Baer, Armitage & Boomsma: Cost of sperm storage1
Supplementary information
Measuring insect immune defences
The encapsulation response is an innate immune reaction of invertebrates that occurs against a wide range of pathogens and parasites 1. It typically results in the formation of multiple layers of dead melanised haemocytes 2, which isolate the parasite from the haemocoel and kill it by asphyxiation or cytotoxic compound production 3. Melanin deposition during the encapsulation response is most commonly initiated by the haemocytes, but it can be activated by melanogenic enzymes (phenoloxidases) circulating in the plasma 4. Melanin is a black pigment that is synthesised from the amino acid tyrosine, a reaction that is in part catalysed by the enzyme phenoloxidase (PO). It has been shown that this enzyme is essential in the melanisation of parasites 5 and it is the reactive intermediates of oxygen and nitrogen produced during melanogenesis, that have been implicated in parasite death 4. In the mosquito Anopheles gambiae, Gorman et al. 6 demonstrated that there is a shared genetic mechanism for melanotic encapsulation of artificial implants and a real pathogen, and previous studies have used the degree of darkening (melanisation) of artificial implants as an assay of immunity e.g. 7,8,9,10. Introduction of an artificial nylon implant in Atta colombica queens results in a darkening of the implant, consistent with melanin production and possibly also involving haemocytes. In our 2004 samples, we measured the degree of darkening of a nylon implant as our assay of immune function, but in subsequent 2005 samples we also made haemocyte counts (see below).
Haemocytes are not only responsible for immune responses such as phagocytosis, nodule formation and encapsulation, but also recognise foreign bodies in the haemocoel and are intimately involved in wound healing 11. The total number of circulating haemocytes has been found to be positively correlated with the ability to encapsulate a parasitoid across six species of Drosophila12,13, and to be genetically correlated with encapsulation volume around a nylon implant 14. The number of haemocytes in the the haemocoel has been frequently measured in the immune literature e.g. 15,16,17 and is likely to reflect the capability of the immune system to deal with pathogens, both in terms of the encapsulation response and because of haemocytic involvement in other immunological processes. We may therefore expect that if the haemocytes are utilised in the melanotic encapsulation response towards a nylon implant that an increase in the encapsulation response (darker implant) would be correlated with an increase in haemocyte numbers. Even if the haemocytes are not involved in the encapsulation response, finding an increase in both encapsulation and haemocyte numbers is likely to indicate a generally up-regulated immune system. We therefore counted the number of haemocytes in a specific volume of haemolymph.
The encapsulation response and haemocyte number may also be influenced by other physiological processes, for example mating can result in increased juvenile hormone titers in insects 18,19,20. Juvenile hormone in turn has been found to have negative effects upon the activity of PO 21, and because this enzyme is in part responsible for the production of melanin in the encapsulation response we may see reduction in encapsulation straight after mating. Furthermore, the developmental stage of the insect can also cause variation in haemocyte numbers: for example, freshly emerged adult scorpion flies (Panorpa vulgaris) have lower haemocyte counts when compared to larvae 22.
We therefore extended our data set in 2005 to test not only the extent to which our immune assays could be reproduced across years, but also to see whether our encapsulation findings would be reflected in the second immune correlate: haemocyte numbers. To this end, on the 21st of May we marked another set of around 300 entrances of queen-burrows in four locations, all within 400 meters of one another and with one location being the same as where the 2004 queens had been collected. Furthermore, we sampled 36 winged virgin queens from four colonies in the direct vicinity (on the 5th, 6th, 7th and 18th of May, respectively) to estimate whether the encapsulation responses and haemocyte number before and after a mating flight would be different. We expected that the encapsulation responses before the mating flight would either be lower than those one day after mating (in the case that up-regulation of the immune response triggered by solitary founding would already be expressed as soon as queens no longer had grooming workers to protect them), or the same (in the case that up-regulation of the immune response would take several days to become effective). Similar to the previous year, one and nine days after the mating flight we excavated 33 and 32 queens, respectively, from a random subset of the marked burrows and placed them in individual petri dishes with moist cotton wool. For each queen we made two immune measurements: encapsulation response and a haemocyte count (per 0.01 µl haemolymph). Whilst digging we found that mortality of founding queens was lower than in 2004, but it followed a similar pattern in both years (Figure S1).
Figure S1. Mortality of founding queens of A. colombica in Gamboa, Panama one and nine days after the mating flight for the years 2004 (black bars) and 2005 (open bars).
For the haemocyte counts, each queen was chilled individually on ice and after she had stopped moving (around 20 minutes later) she was put in the queen holder of the adjusted machinery normally used for artificially inseminating bumblebees 23. A small hole was made with a sterile syringe needle in the intersegmental membrane between the 6th and 7th sternite. Through this hole 0.1 µl haemolymph was allowed to collect in a calibrated glass capillary which had been pulled to a fine point. The haemolymph was immediately added to 20 µl Ringer solution on a piece of parafilm and mixed with the Ringer by sucking it up and down the capillary 10 times. For each queen, 2 µl of this solution was added to a slide which had previously been coated with Poly-D-Lysine. The slide was kept in a humid chamber overnight and then the droplet was allowed to dry out. Up to two months later, the drops were rehydrated for 3-4 hours in 2 µl Ringer. Then 2 µl of DAPI solution was added to each drop and the slide was left in a lightproof humid chamber overnight. The following day the excess DAPI was removed by washing each drop six times with 2 µl Ringer solution. The haemocytes were visualised and counted under a fluorescence microscope at 400x magnification, to give the number of haemocytes in 0.01 l haemolymph.
Just after the haemolymph was removed, a piece of nylon was inserted into the same hole to measure the encapsulation response, as in 2004. The queen was returned to her petri dish and kept at ambient Panamanian temperature for 24 hours and then frozen at –20°C, after which time the nylon was removed and photographed at 62.5x magnification as described in the main paper. A digital camera (Canon EOS 350D) connected to a Leica stereomicroscope was used. In both years, we took special precautions to ensure that illumination conditions around the microscope and camera were the same, by placing the equipment on the same light box and dimming ambient light.
Figure S2. Encapsulation response and haemocyte numbers (means ± SE) for queens of A. colombica sampled in 2005 before the mating flight, and one and nine days after the mating flight. For both immune measures, mean values before the mating flight and one day after the mating flight did not significantly differ, but values for one day and nine days after the mating flight did significantly differ (*p = 0.022, **p = 0.005).
The results (Fig. S2) show that the encapsulation response and the haemocyte number follow similar patterns across the two sampling times (three sampling dates). As some of the analyses involved more than one measurement taken from the same queen (i.e. encapsulation response and haemocyte number), we used a multivariate analysis of covariance (MANCOVA) where appropriate. First, when comparing queens one day and nine days after mating, there was no significant effect of weight (MANCOVA: Wilks’ Lambda = 0.952, F2,58 = 1.478, p = 0.237) or place of origin (the four digging locations: MANCOVA: Wilks’ Lambda = 0.917, F6,116 = 0.855, p = 0.530) upon the combined dependent variables. These predictor variables were thus removed from the analysis. In the remaining model, there was a significant effect of day of digging upon the combined dependent variables encapsulation response and haemocyte number (MANCOVA: Wilks’ Lambda = 0.824, F2,62 = 6.618, p = 0.002). Subsequent univariate ANOVAs revealed that there was a significant increase in both encapsulation response (F1,63 = 5.526, p = 0.022) and haemocyte number (F1,63 = 8.383, p = 0.005) after nine days (Fig. S2). Second, when comparing queens from before the mating flight and one day after mating there was no significant effect of weight (MANCOVA: Wilks’ Lambda = 0.934, F2,65 = 2.31, p = 0.107) or time of digging (MANCOVA: Wilks’ Lambda = 0.951, F2,65 = 1.679, p = 0.195) upon the combined dependent variables (Fig. S2).
Overall, the evidence is quite compelling that the immune responses measured are up-regulated after mating and that it takes a number of days for this to happen. We suggest that the build up of these additional defences most likely happens in response to founding queens being exposed to soil microorganisms. This scenario would be consistent with evidence collected for the mealworm beetle Tenebrio molitor. Moret & Siva-Jothy 24 suggested that if an individual experiences a pathogen it may be indicative of an increased risk of encountering pathogens in the future and that this may help to explain why insects can produce immune responses that last for a long time, the so-called ‘responsive-mode prophylaxis’. They found that T. molitor that were pre-challenged with lipopolysaccharide prior to exposure to a fungal pathogen had higher survival, compared to larvae that did not receive a pre-challenge. Thus we may expect that if A. colombica queens have been exposed to pathogens in the soil, they may have up-regulated their immune responses in such a prophylactic manner, and we suggest that it is this kind of increase we have measured after nine days. This is consistent with a recent gene expression study in the fire ant Solenopsis invicta, where it was shown that two precursors of antibacterial peptides were up regulated 24 hours after queens were inseminated 25.
The cost of mounting an innate immune response
One of the central ideas behind the study of the ecology of immune systems is that their useage results in a cost for the host 26. In the most simplistic scenario, immune costs can be classified into two categories, the first being the physiological costs of maintaining and using the immune system, and the second the costs of having evolved immunity in the first place 26. Life history theory predicts that when two traits compete for allocation of materials and/or energy within a single organism a trade-off will occur 27. Thus, if an immune response is costly in some way it will compete with other energetically demanding processes that are occurring within the organism. We therefore hypothesized that storing sperm and utilising the immune system are both costly in A. colombica queens,and thus trade-off against one another for resources. Evidence for the cost of using the immune system in invertebrates has been accumulating by manipulation studies of specific aspects of the host biology and measuring its effect upon immunity, predicting that individuals undertaking a costly activity will show a reduced immune response e.g. 28, 29. Another fruitful approach has been to challenge the immune system and to measure corresponding changes in other traits, predicting that if producing an immune response is costly, an immune challenge will negatively affect the other measured trait(s) e.g. 30, 31.
Estimating the number of males that contributed sperm
We obtained 2-5 haplotypes from our spermatheca PCR’s, which matches the range of number of fathers reported for A. colombica in Gamboa in a previous study 32, where multiple paternity was inferred from offspring workers. A total of 15 alleles were observed for the locus Etta 5-6TF and 16 alleles for the locus Etta 7-8TF. The sums of the squared allele frequencies (Σp2) were 0.90 and 0.87, respectively. This implies that our analysis underestimated the real number of fathers (1- Σp2) by 10-13% for each marker locus, i.e. one of every ca. 8-10 inferred fathers was in fact two fathers who had the same marker allele by chance, but most of these undetected males were likely to be captured by the second locus 33. However, when alleles of fathers were the same as queen alleles, a more conservative interpretation would be that these were in fact maternal alleles (amplified from remnants of spermatheca tissue), so that a lower number of fathers would be inferred. We believe that this problem has been minor, because “allelic overlap” (a parental marker allele being shared by the queen and her stored sperm) occurred in only 61 of the 391 (15.75%) single locus comparisons of spermatheca-amplified alleles and queen alleles that could be made. This overlap almost exclusively concerned the more common alleles, which are most likely to be “sampled” twice by chance. Of the 18 Etta 5-6TF queen genotypes that showed either double or single allelic overlap with spermatheca DNA, 13 (72.2%) matched in only one of the two queen alleles (91% of the queens were heterozygous). The corresponding figures for Etta 7-8TF were 28 cases of allelic overlap, of which 20 (69%) matched in only one of the two queen alleles (82.6% of the queens were heterozygous). This is consistent with these single alleles representing a father allele identical with a queen allele by chance, but not with the amplification of maternal spermatheca tissue, as that should have recovered both maternal alleles from the sperm sample. In only two cases we found complete overlap of maternal alleles with paternal alleles (for both loci), and these queens were thus excluded from further analysis, as their mate number remained ambiguous.
Statistics
All statistical analyses were performed using SPSS version 11 for Macintosh. Data were tested for normality and homoscedasticity of variances using, respectively, Kolmogorov-Smirnov tests and Levene’s test for equality of variances, and were found to fulfill the conditions for parametric testing.
References
1.Götz, P. & Boman, H. G. in Comprehensive Insect Physiology, Biochemistry and Pharamacology (eds. Kerkut, G.A. & Gilbert, L. I.) 453-485 (Pergamon Press, Oxford, 1985).
2.Gillespie, J. P., Kanost, M. R. & Trenczek, T. Biological mediators of insect immunity. Annu. Rev. Entomol.42, 611-643 (1997).
3.Nappi, A. J. & Vass, E. Melanogenesis and the generation of cytotoxic molecules during insect cellular immune reactions. Pig. Cell Res.6, 117-126 (1993).
4.Nappi, A.J. & Christensen, B.M. Melanogenesis and associated cytotoxic reactions: applications to insect innate immunity. Insect.Biochem. Molec.35, 443-459 (2005).
5.Shiao, S. H., Higgs, S., Adelman, Z., Christensen, B. M., Lui, S. H., & Chen, C. C. Effect of prophenoloxidase expression knockout on the melanization of microfilariae in the mosquito Armigeres subalbatus. Insect Mol. Biol.10, 315-321 (2001).
6.Gorman, M. J., Cornel, A. J., Collines, F. H. & Paskewitz, S. M. A shared genetic mechanism for melanotic encapsulation of CM-Sephadex beads and a malaria parasite, Plasmodium cynomolgi B, in the mosquito, Anopheles gambiae. Exp. Parasit.4, 380-386 (1996).
7.Baer, B., Krug, A., Boomsma, J.J. & Hughes, W.O.H. Examination of the immune responses of males and workers of the leaf-cutting ant Acromyrmex echinatior and the effect of infection. Insect. Soc.52, 298-303 (2005).
8.Baer, B. & Schmid-Hempel, P. Effects of selective episodes in the field on life history traits in the bumblebee Bombus terrestris. Oikos101, 563-568 (2003).
9.Gerloff, C. U., Ottmer, B. K. & Schmid-Hempel, P. Effects of inbreeding on immune response and body size in a social insect, Bombus terrestris. Funct. Ecol.17, 582-589 (2003).
10.Cotter, S.C., Hails, R.S., Cory, J.S. & Wilson, K. Density-dependent prophylaxis and condition-dependent immune function in Lepidopteran larvae: a multivariate approach. J. Anim. Ecol.73, 283-293 (2004).
11.Lavine, M.D. & Strand, M.R. Insect hemocytes and their role in immunity. Insect.Biochem. Molec.32, 1295-1309 (2002).
12.Eslin, P. & Prevost, G. Hemocyte load and immune resistance to Asobara tabida are correlated in species of the Drosophila melanogaster subgroup. J. Insect. Physiol.44, 807-816 (1998).
13.Fellowes, M. D. E. & Godfray, H. C. J. The evolutionary ecology of resistance to parasitoids by Drosophila. Heredity84, 1-8 (2000).
14.Ryder, J. J. & Siva-Jothy, M. T. Quantitative genetics of immune function and body size in the house cricket, Acheta domesticus. J. Evol. Biol.14, 646-653 (2001).
15.Robb, T., Forbes, M. R. & Jamieson, I. G. Greater cuticular melanism is not associated with greater immunogenic response in adults of the polymorphic mountain stone weta, Hemideina maori. Ecol. Entomol.28, 738-746 (2003).
16.Doums, C., Moret, Y., Benelli, E. & Schmid-Hempel, P. Senescence of immune defence in Bombus workers. Ecol. Entomol.27, 138-144 (2002).
17.Alleyne, M. & Wiedenmann, R. N. Encapsulation and hemocyte numbers in three lepidopteran stemborers parasitized by Cotesia flavipes-complex endoparasitoids. Entomol. Exp. Appl.100, 279-293 (2001).
18.Loher, W., Ruzo, L., Baker, F. C., Miller, C. A. & Schooley, D. A. Identification of the juvenile hormone from the cricket, Teleogryllus commodus, and juvenile hormone titre changes. J. Insect Physiol.29, 585-589 (1983).
19.Nijhout, H. F. Insect hormones (Princeton University Press, Princeton, 1994).
20.Cusson, M., Delisle, J. & Miller, D. Juvenile hormone titers in virgin and mated Choristoneura fumiferana and C. rosaceana females: assessment of the capacity of males to produce and transfer JH to the female during copulation. J. Insect Physiol.45, 637-646 (1999).
21.Rolff, J. & Siva-Jothy, M. T. Copulation corrupts immunity: a mechanism for a cost of mating in insects. Proc. Natl. Acad. Sci. USA99, 9916-9918 (2002).
22.Kurtz, J. Phagocytosis by invertebrate hemocytes: Causes of individual variation in Panorpa vulgaris scorpionflies. Microsc. Res. Techniq. 57: 456-468 (2002).
23.Baer, B. & Schmid-Hempel, P. The artificial insemination of bumblebee queens. Insect. Soc.47, 183-187 (2000).
- Moret, Y. & Siva-Jothy, M. T. Adaptive innate immunity? Responsive-mode prophylaxis in the mealworm beetle, Tenebrio molitor. Proc. R. Soc. B270, 2475-2480 (2003).
25.Tian, H.S., Vinson, S.B. & Coates, C.J. Differential gene expression between alate and dealate queens in the red imported fire ant, Solenopsis invicta Buren (Hymenoptera : Formicidae). Insect Biochem. Mol. Biol.39 937-949.