chronic arthritis produced by streptococcal l-forms

C H R O N I C ARTHRITIS P R O D U C E D B Y STREPTOCOCCAL L-FORMS

JOSEPHINE COOK, w. J. FINCHAM AND c. H. LACK Department of Pathology, Institute of Orthopaedics, Royal National Orthopaedic Hospital,

Brockley Hill, Stanmore, Middlesex

PLATES CXXV-CXXVIII

CHRONIC arthritis in rabbits produced by intra-articular injections of streptolysin S (SLS) has already been described by Weissmann et al. (1965) and by Cook and Fincham (1966). This paper reports that a similar arthritis has been produced by the L-form of the strain C 203 S which is the strain of group-A streptococcus used initially for the production of SLS, and by the L-form of C 203 U, a non-haemolytic variant that we obtained from Professor A. Bernheimer.

Some animals were injected intra-articularly with the parent strain C 203 S so that the changes could be compared with those produced by the L-form. Ten animals were injected with protoplast membranes derived from the C 203 S streptococcus.

MATERIALS AND METHODS

New Zealand Red rabbits, bred in our own laboratory, were used. Induction of L-form growth. To induce L-form growth the parent strain was inoculated

on brain heart infusion agar (Difco) with additional 3 per cent. sodium chloride, and 10 per cent. horse serum added at the time of pouring.

The surface of the plates was seeded with an 18-hr broth culture of the parent strain. Plates were then incubated aerobically for 2 hr at 37°C. After incubation, penicillin solution (5000 units per ml) was added to a hole 4 mm in diameter in the centre of the plate. Plates were then incubated aerobically for 2-14 days at 37°C. A few typical L-form colonies were usually found at the edge of the zone of penicillin inhibition.

Maintenance of growth. The medium above was used, but we added penicillin to the medium to a final concentration of 1000 units per ml before pouring, instead of adding penicillin to a well in the plate. Cultures were incubated anaerobically at 37°C. After several cultures the organisms were found to grow equally well when cultured aerobically.

Subcultures from solid medium were made by cutting blocks of agar, approximately 1 cm square, bearing L-form growth, and inverting them on to the surface of a fresh plate. When fluid cultures were required, the above medium without agar was used to subculture the strain. Primary cultures in liquid media were made by cutting blocks of agar as already described and introducing them into the liquid media. Subsequent subcultures were made by transferring heavy iunocla of fluid culture into fresh liquid media. The success of cultures in liquid media was found to be variable.

Preparation of material for injection. Four methods were tried. The first was to harvest the growth from the surface of an Oxoid membrane filter supported on the agar medium. Growth was subsequently washed from the surface of the membrane. The second method

Received 5 June 1969; accepted 26 July 1969. J. PATH.-VOL. 99 (1969) 283

284 JOSEPHZNE COOK, W. J. FZNCHAM AND C. H . LACK

consisted of harvesting the surface growth from 2-in. (5-cm) agar plates. The third method, which was used for all but the first 4 animals in the C 203 S series, was as follows. Growth in liquid media was centrifuged and the deposit resuspended in fresh sterile medium. In- itially liquid medium containing horse serum was used, but after repeated subcultures liquid medium that did not contain horse serum was found to support the growth of the organism. For the C 203 U series of animals a fourth method was used. Slopes of solid Gform medium were poured and agar blocks bearing L-form growth inoculated on them. After 24 hours’ incubation at 37°C each block was moved about all over the slope. Growth was usually established after a further 1 or 2 days’ incubation, at which time 10 ml of liquid L-form medium was added. After a further 24 hours’ incubation the fluid from several cultures was pooled, centrifuged, washed and resuspended in 0.2 ml of 3.6 per cent. saline. Either fresh medium minus horse serum or saline was used for washing the deposits. They were then resuspended in the same solution.

All these methods yielded viable L-forms after washing and resuspension. Injected material was inoculated on solid L-form medium to test the viability of the organism. In addition, some of the material was inoculated on blood agar plates to exclude reversion to the parent strain. Agar blocks bearing C 203 S and C 203 U L-form growth were inoculated on solid L-form medium to which 10 per cent. human blood had been added, and incubated aerobically and anaerobically at 37°C. This procedure was repeated with the same medium incorporating anti-streptolysin 0. The dilutions of the antitoxin used in three plates contained 19, 67 and 80 International Units. Organisms were killed when required by holding them at 80°C for 20 min.

Parent strain. The organism used was the group-A haemolytic streptococcus C 203 S, and the medium for culture was brain heart infusion (Oxoid). The injection material consisted of 0-2 ml of an 18-hr culture previously diluted 1 in 100 in phosphate-buffered saline. Where heat-killed organisms were used they were treated in the same way as the heated L-forms. When unheated streptococci were injected some of the material was inoculated on blood agar and incubated at 37°C to check viability.

Protoplast membranes. These were prepared from the C 203 S streptococcal A-form by the method described by Freimer, Krause and McCarty (1959).

Zntra-articular injections (C 203 S L-form). In all, 32 rabbits were given injections in either one or both knees.

Group Z. Single injection of washed material. Washed suspension of live L-form culture was injected once only into 13 joints. Animals were killed after periods of time varying between 2 and 90 days.

Unwashed live L-form culture was injected jnto 5 joints. One animal was killed on the 8th day, 2 on day 21 and 2 on day 28.

Group IZZ. Two injections of washed material. Washed live L-form culture was injected twice, at an interval of 1 wk, into 3 joints. One of these animals was killed on day 13, and the other on day 24.

Group ZV. Two injections of unwashed material. Unwashed live L-form culture was injected twice into 2 joints, and both animals were killed on day 24.

Group V. More than two injections into the same joint. Two animals received 4 injections of washed live culture in 1 knee only, at weekly intervals, and were killed on day 36. A further rabbit received 5 injections of similar material at weekly intervals and was killed on day 103.

Group VZ. Single injection of heat-killed L-form culture. Four rabbits were given a single injection of an unwashed heat-killed suspension of L-form culture in the right knee, and the same material, but washed, in the left. One animal was killed at 4 wk, and 3 at 7 wk.

Group VZZ. Control joints. Thirteen rabbits were given a single injection of sterile liquid L-form medium containing horse serum and penicillin, into 1 knee. Four of the animals were killed at 1 wk, 4 at 3 wk and the remainder at 4 wk.

Three animals received an injection into 1 knee only for 2 successive weeks and were killed 1, 2 or 3 wk after the last injection.

Group IZ. Single injection of unwashed material.

ARTHRITIS PRODUCED BY STREPTOCOCCAL L-FORMs 285

Intra-articular injections (C 203 U L-forms). In all, 24 animals were used. Group VZZZ. Eight animals received a single injection of live washed C 203 U L-form

culture in the right knee and the same culture, but heated, in the left knee. The rabbits were killed in pairs at 2 days, 1 wk, 2 wk and 3 wk.

Group ZX. Eight animals received 1 injection of live washed L-form culture in both knees. No further injections were made into the left knee, but L-form culture was injected into the right knee on three further occasions at weekly intervals. The animals were killed at weekly intervals from 1 to 9 wk after the last injection.

Group X. Eight rabbits were treated in exactly the same way as those in group IX, except that the culture was heat-killed before it was injected into joints.

C 203 S parent strain. Nine animals were given intra-articular injections. In 4, live diluted streptococcal culture was injected into 1 knee only; these were killed at 2 days, 2 wk, 2 mth and 3 mth. The same material, but heat-killed, was injected into 1 knee of a further 3 rabbits. These were killed at 2 days, 2 wk and 3 mth.

Live culture was injected into 1 knee of 2 more rabbits and heat-killed culture into the opposite knee. The material in this case was washed 3 times in phosphate-buffered saline before being injected to remove any free toxin that might be present. These animals were killed at 1 and 2 wk.

Protoplast membranes. 0.2 ml of a suspension of protoplast membranes was injected intra-articularly into 10 rabbits. Unheated material was injected into the right knee and material that had been held at 75°C for 20 min. into the left. Two animals were killed after 1 wk, one after 2 wk, two at 3, 5 and 7 wk and one at 12 wk.

Necropsies Post-mortem material was treated as described in a previous paper (Cook and Fincham,

1966). Swabs taken from the jointspost mortem were cultured in digest broth. In addition, small pieces of synovium were taken from all joints that had been injected with live L-forms, and these were cultured on solid L-form medium.

Circulating antibodies Animals were bled before the start of each experiment and again post mortem. The sera

of animals given injections of L-form were tested for precipitating antibodies in gel diffusion against a suspension of the relevant L-form in 0.85 per cent. saline. The gel diffusion method has been described in a previous paper (Cook and Fincham, 1966).

Because horse serum was included in some of the culture medium, it was necessary to screen all the rabbit sera for the presence of antibodies against horse serum. This was used at a dilution of 1 in 10 as the antigen in the central well. Any rabbit serum that showed precipitating antibodies in this test was absorbed with horse serum for 1 hr at room temperature. Sera were then spun sufficiently to deposit any precipitate formed, and the clear supernatant was removed for further testing. Antibodies against horse serum were shown to have been removed, and the sera were then tested against the L-forms.

Differential agglutination tests. All post-mortem sera from rabbits treated with L-forms were tested for rheumatoid factor in a sheep cell differential agglutination test (Ball, 1967), and a latex slide agglutination test, by the Hyland RA method (Hyland Laboratories, Inc., Los Angeles, California, USA).

RESULTS In-vitro testing of injected material

Cultures of material injected showed growth of L-form, proving viability of the organism at the time of the injection. No growth of streptococci occurred on blood agar plates on which the same material had been inoculated; this showed that the L-forms were stable.

286 JOSEPHINE COOK, W. J. FINCHAM AND C. H. LACK

When solid L-form medium enriched with 10 per cent. human blood was inoculated with the C 203 S Lform, growth with haemolysis occurred after both aerobic and anaerobic incubation. This haemolysis was inhibited but not abolished by the addition of antistreptolysin 0, and even when the largest amount (80 IU per ml) was used some trace of haemolysis still remained. From this result it appears that the C 203 S L-form produces small amounts of both streptolysin S and streptolysin 0.

When the C 203 U L-form was inoculated on L-form medium containing 10 per cent. human blood no haemolysis occurred after incubation. Viability testing of the injected parent strain of streptococcus showed that the material contained live bacteria at the time of injection.

0

1

TABLE I Grading of joint changes

~~

Nil

Minimal changes only, i.e., lining-cell hyperplasia marginal or scanty cellular infiltration

Category Joint changes

2 Definite localised changes, i.e., Iocalised dense cellular infiltration, lymph follicles, giant cells, abscess formation

3 Generalised arthritis. i.e., synovial hypertrophy,

Any of the above and any of the features listed

Marked fibrosis

villous proliferation, pannus formation

under 2

Post-mortem Jindings Cultures of swabs from joints into which L-form had been injected were

invariably sterile when inoculated on conventional media, and in no case were L-forms recovered from the synovia, when L-form medium was used.

Only on two occasions did cultures takenpost mortem from joints into which live streptococci had been injected show growth of the organism. Streptococci had been injected into these 2 joints 1 wk and 2 wk earlier. Cultures from the other 4 joints similarly treated were sterile.

The histological changes in joints into which L-form had been injected are summarised in table 11, and vary from no detectable change to a generalised arthritis with synovial hypertrophy, dense cellular infiltration with lymphocytes, plasma cells, epithelioid cells and giant cells, and pannus formation with destruction of cartilage. Others show less marked localised changes. These

TA

BL

E I1

C 2

03 S

str

epto

cocc

al L

-for

m a

rthr

itis

Inje

ctio

n pr

oced

ure

Was

hed

or

unw

ashe

d

Expe

rimen

tal

grou

p

b

hist

olog

ical

chan

ges

of c

ateg

ory

2 N

umbe

r of

join

ts s

how

ing

% B

Tot

al n

umbe

r of

join

ts

treat

ed

z 0

1 2

3 b

-.

I

__

__

_

1 -

1 2 -

2

>2

I1

C 2

03 S

L-f

orm

C 2

03 S

L-f

orm

C 2

03 S

L-fO

rm

C 2

03 S

L-f

orm

C 2

03 S

L-f

orm

I11

IV

Was

hed

Unw

ashe

d

Was

hed

Unw

ashe

d

V

2 3

c b

2 0

b

Y 8

2 1

3

2 0

3 @

13 5 3 2

0 0

VI

-

VII

Num

ber

of

inje

ctio

ns

Nat

ure

1 C

203

S L

-fom

t- 1 2

Liqu

id L

-for

m

med

ium

con

- ta

inin

g pe

nici

llin

and

hors

e se

rum

Aliv

e or

he

at-k

illed

Live

Live

Live

Live

Live

Hea

t-kill

ed

.,,

...

I I

4

Unw

ashe

d W

ashe

d 4 4

1 2

0 t-c

$3 I:

2 1

0

f -

5 0

Oh

I

l3

I*

I..

I 3

0 0

io

...


include the presence of lymph follicles in the synovium and small abscesses resembling tuberculous abscesses, with a necrotic centre of caseous-like material.

The results have been tabulated by grading the lesions as shown in table I. Table I1 shows the numbers of animals in each group graded in categories 0-3 for the C 203 S L-form, and table I11 for the C 203 U L-form. Table IV shows the results, graded in the same way, when the parent strain was used.

C 203 S L-form. A closer analysis of the results in group I (table 11) illus- trates the unpredictability of the model. The two joints that showed no abnor- mality had received an injection 21 and 28 days before the animals were killed, whereas one of the three joints in this group showing a gross arthritis post mortem had received an injection only 5 days previously. Such changes are unlikely to be reversible.

Two of the 6 rabbits showing mild changes only, graded as category 1, were killed 2 days after intra-articular injection, and these might have progressed further if the animals had been allowed to survive longer. These early changes consist of a moderate cellular infiltration of the synovium with lymphocytes, epithelioid cells, polymorphs and plasma cells. The joint cavities contain scanty exudate made up of fibrinoid material and cells similar to those seen in the synovium.

A11 the joints treated with C 203 S L-form and graded as category 3 show projection of the cellular synovium over the articular surface, with destruction of underlying cartilage (fig. l), except one of the severely affected joints in group I. This joint received only one injection, 3 mth before the rabbit was killed, and sections show an organised pannus. The superficial layer of the femoral articular cartilage is replaced by a narrow strip of flattened fibroblasts. The synovium also shows large areas of fibrosis, but in other parts of the synovial membrane there are small areas of closely packed lymphocytes (fig. 2). The walls of the arterioles in the synovium are markedly thickened. These findings suggest that an earlier inflammatory process had led to partial healing by fibrosis, but that by some self-perpetuating mechanism, inflammatory activity was still continuing.

An analysis of groups I11 and IV in respect of the time factor shows clearly that relatively early lesions can be severe and presumably irreversible, whereas the same L-form can cause, after a longer period, less striking changes that might be reversible. For example, one of the animals in group 111, which was killed after 13 days, show a category-3 arthritis with pannus formation (fig. 3), whilst the four remaining joints in these two groups, all of which had received an injection 24 days previously, show definite local changes only. Presumably these changes could have progressed later to a generalised arthritis, but they could also have regressed, leaving a normal joint in the course of time. All the joints that had received more than 2 injections, i.e., those in group V, show a severe generalised arthritis.

The findings in the 8 joints given a single injection of heat-killed C 203 S L-form also show an unpredictable pattern. Two joints show no changes, and 3 others show only minimal, insignificant changes.

Three of the 8 joints show category-2 changes, which are localised and well

COOK, FINCHAM AND LACK

ARTHRITIS PRODUCED BY STREPTOCOCCAL L-FORMS

PLATE CXXV

Frc. ].-Right femur and adjacent synovium showing extensive pannus, 11 days after fourth weekly injection of C 203 S L-form. Haematoxylin and eosin. x 225.

FIG. 2.-Right femoral synovium 3 mth after single injection of C 203 S L-form, showing lymphocytic infiltration and extensive fibrosis. HE. X 200.


ARTHRITIS PRODUCED BY STREPTOCOCCAL L-FORMS PLATE CXXVI

FIG. 3.-Right femoral synovium after 2 injec- FIG. 4.-Right femoral synovium showing infil- tions of C 203 S L-form, 6 days after second tration with lymphocytes and giant cells. Single injection. There is early pannus on the injection of heat-killed C 203 S L-form. femoral cartilage. HE. x 125. Rabbit killed 2 wk after injection. HE. X 200.

FIG. 5.-Synovial ‘ I abscess ”, 7 wk after single injection of heat-killed C 203 S L-form. HE. x 75.


ARTHRITIS PRODUCED BY STREPTOCOCCAL L-FORMS

PLATE CXXVII

FIG. 6.-Right femoral synovium showing lining cell hyperplasia. Two weekly injections of sterile liquid L-form medium. One week after second injection. HE. x 300.

FIG. 7.-Right femoral synovium, 2 days after single injection of C 203 U L-form. HE. x 225.

FIG. 8.-Giant-cell infiltration in right femoral synovium 2 wk after single injection of C 203 U L-form. HE. ~ 8 0 0 .

COOK, FINCHAM AND LACK ARTHIU~IS PRODUCED BY STREPTOCOCCAL L-FORMS

PLATE CXXVIII

FIG. 9.-Femoral synovium 1 wk after single injection of heat-killed C 203 U L-form, showing cellular infiltration. HE. x 300.

FIG. 10.-Hypertrophied, cellular right femoral synovium in joint injected 1 wk previously with C 203 S parent strain of streptococcus. HE. X225.

FIG. 1 1 . - Lymphocytic infiltration in right femoral synovium 2 days after 1 injection of C 203 S parent strain. HE. x600.

ARTHRlTtS PRODUCED BY STREPTOCOCCAL L-FORMS 289

defined. After 2 wk one shows a large aggregation of lymphocytes and macrophages, just beneath the synovial lining (fig. 4). A second joint examined at 7 wk shows a similar lesion in the femoral synovium and a small abscess containing Gaseous-like material, walled off by lymphocytes, in the patellar synovium (fig. 5). The third joint shows a localised collection of giant cells in the femoral synovium at 7 wk. There are no significant differences in the severity of the joint changes between those given washed and those given unwashed material.

None of the control joints shows marked changes. Five of the 13 joints given an injection of liquid medium show lining-cell hyperplasia and so are placed in category 1, whereas all three of the joints given two injections of liquid medium show lining-cell hyperplasia and no other changes (fig. 6).

C 203 U L-form (non-huemolytic variant). The results are shown in table 111. Only one of the joints in group VIII (given a single injection of live non- haemolytic L-form in the right knee) shows a generalised arthritis, and none shows pannus formation.

The joint showing generalised arthritis was graded as category 3 (fig. 7). The changes consist of a marked generalised lymphocytic infiltration and peri- vascular cuffing. Macrophages are also present, and the blood vessels are dilated and congested. Polymorphs are present in the exudate. Live C 203 U L-form had been injected into this joint only 2 days before.

Two of the right knees show similar changes, but of a more localised nature. They were classed as category 2. In one of them there are giant cells in the synovium (fig. 8). These animals were killed 2 days and 2 wk respectively after injection.

Of the left knees, given a single injection of heat-killed L-form, only one, examined 1 wk after the injection, shows any significant change. There are patches of lymphocytic infiltration in the synovium, some of them round blood vessels. This was classed as category 2 (fig. 9). The other left knees in this group show insignificant changes only.

In group IX, the left knees (given a single injection of live non-haemolytic L-form) show significant changes in only 2 out of the 8 joints. The first of these, examined at 9 wk, shows a category-3 arthritis, with fairly widespread fibrosis in the synovium, marginal areas of infiltration with lymphocytes and plasma cells and a thin layer of healing pannus. This is a similar picture to the one seen in the joint into which C 203 S L-form had been injected 3 mth before. They both show healing by fibrosis, with patches of cellular infiltration that suggest that the inflammatory process is still active in some areas. The second, examined at 10 wk, shows category-2 changes. There is patchy infiltration with lymphocytes and plasma cells, with some areas of dense lymphocyte aggregation.

All the right knees (given 4 injections of live non-haemolytic L-form) with the exception of the right knee of the animal killed at 12 wk, show significant changes. In 6 they were classed as category 2, and show changes similar to those described above. The eighth, examined 1 wk after the last injection, shows villous hypertrophy with marked plasma cell infiltration in the villi.

Exp

erim

enta

l GO

UP

C 2

03 U

L-fo

rm

C 2

03 U

Gfo

rm

VII

I L

ive

Hea

t-ki

lled

IX

C 2

03 U

L-fo

rm

C 2

03 U

L-f

om

X

Liv

e L

ive

TA

BL

E

I11

C 2

03 U st

rept

ococ

cal L

-form

art

hriti

s

Kne

e

R

L R

L

R

L

Inje

ctio

n pr

oced

ure

Num

ber

of

inje

ctio

ns

1 1 4 1 4 1

Mat

eria

l inj

ecte

d*

Aliv

e or

he

at-k

illed

N

atur

e

C 2

03 U

L-fo

rm

' Hea

t-kill

ed

C 2

03 U

L-fo

rm

1 H

eat-k

illed

* All

inje

ctio

n m

ater

ial w

as w

ashe

d be

fore

use

.

~~

Num

ber

of

join

ts

trea

ted

8 8 8 8 8 8

Num

ber

of jo

ints

sho

win

g hi

stol

ogic

al ch

ange

s of c

ateg

ory

0 1

2 3

2 3

2 1

4 2

2 0

0 1

6 1

4 2

1 1

0 1

4 3

5 2

0 1

ARTHRITIS PRODUCED BY STREPTOCOCCAL L-FORMS 29 I

In the deeper layers of the synovium there is gross cellular infiltration with macrophages, lymphocytes, cells resembling epithelial cells and giant cells. There are several areas of lymphocyte aggregation. Some of the arterioles show gross thickening of their walls, and there is peri-arteriolar fibrosis. The synovium contains deposits of fibrinoid-like material, and there is pannus on the articular surface of the femur. All the changes seen in this joint are identical with those produced by the C 203 S L-form (haemolytic strain).

In the animals in group X, the heat-killed C 203 U L-form was injected once only into the left knee, but 4 times at weekly intervals into the right knee. The right knees show similar changes, and of the same order, as the right knees in group IX that received the same number of injections, but of live instead of heat-killed Gform culture. Of the 8 right knees in both groups IX and X, 7 show significant changes and were classed in either category 2 or 3. In group X there were 3 joints in category 3, whereas in group IX there was only one.

Thus it is apparent that heat-killed C 203 U L-form is as effective in producing chronic arthritic changes, as when it is injected live. Indeed, when the incidence of pannus in these joints is studied, it seems that the heat-killed material is a more potent agent than the live culture. Four out of 8 knees in group X show pannus formation, whereas among the right knees in group IX only one joint has developed pannus.

Among the left knees in group IX that received only one injection of heat- killed material, only one showed significant changes. This joint was examined 6 wk after the injection, and was classed as category 3. It shows gross lining cell hyperplasia, widespread infiltration with lymphocytes and macrophages, and large areas infiltrated with plasma cells. Fibrin and fibrinoid are present in the synovium. There is one area of early fibrosis, containing many fibroblasts, and there is pannus on the sesamoid bone. None of the other 7 joints in this group shows significant changes.

Control knees (not included in the table) are normal in every case. Out of the total of 6 joints treated with live parent strain, it will be seen that 4 show a generalised arthritis, occurring after as short a time as 1 wk, when synovial hypertrophy and infiltration with lymphocytes and plasma cells are conspicuous (fig. 10).

Pannus has not yet formed, but in the other 3 joints showing generalised arthritis, 2 of which were examined at 2 wk and the third at 2 mth, pannus is present on either the femur, patella or sesamoid bone. The joint in this group classified as a category-2 arthritis was the one examined 2 days after injection. Here the synovium shows widespread invasion with lymphocytes, grouped mainly around blood vessels (fig. 11).

It will be seen from table IV that one joint treated with live streptococci shows no changes at all. This joint had been examined at 3 mth. There are no inflammatory changes, nor is there fibrosis suggesting a healed lesion. Apossible explanation is that, in this particular case, the streptococci were destroyed and removed by the defences of the host tissue so that an arthritis was never established.

C 203 Sparent strain. The results are shown in table IV.


The arthritis produced by the parent strain resembles very much that produced by the Lforms but differs in certain features. Firstly a higher proportion of joints injected with the parent strain show category-2 or 3 arthritis, compared with the proportion of the total showing these changes in the L-form series. Secondly, the changes produced by the streptococcus, when they occur, involve the whole of the synovium. In L-form arthritis the changes may be quite localised and patchy.

The cell populations in the two kinds of synovitis consist of lymphocytes, plasma cells and macrophages or cells resembling epithelial cells. It is only

Number of joints

TABLE IV C 203 S streptococcal arthritis

Number of joints showing histological changes of

category

Injection procedure

1

I Material injected

C 203 S 1 Live 1 Unwashed 1 4 1 1 0 1 2 parent strain

1

Number of treated injections 1 Nature 1 Alive or 1 Washed 1

~

heat-killed or unwashed 0 1 2 3

C 203 S 1 Live 1 Washed 1 2 1 0 0 0 2 parent strain

I I I I i

1 C 203 S 3 0 0 0

I- I I 1

I I I I

2 1 0 0 parent strain I c 2 0 3 s

in the L-form arthritis that giant cells are seen. Another feature of L-form arthritis not seen in the joints into which the streptococcal parent strain had been injected, is the small abscess resembling a tuberculous abscess, with a walled-off core of cheese-like material.

As shown in table IV, heat-killed streptococci produced no significant joint changes. This is in contrast to the situation regarding Lform, for when heat- killed organisms of these strains are injected into joints, significant changes are frequently produced.

Protoplust membranes. The results are shown in table V. Only 2 of the animals show significant joint changes, classed as category 2. The first of these shows scattered infiltration with lymphocytes, plasma cells and giant cells in the synovium, and lining-cell hyperplasia in the right knee, into which unheated membranes had been injected. The left knee, treated with heated membranes, shows one area of giant cells and macrophages in the patellar synovium.

ARTHRITIS PRODUCED BY STREPTOCOCCAL L-FORMS 293

This rabbit, killed at 7 wk, shows patchy fibrosis and small areas of cellular infiltration only, in both knees.

Protoplast membranes C 203 S

Circulating antibodies Only 2 of the 32 animals given injections of C 203 S L-form showed anti-

bodies against the suspension of L-form itself. These animals had received multiple injections and showed a category-3 arthritis. The post-mortem sera of eight rabbits showed either 1 or 2 weak lines in diffusion tests against horse serum. Four of these animals show no significant joint changes. Other

Unheated

TABLE V Joint changes afer injection of C 203 Sprotoplast membranes

1

Injection procedure

Material injected Number of

Protoplast Heated membranes C 203 S

I Heated or I unheated injections

Nature

1

Number of joints treated

10

10

Number of joints showing histological changes of category

0 1 2 3

3 5 2 0

4 4 2 0

animals that had no circulating antibodies to horse serum show a more severe arthritis suggesting that the changes are produced by the L-form itself, and not the foreign protein.

None of the 24 animals given injections of C 203 U L-form showed antibodies to horse serum. When the sera were tested against the C 203 U L-form as the antigen, the results were as follows. In group VIII, when single injections only were given, no precipitating antibodies were found. In group IX, in which live L-forms were injected 4 times into the right knee and once into the left knee, post-mortem sera from 6 out of 8 animals gave from 1 to 3 weak lines in diffusion tests. In group X, where the treatment was the same as for group IX, except that the organism was heat-killed before being injected, the results are very similar. Five out of 8 post-mortem sera showed 1 or 2 lines in the diffusion tests. In table VI these results have been correlated with the joint changes. It does appear that there is a slight predominance of severe joint changes in those animals that had circulating antibodies against the L-form.

Diferential agglutination tests, and latex slide tests. None of the post- mortem sera tested by these methods gave a positive result.

294

3 2 2 2 1

JOSEPHINE COOK, W. J. FINCHAM AND C. H . LACK

0 0 1 0 1

DISCUSSION The characteristic histological features of the synovial changes in rheuma-

toid arthritis have been produced in rabbits by the intra-articular injection of streptococcal L-forms and of crude SLS. Weissmann et al. (1965) suggested that their experiments with SLS appeared to demonstrate that an exogenous agent capable of disrupting lysosomes in vitro and in living cells, can produce an experimental arthritis morphologically resembling the local lesions of human disease.

TABLE VI Precipitating antibodies against C 203 U L-forms in post-mortem ,veerurn samples from rabbits

given intra-articular injections of live and hear-killed C 203 U L-form

Experimental group

IX

x

C 203 U L-forms

alive or heat- killed

Live

~

Heat-killed

Category of joint changes in

R. knee* 1 L. knee? I

3 1 0 2 0 2 0

2 2 I ! 2 1 ; 2

l I f

2 0

3 3 1 ;

Number of lines produced in gel-diffusion between

serum and C 203 U L-forms

1 (weak) 2 (weak) 2 3 2 2 (weak) 0 0

1 (weak) 2 (strong) 1 (weak) 1 (weak) 1 (weak) 0 0 0

* Four injections. f One injection.

In a later report on arthritis induced by filipin, Weissmann, Pras and Rosen- berg (1967) indicated that this labiliser of lysosomes caused loss of matrix polyanions and fibrillation of cartilage as well as hyperplasia and villous formation, cellular infiltration and pannus. But filipin did not induce the focal collections of lymphocytes in the synovial tissue and villi that are an important feature of the lesions produced by SLS and by L-forms.

Following the demonstration that vitamin A causes destruction of cartilage matrix by the release of enzymes from the lysosomes of chondrocytes we might expect any " labiliser " capable of diffusing into cartilage to have this disruptive effect. The absence of this feature in the articular cartilage of joints into which SLS has been injected has not been explained. It might be argued that the


chronic synovitis produced by SLS could still be due to labilisation of lysosomes in synovial cells even though insufficient SLS reached the chondrocytes to produce the disorganisation of cartilage that is seen with filipin. Against this is the report by Cook and Fincham (1966) that chloroquine did not protect joints injected with SLS from undergoing the chronic inflammatory changes already listed.

In the experiments reported here we have no evidence that the L-form of the non-haemolytic mutant contains any SLS ; the haemolytic lysosome-labilising fraction of SLS is heat-labile, yet the L-forms were able to produce chronic arthritis after being held at 80°C for 20 min., at which temperature SLS loses all its haemolytic activity. Keiser, Weissmann and Bernheimer (1964) have also reported that the lysosome-labilising material is heat-labile.

The endotoxic properties of the cell-wall mucopeptide of group-A streptococci are unlikely to be responsible for the changes produced by the L-form, as this is a stable L-form and there was no evidence of reversion to cell-wall synthesis at the time these L-forms were injected, nor were streptococci ever recovered from the lesions.

Experiments in this and other laboratories have shown that a wide range of materials, especially when particulate, will cause hyperplasia of lining cells and villous formation when introduced into a joint cavity. These appear to be a non-specific response of the reticulo-endothelial cells engaged in clearing any foreign matter from the joint cavity.

Pannus and the subjacent erosion of cartilage matrix also occur when particulate matter, especially fibrin, accumulates on the cartilage surface adjacent to synovial lining cells. Pannus is usually seen when there is chronic inflammation with deposition of fibrin on the joint surfaces; it was also found in the filipin-treated joints, but in that case there was an abnormal amount of cartilage debris. It has been suggested that excessive endocytosis of inflammatory debris may be responsible for the leakage of enzymes from cells in the pannus to the surrounding matrix, thereby causing the localised cartilage erosion (Hollander et al., 1965; Riddle, Bluhm and Barnhart, 1965).

The formation of lymphoid foci and of giant cells in joints injected with L-forms is noteworthy. Lymphoid foci are regarded as essential features of rheumatoid synovia, and were numerous in the lesions produced by SLS. Giant cells, also a feature of rheumatoid synovitis, were not found in the SLS lesions. Interstitial and surface fibrinoid deposits feature predominantly in rheumatoid synovia and in both models. Plasma cells associated with the production of rheumatoid factor have received much attention in rheumatoid arthritis, but though plasma cells are found in the cellular infiltrates in the synovitis produced by L-forms and by SLS, rheumatoid factor has not been found in the serum.

Arthritis produced experimentally in rabbits by the injection of streptococci has been recorded by Faber (1915), Hadjopoulos and Burbank (1932), Cecil, Angevine and Rothbard (1939) and Norlin (1960). Norlin, who injected live beta-haemolytic streptococci intravenously, produced all the major features of rheumatoid synovitis, so our results in respect of the live parent strain are not


new. What is surprising is that heated streptococci do not produce results similar to those produced by the heated L-form.

A noteworthy characteristic of both strains of L-form is their capacity, even after they have been heated to 80°C for 30 min., to cause transformation of lymphocytes (Cook and Fincham, 1969). This action may be caused by the mitogen which is known to be present in SLS and which Taranta, Cuppari and Quagliata (1969) showed to be unrelated to the haemolytic factor. Taylor, working in this department, has confirmed this separation of haemolytic and mitogenic activity and produced some evidence suggesting that lymphocyte stimulation may be due to the presence of streptococcal ribonuclease (Taylor, 1969). We have not yet determined whether this stimulation of lymphocytes by L-forms is accompanied by a release of a factor chemotactic for lymphocytes or factors cytotoxic to target cells, such as have been described (Moller, 1965). It is conceivable, though unlikely, that the specific stimulation of lymphocytes could set in train a process of lymphocyte infiltration and division as well as reticulo-endothelial cell damage, with subsequent fibrosis, or that cells other than lymphocytes are stimulated directly by the streptococcal ribonuclease or whatever the mitogen is. Phytohaemagglutinin injected into rabbit joints has caused lymphoid " follicle " formation but not the " rheumatoid " arthritis described here.

Another alternative is that the L-form surface may contain an antigen that cross-reacts with a tissue antigen, thereby setting up an auto-immune response. This possibility is being investigated, though it should be pointed out that if this was the basis for the chronic synovitis, protoplast membranes might be expected to induce similar changes, which they did not.

SUhlhlARY

A chronic rnonarticular arthritis is induced by the intra-articular injection of streptococcal L-forms derived from a streptolysin S-producing strain and from a non-haemolytic mutant. The L-forms induce arthritis even after being heated to 80°C for 20 min.

REFERENCES BALL, J. . . . . . . . . . . 1967. Laboratory diagnostic procedures in the

CECIL, R. L., ANGEVINE, D. M., AND 1939. Amer. J. Med. Sci., 198, 463.

COOK, JOSEPHINE, AND FINCHAM, w. J. . 1966. J. Path. Buct., 92, 461. 3, 9 , Y, 3, . 1969. Life Sci., 8, 357.

FABER, H. K. , . . . . . . . 1915. J. Exp. Med., 22,615. FREIMER, E. H., KRAUSE, R. M., AND 1959. Zbid., 110, 853.

HADJOPOULOS, L. G., AND BURBANK, R.. HOLLANDER, J. L., MACCARTY, D. J.,

ASTORGA, G., AND CASTRO-MUWLO, E. KEISER, H., WEISSMANN, G., AND BERN-

MOLLER, E. . . . . . . . . . 1965. Science, N. Y., 147, 873.

rheumatic diseases, London, p. 107.

ROTHBARD, S.

MCCARTY, M. 1932. J. Bone Jt Surg., 14,471. 1965. Ann. Intern. Med., 62, 271.

1964. J. Cell Biol., 22, 101. €ELMER, A. W.


NORLIN, G . . . . . . . . . . 1960. Acta rheum. scand., 6, 309. RIDDLE, JEANNE M., BLUHM, G . B., AND 1965. J. Reticuloendothel. SOC., 2, 420.

TARANTA, A., CUPPARI, G., AND QUAG- 1969. J. Exp. Med., 129, 605.

TAYLOR, A. G. . . . . . . . . 1969. Biochem. J., 114, 32P. WEISSMANN, G., BECHER, B., WIEDERMANN, 1965. Amer. J . Path., 46, 129.

WEISSMANN, G., PRAS, M., AND ROSEN- 1967. Arthritis Rheum., 10, 325.

BARNHART, MARION I.

LIATA, F.

G., AND BERNHEIMER, A. W.

BERG, L.

J. PATH.-VOL. 99 (1969) T

chronic arthritis produced by streptococcal l-forms

Documents