Strain-related effects of fenbendazole treatment on murine experimental autoimmune encephalomyelitis

Abstract

Parasitic infections are a concern in animal facilities, in view of their influence on physiological processes and the immune status of animals. Pinworms are effectively controlled with the anthelminthic fenbendazole (FBZ, [5-(phenylthio)-1H-benzamidazol-2-yl]carbamic acid methyl ester; C₁₅H₁₃N₃O₂S); however, questions remain as to whether prolonged FBZ exposure alters the disease course in specific experimental models, such as those pertaining to the immune system. We report that a three-month regimen of FBZ-medicated feed severely affected the onset and disease severity of murine experimental autoimmune encephalomyelitis (EAE), a disease that mimics multiple sclerosis. Differences were recorded between mouse strains used. Our data suggest that where the use of FBZ is mandatory, its full effect should be verified on the particular EAE variant adopted by the laboratory.

Experimental autoimmune encephalomyelitis (EAE) is a model of neuroinflammation used for investigations of immunopathogenetic mechanisms that may underlie multiple sclerosis (MS). ¹ The disease is induced by active immunization with a myelin protein-derived peptide in complete Freund's adjuvant, in susceptible rodents or non-human primates. ² Protocols also include injection of toxin from Bordetella pertussis, which aids in disease initiation by causing transient breakdown of the blood–brain barrier. ² In rodents, specific combinations of neuroantigen/mouse or rat strain give rise to a range of clinical profiles resembling those observed in MS.^3,4

Studies described here are based on two mouse strains: C57BL/6J (Animal Resources Centre, Canning Vale, WA, Australia) and non-obese diabetic (NOD)/ShiLtJ (Walter and Eliza Hall Institute, Kew, VIC, Australia). Our institutional animal facility complies with regulations set by the National Health and Medical Research Council of Australia (NH&MRC, care of animals site: nhmrc.gov.au), which is the national governing body for overseeing animal experimentation. All procedures require approval from the institutional animal ethics committee. Mice are housed in an air-conditioned room with 12:12 h light-dark cycle and are fed with standard rodent chow (Barastoc Rat and Mouse Feed, Ridley AgriProducts, Melbourne, VIC, Australia) and tap water ad libitum. When required, fenbendazole (FBZ, [5-(phenylthio)-1H-benzamidazol-2-yl]carbamic acid methyl ester; C₁₅H₁₃N₃O₂S)-medicated rodent diet (Speciality Feeds, Glen Forrest, WA, Australia) is used. Our laboratory has generated two variants of the murine EAE model, ⁵ resulting in either (a) a relapsing–remitting clinical profile in NOD/ShiLtJ mice and peptide 35–55 from myelin oligodendrocyte glycoprotein (MOG_35–55) or (b) a chronic progressive profile with MOG_35–55 in C57BL/6J mice. Disease induction is performed in adult mice aged 9–11 weeks. Progression is monitored by scoring of motor symptoms that usually occur in an ascending pattern, ⁵ where 1 = tail weakness, 2 = hind limb weakness, 3 = hind limb paralysis and 4 = paralysis of hind limb plus one forelimb. Mice are culled as soon as score 4 is reached. The first motor symptoms occur at 14–16 days postdisease induction (dpi) in the NOD/ShiLtJ mouse. This first attack lasts for about three days, reaching a maximum score of 2.5–3, followed by remission. The second attack occurs at about 30–40 dpi, but subsequent ones are less predictable. Three to four attacks, over 60 days, can occur before a secondary progressive stage is reached. In the chronic progressive model, a score of 2.5–3 is also observed by 14–16 dpi, but is followed by progression without remission and severe illness within 30 dpi. Two control groups are included, namely vehicle-only mice and mice that are monitored till days 30–40, to ensure reproducibility of disease course (positive controls). In both models an incidence of 99% or greater is observed. In our hands, the above profiles have proven to be highly reproducible. The only deviation from these patterns has been attributable to mild stress, for example noise due to maintenance (unpublished observations), which results in moderate to high reduction in disease intensity. Here we show that administration of the anthelminthic drug FBZ can severely affect disease incidence and severity of EAE.

FBZ is delivered orally by premixing into rodent chow at a rate of 150 mg/kg of diet. This assumes an average daily food intake of 8 mg/kg body weight, sufficient to ensure a therapeutic dose. ⁶ In some animal facilities, FBZ-medicated chow is administered according to a rolling regimen, while in others it is delivered when required, for periods of up to 14 weeks. Following the discovery of pinworm eggs by perianal tape testing and fecal flotation of caecal contents in mice in the containment room, subsequently confirmed to be of the species Syphacia obvelata and Aspiculuris tetraptera, the whole rodent area was switched to FBZ-medicated feed for 13 weeks. Mice appeared asymptomatic; there was no apparent increase in anal prolapse.

Because our studies are focused on early pathological events, it is crucial to monitor the reproducibility of disease pattern. It became apparent that coinciding with FBZ treatment, reduction in disease incidence, clinical onset, disease duration and severity, within positive control groups of both models, was observed (Table 1). Disease incidence was reduced to about 75% overall. At the time coinciding with the first attack in NOD/ShiLtJ mice (14–16 dpi), only 38.6% of mice demonstrated any clinical scores (Group A), which were low (average of 1) and of short duration (1 day). By 30–40 dpi, a further 7.7% first demonstrated clinical scores (Group B), which were again low (1.5) and of short duration (1 day). However, the single mouse within Group B did demonstrate further attacks. A third subgroup experienced their first attack with a long delay, with 23.1% at 50 dpi (Group C) and 7.7% above 80 dpi (Group D). Surprisingly, in the latter two groups some mice had severe attacks with clinical scores of up to 3–3.5, with lesions of similar histological appearance to those normally observed (not shown). The remaining mice never experienced any attacks (Group E). The effect was less severe in C57BL/6J mice, with 62.5% showing overt disease at the anticipated time of 14–16 dpi and reaching scores of 3 or above (Group F), while 12.5% first demonstrated clinical scores at 30–40 dpi (Group G). The remaining 25% never demonstrated overt disease (Group H). The mice used in these experiments were born prior to FBZ treatment. Their weight relative to age at the time of disease induction was normal and weight increase over the period of FBZ treatment appeared normal.

Table 1

Effect of FBZ on disease incidence, clinical onset and disease severity on the EAE model

Mouse strain and treatment	No sick mice/total no. mice (%)	14–16dpi (range of scores)	30–40 dpi (range of scores)	>50 dpi (range of scores)	>80 dpi (range of scores)	Clinical patterns NOD/LtJ	Clinical patterns C57BL/6J
NOD/LtJ Untreated	(99%)	1 st attack (2.5–3.0)	2nd attack (2.5–4.0)	3rd attack (2.5–4.0)	Secondary progressive (3.5–4.0)
NOD/LtJ ± FBZ: Group A	5/13(38.6%)	1 st attack (0.5–1)	-	-	-
NOD/LtJ ± FBZ: Group B	1/13(7.7%)	-	1 st attack (1.5)	-	2nd and 3rd attacks (2.0–3.5)
NOD/LtJ ± FBZ: Group C	3/13(23.1%)	-	-	1 st attack (1.5–3.0)
NOD/LtJ ± FBZ: Group D	1/13(7.7%)	-	-	-	1st attack (2.0)
NOD/LtJ ± FBZ: Group E	3/13(23%)	-	-	-	-
C57BL/6J Untreated	(100%)	Overt disease (2.0–2.5)	Overt disease (3.0–4.0)	Culled
C57BL/6J ± FBZ: Group F	5/8 (62.5%)	Overt disease (2.0–3.0)	Overt disease (2.0–3.0)	Culled
C57BL/6J ± FBZ: Group G	1/8(12.5%)	-	Overt disease (2.0)	Culled
C57BL/6J ± FBZ: Group H	2/8 (25%)	-	-	-	-

FBZ: fenbendazole; EAE: experimental autoimmune encephalomyelitis; NOD: non-obese diabetic

EAE was induced in female NOD/LtJ and C57BL/6J mice that had been switched from standard to FBZ-containing feed. The timing of the first attack, relative to that which is routinely observed (‘untreated’ mice), as well as the occurrence of subsequent attacks if any, was very variable and fell into five (NOD/ShiLtJ) or three (C57BL/6) patterns in the two mouse strains

Following discontinuation of treatment and a lag period of three weeks on normal chow, EAE was induced in a small group of NOD/ShiLtJ mice (n = 3), to determine whether the disease pattern was restored. These mice were confirmed to be free from pinworm, by the same methods as above. All mice showed clinical signs at the anticipated time, with the expected scores of 2.5–3 for the first attack and duration of three days, followed by the normal profile of relapses and remissions (not shown). The same batches of reagents were used in these test mice, as those injected while FBZ treatment was in progress. All conditions in the animal facility remained unchanged.

FBZ acts by inhibiting the polymerization of tubulin into microtubules thereby blocking mitosis, its anthelminth effect resulting from greater affinity for parasite relative to mammalian microtubules. ⁷ The mode of action of FBZ differs from that of other drugs affecting microtubule dynamics, such as paclitaxel, which enhance microtubule polymerization and interfere with normal microtubule formation during mitosis, migration, Chemotaxis and intercellular transport and have been demonstrated to ameliorate EAE.^8,9 The wide use of this drug, resulting from its efficacy as a prophylactic parasite treatment, has caused concern on potential effects on experimental designs. Studies relating to immunological parameters have shown apparently contradictory outcomes. In one, ¹⁰ prolonged FBZ exposure (23 weeks) of NOD/ShiLtJ mice, susceptible to spontaneous development of type 1 diabetes, had no effect on the onset, progression or incidence of the disease. Furthermore, various lymphocyte sub-populations, as well as the proliferative response of T-cells to concavalin A appeared unaffected. Another study ¹¹ using BALB/cN mice reported severe effects on the humoral immune system. mRNA and protein levels for the transcription factor E2A, which regulates expression of several B lineage genes as well as immunoglobulin rearrangements, were significantly reduced in bone marrow precursor B lymphocytes, resulting in decreased function of activated B lymphocytes. In a recent study, a direct effect of FBZ on T-cell proliferation and activation in vitro, together with reduced cytokine production in vitro and in vivo in an asthma model, was demonstrated. ¹² Such an effect would predict reduced inflammation following EAE induction as well. A review ⁶ of findings from toxicological, carcinogenic, immunological and metabolic investigations on FBZ suggests that the ability of B and T lymphocytes to proliferate in the secondary immune response may be suppressed with ongoing FBZ treatment, based on studies on sheep and human lymphocytes.

EAE is mediated by activated T-cells, but parasitic infections promote a T helper 2 (Th2)-type response, which would counterbalance the pathogenic Th1 cells responsible for inflammatory demyelination. It is unlikely, however, that the parasitic infection (or parasitic infection alone) was the basis for the altered profiles, for the following reasons. Firstly, the change in disease profile was clearly coincident with the introduction of medicated feed. By the time this was implemented infection of the whole rodent colony was likely, yet no alterations in clinical profiles was observed. Secondly, the infection was under control within four weeks, but aberrant profiles were observed for the full duration of FBZ treatment. Thirdly, in view of the reported studies in other species, it is likely that FBZ adversely affects selected immune responses in mice.

Variants of EAE clinical profiles result from specific antigen/mouse strain combinations.^2,3 Our data show that NOD/ShiLtJ mice were more severely affected in terms of disease severity and lag time before the first appearance of clinical disease, relative to C57BL/6J mice. The sum of our observations, therefore, suggests that FBZ treatment causes or contributes to an altered EAE profile, with different severity in different mouse strains. This may explain the inconsistent reports on the effects of this drug on the murine immune system and argues in favour of full evaluation of the effect of the drug on the disease course in specific experimental models prior to data collection.