J. Comp. Path. 1991 Vol. 104
A Comparison of Central Nervous Lesions Directly Induced by E s c h e r i c h i a coli Lipopolysaccharide in Piglets, Calves, Rabbits and Mice Y. Nakajima, Y. Ishikawa, E. Momotani, K. Takahashi*, H. Madarame*, A. Itot, H. U e d a ~ M. Wada t and H. Takahashi t Hokkaido Branch, National Institute of Animal Health, Sapporo, 062 Japan, * The Department of Experimental Pathology, Kitasato University, Towada, Aomori, 034 Japan and t Tohoku Branch, National Institute of Animal Health, Shichinohe, Aomori, 039-25 Japan
Summary To evaluate the role ofendotoxin during Gram-negative bacterial meningitis, the nervous lesions of piglets, calves, rabbits and mice were compared by direct inoculation ofEscherichia coli lipopolysaccharide into the central nervous system. Suppurative leptomeningitis was induced in piglets by small doses of lipopolysaccharide. Mice also had a mild suppurative inflammation in the leptomeninges. In contrast, calves showed suppurative pachymeningitis, but no lesions in the leptomeninges. Leptomeningeal inflammation was not induced in rabbits. Induction of the leptorneningitis by endotoxin was compared with sensitivity to intravenous or intraperitoneal endotoxin in these specms.
Introduction Despite antibiotic therapy, bacterial meningitis still remains a fatal and devastating illness in man and animals. Some bacterial components, which can induce meningeal inflammation may complicate diagnosis, therapy or prognosis of the disease (Feigin and Dodge 1976; LeFrock, Smith and Molavi, 1985; Quagliarello and Scheld, 1986; Rings, 1987). Streptococcal cell wall components of teichoic acid and peptidoglycan, or fusobacterial cytoplasm and its culture supernate induce meningeal inflammation in rabbits, while capsular polysaccharide of Haernophilus influenzae type b and Streptococcuspneumoniae does not (Tuomanen, Liu, Hengstler, Zak and Tomasz, 1985; Quagliarello and Scheld, 1986; Nakajima 1988). In Gram-negative bacterial meningitis, the following results indicate some role for lipopolysaccharide (LPS) endotoxin in cerebrospinal fluid (CSF) in the development of meningitis. (1) Bacteria-free LPS, corresponding to the bacterial population, increased in the CSF of rabbits with meningitis due to Escherichia coli (Munford, Hall and Grimm, 1984). (2) Neisseria meningitidis isolated from human clinical cases is significantly more endotoxin liberating than that from carriers in vitro (Anderson, Solberg, Bryn, Froholm, Gaustad, 0021-9975/91/010057 + 08 $03.00/0
© 1991 Academic Press Limited
Y. N a k a j i m a et al.
58
Hoinby, Kristiansen and Bovre, 1987). (3) The lipo-oligosaccharide (shorter saecharide chain than classic LPS) of/-/, influenzae type b induces meningitis in rabbits (Syrongiannopoulos, Hansen, Erwin, Munford, Rutledge, Reisch and McCracken, 1988). However, the effect of LPS in CSF on meningitis development has been inadequately examined (Liehr, 1982; Quagliarello and Scheld, 1986; Nakajima, 1988; Syrongiannopoulos el al., 1988). LPS induces a variety of inflammatory responses in animals, the extent of which differs depending on the species (Berczi, Bertok and Bereznai, 1966; Morrison and Ulevitch, 1978; Culbertson and Osburn, 1980; Liehr, 1982; Rietschel, Schade, Jensen, Wollenweber, Luderitz and Greisman, 1982; Morrison and Ryan, 1987). We compared the histopathology of the central nervous system (CNS) lesions induced by direct inoculation of LPS into the CNS in pigs, calves, rabbits and mice.
Materials and Methods
Piglets, calves, rabbits and mice were used (Table 1). Piglets were fernale or castrated male and of mixed breeding (Landrace and Duroc). Calves were castrated male Japanese Black. Rabbits were male and female Japanese White. Mice inbred strains were LPS low responder, C3H/HeJ (Crea Inc.,Japan) and high responder, C3H/HeN (Charles River Japan Inc., Japan) (Morrison and Ulevitch, 1978; Rietschel el al.,
1982).
Escherichia coli LPS (0111: B4, Difco Lab., USA) was dissolved in l ml (piglets), 0"2 ml (rabbits), 0"025 ml (mice) or 5 ml (calves) of sterile 0.85 per cent saline solution
Table 1 Lesions produced by inoculation of E. coli lipopolysaccharide into the CNS of animals Animal
Age (months)
Weight (kg)
Number
Piglet Piglet
2 2
25 20
1 3
0'2 0"02
Piglet Piglet
1 l
I2 to 14 10
3 3
Calf Calf
12 12
230 to 250 200
2 to 3 2 to 3 1.5 1.5 1,5 1.5 1'5 1.5 1.5 1.5
Rabbit Rabbit Mice C3H/HeN C3H/HeN C3H/HeN C3H/HeN C3H/HeJ C3H]HeJ C3H/HeJ C3H/Hej
Dose Terminatedon post(mg per kg) inoculation day
Lesions
0"0002 Saline
1 (Died) 1 (2 of 3 Died) 2 (1 of 3 Killed) 2 (Killed) 1 (Killed)
Meningitis Meningitis Meningitis Meningitis No lesions
5 I
0,1 Saline
2 to 4 (Killed) 2 (Killed)
Pachymeningitis* No lesions
2"5 to 3.5 2.5 to 3.5
3 6
3.3 0.33
i 2
Haemorrhage No apparent lesions
0'025 0"025 0'025 0'025 0'025 0'025 0.025 0'025
4 4 4 4 4 4 4 4
I0 1 0,1 Saline 10 1 0' 1 Saline
1 (3 of 4 Died) 1 (Killed) 1 (Killed) 1 (Killed) 1 (Killed) 1 (Killed) 1 (Killed) 1 (Killed)
* Lesion in the spinal cord.
(Died) (Killed)
Meningitis Meningitis Meningitis No lesions Meningitis Meningitis Meningitis No lesions
(severe) (severe) (mild) (slight)
(mild) (mild) (mild) (mild) (mild) (mild)
M e n i n g i t i s Due to E n d o t o x i n
59
for inoculation (Table 1). Because the sensitivity to LPS differs among animal species, the doses were adjusted to a range that included the lethal dose by the intravenous or intraperitoneal route, except for piglets. The lethal dose for calf, piglet or rabbit is 25 ~tg per kg, 5 nag per kg or more, and 3 mg per kg, respectively (Berczi et al., 1966; Kurtz and Short, 1976; Culbertson and Osburn, 1980). Under anaesthesia with ketamine hydrochloride (40 mg per kg, Ketalar 50, Sankyo K.K., Japan) for piglets, or with ether anaesthesia for rabbits and mice, these animals were intracerebrally inoculated with LPS solution by a 26-gauge needle as previously described (Nakajima, 1988). Calves were inoculated with LPS solution into the lumbosacral spinal cord by an 18-gauge lumbar puncture needle, after 5 ml of CSF was aspirated. Sterile 0"85 per cent saline solution was used as the control. At necropsy (Table 1), samples taken from the brain, 3 to 10 portions &spinal cords, thymus, lungs, heart, liver, spleen, kidneys, stomach, small and large intestines and lymph nodes were fixed in 10 per cent buffered formalin. The brain specimens were cultured in blood agar plates and brain heart infusion broth (Difco Lab., U.S.A.) at 37°C for 2 days, in order to rule out accidental infection. Paraffin wax sections were stained with haematoxylin and eosin (HE). The sections from the brain and spinal cord of the animals other than rabbits were stained to detect LPS antigen by indirect peroxidase anti-peroxidase methods (PAP). Rabbits were immunized subcutaneously with LPS in complete Freund's adjuvant (Freudenberg, Freudenberg and Galanos, 1982). Anti-rabbit IgG Fc sheep serum (kindly provided by Dr S. Takeuchi, National Institute of Animal Health, Tsukuba, Japan) and horseradish peroxidase antiperoxidase complex (Miles Lab., U.S.A.) were used. Normal rabbit, sheep and calf sera served as controls to evaluate specific reactions. Results Piglets inoculated with 0"02 mg per kg of LPS developed nervous signs including trembling, an inability to support themselves on the hind limbs, incoordination, paddling movements or coma and all but one died within 24 h of inoculation. At necropsy, the leptomeninges were oedematous and accompanied by serofibrinou~ exudate. The surface of the brain was cloudy and pale. In the piglets which died, the leptomeninges of all the samples from the brain and cervical, thoracic and lumbar spinal cords were infiltrated by serofibrinous exudates , neutrophils, eosinophils and macrophages (Fig. 1). Blood vessels were congested and oedematous with haemorrhage. The parenchyma next to the leptomeninges was necrotic and spongy with infiltration of degenerated neutrophils. Extension of the leptomeningitis was seen as mild cuffing composed of mononuclear cells and neutrophils in the medulla. Suppurative inflammation often involved the ventricles, spinal nerves, dorsal root ganglia and spinal dura mater. Organs other than the CNS had no marked lesions. Specific LPS antigen was seen in the leptomeningeal membrane and infiltrated macrophages (Fig. 1 inset). The result of bacterial culture was negative. Piglets inoculated with 0"2 gg per kg of LPS survived without nervous signs. Microscopic lesions were infiltration by neutrophils, eosinophils and macrophages into the leptomenlnges or choroid plexus of the brain. A few leucocytes were present in the leptomeninges of the spinal cord with occasional involvement of the spinal nerves. The controls showed no evidence of meningitis. Calves did not show clinical signs except for transient and mild pyrexia. Infiltration by neutrophils and macrophages with fibrinous exudate was
60
Fig. 1.
Y. Nakajima
e t al.
Marked infiltration by neutrophils and macrophages in the leptomeninges of neerotic cerebral cortex is seen in a piglet inoculated with 25 gg per kg of llpopolysaccharide (LPS). HE x 115. Inset: LPS antigen is present on the leptomeningeal membrane and in some of the infiltrated leucocytes. Peroxidase anti-peroxidase (PAP) x 230.
present around the spinal dura mater and on the epineurium of the spinal nerves (Fig. 2a). LPS antigen was seen in the adipose tissue and infiltrated macrophages around the dura mater (Fig. 2b). Neither inflammatory changes nor LPS antigen were observed in the leptomeninges and p a r e n c h y m a of the spinal cords and brains (Fig. 2c). Organs other than the CNS had no marked lesions. The control had no inflarnmatory changes in the brain, spinal cord or spinal d u r a mater. Rabbits inoculated with 3"3 mg per kg of LPS manifested similar signs of endotoxaemia such as nasal and ocular discharge, diarrhoea and depression, and died within 24 h of inoculation. Microscopically, the leptomeninges of the brain were oedematous with mild haemorrhage, however, suppurative inflammation was not seen. Disseminated intravascular coagulation involving the brain or spinal cord was seen in all three rabbits. Rabbits inoculated with 0"33 mg per kg of LPS survived with no apparent suppurative inflammation in the brain and spinal cord. O e d e m a and a few granular leucocytes and macrophages were seen in the leptomeninges of three of the rabbits. The other organs were normal. T h e result of bacterial culture was negative. Three of fbur C 3 H / H e N mice inoculated with 10 mg per kg of LPS died (Table 1). One of four C 3 H / H e N and C3H/HeJ mice in this group had focal ischaemic necrosis with haemorrhage in the p a r e n c h y m a near the ventricles or around the needle track. In all groups, other than controls, there was mild
Meningitis Due to Endotoxin
Fig. 2.
61
In a calf inoculated with 0' 1 mg per kg of LPS, fibrinous and cellular infiltrates around the dm'a mater are seen in (a) (HE x 48). LPS antigen is present in some of the infiltrated leucocytes (PAP x 240) in (b). However, the leptomeninges show no inflammatory changes in (e) HE x 48.
infiltration of neutrophils and macrophages with haemorrhage in the leptomeninges (Fig. 3) and necrotic nerve cells with slight emigration ofneutrophits around the ventricles. Severity of the infiltration appeared not to be dosedependent. LPS antigen was present in the leptomeninges, ventricles and nerve tissues near the ventricles. The control mice had no significant lesions. The result of bacterial culture was negative.
62
Fig. 3.
Y. N a k a j i m a
et al.
/vlild cellular infiltration in the leptomeninges and presence ofneutrophils in the cerebral cortex are seen in a mouse inoculated with 10 mg per kg of LPS. HE x 240.
Discussion
The effects of LPS on the CNS include pyrogenic activity and induction of slow-wave sleep (Morrison and Ulevitch, 1978; Krueger, Kubillus, Shoham and Davenne, 1986; Morrison and Ryan, 1987). We have shown another effect of LPS, i.e. induction of meningitis, on the CNS of piglets. Severity of the meningitis by LPS appeared to be dose-dependent in piglets. I n an experimental rabbit meningitis model produced by E. coli, LPS (from 0" 1 to 100 gg per ml) is released into the CSF, corresponding to a bacterial population of from 10:~ to 109 (Munford et al., 1984). It is likely that the present amount of LPS (0"2 gg per kg or more) may be released in piglets during meningitis caused by E. coli. The present results mean that bacterial lysis, which m a y release LPS into CSF, enhances meningeal inflammation and produces sequelae in piglets. Since LPS is composed of lipid A, R-core and O antigen regions (Morrison and Ulevitch, 1978; Culbertson and Osburn, 1980; Rietschel et al., 1982), effects of each separate eomponent on the induction of meningitis should be examined in piglets. Most of the biological activities of LPS are due to lipid A, but the polysaccharide of LPS has activities independent of lipid A (Morrison
Meningitis Due to Endotoxin
63
and Ryan, 1987). In the case of the rabbit pneumococcal meningitis model, cell wall components of peptidoglycan or teichoic acid induce meningeal inflammation, but not when degraded (Tuomanen el al., 1985). In calves, the pachymeningitis was due to escape of'the inoculum around the dura mater. Differences in the lesions between the dura mater and leptomeninges showed that induction of inflammation by LPS differed among tissues of calves. Though the route of inoculation was different from the other species, CSF and leptomeningeal tissue are common to the spinal cord and brain. There might be no great difference in induction of leptomeningeal inflammation by LPS between the spinal cord and brain of calves. Response to LPS is genetically regulated in mice, and is due to differentiation of macrophages in mouse strains (Rietschel el al., 1982). LPS antigen was seen in macrophages in the leptomeninges. However, the similarity of the lesions in two mice strains suggested that the genetic difference in macrophage generation did not significantly contribute to the induction of meningitis in mouse brain. Differences in lethality, which corresponded to their sensitivity, might be due to intravenous escape of the inoculum, as observed in the rabbit (Mims, 1960). In the rabbit, LPS in CSF-induced similar lesions as previously described (Nakajima, 1988). Lipo-oligosaccharide of Haemophilus influenzae type b caused meningeal inflammation in rabbits, by a mechanism which is considered to be due to lipid A (Syrongiannopoulos el al., 1988). On the other hand, concentration of LPS in the CSF does not correlate with the intensity of meningeal inflammation in the rabbit E. coli meningitis model (Munford el al., 1984). The present resul~ showed no evidence that lipid A causes meningitis, but confirmed the observation by Munford et al. (1984). Evaluation of effects of LPS in the rabbit meningitis model by E. coli should be done with care because of interspecies differences in reaction to LPS. The present variations could be involved in the different pathogenesis of Gram-negative bacterial meningitis in these animals. Calves and rabbits are sensitive to LPS, while piglets and mice are not, which correlates with the phylogenetic development of these species (Berczi el al., 1966; Culbertson and Osburn, 1980). The present difference in induction ofleptomeningitis by LPS contrasts with the sensitivity in these animals. Though the reason for the present difference was unclear, a similar difference among animal species is known in pyrogenicity response to LPS (Ogawa and Kanoh, 1982).
References
Anderson, B.M., Solberg, O., Bryn, K., Froholm, L.O., Gaustad, P., Hoinby, E.A., Kristiansen, B-E. and Bovre, K. (1987). Endotoxin liberation from Neisseria meningilidis isolated from carriers and clinical cases. Scandinavian Journal of Infectious Diseases, 145, 409-419. Berczi, I., Bertok, L. and Bereznai, T. (1966). Comparative studies on the toxicity of" Escherichia coli lipopolysaccharide endotoxin in various animal species. Canadian Journal of Microbiology, 12, 1070-1071. Culbertson, R., Jr. and Osburn, B.I, (1980). The biologic effects of bacterial endotoxin: A short review. Velerinary Science Communications, 4, 3-14.
64
Y. Nakajima
et al,
Feigin, R.D. and Dodge, P.R. (1976). Bacterial meningitis: Newer concepts of pathophysiology and neurologic sequelae. Pediatric Clinics oat".North America, 23, 541-556. Freudenberg, M,A., Freudenberg, N. and Galanos, C. (1982). Time course of cellular distribution of endotoxin in liver, lungs and kidney of rats. British Journal of Experimental Pathology, 56, 56-65. Krueger, J.A., Kubillus, S., Shoham, S. and Davenne, D. (1986). Enhancement of slow-wave sleep by endotoxin and lipid A. American Journal of Physiology, 251, R591-R597. Kurtz, HJ. and Short, E.C. (1976). Pathogenesis of edema disease in swine: Pathologic effects of hemolysin, autolysate, and endotoxin of Eseherichia coli (0141). American Journal of VeterinaryResearch, 37, 15-24. LeFrock, J.L., Smith, B.R. and Molavi, A. (1985). Gram-negative bacillary meningitis. Medical Clinics of North America, 69, 243-256. Liehr, H. (1982). Endotoxins and the pathogenesis of hepatic and gastrointestinal diseases. Ergebnisseder Inneren Medizin und Kinderheilkunde, 48, (117-193). Mims, C.A. (1960). Intracerebral injections and the growth of viruses in the mouse brain. British Journal of Experimenlal Pathology, 41, 52-59. Morrison, D.C. and Ryan, J.L. (1987). Endotoxlns and disease mechanisms. Annual Review of Medicine, 38, 417-432. Morrison, D.C. and Ulevitch, RJ. (1978). The effects of bacterial endotoxins on host mediation systems. AmericanJournal qfPathology, 93, 525-617. Munford, R.S., Hall, C.L. and Grimm, L. (1984). Detection oI' fi'ee endotoxin in cerebrospinal fluid by the Limulus lysate test. Infection and hnmunity, 45, 531-533. Nakajima, Y. (1988). Shwartzman reaction in the brain induced by fi'actions of Fusobacterium necrophorum and Escherichia coli lipopolysaccharide in rabbit,~. Acta Pathologica Japonica, 38, 541-547. Ogawa, Y. and Kanoh, S. (1982). Interactions between bacterial pyrogen and proteolipid extracted from the cerebrum (tII). Variations in Ml3nity ot" proteolipid proteins derived from rabbit, rat and chicken cercbrums to bacterial pyrogen. Japanese Journal of Pharmacology, 32~ 189-194. Quagliarello, v j , and Seheld, W.M. (1986). Recent advances in the pathogenesis and pathophysiology of" bacterial meningitis. American Journal oJ' Medical Science, 292, 306-309, Rietschel, E. T-H., Schade, U., Jensen, M., Wollenwebet', H-W., Luderitz, O. and Griesman, S. G. (1982). Bacterial endotoxlns: Chemical structure, biological activity and role in septicaemia. Scandinavian Journal of Infeclious Diseases, 31 (Suppl.), 8-21. Rings, D. M. (1987). Bacterial meningitis and diseases caused by bacterial toxins. Veterinary Clinics of North America: Food Animal Practice, 3, 85-98. Syrongiannopoulos, G.A., Hansen, EJ., Erwin, A.L., Munford, R.S., Rutledge, J., Reisch, J.S. and McCraeken, G.H. Jr. (1988). Haemophilus influenzae type b lipooligosaccharide induces meningeal inflammation. Journal of Infectious Diseases, 157, 237-244. Tuomanen, E., Liu, H., Hengstler, B., Zak, O. and Tomasz, A. (1985). The induction of meningeal inflammation by components of the pneumococeal cell wall. Journal of Infectious Diseases, 151, 859-868.
~ Received, April 17th, 1990-] Accepted, June 26th, 1990]