The female accessory glands

Some commonly stated functions of female accessory sex glands in insects are to synthesize material for the egg capsules, to produce the lipid waterproofing layer, to lubricate the egg while it moves through the ovipositor, and to digest the spermatophore by means of enzymes secreted into the reproductive tract (Gillott, 1988). In ticks, the female possesses three such glands: (1) a tubular accessory gland (argasid and ixodid ticks) produces a protein-rich secretion, the function of which is unknown (Diehl, Aeschlimann & Obenchain, 1982; Sonenshine, 1991), (2) a lobular accessory gland (ixodid ticks only). Lees & Beament (1948) demonstrated that it provides the initial waterproofing layer for the egg. Sonenshine (1991) suggests it is also the most likely source of the genital sex pheromone, and (3) Gené's organ (found only in ticks) deposits wax on each egg as it is delivered from the vagina. The wax coat waterproofs the eggs, causes them to stick together, and probably protects them from soil fungi and microbes (Lees & Beament, 1948; Schöl et al. 2001).

The male accessory glands

Semen consists of spermatozoa bathed in a seminal fluid, the latter comprising secretions from the testis and secretions from the male accessory glands (MAGs). Commonly stated functions of MAG secretions in insects (reviewed by Gillott, 1988, 1996 and 2003) are: (1) to nourish/protect the spermatozoa within the male tract, (2) to secrete the spermatophore, (3) to alter the behaviour of the female in such ways as to (a) hasten oviposition and/or increase egg production and (b) attenuate mating behaviour in the female, (4) to secrete components of the mating plug, a coagulum which helps ensure P1 sperm precedence, (5) to secrete antibacterial and antifungal agents which protect the gametes within the female reproductive tract (Lung, Kuo & Wolfner, 2001b), (6) to provide a nutritional supplement to the female (Friedel & Gillott, 1977), although Vahed (1998) comments that it is usually difficult to determine the extent to which spermatophore contents contribute to the overall nutritional budget of the female, and finally, (7) to extend the duration of sperm viability within the female genital tract (Tram & Wolfner, 1999; Chapman et al. 2000), a function also suggested for ticks (Feldman-Muhsam, 1991).

Friedel & Gillott (1977) coined the terms ‘fecundity-enhancing substances’ (FES) and ‘receptivity-inhibiting substances’ (RIS) in reference to some MAG secretions. An FES is one that (directly or indirectly) increases egg production, oviposition or egg fertility. An RIS is one that renders the female less likely to copulate with further males, and thus constitutes the mechanism for remote mate-guarding. FESs and RISs are well-documented for numerous insects (Gillott, 1988, 2003), and in this article I shall discuss a few examples.

The blood-sucking bug, Rhodnius prolixus

In Rhodnius, transfer of a spermatophore influences fecundity in at least two ways: (1) by promoting migration of sperm to the spermathecae and (2) by stimulating ovulation and/or oviposition. Each of these effects is regulated by a distinct mechanism. (1) A spermatophore in the female's bursa copulatrix stimulates rhythmic contractions in the common oviduct, and within 5–10 min spermatozoa are found in the spermathecae. The MAG comprises three transparent lobes (which produce the spermatophore) and one opaque lobe. Davey (1958) demonstrated that a secretion from the opaque lobe stimulates peristalsis in the common oviduct, thus effecting migration of sperm to the spermathecae. This MAG secretion is not itself a myotropin, but exerts its peristaltic effect on the oviduct via the nervous system (Davey, 1958). (2) The presence of sperm within the spermathecae leads to a doubling of the rate of oviposition compared to that of fed virgins (Davey, 1965). Thus, if fed females are mated to males lacking either seminal vesicles (no sperm) or the opaque accessory gland lobes (sperm migration inhibited), the rate of oviposition equals that of normally fed virgins (Kuster & Davey, 1986). Likewise, the rate of oviposition in fed mated females surgically deprived of their spermathecae is similar to that of normally fed virgins. Finally, fed virgins receiving implants of spermathecae from mated females lay significantly more eggs than fed virgins receiving implants of virgin spermathecae. Taken together, these experiments indicate that the spermathecae secrete a factor that stimulates ovulation and/or oviposition, but only when they contain sperm.

Further experiments (Davey, 1967) demonstrated that the spermathecal factor has its accelerating effect on oviposition by stimulating medial neurosecretory cells (mNSCs) in the brain to release a myotropin. This myotropin augments contractions in the ovarian sheath and lateral oviducts, and hence ovulation and oviposition follow (Kriger & Davey, 1982, 1983). Moreover, the release of the myotropin from the mNSC depends on the presence of ecdysteroids from the ovary (Ruegg et al. 1981). Thus, the mNSCs require at least two inputs to release myotropin: a signal from mating (the spermathecal factor), and ovarian ecdysteroid. The myotropin of Rhodnius appears to be a peptide (~8·5 kDa) in the FMRFamide family (Sevala et al. 1992).

Mosquitoes

Insemination produces profound effects in the female mosquito, including host-seeking, oviposition, receptivity and circadian flight activity (Klowden & Chambers, 1991). The MAGs of Aedes aegypti and other mosquitoes produce a secretion, originally named ‘matrone’ because, when injected into virgins, it elicits numerous characteristics of mated females (Fuchs, Craig & Hiss, 1968). According to earlier literature, matrone was composed of α- and β-fractions. The α-fraction alone stimulated oviposition, but both together were required to inhibit receptivity (Hiss & Fuchs, 1972; see further below). More recently, Klowden (1999) suggested that the term ‘matrone’ is now obsolete, since MAG extracts having matrone-activity contain at least several distinct substances which may vary from one species to another. For example, heterologous MAG substances activate different subsets of behaviour in different species, and active material from different sources come in a wide range of molecular size.

A series of elegant experiments by Adlakha & Pillai (1975) demonstrated that MAG secretions are also necessary for normal fertilization to occur, perhaps by activating the sperm or by altering the egg surface so as to facilitate penetration of the sperm.

One important behavioural change attributed to MAG secretions in mosquitoes is a switch from swarming (=mating) to host-seeking flight behaviour (Gillott, 1988). The switch to host-seeking behaviour can be regarded ultimately as a fecundity-enhancing effect, because attaining a blood meal is required for producing a batch of eggs.

It has recently been shown in Anopheles gambiae that oviposition does not occur if the spermathecae of blood-fed females are removed either at emergence or after mating (Klowden, 2001). It is not known whether the spermathecal signal is neural or is a humoral factor, reminiscent of the system just described for Rhodnius (M. J. Klowden, personal communication). Whatever the nature of the spermathecal effect, it is not mimicked by implantation of a MAG (Klowden, 2001).

Drosophila

Most experimental studies on the MAGs of insects depend on delicate surgical techniques, and the implantation of tissues or injection of homogenates into the haemocoel (most often) or into the female genital tract (less often). Even though we have learned much from these studies, they obviously do not mimic the normal mating situation. In contrast, the relative ease of generating mutant strains of Drosophila that lack specific components of the MAG, or that do not produce sperm, enable an experimenter to study the effects of these genetic lesions under more natural conditions. Thus, it is not surprising that we know far more about the functions of MAG secretions in Drosophila than we do for any other insect. To date, no fewer than 83 gene products are predicted to be accessory gland proteins (Acps), all of them produced in the ‘main cells’ (MC; 96% of total) of the MAG (Wolfner, 2002).

Extracts of the MAG, when injected into the haemocoel of virgin Drosophila females, markedly increase the number of eggs laid (Leahy & Lowe, 1967; Chen & Bühler, 1970). Kalb, DiBenedetto & Wolfner (1993) produced strains of flies in which males lacked spermatozoa or in which the MCs of the MAG were destroyed by directed cell ablation. They observed the effects on oviposition after selective cross-matings, and compared these to normal mated controls (average of 108 eggs in 5 days) and virgin controls (average of 12·2 eggs in 5 days). When mated with males producing neither spermatozoa nor MC-secretions, females laid an average of 9·3 eggs in 5 days. When mated with males producing no spermatozoa (but having functional MCs), females laid an average of 40 eggs in 5 days. Thus, MC-secretions cause only a partial, short-term fecundity-enhancing effect.

Acp62F shows sequence similarities to a neurotoxin (PhTx2-6) produced by a Brazilian spider, Phoneutria nigriventer (Wolfner et al. 1997). PhTx2-6 prolongs action potentials and increases membrane excitability, suggesting that Acp62F may behave like the myotropin of Rhodnius and stimulate oviposition or regulation of sperm release from the seminal receptacle. Acp62F also has significant sequence homology to a class of serine protease inhibitors from the nematode, Ascaris (Lung et al. 2002).

Mosquitoes

Female Aedes aegypti and other mosquitoes mate once only, and virgin females injected with MAG homogenates remain forever virgin (Craig, 1967). Extracts of Aedes MAG have at least some activity across 12 species of mosquitoes from 3 genera (Craig, 1967). In fact, even the sex peptide of Drosophila (Chen et al. 1988) has biological activity in Aedes similar to that of α-matrone. Thus, injections into Aedes of extracts containing Drosophila sex peptide stimulates oviposition. In order to inhibit receptivity in Aedes, however, Drosophila sex peptide has to be injected together with β-matrone (Hiss & Fuchs, 1972). More recently, Klowden (2001) demonstrated that, in contrast to aedine and culicine mosquitoes, MAG substances from several anopheline mosquitoes do not control sexual receptivity as had been suggested earlier (e.g. Craig, 1967).

Drosophila

Whereas 95% of virgin females from a normal population will mate, only 3% of mated females will accept a male 1 day later (Kalb et al. 1993). When mated with males producing neither spermatozoa nor MC-secretions, 86% of the females mated again 1 day later. When mated with males producing MC-secretions but no spermatozoa, 43% mated again 1 day later. These partially inhibitory effects were only short-lived, however, because 90% of females mated with spermless males remated 2–7 days later. Thus, stored spermatozoa are necessary for a long-term receptivity-inhibiting effect in Drosophila.

The grasshopper, Gomphocerus rufus L.

The situation in Gomphocerus is unusual in that mating inhibits receptivity but does not increase the rate of egg production or oviposition (reviewed by Hartmann & Loher, 1999). A sexually immature female grasshopper will reject a male's attempt to copulate with her. This behaviour is called ‘primary defence’ (Hartmann & Loher, 1996). ‘Secondary defence’ is the rejection behaviour elicited by sexually mature females after they have mated (strong directed kicks at the male which make copulation impossible). Secondary defence endures until oviposition terminates, about 3–4 days following mating. Denervation of the spermathecae or severance of the ventral nerve cord immediately after mating abolishes secondary defense (Hartmann & Loher, 1996). Experiments by Hartmann & Loher (1996, 1999) have established that secondary defence is elicited by chemical and mechanical stimuli. The chemical stimulus comes from a specific MAG tubule (containing a white secretion), and the mechanical stimulus arises from stretch in the lateral oviducts as eggs move through them. Secondary defence in the grasshopper does not occur if the spermathecae are surgically removed prior to mating. This is because receptors for the white secretion (an identified group of sensory bristles) are found within the spermathecal bulb. In several insects discussed so far, on the other hand, the MAG secretion is effective even if injected into the haemocoel.

Mating plugs

In numerous insects, a mating plug constitutes a physical barrier deposited in the female genital tract by the male. The evolutionary advantage of reduced competition from other males is clear. The spermatophore itself can serve as an effective mating plug, provided it remains in place until the first male's sperm has fertilized the eggs. But as Gillott (2003) points out, the spermatophore is frequently ejected or digested in situ within a few days. In ticks, the (ecto)spermatophore is not even internalized within the female, and the endospermatophore, when evaginating into the female, forms a capsule around the prospermia which follow closely behind (Feldman-Muhsam, 1986). These sperm capsules can remain intact for extended periods, at least in O. savignyi, in which starved females still contained viable sperm within their capsules 1 year after mating (Feldman-Muhsam, 1986). However, it is doubtful that such sperm capsules are effective as mating plugs, because at least some ticks show P2 sperm preference (see below; Yuval & Spielman, 1990).

In many insects, the mating plug is a coagulum of the seminal fluid rather than the spermatophore. In Drosophila, a major component of the mating plug is a 38 kDa protein (PEB-me) from the ejaculatory bulb (Lung, Kuo & Wolfner, 2001a). The plug also contains at least one other Acp (Acp76A; see below).

Antimicrobial proteins

In numerous organisms, antibacterial and antifungal peptides constitute a major defence mechanism against pathogens. Male Drosophila transfer at least three antimicrobial peptides to the female during copulation, one of which, produced by the male ejaculatory duct, is the 28 kDa andropin. These peptides are believed to protect egg and sperm while in the female genital tract, for they are likely to be expelled during oviposition, several hours after mating (Lung et al. 2001b). Although ticks do produce antimicrobial peptides (e.g. van der Goes van Naters-Yasui et al. 2000; Johns, Sonenshine & Hynes, 1998, 2001), it is not known whether they are a component of the seminal fluid.

Protease inhibition and sperm storage

The seminal fluid of many organisms contains protease inhibitors, which affect male fertility (see Lung et al. 2002, for references). Wolfner et al. (1997) identified genes encoding a dozen Acps in Drosophila, eight of which begin with putative secretion signals, and therefore are likely to be included in the semen. Acp76A, a 388 amino acid pro-protein shows a sequence homology to the serpin class of protease inhibitors. Wolfner et al. (1997) hypothesized the following: Acp26Aa and Acp36DE may be activated in the female by means of regulated proteolytic action. It is conceivable that this proteolytic activation does not occur in the male genital tract because Acp76A inhibits proteolysis there. Acp76A itself may be inactivated following copulation, and hence female-derived proteases would be able to activate the other Acps. Acp76A may also play a role in forming the mating plug, because it can be recovered from plugs extruded by the female (Wolfner, 1997).

Finally, Acps are also necessary for sperm to be stored in the spermathecae. Tram & Wolfner (1999) demonstrated that male Drosophila, deficient in MAG secretions through directed cell ablation, produced a normal amount of sperm and transferred sperm successfully. Females mated to these males, however, stored only 3–10% the amount of sperm after 6 h compared to those mated to control males. The specific peptide mediating this effect is Acp36DE (Chapman et al. 2000). Sperm storage in Rhodnius is mediated by the spermathecal factor rather than a MAG secretion (Kuster & Davey, 1986). Ticks can also store sperm for extended periods: for a year or more in the case of O. savignyi (Feldman-Muhsam, 1986) and for at least several months in the case of Amblyomma americanum (Gladney & Drummond, 1971), although in neither case is the mechanism of long-term storage known.

With the foregoing as background, let us now focus more directly on the feeding and mating strategies of ticks, and consider how information from various insect models might point the way to future studies.

There are two major families of ticks: Argasidae and Ixodidae. A third family (Nuttallielidae) consists of a single species (Nuttalliella namaqua), about which very little is known, because only a handful of specimens have ever been found (in Namibia and Tanzania; Keirans et al. 1976; Sonenshine, 1991). The Argasidae comprise five genera, in five corresponding sub-families. The Ixodidae are divided into two phyletic lines: Prostriata and Metastriata, the former comprising a single genus (Ixodes), and the latter comprising 12 genera in 4 sub-families (Sonenshine, 1991; and see chapter by Barker & Murrell in this Supplement). The major physiological difference between the prostriate and metastriate ticks, at least for our purposes, is that the former can mate off the host before taking a blood meal, whereas the latter mate only while on the host, and only after they have fed for at least a few days.

Feeding in ticks

In all haematophagous arthropods, the taking of a blood meal constitutes the signal for profound developmental and physiological changes. The blood-sucking insects, Rhodnius and Aedes, feed on a host for only a matter of minutes. The argasid tick, O. moubata, feeds for about 30–60 min (Kaufman, Kaufman & Phillips, 1981, 1982). The latter species can lay a number of egg batches, each one requiring a blood-meal. Ixodid ticks, on the other hand, remain attached to a host for 5–15 days, the specific duration depending on species, the developmental stage and other factors. Mating causes distinct behavioural and physiological changes in ixodid females (reviewed by Kaufman & Lomas, 1996; Lomas & Kaufman, 1999), and I shall return to these further below.

The extended feeding period of female ixodid ticks is divided into three phases. (1) The ‘preparatory feeding phase’ occupies the first day or two, when the female anchors itself in place by means of a cement-like substance, and the feeding lesion is formed. (2) The ‘slow feeding phase’ occurs over the subsequent 4–8 days, when the female mates, and increases its weight approximately 10-fold; mating normally occurs during this period. (3) The ‘rapid feeding phase’ occupies only about 24 hours, during which the female increases its weight a further 10-fold and then detaches from the host.

The salivary glands (SGs) of Amblyomma hebraeum degenerate within 4 days following engorgement (Harris & Kaufman, 1981), a process controlled by 20E (Harris & Kaufman, 1985; Kaufman, 1991; Mao & Kaufman, 1999). Oocyte maturation begins during the feeding period, and yolk synthesis (also controlled by 20E) and yolk uptake are underway by the 4th day post-engorgement. Oviposition begins around day 10, and continues for about 3 weeks, after which the female dies (Friesen & Kaufman, 2002).

The transition between the slow and rapid phases of feeding occurs at about 10-fold the unfed weight (named the ‘critical weight’ (CW) by Harris & Kaufman (1984)). Females forcibly removed from the host below the CW do not lay eggs, do reattach to a host if given the opportunity, and their SGs do not degenerate. Females forcibly removed from the host above the CW do lay eggs (the size of egg mass being proportional to the extent of engorgement), do not reattach to a host if given the opportunity, and their SGs degenerate (Kaufman & Lomas, 1996). Moreover, there are slightly different CWs for reattachment to the host, SG degeneration and egg development (Weiss & Kaufman, 2001).

Changes in sensory physiology as a result of feeding are probably responsible for the cessation of host-seeking behaviour in ticks above the CW. Anderson, Scrimgeour & Kaufman (1998) demonstrated that the ability of CO₂ to induce questing and walking in A. hebraeum changes dramatically with age, and feeding state. In general, ticks responded to CO₂ during periods in which they would normally be attracted to a host (e.g. when partially fed but below the CW), but were refractory to CO₂ at other times (e.g. when partially fed but above the CW, or when fully engorged).

Sperm development

In the Metastriata, spermatogenesis (the production of haploid spermatids) begins in the nymph with the taking of a blood-meal, and attains only the primary spermatocyte stage at the time of ecdysis to the adult. When the adult feeds, spermatogenesis is completed and spermiogenesis follows. In ticks, the two phases of spermiogenesis are: (1) growth and elongation of spermatids to form prospermia (the form in which sperm are packaged in the spermatophore and conveyed to the female), and (2) capacitation (precedes migration to, and fertilization of, the ova); capacitation occurs only after the prospermia are transferred to the female. Capacitation is triggered by a MAG substance which is discussed below under ‘Mating factors in ticks’. After capacitation, the prospermium is called a ‘spermiophore’ (synonymous with ‘spermatozoon’). A detailed description of spermiogenesis and spermatozoan structure is presented by El Said et al. (1981).

In contrast to the Metastriata, sperm development in the Argasidae and Prostriata advances to the prospermia stage by the time of adult ecdysis, thus accounting for these ticks being able to copulate prior to feeding.

Sperm transfer

The following pertains generally to argasid and ixodid ticks. Once the male and female are in the copulatory position, the male pushes its mouthparts through the female gonopore into the vagina, and may maintain this position for 1–2 h. At least one function for this probing behaviour is to detect the genital sex pheromone (Sonenshine, 1991; see also chapter by Sonenshine in this Supplement). But perhaps this is not the only function because (1) the duration of the behaviour is so long and (2) probing occurs in both argasid and ixodid ticks, and only the latter are known to produce a genital sex pheromone. Whether or not the male can use its mouthparts to remove sperm deposited by a previous male has not been reported, although the frequent occurrence of several spermiophore masses (from multiple males) in the female genital tract (see below under ‘Inhibition of receptivity in ticks’) suggests that sperm removal does not occur. The spermatophore is then formed externally in less than a minute.

The ectospermatophore is a shell-like structure formed from a coagulum of mucopolysaccharides and proteins produced by the MAG (Feldman-Muhsam, 1986). Soon after the ectospermatophore is produced, the prospermia are ejaculated into it, followed immediately by a droplet containing yeast-like sperm symbionts (Adlerocystis spp.; Feldman-Muhsam, 1991). The last droplet secreted, the endospermatophore, seals off the ectospermatophore (Feldman-Muhsam, 1986).

Male ticks have no specialized copulatory organ, so their mouthparts push the tip of the spermatophore into the female's gonopore. While manipulating the spermatophore, the male also secretes saliva, which is believed to serve as a lubricant. Once the tip of the spermatophore has been inserted into the vagina, the endospermatophore evaginates into the female genital tract. There it forms capsules into which the prospermia and sperm symbionts are ultimately received. The generation of CO₂ within the ectospermatophore constitutes the force causing the evagination of the spermatophore contents into the female, although the mechanism of CO₂ production is not known (Feldman-Muhsam et al. 1973).

A ‘sperm capacitation factor’ from argasid and ixodid ticks

As mentioned earlier, spermiogenesis comprises two phases: formation of the prospermia, and capacitation (also called ‘spermateleosis’ by some authors), which takes place in the female. After copulation, an operculum at the end of each prospermium ruptures, and the prospermium evaginates to about double its original length (Sahli, Germond & Diehl, 1985) within ~60–90 min (Shepherd, Oliver & Hall, 1982). The latter authors studied capacitation of prospermia in hanging drop cultures. They exposed prospermia of Dermacentor variabilis and O. moubata for 24 h to extracts of MAGs from fed males and found a detectable capacitation response with as little as 0·01–0·03 MAG per hanging drop (2–4 μl). The capacitation factor in both species was heat-stable but destroyed by trypsin. Gel filtration revealed a single peak of activity at about 12·5 kDa for both species. In spite of the above similarities, the MAG extract of one species (tested at one concentration) was inactive on the prospermia of the other.

Because capacitation involves rupture of the operculum, Shepherd et al. (1982) hypothesized that the factor might be a proteolytic enzyme. However, treatment of prospermia of O. moubata with trypsin and pronase for several hours did not cause significantly more capacitation than observed in the controls. Moreover, these treated prospermia underwent normal capacitation when exposed to MAG extracts. They then exposed fresh prospermia to a MAG extract of O. moubata for 0–50 min and examined the cultures at 24 h. Percent elongation was then compared to that of similar cultures exposed to a MAG extract continuously for 24 h. As little as 2 min exposure to the MAG extract caused a significant degree of prospermial elongation (20%); elongation was 40% following 15 or 50 min exposure and was 97% after continuous exposure. Taken together, the results suggest that the capacitation factor acts more as a signalling molecule than as an enzyme. Only a brief exposure is necessary to trigger an effect which is not manifested until later. Capacitation of tick sperm is somewhat reminiscent of the acrosome reaction in sperm of vertebrates and marine invertebrates, which is triggered by Ca²⁺, and Shepherd et al. (1982) speculated that a similar mechanism may apply to tick sperm.

A ‘vitellogenesis-stimulating factor’ (VSF) from argasid ticks

When virgin O. moubata feed, they undergo egg development for about a month, and then begin resorbing the yolk, a process called ‘abortive vitellogenesis’ (Connat et al. 1986). Within about 100 days after feeding, the ovary again resembles that of an unfed virgin (Sahli et al. 1985; Connat et al. 1986). If ticks are mated in the absence of a blood-meal, oviposition does not occur. However, if a fed virgin is mated, even 4–5 months following the meal, eggs are produced and laid; however, the egg mass is only about half that laid by females which mate at the time of feeding (Sahli et al. 1985; Connat et al. 1986).

Small metal beads inserted into the genital tracts of ‘abortive virgins’ stimulated a resumption of oogenesis, and many of these females oviposited an egg-batch (only 30% smaller than normal) after a slightly prolonged pre-oviposition period (20–25 days; Connat et al. 1986). Thus, in O. moubata, mechanical stimulation of the female's lower reproductive tract appears to be sufficient for a significant degree of egg development and oviposition.

The following summarizes the evidence suggesting that a chemical substance from the male (called VSF) also contributes to egg development. Homogenates of spermiophores taken from the seminal vesicles of O. moubata induce oviposition when injected into the haemocoels of abortive virgins (Ducommun, 1984); but homogenates of testis or MAGs do not have this effect (Connat et al. 1986). Mature spermiophores, washed for 20 h in a physiological medium prior to injection, retain VSF-activity, indicating that the factor is not a component of the testicular fluid. The wash medium, however, also contained some VSF-activity.

Sahli et al. (1985) described the details of capacitation in O. moubata and proposed a potential secretory mechanism for VSF. Between 1–48 h after copulation, endospermatophores were dissected out of females, and the sperm masses were washed and incubated for 12 h in a physiological medium containing antibiotics. The sperm masses were centrifuged and the supernatants tested for VSF-activity in a bioassay. No activity was found in the supernatant when the spermiophores were collected 1 h post-copulation; up to this time, evagination of the prospermia was incomplete. VSF was present in the supernatant 12–48 h post-copulation (capacitation complete), with the highest activity at 12 h. Also, VSF-activity was destroyed by proteinase-K. Sahli et al. (1985) concluded that VSF is a protein synthesized by and contained within the prospermia while in the male, but liberated into the female genital tract during capacitation. The supernatants from the above experiments were subjected to electrophoresis; two proteins (~100 kDa and ~200 kDa) were particularly abundant in the 12 h supernatant. Sahli et al. (1985) proposed that these proteins constituted VSF.

A ‘male factor’ (MF) from ixodid ticks

The discovery of a protein (MF) from the male gonad of Amblyomma hebraeum, which exerts specific changes in female behaviour and physiology, arose from the following observations. Only a small minority of virgin females spontaneously feed beyond the CW, and none of these attain full engorgement even if left on a host for several weeks (Snow, 1969; Harris & Kaufman, 1984; Lomas & Kaufman, 1992a). SGs of mated or virgin females removed from the host below the CW do not degenerate within 14–21 days (Kaufman, 1983; Harris & Kaufman, 1984). SGs of mated females removed from the host above the CW degenerate within 4 days, whereas those of virgin females above the CW require 8 days to degenerate (Lomas & Kaufman, 1992a). Injecting an homogenate of male gonad (including or excluding the MAGs) into the haemocoel of virgin females above the CW, induces SG degeneration within 4 (reduced from 8) days (Lomas & Kaufman, 1992a).

MF acts by stimulating the synthesis of 20E once females exceed the CW. Thus, injecting a male gonad homogenate into virgin females above the CW increases the haemolymph titre of 20E to levels characteristic of weight-matched, normal mated females (Lomas & Kaufman, 1992b). MF is a protein (heat labile and inactivated by proteinase-K; Lomas & Kaufman, 1992a), in the range 20 to 100 kDa (Kaufman & Lomas, 1996). MF is a testicular protein, because the MAG has no MF-activity, and when partially purified homogenates of testis/vas deferens (T/VD) are layered on a discontinuous sucrose density gradient, MF-activity is not associated with the spermatozoa, which sediment to the bottom (Lomas & Kaufman, 1992a). MF-activity is found in abundance in the gonads of fed males, but is almost undetectable in the gonads of unfed males (Harris & Kaufman, 1984; Lomas & Kaufman, 1992a). MF-activity is detectable in the haemolymph of mated females above and below the CW, but not in the haemolymph of virgin females of any size (Harris & Kaufman, 1984). It is not known whether the MF-activity found in the haemolymph is due to MF itself or, analogous to the spermathecal factor of Rhodnius (Davey, 1967), to a substance secreted by the seminal receptacle after it receives a spermatophore.

An ‘engorgement factor’ (EF) from ixodid ticks

It has long been known that virgin females of ixodid ticks feed to only a fraction of the normal engorged weight range of mated females (e.g. Gregson, 1944; Sonenshine, 1967; Snow, 1969; Pappas & Oliver, 1971). Pappas and Oliver (1972) showed that female D. variabilis did not feed to repletion when exposed to males which had been irradiated or in which the genital aperture had been blocked. These experiments suggested that an EF, produced by the male, promotes feeding to repletion.

The observations of Pappas & Oliver (1972) have recently been extended to A. hebraeum (Weiss et al. 2002). We made a cDNA library from the T/VD portion of the male gonad and used a differential cross-screening approach to identify feeding-induced genes in this tissue. Thirty-five feeding-induced transcripts were identified, only two of which (AhT/VD16 and AhT/VD146) exhibited significant homologies to sequences in the Genbank. The predicted amino acid sequence of AhT/VD16 is 53% similar to human muscle-type acylphosphatase 2, and the predicted amino acid sequence of AhT/VD146 is 44% similar to a 9 kDa basic protein (of unknown function) from Drosophila.

We have recently produced recombinant proteins (recproteins) from those feeding-induced transcripts which had full open reading frames, and have identified recAhEF using a specific bioassay (Weiss & Kaufman, 2004); recAhEF is neither AhT/VD16 nor AhT/VD146. recAhEF comprises two peptides: recAhEFα (16·1 kD) and recAhEFβ (11·6 kD). Neither peptide has EF-activity on its own. Virgin females injected with recAhEF engorged normally, achieved ovary weights significantly beyond those of control virgins (though less than those of normal mated females), and their SGs degenerated within 4 days. Virgin females injected with any other of the recproteins did not feed beyond the CW, did not increase their ovary weights beyond those expected for small, partially-fed virgins, and their SGs did not degenerate within 4 days.

On the following grounds, we hypothesize that MF and EF are the same protein and that they differ from the two other known male factors in ticks. (1) Both are found in homogenates of T/VD, whereas VSF is believed to be secreted by the prospermia and capacitation factor by the MAG. (2) Both are virtually undetectable in the gonads of unfed males, but are detectable at high levels in the gonads of fed males. (3) The MW of MF (20–100 kDa) is different from that of the ixodid and argasid capacitation factors (12·5 kDa) and VSF (100–200 kDa). The MW of natural EF has not yet been determined, but the combined MW of recAhEFα and recAhEFβ is 27·7 kDa (Weiss & Kaufman, 2004). (4) As shown immediately above, recAhEF (but none of the other recproteins) also triggered SG degeneration and stimulated ovary growth – two known effects of MF. We propose the name ‘voraxin’ for the natural EFs/MFs of ixodid ticks (from the Latin, vorax, ‘gluttonous’, ‘voracious’).

Site of action of the male factors

The sperm capacitation factor acts directly on the prospermia. The fact that VSF is effective when injected into the haemocoel suggests that the target tissue is outside of the female genital tract. Whether VSF acts directly on the fat body (which synthesizes vitellogenin) or on the neuroendocrine system is not known. However, the synganglion contains at least several groups of NSCs, some of which appear to become active during egg development (Shanbaky, El-Said & Helmy, 1990). Also, homogenates of synganglion from mated fed females (but not from virgin fed or unfed) stimulate oogenesis, but not oviposition (Connat et al. 1986).

As MF in A. hebraeum stimulates ecdysteroid synthesis (Lomas & Kaufman, 1992b), and as ecdysteroid synthesis by the epidermis of A. hebraeum is controlled by a neuropeptide (Lomas, Turner & Rees, 1997), MF also may act via NSCs in the synganglion. Alternatively, MF might act directly on the epidermis to stimulate ecdysteroid synthesis. Finally, it is conceivable that the EF-effect of voraxin is mediated by NSCs and the MF-effect via the epidermis.

Inhibition of receptivity in ticks

The weight of current evidence indicates that inhibition of receptivity does not occur in mated female ticks. As early as 1948, Lees & Beament reported finding up to 12 sperm capsules in the genital tract of O. moubata females. When a number of virgin O. moubata were kept for 24 h with 5 males each, the females engaged in multiple copulations, in most cases with several partners, and O. tholozani females are said to be able to copulate even while ovipositing (Feldman-Muhsam, 1986). Most argasid ticks require a meal in order to produce each batch of eggs. They will readily copulate before or after each blood-meal, although only the initial copulation is necessary to produce several batches of fertile eggs (Aeschlimann & Grandjean, 1973; Connat et al. 1986).

Ixodid females produce only a single large batch of eggs and then die, so they probably require only a single copulation. Although the seminal receptacle may contain multiple sperm capsules (Feldman-Muhsam, 1986), it is not clear whether these come from the same or different males. The following study, however, demonstrates that ixodid females can also mate with multiple partners.

Sperm precedence in ticks

Yuval & Spielman (1990), taking advantage of the fact that prostriate ticks are able to mate prior to feeding (preprandial) as well as during feeding (perprandial), conducted a study on sperm precedence in Ixodes dammini (more commonly known as I. scapularis). A group of males were sterilized by exposure to cobalt-60 irradiation. When 22 such males were fed in the company of females, none of the laid eggs hatched. One group of virgin females was then kept (off the host) with irradiated males and another group with normal males for 2 weeks. Following this, the females were placed on a host, but this time in the company of the opposite group of males. Thus, females mated at least twice, each time with a different male. The results indicated that mated females remain receptive to subsequent males, and that sperm from a second mating takes precedence over that of the first. Thus, when the first male was irradiated and the second normal, more than 75% of the eggs of 78% of the ovipositing females hatched, whereas less than 25% of the eggs of the remaining females hatched. However, when the first male was normal and the second irradiated, less than 25% of the eggs of 86% of the ovipositing females hatched, and all of the eggs of the remaining females hatched.

Yuval & Spielman (1990) also showed that irradiated male I. scapularis can copulate normally. Twenty-five virgin females were placed on each ear of a rabbit with 20 normal plus 20 irradiated males. Of the egg batches laid by the engorged females, 49% showed no hatching, in 43% the hatch was greater than 75% and in the remaining 8% the hatch was 25–50%, results to be expected if all males were equally competitive. Irradiated males of Argas persicus likewise compete successfully for females (Sternberg, Peleg & Galun, 1973).

The results of Yuval & Spielman (1990) suggest that a male I. scapularis engaging in preprandial mating will be successful in passing his genes to the next generation only if the female, on finding a host, does not encounter another male. Otherwise, his paternity will pass to only a minority of her eggs. (By the same token, the female benefits from preprandial mating whether or not she encounters another male on the host.) Also, the viability of sperm in the female reproductive tract falls progressively over 8 weeks (Gladney & Drummond, 1971). The foregoing arguments perhaps explain Yuval & Spielman's (1990) observation that the incidence of insemination in female I. scapularis collected from vegetation in the wild, though significant (30–72% had copulated preprandially), was somewhat lower than that in females collected from deer (90–100% had copulated perprandially). These results also explain the male's tendency to remain in close contact with the female throughout the feeding period (an example of proximate mate-guarding): his physical proximity would preclude the advances of other males (Kiszewski, Matuschka & Spielman, 2001). Perhaps the practice of proximate mate-guarding by male ticks accounts for the lack of remote mate-guarding (i.e. an RIS). A similar situation occurs in the male desert locust which, during the gregarious phase, breeds under crowded conditions. The male remains on the female throughout the period of oviposition, for if the pair is separated before then, the female will accept another male immediately (Seidelmann & Ferenz, 2002).

Voraxin (EF/MF) is clearly an FES, because egg production is much reduced in females which exceed the CW but which do not engorge (Kaufman & Lomas, 1996). Voraxin and the VSF in argasid ticks stimulate vitellogenesis (Weiss & Kaufman, 2004; Connat et al. 1986), and following stimulation, some days are required for oviposition to occur. Most of the insect FESs described earlier stimulate ovulation and/or oviposition, and so their effects are often manifested within only a matter of hours. A notable exception was the effect of MAG extracts on mosquitoes, which also acts at the level of vitellogenesis (Borovsky, 1985; Klowden & Chambers, 1991). Does this suggest that the blood-sucking lifestyle may be a factor in determining the level (vitellogenesis vs. ovulation or oviposition) at which mating factors act? A few prominent exceptions suggest otherwise. We have already seen that in Rhodnius, the male's influence is not manifested at the level of vitellogenesis but at ovulation and/or oviposition (Davey, 1967; Kriger & Davey, 1982, 1983). Also, there are at least a few instances among non-haematophagous insects from several orders, including Drosophila, in which MAG secretions promote egg development, apparently by stimulating JH-synthesis by the corpora allata (reviewed by Gillott, 2003).