rapid, it requires
no anesthetic, and it causes no discomfort. It also
avoids disturbance of the feet and the tail, where
clinical observations will be made. Large volumes of
inocula can be injected, and repeated doses can be
tolerated. For the occasional study in which the IP
route of inoculation is desirable, the present findings
indicate that there are a number of effective methods
of arthritis induction.
We thank Drs. Jan Schwarz and S. D. Chaparas for
providing the mycobacteria, and Drs. S. S. Kerwar and A.
Weinstein for providing the collagen.
1. Levine S: Postinflammatory increase of pathogenicity of
lymphocytic inocula in the peritoneal cavity. Proc SOC
Exp Biol Med 178:462-467, 1985
2. Levine S: Postinflammatory increase of absorption from
the peritoneal cavity into lymph nodes: particulate and
oily inocula. Exp Mol Pathol 43: 124-134, 1985
3. Levine S: A relapsing form of allergic encephalomyelitis
in rats due to enhanced absorption of inoculum. J
Neuropathol Exp Neurol45:247-257, 1986
4. Levine S, Saltzman A: Increased hemagglutinins after
intraperitoneal immunization in postinflammatory state.
J Immunol Methods 105: 179-182, 1987
5. Levine S, Saltzman A: The role of antigen absorption in
the resistance of Brown-Norway rats to experimental
allergic encephalomyelitis. Immunol Lett 19: 103-108,
Levine S, Saltzman A: The hyperacute form of allergic
encephalomyelitis produced in rats without the aid of
pertussis vaccine. J Neuropathol Exp Neurol 48:255262, 1989
Levine S, Saltzman A: Lymphatic metastases from the
peritoneal cavity are increased in the postinflammatory
state. Invasion Metastasis (in press)
Levine S, Saltzman A, Deibler GE: Encephalitogenicity
for rats of myelin basic protein without the aid of
water-in-oil emulsions. J Neuropathol Exp Neurol 49:
480-485, 1990
Levine S, Saltzman A, Deibler GE: Peptides of myelin
basic protein are encephalitogenic in rats without the aid
of water-in-oil emulsions. Proc SOCExp Biol Med 195:
Ward JR, Jones RS: Studies on adjuvant-induced polyarthritis in rats. I. Adjuvant composition, route of injection, and removal of depot site. Arthritis Rheum 5557564, 1962
Pearson CM: Development of arthritis in the rat following injection with adjuvant, Mechanisms of Hypersensitivity. Edited by JH Shaffer, GA Logrippo, MW Chase.
Boston, Little, Brown & Co., 1959
Glenn EM, Gray J: Adjuvant-induced polyarthritis in
rats: biological and histologic background. Am J Vet Res
26: 1180-1 194, 1965
Beck FJ, Whitehouse MW: Impaired drug metabolism in
rats associated with acute inflammation: a possible assay
for anti-injury agents. Proc SOCExp Biol Med 145: 135140, 1974
Debrito FB, Corry DG, Moore AR, Howat DE,
Willoughby DA: Polyarthritis and the air pouch reaction: dissimilarity of adjuvant and collagen models.
Immunopharmacology 15: 123-130, 1988
Stuart JM, Cremer MA, Kang AH, Townes AS: Collagen-induced arthritis in rats: evaluation of early immunologic events. Arthritis Rheum 22:1344-1351, 1979
Liyanage SP, Currey HLF, Vernon-Roberts B: Influence of tubercle aggregate size on severity of adjuvant
arthritis in the rat. Ann Rheum Dis 34:49-53, 1975
Dolbeare FA: High incidence of adjuvant arthritis as a
secondary lesion to peritonitis in rats. Proc SOCExp Biol
Med 138:312-316, 1971
Petersen BH, Dolbeare FA: Immunoglobulins, rheumatoid factor and antibody in adjuvant induced peritonitispolyarthritis of the rat (abstract). Fed Proc 32:982, 1973
Dall L, Long L, Stanford J: Poncet’s disease: tuberculous rheumatism. Rev Infect Dis 11: 105-107, 1989
‘Torisu M, Miyahara T, Shinohara N , Ohasto K,
Sonozaki H: A new side effect of BCG immunotherapy:
BCG-induced arthritis in man. Cancer Immunol Immunother 5:77-83, 1978
Newbould BB: Lymphatic drainage and adjuvant induced arthritis in rats. Br J Exp Pathol45:375-383, 1964
Без категории
Размер файла
1 425 Кб
without, induction, mycobacterium, arthritis, oil, aqueous, suspension, rats, use
Пожаловаться на содержимое документа