The production of prostaglandins in response to experimentally induced osteomyelitis in rabbits

Prostaglandins

and Medicine

2: 403-412,

1979

THE PRODUCTI ON OF PROSTAGLANDINS IN RESPONSE TO EXPERIMENTALLY INDUCED OSTEOMYELITIS IN RABBITS M. Corbett, S. Dekel, B. Puddle, R. A. Dickson and M. J. Nuffield Orthopaedic Centre, Headington, Oxford OX3 7LD. (reprint requests to MJOF).

0. Francis.

ABSTRACT Osteomyelitis was induced in the tibiae of rabbits by injection of staphylococcus aureus and sodium tetradecylsulphate (STD); additional rabbits were injected with STD alone. Confirmation of osteomyelitis was based on positive cul ture of the same phage type bacteria from the tibiae and on the characteristic radiographical and histological appearance of osteomyel i tis. Only tibiae which proved to be infected by the above criteria showed significantly increased in vitro release and content of Prostaglandin E and Prostaglandin F2Acompared with tibiae injected with STD (PcO.05). After two weeks infection, infected tibiae released nine times more Prostaglandin E and five times more Prostaglandin F &than tibiae injected with STD alone. After four weeks infection,infec released less Prostaglandin E (PCO.05) than t ed tibiae after two weeks infection but the release of Prostaglandin F$was similar. The production of large amounts of prostaglandins by bones in response to infection may be the cause of the rapid bone resorption and sequester formation observed in osteomyel i tis. INTRODUCTION Prostaglandins are important mediators of inflammation (1). They are present in inflammatory exudates (2) and their levels are increased in the synovial fluids of patients with inflammatory joint disease (3). In general the relationships of the prostaglandins to the consequent inflammation are complex and the different series of prostaglandins may be mutual antagonists Thus the acuteness of an inflammatory reaction may depend on the (1, 4). ratio of Prostaglandin E to Prostaglandin F produced during the inflammation (2, 5). Prostaglandins also stimulate bone resorption, being active in the nanomolar range (6). Bone loss is observed in inflammatory conditions in which high levels of prostaglandins have been produced adjacent to bone as by inflamed bone cysts (8) or the joints of patients with gingival tissue (7)) dental

403

rheumatoid

arthritis

(3).

The present experiments have been performed to determine if prostaglandins are also produced during osteomyelitis, the response of bone to bacterial infection. Confirmation of elevated prostaglandin production as a result of bacterial infection of bone would not only explain the gross bone destruction that can result from osteomyelitis but also suggest a new rationale for the treatment of this distressing and painful condition (9). METHODS Animals New Zealand white rabbits, weighing 2 - 2.5 kg., were experiments. They were caged individually and allowed and a pelleted rabbit food. Production

of

experimental

used free

in all access

Ten rabbits were used for this group. The (ii) Control rou tibiae - of contra rabbits were injected with 0.4 ml. of 3% (w/v) saline. followed by 0.1 ml. of physiological of

water

osteomyelitis

(i) Osteomyelitis group. Twenty rabbits were included in this osteomyelitrs of the right tibia was established by a modification method of Norden (10). In summary osteomyelitis was induced by into the right upper tibia of 0.4 ml. of an aqueous solution of sodium tetradecylsulphate (STD) followed by 0.1 ml. of a suspension containing about one million staphylococcus aureus (S. Aureus; type 80) and finally 0.1 ml. physiological saline. The rabbits al lowed to recover and kept for up to four weeks before sacrifice.

Confirmation

to

group. An of the injection 3% (w/v)

were

right STD

osteomyelitis

At weekly intervals lateral uninjected

all tibiae

rabbits were weighed and examined by radiography.

both

injected

and

contra-

At sacrifice each tibia was separated from adherent soft tissues and the proximal end, measured from the junction of the fibula with the tibia, split through the sagittal plane into two halves. Samples were taken from The the tibiae and cultured for subsequent bacteriological examination. tibiae were then washed in physiological saline, weighed and used for measurement of prostaglandin production (see below). After this measurement half the bone was fixed in 10% formaldehyde, stained with haematoxylin-eosin and examined histologically, the other half was used for measurement of total prostaglandin content. Rabbits injected

have with

been included in the osteomyelitis S. Aureus and STD satisfied the

(a)

Characteristic

appearance

(b)

Histologically

proven

of

group following

osteomyelitis

osteomyelitis 404

(11).

only if the criteria:on

radiography

tibiae

(Fig.

1)

(c)

Isolation on culture.

of S. Aureus

of

same phage

About three quarters of the rabbits whose right with S. Aureus and STD developed osteomyelitis by the above three criteria.

type

(i.e.

Oxford

80)

type

tibiae had been injected in these tibiae, as defined Figure 1. Radiographic changes of an infected rabbit upper tibia one, two and four weeks after initial infection by injection of S. Aureus and STD. Note the significant bone loss after one week and new bone and sequester formation after two and four weeks.

Measurements

were

thus

(a)

established

(b)

S. Aureus

(c)

control group STD alone).

In contrast

four

obtained

from

osteomyelitis injected

types

of

not

that

those

tibia

groups

(‘infected’)

group

(i.e.

three

did

whose

were

not right

develop tibiae

tibia1 were

osteomyelitis injected

with

examined:-

‘infected’.

(b)

injected

but

(c)

injected

with

(d)

normal (no injection). Normal tibiae tibiae of rabbits whose right tibiae S. Aureus and STD or with STD alone. of

rabbits:-

group.

(a)

Measurement

of

infected.

STD alone.

prostaglandin

were the contralateral had been injected either

left with

production

The washed tibiae, weighing approximately 5 g., were placed in 20 ml. of a Ringer-bicarbonate buffer solution (12) and incubated in shaking water bath at 370 C for up to 4 h. Samples for prostaglandin assay were withdrawn at 30, 60, 120, 180 and 240 min. and itnnediately frozen to -2OO. At the end of the incubation period the half of the bones not taken for histological 405

examination (see above) were weighed, crushed in liquid nitrogen in a Spex freezer mill and prostaglandins extracted from the resulting bone powder after acidification with a mixture of ethyl acetate and cyclohexane (1: 1; v/v). The Prostaglandin E and F2dand contents of the incubation media and bone extracts were estimated using the radio-immunoassay and antibodies developed by Bauminger (13) and Kirton (14). RESULTS PGE release

two weeks after

induction

of osteomyelitis

Two weeks after injection a low amount of PGE was released during 4 h, incubation by tibiae injected with STD alone. PGE release by infected tibiae was significantly increased even after 30 min. incubation and reached nine times that of the STD injected tibia after 4 h. incubation (Fig. 2).

w (s,“TIIaytn$ !$ $D injected (broken line) tibiae incubated in vitro for four hours following twos infection in vivo. Mean * ‘SEM; 0 =NS; *PC 0.05; +P
200 POE w/o bor*lso

The increased PGE release by the infected t incubation ranged from six to 22 times that significant differences in PGE release were with STD and S. Aureus that did not become

406

ibiae was variable and after 4 h. of the STD injected tibiae. No observed between tibiae injected infected and normal (uninjected) tibiae

PGE release

four

weeks

after

induction

of osteomyelitis

PGE release by infected tibiae was significantly increased (P(O.01) that released by the tibia injected with STD alone only after 2 h. (Fig. 3). Figure

3.

As for

Figure

2 but

following

four

weeks of

infection

over incubation

in vivo.

60-

I Incubation

%ne (mins)

I

I

I

120

180

210

In general, PGE release by infected tibiae was significantly decreased compared to the PGE released by infected tibiae two weeks after induction of osteomyelitis (Fig. 4). Figure 4. Comparison of release of PGE by infected tibiae incubated in vitro following two (solid line) -((broken 1 ine) following two weeks of infection in vivo. Legen d otherwise as in Figure 2.

--•

tins1

120

la0

I

7.4

407

PGF2urelease Both than

two

two and four those injected

and

four

weeks with

weeks

after

induction

after injection, STD alone (Fig.

of

osteomyelitis

infected tibia 5 and 6).

released

more

PGF2c(

i+&&&y;;;::::d (sol id 1 ine) and STD injected (broken 1 ine) tibiae incubated in vitro for four hours following two weeks of infection in viva, Legend otherwE in Figure 2.

- -

bone

xxx

POFd” Wg bone 6fol lowing infection

6-

l Incubation

%e

1

(mins)

120

I

l60

408

I

2&I

four weeks in vivo.

In contrast to the release four weeks after induction

w

In v tro infection Tibia1

of PGE the release from tibiae obtained of osteomyelitis was similar (Fig. 7).

Comparison of release of PGF &by infected following two (sol id line) or 3our (broken in vivo. Otherwise as in Figure 2. content

of

two and

tibiae incubated 1 ine) weeks

PGE and PGF2

The tiblal content of PGE and PGF20(in general followed the same pattern as the release of PGE and PGF2d. Infected tibiae contained significantly more PGE and PGF2dthan uninfected tibiae (PcO.01) both two and four weeks after induction of infection but the PGE content was significantly reduced after four weeks compared to two weeks after infection (PC 0.001; Fig. 8). No difference in bone content was found between the STD injected and uninjected normal bones (Fig. 8).

409

PGE

800-f7

Wmo

bone

\

600

P< o-001

Control

ST0

I‘4 JInlfected

Control

ST0

infected

Control

ST0

infected

Control

ST0

Infected

PGF2a Wmg bone

Figure 8. PGE and PGF2d content of control, tibiae after four‘hours incubation in vitro, and four (PGE: B; PGF20’ : D) weeks tnfection

STD injected and infected fol lowing two (PGE: A; PGF20’ in vivo. Mean * SEM.

: C)

DISCUSSION The present results demonstrate that one response of bone to infection is the release of large amounts of prostaglandins (Figs. 2 - 7). Injection of sclerosing agent (STD) alone did not result in significantly elevated Elevated prostaglandin release was strictly confined prostaglandin release. histological and bacteriological to those bones which had radiographical, Importantly those tibiae which had been evidence of continuing infection. injected initially with both sclerosing agent and S. Aureus but that did not become infected, by the criteria given in the ‘Methods’ section, did not release additional prostaglandin compared to the control tibiae. The time course of prostaglandin release changed during the course of the infection. After two weeks infection significantly greater, though variable, amounts of prostaglandins of the E series were released from the infected After four weeks compared to the uninfected tibiae at all incubation times. much less additional prostaglandin E was released by the infected tibiae (Fig. 4) and the increase, over the controls was only significant at the In contrast the in vitro release two-hour time point of incubation (Fig. 3). of PGF2dby the infected tibiae at two and four weeks after infection was similar (Fig. 7). Thus the ratio of PGE/PGF2Kdeclined during the course of

410

the infection. This may be related to changes in infection as has earlier been postulated (2, 5).

the acuteness

of

the

Establishment of a chronic infection in bone following an acute infection is notorious and the observed changes in ratio of prostaglandins produced by the bone may reflect this. However, in the inflammatory response of tissues prostaglandin production seems to be most important in mediating the initial acute tissue responses (1, 5). The potent stimulation of bone resorption by prostaglandins, especially PGE2 is well documented (6). The release of PGE (probably PGE2) by bone in response to the establishment of an infection within bone demonstrated here could therefore account for the bone destruction often observed clinically. The osteomyelitis achieved by the method used here resembles that of humans both radiographically and pathologically (9). Thus in humans also the bone destruction that occurs in osteomyelitis could result from the prostaglandin produced locally in response to the osteomyelitis. Bone destruction in acute osteomyelitis in children can lead not only to fracture but also to permanent deformity (15). Treatment with drugs that inhibit prostaglandin production, i.e. the anti-inflammatory drugs (4) could be of benefit in the prevention of such traumatic complications.

ACKNOWLEDGEMENTS We are grateful to Dr. S. Bauminger and Dr. K. Kirton supplies of their prostaglandin antibodies available supported by a Girdlestone fellowship of the Nuffield

for kindly making to us. M.C. was Orthopaedic Centre.

REFERENCES 1.

Prostaglandins as mediators of inflammation. Vane JR. Prostaglandin and Thromboxane Research (B Samuelsson, Raven Press, New York, 1976.

2.

Velo GP, Dunn CT, Girand JP, Timsit J, Willoughby Prostaglandins in inflammatory exudate. Journal

1949,

DA. Distribut ion of of Pathology II I :

1973.

3.

Robinson DR, Levine L. in rheumatic diseases: Prostaglandin Synthetase Pess, New York 1974.

4.

Horrobin DF. significance.

5.

Crunkhorn P, Willis given intrademrally

507,

In Advan ces in R Paoletti eds)

fluid Prostaglandin concentrations in synovial Action of indomethacin and aspirin. In Inhibitors (HJ Robinson, JR Vane eds.) Raven

Prostaglandins: Eden Press,

physiology, Montreal 1978.

pharmacology

and clinical

AL. Interaction between prostaglandins E and F in the rat. British Journal of Pharmacology 41:

1971. 411

6. 7. 8. 9. 10. 11.

12. 13.

Dietrich Clinical

JW, Raisz Orthopaedics

Goodson peridontal Harris bone

JM,

LG. III:

Dewhirst disease.

M, Jenkins resorption

by

FE, Brunetti Prostaglandins

Acute 59B:

Floman arthritis

Y,

Bauminger prostaglandin

Okon E, Zor in the rat. S,

6:

Kirton KT. Communication.

Upjohn

15.

Morrey BF, Orthopaedic

Peterson Clinics

and

Prostaglandin 81, 1974.

bone E2

metabolism.

levels

Prostaglandin MR. 245: 213, 1973. in

children.

osteomyelitis. Diseases 122:

410,

and

human

production Journal

and of

A description 1970.

of

JN,

Silberberg PK, Lerner AM. Xeroand pathologi . studies in experimental (39926). Proceedings of the Society and Medicine 156: 303, 1977.

The U. Clinical

Company,

calcium

osteomyelitis 2, 1977.

role of prostzglandins Orthopaedics 125:

214,

Radioimmunological Prostaglandins

4:

Zor U, Lindner HR. synthetase activity.

14.

Willis Nature

E. Experimental of Infectious

Crane LR, Kapdi CC, Wolfe bacteriologic radiographic, staphylococcus osteomyelitis for Experimental Biology

in

A.

MV, Bennett A, dental cysts.

Mollan RAB, Piggot J. Bone and Joint Surgery Norden CW, Kennedy the model. Journal

Prostaglandin 228, 1975.

Kalamazoo,

Hematogenous HA. of North America

412

Michigan,

6:

pyogerric 935,

in

experimental

1977.

assay of 313, 1973.

U.S.A. osteomyelitis 1975.

Personal in

children.