Microscopio electrónico de transmisión (MET)

Models of inflammation have been used extensively for the screening and

evaluation of anti-inflammatory drugs. Various agents have been used in the past including brewers yeast, formalin, dextran and egg albumin. Unfortunately the anti-inflammatory drug effects (measured as a reduction in oedema induced by these agents) are often non­ specific and have failed to demonstrate clear dose-response relationships at clinically relevant concentrations. During the sixties several newer agents (carrageenan, kaolin, mustard and formalin) were tested and carrageenan was found to possess distinct advantages over previously used inflammatory agents ( Winter et al., 1962; Woolf and Thompson, 1991).

Three of the more common models of inflammation in animal experimental work are described below. A brief description of formalin and Freund’s adjuvant is followed by a fuller description of carrageenan.

4 .2 .1 Form alin

The use of formalin grew out of experiments that utilised hypertonic saline in human experimental subjects to produce brief, intense laboratory pain (Lewis and Kellgren

1939; Frankstein, 1947). Dubuisson and Dennis gave the test its definitive form and used it to evaluate the analgesic effects of morphine, pethidine and periaqueductal grey matter stimulation (Dubuisson and Dennis, 1977).

Formalin is the aqueous solution of 37% (weight/weight) formaldehyde often with methanol as preservative. It produces a sterile injury if injected into living tissue and has been used in many animal models of pain. Injected into human subjects it produces a poorly localized pain with burning and stinging qualities which then gives way to a steady throbbing ache lasting up to 60 minutes. In rats, the behavioural response takes a biphasic pattern. Initial pain behaviours after paw injection (elevation, licking and shaking) last about 5-10 mins and are followed by a brief period (5-15 mins) during which the rat ignores the paw. A second phase of pain behaviour then re-emerges and lasts for up to 2 hours (Dubuisson and Dennis, 1977; Guy and Abbott, 1992).

The test has the following advantages over tests of acute nociception (such as the tail flick test):

a) the relatively continuous nature of the induced pain compared to transient laboratory pain induced by heat, electric shock or skin deformation. The pain being modelled is tonic rather than phasic (short lasting). This may be of greater clinical utility.

b) its convenience as the test obviates the need for animal restraint, allowing unhindered observation of the behavioural responses. This represented a major advance from the use of the tail-flick test, the flinch-jump test and the hot-plate test.

c) the use of formalin concentrations between 0.05 and 0.2% avoids major permanent tissue damage. (Although originally suggested to be repeatable, formalin can lead to blister formation and is now rarely applied more than once in any animal).

d) the formalin test appears to distinguish between acute nociception (first phase) and sensitization (second phase) and much has been made of this distinction in

pharmacologic studies utilizing the formalin test (Yamamoto and Yaksh, 1992). A further advantage of the formalin test is its ability to model spontaneous pain whose behavioural correlate is flinching and licking (Tjolsen et al., 1992).

e) some correlation with human experience is possible (at least in volunteers) in contrast to the models like the “quinolone writhing test”.

f) it does not appear to induce a systemic illness (in contrast to complete Freund’s adjuvant ).

g) the test has a limited duration (1 hour) and is therefore ethically acceptable.

Disadvantages of this model include: a) tissue temperature dependence- especially in smaller rodents (mice), b) susceptibility to environmental cues and stress and c)

requirement for scoring of multiple behavioural parameters.

4 .2 .2 Com plete Freund’s Adjuvant (CFA)

This is a suspension of heat killed and dried Mycobacterium Tuberculosis (H 37Ra 25177 ). It is made up in mineral oil and mannide monooleate and can be emulsified in saline. It appears as a clear amber liquid containing brown particles.

Adjuvants are admixtures of compounds which are capable of stimulating the immune system in a non-specific fashion (Allison and Byars 1991). Adjuvants may act in any of the following ways:

i) stimulation of macrophages to secret cytokines such as IL l and IFN. This follows either direct (for particles and emulsions) or indirect (requiring interaction with complement) phagocytosis by tissue macrophages.

ii) as an antigen vehicle to facilitate long-term presentation of an antigen or to protect an antigen from degradation.

iii) direct stimulation of T cell carrier function. This is achieved by inclusion of T cell epitopes from bacterial proteins in the admixture.

iv) direct stimulation of B cells. This is achieved by incorporating mitogens e.g. muramyldipeptide, lipid A or bacterial polysaccharides.

v) reduced lymphocyte circulation with resulting improvement in lymph node responses (Allison and Byars 1991).

Freund’s adjuvant is prepared by the addition of heat killed mycobacteria (+/- 0.5 mg/ml) to a water in oil emulsion (WIO). This emulsion has, in the past included clear mineral oils such as paraffin or plant oils such as peanut oil. When mixed with an emulsifier and water the result is known as Freund’s incomplete adjuvant (FIA). WIO emulsions have been used in veterinary practice for some time either alone or as an adjuvant with various microorganisms. Beneficial effects are noted with a large range of adverse effects, the two being difficult to separate. Local inflammation is consistently present at sites of inoculation and may be crucial to proper functioning as an adjuvant. The nature of the oil phase and the emulsifier appear to determine the degree of local irritation and is similar across species.

The addition of microorganisms (e.g. M. mycoides, ActinoB pleuropneum.. Bovine rota virus) and bacterial proteins and cell wall components increases the immune response to specific determinants thereby creating a vaccine. CFA enhances the immune response aspecifically and is still considered an adjuvant and not a vaccine. It finds regular use in the manufacture of specific immune system derived products (antibodies, cytokines, T cells etc.).

Inflammation induced by CFA (complete Freund’s’ adjuvant) also involves joints and deep tissues (Calvino et al., 1987) i.e. it induces a local and systemic arthritis (Fawcett

1990). Significant side effects of CFA evident in various animal models (both commercial and experimental) have prompted reviewers to demand careful consideration of ethical issues and animal welfare regulations (Claassen et al., 1992).The range of local side effects inlclude: granuloma and sterile abscesses,

ulceration, fistulous tracts and necrotizing dermatitis (Johnston 1991), sterile peritonitis (Toth 1989), splenomegaly (Toth 1989) and muscle atrophy.

Systemic side effects include: “metastatic granuloma” e.g. pulmonary (Schiefer 1979), lymphoid hyperplasia, autoimmune polyarthritis and uveitis (Petty 1989), peripheral nerve demyelination (rabbits) (Mizisin 1987).

Despite these limitations, the injection of CFA into the root of the tail

subcutaneously became established as a model of polyarthritis in the rat (Pearson and Wood 1959). This adjuvant-induced arthritis (ALA) model was subsequently validated in a variety of behavioural and pharmacological studies (Calvino et al., 1987; Colpaert, 1987).

Because the complexity of the polyarthritic model makes it less useful for the study of chronic pain states (i.e. observed changes in behaviour could not be definitively and solely ascribed to nociceptive input) an attempt to develop a model of unilateral, localized inflammation using CFA was made. An intraplantar injection of CFA in to the hindpaw (rather than into the root of the tail as in the ALA model) was reported to produce just such a localized inflammatoiy insult (Stein et al., 1988). This report remains unconvincing however, as a systemic response was still recorded (reduced weight gain, reduced food and water intake and a disruption of circadian temperature regulation), follow up was only to 34 days, neither clinical nor histological study was made of supposedly un-involved joints, data on sensory thresholds involving un-involved joints was inadequate and the researchers admitted that in a proportion of animals the arthritis did appear in other joints. The persistence of this model in studies of nociception may partly be a result of uncritical acceptance of this study.

Several more recent studies using the CFA model have shown long term

consequences of the induced inflammatory response. The sensory effects of CFA persist well beyond those of carrageenan; cold allodynia for example persists for 70 days in rats (Perrot, 1993). A few studies have documented long term changes in nociceptive

responses to mechanical stimuli (Butler, 1992) and central changes that persist for 30 days have been reported (Goff, 1998; Jasmin et al., 1998).

Recently , long lasting changes in central afferent terminal fields and behaviour in adult rats following neonatal inflammation with CFA have been reported (Ruda et al., 2000). Some of these changes may be due to non-neural processes (Anand, 2000), but it is equally likely that an intense inflammatory response may result in neural damage and that the behavioural changes are a result of neonatal neuropathy.