Paralysis is associated with definite feeding phases

In experimentally-induced paralysis, secretion of neurotoxin coincides with a definite repletion phase and in hard ticks is limited to females only. For R. evertsi evertsi paralysis, toxicity is associated with a short period between day 4 and 5 of feeding and a tick body mass of 15–21 mg (Gothe & Lämmler, 1982; Neitz & Gothe, 1986). Paralysis induced by the tick I. holocyclus sets in after 4–5 days of feeding, while paralysis with D. variabilis is only detected after approximately 6–8 days after attachment (Gregson, 1973; Masina & Broady, 1999). In the soft tick A. walkerae, it is only larvae that cause paralysis, which occurs after 5–6 days of feeding (Gothe, 1984; Gothe & Neitz, 1991). In all instances, paralysis coincides with the rapid engorgement phase that is marked by the production and secretion of numerous protein products by the salivary glands. Paralysis is exhibited as an ascending flaccid tetraplegia due to an impaired functioning of the nervous system (Gothe & Neitz, 1991). While these are generally observed symptoms, most neurotoxins have specific characteristics not necessarily shared with toxins from other tick species.

Paralysis by A. walkerae

Electromyographical studies indicated that the fast conducting nerve fibres of the peripheral nervous system are affected and the paralysis can be classified as a motor polyneuropathy that does not affect the afferent paths (Gothe et al. 1971; Gothe & Kunze, 1971; Gothe & Neitz, 1991). The toxin seems to affect the liberation of acetylcholine as well as its receptor's sensitivity at the myoneural synapse (Gothe & Neitz, 1991). A monoclonal antibody directed against the toxin from R. evertsi evertsi also recognizes protein complexes of molecular mass 60–70 kDa from crude A. walkerae extracts and prevents paresis (partial paralysis) of day-old chicks (Crause et al. 1994). The toxic fraction was purified using a bioassay to detect toxic activity, based on injection of day-old chicks (Viljoen et al. 1990). The purified fraction showed two bands with molecular masses of 32 and 60 kDa using reducing SDS–PAGE while one band (pI~4·5) was obtained by iso-electric focusing. Macromolecular complexes (80–100 kDa) were observed using size exclusion chromatography (Viljoen et al. 1990). Recently, the monoclonal antibody directed against the toxin from R. evertsi evertsi was used in an attempt to purify the neurotoxin from A. walkerae extracts and whilst a 68 kDa toxin was detected using Western blot analysis, an 11 kDa protein was purified that showed cross-reactivity with the mAb using enzyme linked immunosorbent assay (ELISA), although not being detected during Western blot analysis (Maritz et al. 2000). These conflicting results between purification using detection methods of bio-assay and monoclonal antibody have not been resolved yet. Crude A. walkerae larval extracts inhibited potassium-stimulated and veratridine-evoked release of [³H] glycine from rat brain synaptosomes, suggesting that the toxin might be targeting ion channels involved in depolarization (Maritz et al. 2001).

Purification of sand tampan toxins

It was subsequently shown that salivary gland secretion (SGS) obtained from O. savignyi by pilocarpine stimulation has an LD50 of ~200 μl/kg when injected subcutaneously into 10 g albino mice (Howell, Neitz & Potgieter, 1975). Undiluted SGS (50 μl~5 salivary gland equivalents) kills a mouse within 8 minutes. The toxic activity is temperature stable for at least 15 minutes at 80 °C. An acidic toxin (pI~5) was purified from salivary gland secretion and shown to be a highly abundant protein (~9% of the total salivary gland secretion protein) (Neitz, Howell & Potgieter, 1969). It had a molecular mass of ~15400 Da and in this purified form could kill a mouse within 90 minutes if injected at a concentration of 400 μg/10 g mouse. A non-toxic component (Mr~16000 Da) that showed N-terminal amino acid sequence similarity to the toxin was also described (Neitz, 1976). Recently, the acidic toxin (tick salivary gland protein 2-TSGP2), a non-toxic homologue (TSGP3) and a basic toxin (TSGP4, pI~8) were purified from salivary gland extracts (SGE) (Mans et al. 2001, 2002). The toxins, the non-toxic homologue and another non-toxic protein (TSGP1), are the most highly abundant proteins in the SGE (TSGPs) and comprise individually ~4–5% of the total SGE protein (Mans et al. 2001). Based on this high abundance it was suggested that the TSGPs may be involved in tick salivary gland granule biogenesis. The N-terminal sequences and molecular masses obtained for the acidic toxin and its homologue corresponded very well with each other and the sequences obtained before (Fig. 2). It was also shown that 24 μg of the acidic toxin or 34 μg of the basic toxin in purified form are sufficient to kill a 20 g mouse within 30 minutes (Mans et al. 2002). The difference observed between the toxicities of the original and more recent toxin preparations could be ascribed to the fact that the toxins in purified form seem to be labile and that the current success in obtaining such active toxic fractions can be ascribed to modern high performance liquid chromatography technology that allows purification of the toxins within a few hours, compared to several days for the previous isolation attempts.

Fig. 2. N-terminal sequences obtained for the sand tampan toxins (TSGP2 and TSGP4) and a non-toxic homologue (TSGP3). The sequence for the non-toxic TSGP1 is also indicated, which shows sequence similarity to a highly abundant protein from O. moubata (20A1). The Toxin* and non-toxin* shown derive from sequences previously obtained (Neitz, 1976). Data were taken from Mans et al. (2001).

Clinical pathology of sand tampan toxicosis

In controlled experiments where animals (sheep, rats and mice) were injected subcutaneously with SGS, visible symptoms were minimal (Mans et al. 2002). This was also observed in an experiment where a 300 kg bull was confined to a camp that contained approximately 2000 tampans for two hours on four consecutive days. Most animals show symptoms of shock just prior to death. In large animals the histopathology indicates congestion and oedema in the myocardium, integument, spleen, kidneys, lungs and lymph glands. Haemorrhage within the lungs also occurs in smaller animals such as rats and mice. No other pathological changes can normally be observed and Howell suggested that animals die of heart failure (Mans et al. 2002). Addition of SGE to a rat heart perfusion system causes arrhythmia and bradycardia, followed by cardiac arrest. Monitoring of mouse electrocardiograms after subcutaneous injection of purified toxins showed that ventricular tachycardia (TSGP2) and a Mobitz-type ventricular block (TSGP4) are induced (Mans et al. 2002). This suggests that the pathogenicity of sand tampan toxicosis is due to a targeting of the hosts cardiac system and as such is distinct from tick paralysis.

The origins of sand tampan toxicosis

It was shown that the TSGPs belong to the lipocalin protein family (Mans, Louw & Neitz, 2003). In haematophagous organisms lipocalins perform a variety of functions that include the regulation of inflammation during feeding of the hard tick R. appendiculatus by sequestration of histamine (Paesen et al. 1999, 2000). Lipocalins from the blood-sucking bugs Rhodnius prolixus and Triatoma pallidipenis inhibit various haemostatic processes within the host. The toxic TSGP2 and non-toxic TSGP3 show high sequence identity (46%) with moubatin, a platelet aggregation inhibitor from the closely related non-toxic tick, O. moubata (Waxman & Connolly, 1993; Keller et al. 1993). No anti-platelet activity or anti-blood coagulation capabilities are associated with the TSGPs. Phylogenetic and western blot analysis suggests that the toxins and the non-toxic TSGP3 only evolved after the divergence of O. savignyi and O. moubata by gene duplication from existing lipocalins. This gives convincing evidence of a recent origin for specific tick toxins within a single genus and species. This has important implications as to the origins of all types of tick toxicoses and suggests that various forms of tick toxicoses originated at different times independently. Of interest is the fact that SGE from O. moubata also shows a high abundance of proteins with molecular masses that correspond with the lipocalins from O. savignyi. Their distribution in terms of mass and iso-electric point are, however, slightly different (Fig. 3). This indicates that O. moubata probably has its own set of lipocalin proteins that are non-toxic and which might be more closely related to the original toxin ancestral lipocalins.

Fig. 3. Proteomes obtained for salivary gland extracts from O. savignyi and O. moubata. In each case, the salivary gland extracts are those from 10 ticks. Indicated are TSGP1-4 from O. savignyi. At least three highly abundant proteins in the SGE from O. moubata have similar molecular masses and iso-electric points as TSGP1-3.

Ixodes holocyclus

I. holocyclus has been reported to paralyse dogs, cats, cattle, horses and humans. Adult ticks and nymphs have been associated with paralysis, while larvae cause local irritation only (Masina & Broady, 1999). SGE affects dogs and mice, although only suckling mice (4–5 g) are generally affected, while adult mice (20–25 g) do not show paralysis symptoms (Stone & Binnington, 1986). An increase in blood pressure occurs during paralysis, in contrast to paralysis caused by D. andersoni where blood pressure remains normal. Original purification attempts of the neurotoxin from I. holocyclus showed that the paralysis toxin was associated with high molecular mass complexes (40–80 kDa) (Masina & Broady, 1999). A toxic lethal fraction (Mr<20 kDa) without paralysing activity was also identified in I. holocyclus and it was suggested that this might be the agent of cardiovascular failure previously attributed to holocyclotoxin (Stone, Doube & Binnington, 1979). Recently, the paralysis toxin was purified and was shown to have a molecular mass of ~5 kDa that binds to rat synaptosomes in a temperature-dependent manner (Thurn, Gooley & Broady, 1992). This temperature dependence coincides with earlier observations that there is a temperature-dependent inhibition of evoked acetylcholine release during paralysis (Cooper & Spence, 1976). It was also shown that three different isoforms (HT-1, HT-2 and HT-3) of the toxin exist (Thurn, Gooley & Broady, 1992). The cloning of the holocyclotoxin gene for HT-1 was reported recently (Masina & Broady, 1999). The sequence includes a signal peptide (18 residues), start and stop codons, the polyadenylation signal and a polyA tail. The calculated molecular mass (5·9 kDa) corresponds to the mass obtained for the native toxin and a basic iso-electric point (pI~8·86) is predicted. The arrangement of cysteines suggests that the disulphide bond pattern of holocyclotoxin might be similar to scorpion toxins and could indicate structural similarity (Masina & Broady, 1999). Antibodies raised against the recombinant HT-1 protected neonatal mice against the native toxin thereby confirming the sequence identity (Masina & Broady, 1999). The reader is referred to the recent review for a more thorough account of the biology of I. holocyclus (Masina & Broady, 1999) and to Gothe (1999) for a historical overview of the paralysis induced by this tick species.

Paralysis caused by Dermacentor andersoni

Infestation by D. andersoni affects motor neurons of the efferent pathway but not the afferent. The neuromuscular junction of peripheral nerves is targeted through inhibition of acetylcholine release from the synapse, suggesting a pre-synaptic target (Gothe & Neitz, 1991). Feeding ticks affect dogs, sheep, cattle, guinea pigs, hamsters and man, but not cats, rabbits, rats and mice. In those animals affected the symptoms could be fairly rapidly reversed on removal of ticks, except for marmots that frequently do not recover (Emmons & McLennan, 1980). Gross symptoms of paralysis in marmots include the loss of the animal's normal piercing cry (paralysis of the vocal cords) followed by an ataxia and weakness of the hind limbs. The condition progressess until the fore limbs are paralysed and the animals are unable to move and lie on their sides; there is retention of urine and faeces (Emmons & McLennan, 1980). No paralysis could be observed when SGE from fast feeding ticks was injected subcutaneously into mice and lambs and intraspinally into puppies (Gregson, 1943). Fractionated extracts also failed to produce any symptoms. However, paralysis was observed when saliva from fed females was continuously injected subcutaneously into marmots and hamsters (Gregson, 1973). Variation in the ability of individual ticks to cause paralysis was also observed. This highlights the problems associated with paralysis by D. andersoni. It also points out why progress on this specific toxicosis has been so slow in recent years. No development of immunity has yet been reported for D. andersoni, while immunity to the holocyclotoxin from I. holocyclus is well established. Hyper-immune serum against holocyclotoxin also fails to relieve D. andersoni paralysis (Gregson, 1973). This indicates that no cross-reactivity exists between these toxins and suggests that D. andersoni and I. holocyclus-derived toxins are evolutionarily distant or not at all related.

Sweating sickness, Mhlosinga and Magudu

Sweating sickness occurs in Central, Eastern and Southern Africa and has been recorded in Sri Lanka and India. In nature it affects only cattle, especially young calves, but can also be transmitted to other artiodactyls such as sheep, goats or pigs (Neitz, 1959, 1962). The latter is particularly useful due to the ease of diagnosis of sweating sickness symptoms such as hyperaemia of the skin and the development of pharyngeal and laryngeal lesions that manifest as changes in the tone and pitch of squeals during handling (Neitz, 1956). H. truncatum has been implicated as being responsible for sweating sickness in the Kwa-Zulu Natal region of Southern Africa (Neitz, 1954, 1956). Sweating sickness positive (SS+) and sweating sickness negative (SS−) tick strains are found. Exposure and recovery from the disease leads to immunity. Two similar, but milder forms of toxicosis named Mhlosinga and Magudu are also induced by H. truncatum. It appears as if Magudu is more closely related to typical sweating sickness as animals recovered from sweating sickness were resistant to Magudu but not to Mhlosinga. In Kenya, sweating sickness and Hyalomma ticks go by the same name of ‘Ol masheri’. Other names used to describe sweating sickness include wet calf disease, vuursiekte (Afrikaans), schwitzkrankheit (German), la hydrose tropicale (French) and Foma (Swazi) (Neitz, 1959).

Sweating sickness symptoms and clinical pathology

The name sweating sickness derives from the profuse moist eczema that is the most noticeable symptom. Symptoms are normally observed after 5–7 days of tick feeding. Other symptoms include fever, pyrexia, anorexia, hyperaemia and hyperaestasia of the skin and mucous membranes, salivation, lachrymation, serous or crupous rhinitis, epitaxis, diarrhoea and diphtheroid stromatitis, pharyngitis, laryngitis, oesophagitis, vaginitis or posthitis. Subepicardial and subendocardial petechia, hyperaemia and oedema of the lungs, liver and kidney congestion, atrophy of the spleen, fatty degeneration of the liver and general inflammation and necrosis of the mucous membranes of the buccal cavity, larynx, pharynx, and oesophagus have been observed (Neitz, 1959). During the symptomatic period the fibrin content in the plasma increases but decreases again with recovery (Neitz, 1959). Addition of a few drops of citrated plasma of sick animals to Hayem's fluid (mercuric chloride, sodium chloride, sodium sulphate) causes haemaglutination. Magudu and Mhlosinga are characterized by pyrexia, anorexia and listlessness. The fibrin content of the plasma also increases over time and haemaglutination of citrated plasma is also observed as seen for sweating sickness (Neitz, 1962). The only difference between Magudu and Mhlosinga is that no immunological cross-reactivity can be observed (Neitz, 1962).

Sweating sickness as a tick-derived toxicosis

Tick-derived toxins were suspected as the causal agents since the disease could not be transferred from sick to healthy animals. Animals were, however, protected by hyperimmune sera. The disease could be transmitted over 15 generations of ticks (Neitz, 1962), after which toxicity was suddenly lost (W. O. Neitz, personal communication). Animals injected with salivary gland secretions from partially or fully engorged SS+ ticks, before challenge with SS+ ticks failed to protect against sweating sickness. Animals injected with partially fed or fully engorged SS+ tick suspension before challenge with SS+ ticks, were protected against sweating sickness, although the tick suspension itself did not induce any sweating sickness symptoms (Bezuidenhout & Malherbe, 1981). This suggests that the salivary glands are not the organs from which the toxins originate and that tick suspensions in themselves probably do not contain the active toxin or sufficient quantities of toxin to induce sweating sickness, although enough of the toxin must be present to allow development of a toxin neutralizing response. Toxicosis could be associated with several novel proteins found in the positive strain which are absent in the negative strain (Burger et al. 1991; Spickett et al. 1991). These include three non-immunogenic proteins with molecular masses of approximately 24, 26 and 42 kDa and three immunogenic proteins with molecular masses of 20, 30 and 32 kDa (Fig. 4). Immune responses against these unique immunogenic proteins were observed in cattle that were treated with hyperimmune sera at an advanced stage of sweating sickness (day 7) and rechallenged on day 27. The immunogenic bands became prominent at day 35 (Spickett et al. 1991). Treatment with hyperimmune sera protected against sweating sickness and the recovered animals showed no sweating sickness symptoms upon re-challenge.

Fig. 4. Novel proteins associated with sweating sickness positive strains of H. truncatum. A. Western blot analysis of salivary gland extracts (SGE) obtained from positive and negative strains of H. truncatum. Lane A is naïve serum on SS+ SGE. Lane B is naïve serum on SS- SGE. Lane C is SS+ anti-sera on SS- SGE. Lane D is SS- anti-sera on SS- SGE. Lane E is SS+ anti-sera on SS+ SGE. B. SDS–PAGE analysis of SS- and SS+ SGE. Indicated are unique protein bands observed in the SS+ SGE.

The presence of rickettsiae in sweating-sickness-inducing strains (SS+) and their absence in negative strains, cast doubt on a tick-derived origin for the toxins and suggested that toxicosis might be associated with this pathogen or symbiont (Bezuidenhout & Malherbe, 1981). As yet, the involvement of the novel proteins identified for sweating sickness positive strains or their association with rickettsial organisms identified in SS positive ticks, have not been ascertained.

Brown tick toxicosis

R. appendiculatus causes a leucocytotropic disease in cattle. Symptoms are prolonged fever, oedema of subcutaneous tissues of the ears, eyes, jowls and dewlap, a swelling of the palpable lymphatic glands, anorexia, lachrymation, serous nasal discharge, listlessness and general weakness (Neitz, 1962). This is normally followed by relapses of other tick-borne diseases such as babesiosis, spirochaetosis, anaplasmosis and heartwater. Recovered animals are immune against this toxicosis. Animals resistant to tick-borne diseases from R. appendiculatus-free areas that are introduced into areas where R. appendiculatus is prevalent, succumb to this disease known also as ‘Tzaneen disease’ (De Kock et al. 1937). Treatment with antibiotics known to counter heartwater and anaplasmosis failed to protect animals against this relapse (Thomas & Neitz, 1958). It was suggested that this form of toxicosis weakens the immune system of the affected host to such a degree that no protection against parasitic relapse is present.

Toxicosis induced by B. microplus

Cattle exposed to B. microplus show a weight-loss related to anaemia and loss of appetite (Gothe, 1999). It was considered that this may be due to a disruption in the metabolic processes of the hosts. The possible presence of egg-derived toxins secreted by larvae during feeding via the salivary glands may also have an influence on this phenomenon (see section on tick egg toxins).

Paralysis induced by R. evertsi evertsi

R. evertsi evertsi affects the peripheral nervous system by inducing a motor polyneuropathy in sheep and is appropriately known as spring lamb paralysis (Gothe & Kunze, 1982). Ticks fed on laboratory animals affect mice, rats, hamsters, guinea pigs and rabbits only slightly or not at all (Gothe & Lämmler, 1982). Injection of SGE into sheep, mice and chickens failed to elicit a paralysis response (Viljoen et al. 1986). A very sensitive in vitro assay using a sciatic nerve-gastrocnemius muscle preparation was developed to characterize this toxin (Viljoen et al. 1986). In this assay the dissected nerve was bathed with SGE or purified neurotoxin in a specially constructed nerve bath with a volume of 60 μl. In contrast to the in vivo tests, both SGE and neurotoxin preparations effectively paralysed muscle contraction (Viljoen et al. 1986; Crause et al. 1994). It could thus be argued that the neurotoxin affects the nerve and not the neuromuscular junction. Large quantities of SGE (400–900 μg protein) were used to elicit a response and it could be disputed whether this was a truly specific response. However, total inhibition of nerve impulse propagation was observed with purified preparations (74 μg/ml), which is probably much closer to physiological conditions (Viljoen et al. 1986). The toxin from R. evertsi evertsi that purified as a ~68 kDa protein (Viljoen et al. 1986) was later shown to be the trimeric form of a ~20 kDa protein (Crause et al. 1994). A monoclonal antibody directed against this toxin showed cross-reactivity with both non-paralysis (R. appendiculatus, Hyalomma marginatum rufipes, Boophilus decoloratus and a non-paralysing strain of R. evertsi evertsi) as well as paralysis-inducing ticks (I. rubicundus, A. walkerae and a paralysis-inducing strain of R. evertsi evertsi). Significant was the fact that only paralysis-inducing ticks seem to possess a ~68 kDa reactive antigen (Crause et al. 1994).

Clinical symptoms and histopathology of ixovotoxins

The pathological effects of ixovotoxins on guinea-pigs have been studied in various species (Riek, 1957; Neitz et al. 1981; Vermeulen et al. 1984; Vermeulen & Neitz, 1987; Viljoen et al. 1985). Mortality of mice and rabbits injected with ixovotoxins was also demonstrated (Riek, 1957). All toxins cause similar clinical symptoms within guinea-pigs and include hyperaestasia and anorexia, serous nasal and eye discharge accompanied by conjunctivitis and rhinitis, apparent paresis (that differ from ascending paralysis) and a loss of voice over a 15–36 hour period. Histopathology includes necrosis of the liver and kidneys and oedema of the urinary bladder, lungs and skin, at the site of injection. Necrosis is accompanied by elevated calcium levels within the cytoplasm that leads to mineralization within the cytoplasm and mitochondria. This suggests that the ion permeability of these cells is compromised and that lesions are probably all of vascular origin. Lipofuscin also accumulates within necrotic cells within the 36 hour period of induced toxicosis. This is of interest as lipofuscin is normally associated with ageing and an impaired lysosomal function, which is intimately associated with damaged mitochondria (Brunk & Terman, 2002). In the case of tick egg toxins, it probably indicates that disruption of selective membrane permeability to calcium leads to mitochondrial damage, which in turn places an oxidative stress on the system and impairs lysosomal function, thereby allowing rapid accumulation of lipofuscin.

Ixovotoxins are protease inhibitors

All egg toxins investigated so far function as serine protease inhibitors although no causal link has yet been established between protease inhibition and toxicosis (Vermeulen & Neitz, 1987; Vermeulen et al. 1988). Vertebrate proteases are involved in various important physiological functions such as digestion, maturation of hormones, immune responses, inflammation, blood coagulation, fibrinolysis and morphogenic responses (Holzer & Heinrich, 1980). It should therefore come as no surprise that inhibition of proteases within the vertebrate host will lead to pathological states. The concentrations of the tick egg toxins and protease inhibitors within tick eggs are more than a 1000 times higher than their respective K_i values with trypsin and chymotrypsin (Vermeulen & Neitz, 1987; Vermeulen et al. 1988). This indicates that they probably have a physiological function within the eggs. Several functions have been proposed which include involvement in egg development via the regulation of egg proteases during hatching, as anti-haemostatic components and as anti-microbial agents that protect the egg against microbial invasion (Willadsen & Riding, 1979, 1980; Neitz et al. 1981; Viljoen et al. 1985; Vermeulen & Neitz, 1987; Vermeulen et al. 1988). Toxins and anti-proteases within tick eggs could even protect them against microbial, insect and arthropod scavengers and predators.

Anti-protease kinetics of ixovotoxins

The ixovotoxin (Mr~10 kDa) from A. hebraeum shows specific non-competitive fast-binding inhibition of trypsin (K_i~255 nM), but not chymotrypsin (Neitz et al. 1981; Vermeulen & Neitz, 1987; Vermeulen et al. 1988). A non-toxic inhibitor (Mr~8400 Da) from A. hebraeum shows specific fast-binding competitive inhibition for trypsin (K_i~25 nM) (Vermeulen et al. 1984; Vermeulen & Neitz, 1987; Vermeulen et al. 1988). Ixovotoxin from R. evertsi evertsi (Mr~5–6 kDa) is a specific competitive fast tight-binding inhibitor of trypsin (Viljoen et al. 1985; Vermeulen & Neitz, 1987). The purified ixovotoxins from B. microplus (Mr~30–35 kDa), B. decoloratus (Mr~40 kDa) and H. truncatum (Mr~27 kDa) are all competitive slow binding inhibitors of trypsin. The toxins from B. decoloratus and H. truncatum also inhibit chymotrypsin competitively via a fast tight-binding mechanism, while B. microplus toxin does not inhibit chymotrypsin (Viljoen et al. 1985; Vermeulen & Neitz, 1987; Vermeulen et al. 1988). The toxins from B. microplus and B. decoloratus showed absolute identity during Ouchterlony-double-diffusion, while toxins from R. evertsi evertsi and H. truncatum showed no cross-reactivity with any of these toxins (Vermeulen et al. 1988). This is particularly interesting, considering the molecular mass and kinetic differences between B. microplus and B. decoloratus toxins. Immunization against egg extracts from H. bispinosa induced sensitivity that could also be observed with egg extracts from H. dromedarii and A. variegatum, while B. microplus, I. ricinus, I. holocyclus and I. hexagonus did not show as large a sensitivity reaction (Riek, 1958). Immunization did give cross-protection against egg extracts from B. decloratus, B. microplus, I. holocyclus, I. ricinus and H. dromedarii, but not against paralysis induced by I. holocyclus (Riek, 1957).

Ixovotoxins and protease inhibitors from B. microplus

The toxic component from the eggs of B. microplus was shown to occur in larvae, but not in nymphs and adults (Riek, 1957). Presence of egg toxins in larvae was also confirmed when immunization with larval extracts protected against toxicoses by egg extracts (Riek, 1958). Multiple toxic fractions that cross-reacted with larval proteins were subsequently identified in tick extracts (Riek, 1959). Toxic as well as non-toxic fractions from larvae induced hypersensitivity reactions (Riek, 1958). A double-headed protease inhibitor, that inhibits trypsin (K_i (<0·002 μM) and chymotrypsin (K_i~0·2 μM) was purified from larvae of B. microplus and caused an immediate hypersensitivity reaction when injected intradermally into bovines exposed to this tick species (Willadsen & Riding, 1979). It was subsequently shown that this inhibitor occurs in large amounts in tick eggs and in the initial stages of the larvae before disappearing in the later stages (Willadsen & Riding, 1980). The proteins from tick eggs and larvae, while closely related, differ significantly in terms of their kinetic properties (Willadsen & McKenna, 1983). Two different trypsin-chymotrypsin inhibitors (20800 and 15800 Da) were isolated from tick eggs (Willadsen & McKenna, 1983). A toxic component (Mr~30–35 kDa) was also purified from B. microplus eggs that specifically inhibit trypsin (K_i~4·6 nM) but not chymotrypsin (Viljoen et al. 1985; Vermeulen & Neitz, 1987). A larval inhibitor (22500 Da) was given functional significance during feeding as it was shown that this inhibitor can inhibit bovine plasmin (K_i~0·1 μM) and pig pancreatic kallikrein (K_i~0·33 μM), that it inhibited both extrinsic and intrinsic pathways of the coagulation cascade as well as complement-dependent cell-lysis but did not stimulate lymphocyte proliferation (Willadsen & Riding, 1980). The fact that this inhibitor disappears after the initial larval stage suggests that its biological roles are important within the egg and for the larvae during their initial feeding stages. Recently, a double-headed member (BMTI-A) of the basic pancreatic trypsin inhibitor (BPTI) family was described from B. microplus larvae (Tanaka et al. 1999). It inhibited trypsin (K_i~3 nM), chymotrypsin (K_i~33 nM), elastase (K_i~1·4 nM), plasmin (590 nM) and human plasma kallikrein (120 nM) as well as the intrinsic blood coagulation cascade. It did not, however, inhibit porcine pancreatic kallikrein. These studies indicate that tick proteins may have more than one function within the tick and the host. Immunization of cattle with BmTI-A led to a reduction in total tick number, egg weight and female engorged weight although the number of eggs was not affected (Andreotti et al. 2002). Another trypsin inhibitor (BmTI-B-10 kDa) was also purified from larvae (Tanaka et al. 1999).

While specific egg toxins have been isolated from different tick species, it appears that egg extracts are complex mixtures in which more than one toxic principle might reside. Cross-reactivity assays and similar biochemical data suggests that more than on protein family is involved in toxicosis and anti-proteolytic activity. The identification of these components could assist in the study of the origins of all forms of toxicosis, if considered that tick toxins might be evolutionarily related. The study of protein structure, function and evolution has shown that proteins can exist as single domain or multiple domain structures. As such, a protein fold involved in tick toxicoses can occur within a small protein or even as part of a larger protein. It can be concluded that tick egg toxins seem to be distinct from other forms of toxicosis and due to its source cannot be described as being venomous. They thus serve as evidence of toxic molecules derived from ticks that can influence potential hosts in deleterious ways, although toxicity may not be their main property.