Territorial behaviour and immunity are mediated by juvenile hormone: the physiological basis of honest signalling?
*Correspondence author. E-mail: email@example.com
- 1The role of the juvenile hormone (JH) as a potential mediator in the trade-off between male–male competition and immune response has not been tested, but its study could reveal a potential mechanism that mediates resource allocation between these two traits.
- 2Controlling for body size, we tested whether males of the territorial damselfly Calopteryx virgo administrated with methoprene acid, an analog of the JH (JHa), compared to control males, increased their aggression and occupation time on territories but decreased their phenoloxidase (PO) activity (a key enzyme used during immune response after a bacterial challenge). We found an increase in aggression in JHa treated males compared to control males, but the opposite was found for PO activity.
- 3As fat load and muscle mass are also important traits during a contest, we tested whether JHa males compared to control males showed more fat and muscle content 2 h after JHa administration. Our results did not show a significant difference between both male groups, suggesting that JHa only increased aggression.
- 4These results and a review of other published articles, which have documented an effect of JH on a variety of functions in insects, suggest that JH may be a target of sexual selection: this hormone not only promotes the expression of secondary sexual characters but also seems condition-dependent and so its titers may indicate male condition.
Despite the increasing number of studies on sexual selection (reviewed by Andersson & Simmons 2006), we are still far from understanding the physiological mechanisms that underlie its processes (Lailvaux & Irschick 2006; Irschick et al. 2007). One missing gap is the supposed link between the production and/or maintenance of secondary sexual characters (SSCs) and immune response (Folstad & Karter 1992; Rolff & Siva-Jothy 2003; Schmid-Hempel 2003; Staszewski & Boulinier 2004). One hypothesis, which only applies to vertebrates, suggests that males convey information about their immune ability via the expression of their SSCs with the link being the testosterone production: the hormone promotes the expression of SSCs, but exerts a negative effect on immune response (Folstad & Karter 1992). The logic of this compromise is that only those males in better condition will be able to deal successfully with pathogens while still producing highly expressed SSCs. Unfortunately, not all studies in vertebrates have revealed a clear relationship between immune response and SSCs mainly because the physiological relationship between these traits is not as simple as originally assumed (Roberts, Buchanan & Evans 2004). Interestingly, in invertebrates, despite the fact that testosterone is absent (Chapman 1998), there is still a link between immune response and SSCs: animals in better condition are also good at producing both character types (e.g. Siva-Jothy 2000; Rantala et al. 2002, 2003; Rantala & Kortet 2004; Contreras-Garduño, Canales-Lazcano & Córdoba-Aguilar 2006; Pomfret & Knell 2006; Contreras-Garduño, Lanz-Mendoza & Córdoba-Aguilar 2007a, Kurtz 2007; but see Rantala, Roff & Rantala 2007; Vainikka et al. 2007).
An alternative hypothesis (and more general than that of testosterone), based on resource allocation theory, suggests that SSCs and immune response are energetically costly to produce and maintain and that given such costs, both attributes may compromise each other if the animal has not acquired enough resources (via the ingested food) (Sheldon & Verhulst 1996). Although this mechanism may sound contraintuitive, given the trade-off between SSC and immune response, it is based on the principle that only males with more resources may invest more in both the production of SSCs and immune ability (Sheldon & Verhulst 1996). According to this, not only testosterone but different candidate factors may also be linked to and conflict the SSCs and immune response. This would even explain the observed results of possible compromises in invertebrates whose endocrine system does not rely on testosterone. In calopterygid damselflies, for example, one such candidate is melanin which has been associated to the production of a SSC, a wing pigmented spot (Hooper, Tsubaki & Siva-Jothy 1999), and immune response via melanization/encapsulation against experimentally induced threats (Rantala et al. 2000; Siva-Jothy 2000; Koskimäki et al. 2004; Córdoba-Aguilar, Lesher-Treviño & Anderson 2007); or via the phenoloxidase (PO) cascade, an enzyme that initiates immune response via melanin formation (Cerenius & Söderhäll 2004), which is a good indicator of immune ability (Siva-Jothy 2000; Contreras-Garduño et al. 2007a, 2008). Interestingly, melanin is derived from phenylalanine (Riley 1997) and this amino acid is only gathered from food (Chapman 1998). Thus, only males that are better able to obtain phenylalanine will be able to devote more resources to pigmentation and immune defence against gregarine parasites via the PO cascade (Siva-Jothy 1999, 2000).
The above mechanism using wing pigmentation and immune response would fit well with Sheldon and Verhulst's (1996) hypothesis. However, pigmentation is just one type of SSC in insects and many other SSCs do not rely on melanin. A general principle that explains this example and male–male competition in invertebrates should be then put forward. One candidate factor underlying the production and/or maintenance of SSCs, immune response and condition in general in insects is the juvenile hormone (JH; Flatt, Tu & Tatar 2005). JH is a non-sex-specific hormone (Chapman 1998) that controls a plethora of functions which include physiological or developmental processes, behaviour and life-history, and fitness and fitness-related traits (Teal, Gomze-Simuta & Proveaux 2000; Flatt et al. 2005; Scott 2006b; Crook, Flatt & Smiseth 2008; Trumbo & Robinson 2008). Related to immunity, the expression of JH has also been shown to affect negatively PO activity and the genes related to antibacterial peptide expression (Rolff & Siva-Jothy 2002; Flatt et al. 2005). An interpretation of such relationships is that the JH may drive resource allocation by increasing the expression of SSCs at the expense of immune response and vice versa.
In this article we have tested this hypothesis by administrating a JH analog, methoprene acid (referred hereafter as JHa) to territorial insect males, expecting to increase male aggression during male–male competition for access to females at the expense of immunity. Experimental studies of JH in insects are not new and the interpretation of the positive effect of JH on aggression is not at all clear. For example, early studies as that of Barth et al. (1975) found that JH administration disrupted the social colony structure, via enhanced aggression, in wasp workers. In the honey bee, JH levels correlated positively with aggression within a colony but, despite inter-colony differences in JH levels, these did not correlate with aggression (Pearce, Huang & Breed 2001). Further studies investigating JH levels have actually found no negative effects on aggression (e.g. Brent et al. 2006). On the other hand, aggression has been widely admitted as a SSC in a variety of animals (e.g. birds, Wheatherhead 1990; fish, Bergman & Moore 2003; insects, Kemp & Wiklund 2001; spiders, Riechert & Johns 2003). In insects it has been suggested a link between male–male competition and immunity via melanine production (Koskimäki et al. 2004) although the nature of this link is unknown. However, as JH may promote aggression, it is possible that more aggressive males may bear a cost of an impaired immune response due to the JH titers. We have used Calopteryx virgo males to take advantage of their highly aggressive behaviour shown during territorial competition for access to females (reviewed by Córdoba-Aguilar & Cordero-Rivera 2005). Soon after emergence but before reproduction, calopterygid males feed extensively to build up fat reserves and muscular mass (Siva-Jothy & Plaistow 1999; Matsubara, Tojo & Suzuki 2005). Then, males search for a place to defend with aquatic vegetation substrates which is a key pre-requisite that females require to mate and lay eggs (e.g. Waage 1987; Meek & Herman 1991).
Using the rationale that the JH may mediate the expression of mating competition and immunity, we treated territorial males (i.e. individuals defending a territory) with JHa and recorded how long they spent in fighting and the number of days they remained in their territories, expecting higher values in these two variables compared to control territorial males. For males treated with JHa, we predicted a reduction in PO activity. To our knowledge, there is no information related to a positive or negative effect of JH on fat reserves (another key trait to win a contest in damselflies; Marden & Waage 1990; Plaistow & Siva-Jothy 1996; Marden & Cobb 2004) so that we compared the fat reserves between JHa and control males after JHa administration to know if more aggressive males behaved in that way due to JHa per se but not due to a positive effect of JHa on fat reserves and/or muscle content.
We observed a population of C. virgo near Pontevedra, NW Spain from June to July 2006. A total of 350 males were captured (using an aerial net) and individually marked (using an indelible ink pen) with a three-digit number on the right anterior wing to ascribe male identity. The length of the right anterior wing of all males was measured. Previous to experimental use but once males were marked, individuals had to be distinguished in terms of being territorial or non-territorial by doing the following. Three days after marking we recorded the number and behaviour of males from 10.00 to 15.00 h (the hours of maximum reproductive activity). For the 3 days and for 1 h (within the 10.00–15.00-h range), we observed males for which we had daily recordings and recorded the places where these males were observed. Behavioural data were essentially of aggressive nature and directed to those flying conspecific males that passed close to the focal males in which the latter chased the passing male. With these data of aggression and site fidelity, we were then able to ascribe territorial status: a territorial male was that animal that had remained in the same place after the aggressive encounters it incurred during the hour of observation. Males that did not show the territorial behaviour were classified as non-territorial (see also Córdoba-Aguilar 2000 for the use of a similar rationale). Only territorial males were used in the experiments. In addition, we controlled for possible energy and muscle differences in males, related to age, by only using middle-aged adults: those individuals that had hard wings, flexible at the distal tip, and whose body and wing colour was not pale (for a similar definition of middle-aged males see Plaistow & Siva-Jothy 1996).
jha preparation and experimental groups
We experimentally increased levels of JHa by using methoprene acid, an analog of JH III. Five milligram of methoprene acid (Sigma) were dissolved in 1000 µL of distilled water and a dilution of 1 : 1000 (one part of methoprene-water was diluted in one thousand parts of acetone) was used as JHa. Using a micropipette, 3 µL of the methoprene-acetone mixture (15 ng of JH/per organism) were provided to a set of 35 territorial males (hereafter referred as the ‘JHa increased’ group). The approximate mean mass of each adult is 69·4 mg (N = 3, range 67·8–68·5) for dry specimens (after 1 h at 30 °C) and 94·6 mg (N = 19, range 82·8–110·9) for alive specimens. The mixture was added on the cuticle between the head and the thorax (close to the corpora allata, the insect organ where JH is released; Nijhout 1994). As control groups, another set of 34 animals were treated with 3 µL of acetone only, added between the head and thorax. This control was used because acetone may damage the animals. We simultaneously observed one JHa and one control male that occupied territories located next to each other to avoid any effect due to a difference in the number of intruders entering in their territories.
As far as we know the JH titers in odonate haemolymph have not been reported. Due to this we used nanograms of JHa as this titer concentration has been reported in orthopterans, some dipterans, beetles and lepidopterans (Orth et al. 2003; Elekonich et al. 2003; Panaitof, Scott & Borst 2004; Min et al. 2004; Scott 2006a; Trumbo & Robinson 2008) and lower titers have been reported in smaller insects as dipterans, a coleopteran and isopterans (picograms; Brent & Vargo 2003; Cole, Eggleston & Hurd 2003; Liu et al. 2005). We did not use higher concentrations as they are unlikely to be similar to natural concentrations in insects (Contreras-Garduño & Lanz-Mendoza, unpublished review).
One hour after treatment, we recorded two variables which have been related to aggression in calopterygids: the time spent in fighting and the number of days that animals remained in their territories (e.g. Marden & Waage 1990; Marden & Rollins 1994; Plaistow & Siva-Jothy 1996). Observations were carried out for 3 days.
Nine JHa increased and eight control males of C. virgo were grouped and used to record PO activity. These males were territorial but independent of the sample used for aggression recordings. JHa and acetone were provided as indicated before. All animals were then immune challenged by using a 10 µL syringe (Hamilton; model 80330) to inoculate 2 µL of phosphate buffer saline (PBS, 140 Mm NaCl, 2·6 mM KCl, 1·5 Mm KH2PO4, pH 7·4) with approximately 10 000 µL−1 bacteria of Serratia marcescens inside the thorax of each damselfly. This bacterium is highly pathogenic in insects (Adamo 2004), and has been shown to activate the PO of another calopterygid species (Contreras-Garduño et al. 2007a). Furthermore, although not documented in C. virgo, this bacterium has been found in another calopterygid species, Hetaerina americana (A. Córdoba-Aguilar & H. Lanz-Mendoza, unpubl. data). We used the rationale of infecting the animal prior to PO assessment for two reasons. First, we were interested in how the JHa increased animals would actually be dealing with an infection and, second, as basal PO values are much lower compared to those after an immune challenge (Contreras-Garduño et al. 2007a), the former are less realistic. At all times when animals were not manipulated, they were individually placed in plastic, transparent containers (4·5 × 1·4 cm2) with a wooden piece as a perching place and a humid cotton. Such conditions would reduce animals’ activity, avoiding energetic exhaustion and mortality not related to our immune challenge (for a similar methodology see Contreras-Garduño et al. 2007a).
Two hours after treatment and bacterial challenge, the haemolymph was obtained by perfusion as follows. Five micro-litres of PBS were inoculated inside the thorax, and the heads were immediately separated from the thorax. Then this latter structure was pushed gently to obtain 2 µL of haemolymph/PBS per animal. We used the 2-h time point as after 30 min and following a bacterial challenge, animals substantially elevated their PO activity in a closely related damselfly (Contreras-Garduño et al. 2007a). The 2 µL were added to 100 µL of PBS and these 102 µL were used to record PO activity which was measured spectrophotometrically by recording the formation of dopachrome from l-dihydroxyphenylalanine (l-DOPA, Sigma). Twenty five micro-litres of sample with a concentration of 10 µg µL−1 of protein (see protein determination) were added to 150 µL of PBS and mixed on a micro-well plate with 25 µL of l-DOPA (3 mg mL−1 of PBS) as substrate (in total 200 µL of sample, PBS and substrate, were added to each micro-plate per animal). Optical density was recorded at 490 nm using a micro-plate reader (Model 350, Bio-Rad). As blanks, 175 µL of PBS was mixed with 25 µL of l-DOPA. The mean of PO activity at two different time points (20 and 90 min) was recorded (see Contreras-Garduño et al. 2007a; Ryder 2007). Enzyme activity was expressed as units, where one unit represents the change in absorbance per minute (Söderhäll & Hall 1984). Although it has been suggested that more than one parameter should be recorded in studies of ecological immunity (Adamo 2004; Contreras-Garduño et al. 2007a), previous research in another insect has shown a negative effect of JH on PO activity (Rantala, Vainikka & Kortet 2003) and that PO activity increases after a challenge with S. marcescens in damselflies (Contreras-Garduño et al. 2007a).
The Pierce method was used to determine protein concentration in haemolymph samples to control for PO differences between treatments. We followed the instructions indicated by the BCATM protein assay commercial kit (PIERCE, number 23225). In short, we added 10 µL of sample to 40 µL of PBS and 150 µL of pierce reagent. As a standard curve, we used a known concentration of albumin provided in the kit. Given that protein concentration may vary among individuals, we adjusted such concentration to10 µg µL−1 of protein (see Contreras-Garduño et al. 2007a for details). Such adjustment is necessary to ensure that the differences in PO activity between the two treatments are not due to differences in the amount of total protein but only due to PO activity (Contreras-Garduño et al. 2007a).
fat load and muscle mass
Another independent set of 15 JHa increased and 15 control males were used to measure fat and muscle content. Methoprene + acetone and acetone were provided as indicated above. Two hours after substance administration, males were preserved in ethanol (70%). Following the protocol of Plaistow and Siva-Jothy (1996), thoracic fat was extracted via chloroform immersion for 24 h, and the thoracic weight was recorded (in g) previous and after the chloroform extraction. The difference between both weight measures was interpreted as total thoracic fat (fat load). Again, following the method of Plaistow and Siva-Jothy (1996), muscle mass was measured by immersing the thorax in potassium hydroxide (0·8 M) for 48 h. The weight of this body region was measured (in g) previous and after the extraction being the difference interpreted as total muscle mass.
Whenever it was required, data were transformed using the formula √(datum + 0·5) to reach the criteria of normality. We compared aggression between groups. Wing size was also compared to see whether it explains differences between groups as this is the case at least in another calopterygid, H. americana in which larger males are more successful during territorial fights (Serrano-Meneses et al. 2007). Similarly, PO activity and protein, and fat load and muscle content were compared between experimental and control males by using two GLM.
Results are given as mean ± SE unless stated otherwise. Aggression time is provided in min for time spent in fighting and in days for the number of days defending a territory. PO values are reported as U/µg of protein, while protein concentration and fat are provided in grams and milligrams respectively.
Administration of JHa significantly increased the contest-related flying behaviour (JHa increased males: 1898·32 ± 338·06 sec, n = 19; control males: 680·88 ± 183·89 sec, n = 16; t-test = 3·16, P = 0·003). In addition, JHa increased males remained more days in their territories (1·80 ± 0·25 days, n = 25) compared to control males (1·06 ± 0·06 days, n = 16; t-test = 2·77, P = 0·009).
After treatment, more JHa increased males returned to their territories (25 out of 34), compared to control males (16 out of 35; χ2 = 5·5, P = 0·02). This difference was not due to a correlated effect of size as wing length did not differ between JHa increased (28·13 ± 0·14 mm, n = 30) and control males (28·10 ± 0·13 mm, n = 31, t-test = 0·18, P = 0·85).
PO activity was affected by JHa (F = 4·76, d.f. = 1,26, P = 0·03). The administration of JHa negatively affected the activity of PO in experimental males (0·28 ± 0·09 U µg−1 of protein, n = 9) compared to control males (0·85 ± 0·19 U µg−1 of protein, n = 8, Fisher LSD P = 0·0001), however, no differences were found in protein content between JHa increased (0·94 ± 0·03 g, n = 9) and control males (0·88 ± 0·02 g, n = 8, Fisher LSD P = 0·7).
Fuel content did not differ between JHa increased and control males (F = 0·60, d.f. = 1,56, P = 0·46) in both fat load (JHa increased: 0·62 ± 0·1 mg, n = 15; control: 0·54 ± 0·07 mg, n = 15, Fisher LSD P = 0·7) and muscle mass (JHa increased: 7·4 ± 0·1 mg, n = 15; control: 7·3 ± 0·1 mg, n = 15; Fisher LSD P = 0·14).
Calopteryx virgo males treated with a JHa, showed higher aggression at the cost of decreased PO activity compared to non-treated animals. Defending a territory via enhanced aggression conveys high mating benefits for territorial males as they gain a disproportionately higher mating success compared to non-territorial males as has been shown in C. splendens xanthostoma (e.g. Plaistow & Siva-Jothy 1996). Previous results in C. virgo have documented that winners of territorial contests had a larger immune response in the form of encapsulation/melanization to a nylon filament and more fat reserves (Koskimäki et al. 2004). These results were interpreted as only males in good condition being able to deal with the implicit energetic costs that aggression and immune defence entail. In our study, despite the small sample size but according to other studies, the JHa treated males had their PO activity expression reduced (Rolff & Siva-Jothy 2002; Rantala et al. 2003) but their aggression enhanced. We interpret our results and those of Koskimäki et al. (2004) as males in good condition indeed having a more robust immune response but if they have to allocate more resources to aggression this can impair immune function. This would be similar to when territorial males lose their territory and their immune ability gets reduced to values that are lower than those of non-territorial males (e.g. Contreras-Garduño et al. 2006). Our results also indicate that males that allocate more resources to aggression will end up being more susceptible to be invaded by pathogens. Males possibly may be balancing how much they invest not only in the functions we investigated here but also in those other functions that the JH has been found to control (Flatt et al. 2005).
Evidence from different damselfly species suggests that more wing pigmented males are more successful during territorial competition (Córdoba-Aguilar 2002; Koskimäki et al. 2004; Contreras-Garduño et al. 2006, 2007a,b) and female choice (Siva-Jothy 2000). This has been used to suggest that melanin, the basic pigment for those species where black or black-like wing pigmentation is present, may function to communicate male condition to conspecifics (Siva-Jothy 1999, 2000). Interestingly this communication principle also applies to species where red (e.g. Hetaerina americana; Contreras-Garduño et al. 2006, 2007a,b, 2008; Serrano-Meneses et al. 2007) rather than black pigmentation, is present. Thus, such melanin rationale is unlikely to be a general mechanism so that the principle may operate at a higher scale. Supporting previous claims by other authors (e.g. Rolff & Siva-Jothy 2002; Rantala et al. 2003; Flatt et al. 2005), our results suggest that the JH is a link that mediates resource allocation and, therefore, trade-offs in insects. This claim is further supported by the evidence that JH favours a wide range of key traits also shaped by sexual selection such as mating behaviour and activity (Teal et al. 2000; Rolff & Siva-Jothy 2002), the development of the stalk eyes (another SSC) in flies (Fry 2006) and sclerotized structures such as the horns used during male–male contests in beetles (Emlen & Nijhout 1999) and pheromone production (Rantala et al. 2002, 2003). Melanin may be just one resource under control of JH but several other components may be also controlled by JH. For example, JH apparently compromises pathogen defence when mating activity is increased (Rolff & Siva-Jothy 2002).
Our results suggest that JH does not have an effect on fat reserves and muscle mass after 2 h of JH administration. A long-time effect (i.e. more than 2 h) of JHa on fat reserves and muscle fat is not discarded and, in fact, an ongoing study is addressing this question. However, our findings suggest that the increase of aggression due to JHa administration was not due to a positive effect of JHa on fuel content. Related results in H. americana have detected a trade-off between fat content and melanization-based immune response after an extremely long-lasting territorial contest (Contreras-Garduño et al. 2006). Given that PO activity is necessary for melanin formation (e.g. Cerenius & Söderhäll 2004), the expression of JH during contests may be the proximate mechanism to explain the reduction in melanization activity.
Interestingly, JH production itself may be compromised by the amount of food consumed (Trumbo & Robinson 2004; Noriega 2004; Hernández-Martínez et al. 2007). In burying beetles, females supplemented with high quality food, produced higher JH titers which ultimately led to higher egg fecundity and viability (Trumbo & Robinson 2004). Food intake may provide the means to have a proper balance of the different functions (such as SSC production and immunity) that JH controls. Probably only males in good condition can carry out such a good balance as their JH titer levels are equally good which would not be the case for a male in bad condition. As our results and those in other insects suggest, it may be that JH itself is sexually selected (Emlen & Nijhout 1999; Rolff & Siva-Jothy 2002; Rantala et al. 2002, 2003; Fry 2006). Furthermore, recent studies have found that changing JH titers can be reflected in cuticular hydrocarbons in other insects (e.g. Lengyel, Westerlund & Kaib 2007). Cuticular hydrocarbons have been implicated as a communication mean during intra- and inter-specific recognition (e.g. Ferveur 2005; Hefetz 2007). It may well be that conspecific males communicate their aggressiveness and/or determination to fight via the cuticular hydrocarbons in damselflies.
Scott (2006a) suggested that JH in insects may be an analog of testosterone in vertebrates during the increase of aggression (this analogy could be established in function but not in structure as JH and testosterone likely evolved independently due to similar selection pressures; reviewed in Crook et al. 2008), but another study in beetles fails to support this idea as JH did not affect the outcome of aggressive events (Trumbo 2007). Trumbo (2007) suggested that the role of JH may or may not increase during aggression and that these studies may be performed in more species as JH may have a minimal effect on dominance and competitive ability. Our data and that on bees (Pearce et al. 2001 and references therein) still support Scott's suggestion about the positive effect of JH on aggression and specifically during male–male competition.
To Michelle Scott and Markus Rantala for their invaluable comments. One referee provided the idea of cuticular hydrocarbons being used to communicate aggression. The Doctorado en Ciencias Biomédicas provided grant to J.C.-G. for the field work in Spain. J.C.-G. is in debt with the following people: Mónica Azpilicueta Amorín for her hospitality and help in the field and laboratory; Inés González de Castro, Rosa Ana Sánchez Guillén, Guillermo Velo Antón and Silvia Sánchez Guillén for their help and company in Spain; Alberto Velando for his logistic help for recording PO activity; Salvador Flores for his help to establish the doses of methoprene; and Benito Álvares, Citlalli Gomez, Santos and Mariluz Escarabajal, and Carlos López (members of the government of Tenango del Valle, México) for their logistic support. Raúl Iván Martínez Becerril assisted during manuscript preparation. A.C.-A. was financially supported by two PAPIIT grants (Projects No. IN211506 and IN216808-2; Universidad Nacional Autónoma de México).