Downloaded from rspb.royalsocietypublishing.

org on October 15, 2014

Adaptive innate immunity? Responsive-mode prophylaxis in

the mealworm beetle, Tenebrio molitor
Yannick Moret and Michael T. Siva-Jothy

Proc. R. Soc. Lond. B 2003 270, doi: 10.1098/rspb.2003.2511, published 7 December 2003

References Article cited in:

http://rspb.royalsocietypublishing.org/content/270/1532/2475#related-urls

Email alerting service Receive free email alerts when new articles cite this article - sign up in the box at the top
right-hand corner of the article or click here

To subscribe to Proc. R. Soc. Lond. B go to: http://rspb.royalsocietypublishing.org/subscriptions

 Downloaded from rspb.royalsocietypublishing.org on October 15, 2014

Received 2 April 2003

Accepted 8 July 2003
Published online 2 September 2003

Adaptive innate immunity? Responsive-mode

prophylaxis in the mealworm beetle, Tenebrio molitor
Yannick Moret* and Michael T. Siva-Jothy
Department of Animal and Plant Sciences, University of Sheffield, Sheffield S10 2TN, UK
A primary infection by a parasite may indicate a higher risk of being reinfected in the near future (since
infection may indicate that enemies are becoming more abundant). Acquired immunity does not exist in
invertebrates despite the fact that they also face increased risks of reinfection following primary exposure.
However, when subjected to immune insult, insects can produce immune responses that persist for long
enough to provide prophylaxis. Because these immune responses are costly, persistence must be main-
tained through a selective advantage. We tested for the possibility that these long-lasting immune responses
provided increased resistance to later infections by experimentally mimicking a primary immune insult
(pre-challenge) in larvae of the mealworm beetle, Tenebrio molitor, with lipopolysaccharides (LPS) prior
to early or late exposure to spores of the entomopathogenic fungus Metarhizium anisopliae. We found that
pre-challenged larvae produced a long-lasting antimicrobial response, which provided a survival benefit
when the larvae were exposed to fungal infection. These results suggest that the observed response is
functionally ‘adaptive’.
Keywords: ecological immunology; long-lasting immunity; antimicrobial peptides; phenoloxidase

1. INTRODUCTION immune defence (Reeson et al. 1998; Huang & Song

1999; Barnes & Siva-Jothy 2000; Moret & Schmid-
Encounters with pathogens or parasites are often tem-
Hempel 2001; Wilson et al. 2001; Little et al. 2003).
porally unpredictable (Combes 1995; Schmid-Hempel
Arthropods subjected to immune insult also appear to pro-
1998). However, once a naive organism has received an
duce immune responses that persist after the pathogen is
immune insult it might subsequently be exposed to a
neutralized (Söderhäll 1982; Ratcliffe et al. 1985; Kato et
higher probability of being reinfected because pathogens
al. 1994; Lemaı̂tre et al. 1996). So far, no explanation has
that are now present in the environment are likely to
been provided for the occurrence of such long-lasting
increase in frequency because of their reliance on hosts for
immune responses. Because immune responses impose
transmission (e.g. Lui 2000). In short, experience of an
fitness costs (Moret & Schmid-Hempel 2000, 2001) there
infection may predict a higher risk of becoming infected
must be some selective advantage for maintaining these
in the near future. This is likely to have been part of the
long-lasting immune responses. We propose that such
selection behind the evolution of the acquired immune
long-lasting immune responses are functionally adaptive.
response in vertebrates (Rinkevich 1999; Klimovich
They can be described as providing ‘responsive-mode
2000). Acquired immunity, based on immunoglobulins,
prophylaxis’ to later parasite attacks, which are now eco-
improves the efficiency of the host’s immune response
logically more likely. Under this hypothesis the first insult
during the second encounter by being directed at a spe-
received by a naive individual indicates an elevated likeli-
cific, repeated, insult and provides long-lasting ‘responsive-
hood of parasite insult.
mode prophylaxis’ by maintaining effector systems on
As in other arthropods, immune defence in insects relies
standby after the initial insult has been neutralized.
on both constitutive and inducible mechanisms that pro-
Acquired immunity does not exist in invertebrates
vide non-specific and fairly specific immunity, respectively
(Hoffmann et al. 1996), but both vertebrates and invert-
(Hoffmann et al. 1996; Gillespie et al. 1997). Infection
ebrates possess innate immunity. Because the vertebrate
activates multiple systemic responses, including phago-
acquired immune system is adaptive, flexible and specific,
cytosis and encapsulation by haemocyte blood cells
we tend to view the invertebrates’ reliance on an innate
(Ratcliffe et al. 1985; Hoffmann et al. 1996), and
system as a poor alternative in terms of the breadth of
accompanying melanization reactions (Söderhäll &
response options. However, recent inclusion of ecological
Cerenius 1998). These latter reactions are based on the
considerations in the understanding of immune system
pro-phenoloxidase (proPO) cascade, which is a common
function is changing the entrenched view that invertebrate
and generalized response to immune insult. It can be
innate systems are limited simply by the number of avail-
experimentally triggered by non-pathogenic elicitors such
able effector systems. Arthropods show some fairly sophis-
as lipopolysaccharides (LPS), important antigens that
ticated integration of their defence physiology to produce
characterize the surface of some micro-organisms
some very effective functional outcomes. For example,
(Söderhäll & Cerenius 1998). The proPO system involves
when the risk of parasitism increases, even in the absence
numerous enzymes that are constitutively synthesized and
of parasites, insects show prophylactic investment in
located both in the haemolymph and in circulating haemo-
cytes (Gillespie et al. 1997; Söderhäll & Cerenius 1998).
In addition, the recognition of micro-organism cell walls
*
Author for correspondence ([email protected]). induces the production of antifungal or antibacterial

Proc. R. Soc. Lond. B (2003) 270, 2475–2480 2475  2003 The Royal Society
DOI 10.1098/rspb.2003.2511
 Downloaded from rspb.royalsocietypublishing.org on October 15, 2014

2476 Y. Moret and M. T. Siva-Jothy Responsive-mode prophylaxis in the mealworm beetle

peptides by the fat body. These peptides are secreted into a broad range of microbial agents (Gillespie et al. 1997). We
the haemolymph (Hoffmann et al. 1996). These anti- therefore expected the antimicrobial response to LPS challenge
microbial pathways are induced within the hour following to be efficient against M. anisopliae.
the challenge and the synthesis of antimicrobial peptides Preparation of M. anisopliae spores and assays was performed
can persist for many days post-challenge (Söderhäll 1982; as described by Goettel & Inglis (1997). Fungus was cultured
Ratcliffe et al. 1985; Kato et al. 1994; Lemaı̂tre et al. on potato dextrose agar plates and the dose was prepared by
1996). Moreover, some of these insect antimicrobial scraping fungal spores into 0.05% Triton X-100 solution. The
responses are induced in the absence of microbial antigens spore concentration was determined by counting spores using a
by a trauma alone, for example during wounding haemocytometer with a compound microscope (Leitz Diaplan)
(Lemaı̂tre et al. 1997), presumably as a prophylactic and was then adjusted to 2 × 106 spores ml⫺1 of Triton X-100
response. solution for inoculation. Inoculates were never older than 24 h
In this study, we mimicked a primary immune insult by and were stored at 4 °C when not used.
experimentally challenging larvae of the mealworm beetle, Tenebrio molitor larvae were exposed to the fungus by dipping
Tenebrio molitor, with LPS. The larvae were then exposed them in 1 ml of spore solution for 5 s. Control larvae were
to early or late infection with spores of the entomopathog- dipped in 0.05% Triton X-100 solution without fungal spores.
enic fungus Metarhizium anisopliae to test for the survival Larvae were then allowed to crawl over tissue paper to remove
costs and benefits of producing a long-lasting prophylactic excess moisture and placed back in their box. Mortality was
immune response. If the primary insult provides long- scored over 50 days after fungal inoculation.
lasting protection, we expected to observe a survival bene-
fit to the fungal infection in previously challenged larvae. (d ) Immune parameters
By contrast, because mounting an immune response is Six larvae were sampled at random from each immune treat-
costly (Moret & Schmid-Hempel 2000) we expected to ment at day 4 and day 7 after the pre-challenge but prior to the
see a survival cost in previously challenged larvae that were treatment exposure to fungal spores to test for the antibacterial
not exposed to fungal infection. and the phenoloxidase (PO) activity in their haemolymph. Each
larva was chilled on ice for 10 min and the pleural membrane
between the second and the third abdominal segments was
2. MATERIAL AND METHODS
punctured with a sterile hypodermic needle. The droplet of
(a) Mealworm cultures haemolymph that came out of the wound was collected into a
The beetle larvae used in this study were taken at random sterile, pre-chilled glass capillary. For each insect, 5 µl of haemo-
from a stock culture maintained at the University of Sheffield. lymph were collected and flushed into a 1.5 ml microcentrifuge
Nine cultures, each containing 75 to 80 larvae at the same devel- tube containing 50 µl of cold sodium cacodilate/CaCl2 buffer
opmental stage (10 to 15 mm in length), were set up for the (0.01 M Na-Cac, 0.005 M CaCl2, pH 6.5). A 15 µl subsample
experiment. Cultures were fed ad libitum and kept at 28 ± 1 °C was kept in a 0.5 ml microcentrifuge tube and stored at –80 °C
for the duration of the experiment. until later examination for antibacterial activity using a zone-of-
inhibition test. The methods for this test were as described in
(b) Immune treatments Moret & Schmid-Hempel (2000) except that the assay was per-
Experimental cultures were assigned to one of the three treat- formed using haemolymph diluted 12 times with sodium
ments: ‘naive’, ‘Ringer’ and ‘LPS’. Larvae in the ‘naive’ group cacodilate/CaCl2 buffer. The remaining haemolymph solution
were chilled on ice prior to inclusion in the experiment. Larvae was diluted with 30 µl of cold sodium cacodilate/CaCl2 buffer
in the ‘Ringer’ group received a single injection of 5 µl of saline and immediately stored at –80 °C for later measurement of PO
solution (Ringer solution) after being chilled on ice. Larvae in activity. PO activity was assayed by thawing samples of frozen
the ‘LPS’ group received a 0.5 mg ml⫺1 dose of LPS (Sigma: haemolymph solution (dilution 1 in 20; haemolymph/
L8274) in 5 µl Ringer solution after being chilled on ice. LPS sodium cacodilate/CaCl2 buffer) on ice and then adding 20 µl
is the non-pathogenic, non-living surface molecule derived from to a microplate well containing 140 µl of distilled water, 20 µl
E. coli. It is highly immunogenic (Söderhäll 1982; Ratcliffe et al. of phosphate buffer saline (PBS: 8.74 g NaCl; 1.78 g Na2HPO4,
1985; Jomori et al. 1990) and elicits the production of antimicro- 2H2O; 1000 ml distilled water; pH. 6.5) and 20 µl of l-DOPA
bial peptides over many hours (Söderhäll 1982; Ratcliffe et al. solution (4 mg ml⫺1 of distilled water). The reaction was
1985; Kato et al. 1994; Lemaı̂tre et al. 1996). All injections were allowed to proceed at 30 °C in a microplate reader (Versamax,
made through the pleural membrane between the second and Molecular Devices) for 40 min. Readings were taken every 10 s
the third abdominal segments, using sterilized glass capillaries at 490 nm and analysed using SoftMax Pro 4.0 software
that had been pulled out to a fine point. (Molecular Devices, Sunnyvale, CA). Enzyme activity was mea-
sured as the slope (Vmax value) of the reaction curve during the
(c) Fungal infection linear phase of the reaction (Barnes & Siva-Jothy 2000; Plaistow
Larvae of each immune treatment were then assigned to one et al. 2003).
of the three infection treatments with the entomopathogenic
fungus M. anisopliae: ‘control’ (larvae were not exposed to the (e) Statistics
fungi), ‘early’ infection (larvae were exposed to the fungi 4 days Antibacterial and PO activities were analysed using a multi-
post-challenge) and ‘late’ infection (larvae were exposed to the variate analysis of variance (MANOVA) with pre-challenges and
fungi 7 days post-challenge). The antimicrobial pathway time post-injection as factors. Antibacterial activity was log
induced by the recognition of LPS in insects is different from transformed to homogenize the variance.
the one that would be induced through the recognition of fungi We used a Cox regression with a time-dependent covariate to
(Gottar et al. 2002; Tzou et al. 2002). However, the resulting analyse the differences in survival rates with respect to fungal
antimicrobial peptides that are produced work efficiently against and pre-challenge treatments. This analysis allows the

Proc. R. Soc. Lond. B (2003)

 Downloaded from rspb.royalsocietypublishing.org on October 15, 2014

Responsive-mode prophylaxis in the mealworm beetle Y. Moret and M. T. Siva-Jothy 2477

(a) pre-challenges × time post-injection: F4,60 = 1.55, p = 0.185).

10 As expected from the pre-challenge treatments, the hae-
zone inhibition surface (mm2) + s.e.

molymph of LPS-treated larvae had a higher antibacterial

8 activity than that of Ringer-treated or naive larvae (figure 1a;
ANOVA for zone of inhibition: F 2,30 = 12.27, p ⬍ 0.001).
We conclude that LPS-treated larvae had activated their
6 antibacterial immune pathway for at least 7 days post-
insult. Ringer-treated larvae also had higher antibacterial
4 activity compared with naive larvae but the level of their
antibacterial response did not reach that of LPS-
challenged larvae (figure 1a). Antibacterial activity levels
2 of each pre-challenge treatment (Ringer and LPS pre-
challenges) seemed to converge towards a similar value
0 at day 7 (figure 1a), but these changes across time were
day 4 day 7 only marginally significant (pre-challenges × time post-
(b) injection: F 2,30 = 3.23, p = 0.054). PO activity did not sig-
700 nificantly differ between LPS-treated and naive larvae
phenoloxidase activity (Vmax + s.e.)

(figure 1b), but the PO activity of Ringer-treated larvae

was enhanced (figure 1b; F 2,30 = 6.58, p = 0.004). Time
600
post-injection did not affect PO activity (figure 1b;
F 2,30 = 0.77, p = 0.783) and the effect of the immune chal-
500 lenges on the enzyme activity remained the same across
time (figure 1b; F 2,30 = 0.109, p = 0.897).
400
(b) Survival to pre-challenges and fungal
infections
300 Insect survival was monitored for 50 days after exposure
to fungal spores (figure 2). While exposure to fungal infec-
tion increased insect death rate by a factor of 1.5–2 (see
200
day 4 day 7 odds ratio for infection (4) and infection (7) in table 1),
pre-challenge with Ringer or LPS provided a survival
days post-challenge
benefit for the duration of the experiment by reducing the
Figure 1. (a) Antibacterial and (b) PO activities in the death rate by 33% and 58%, respectively (table 1; figure
haemolymph of Tenebrio molitor larvae from naive (open 2). However, this survival benefit decreased across time
bars), Ringer (grey bars) and LPS (black bars) pre- (table 1: T × Ringer, and T × LPS). The negative survival
challenges tested at days 4 and 7 post-injection. effect of the fungal infections remained constant except
for the early fungal exposure that induced less mortality
across time (table 1: T × infection (4) and T × infection
simultaneous comparison of entire survival curves against an (7)).
assigned reference survival function (baseline function) for dif- The survival benefit of pre-challenge was dependent on
ferent variables (Norusis & Inc. 1989). Our statistical model the fungal treatment to which the beetle larvae were
used a stepwise procedure and the reference survival function exposed. There was no survival benefit from the Ringer
was generated from the control data derived from the pre- challenge when the larvae were exposed to fungal infec-
challenge and fungal treatments (e.g. naive and no fungal tions, regardless of time after injection (see interaction
infection). The pre-challenges were coded as categorical vari- terms: infection (4) × Ringer and infection (7) × Ringer in
ables (0, 1, 2) for naive, Ringer and LPS, respectively, and with table 1). By contrast, the LPS challenge provided a sur-
the fungal treatments (0, 1, 2) for control, early (at day 4 post- vival benefit when the larvae were exposed to early or late
injection) or late inoculation (at day 7 post-injection), respect- fungal infections (figure 2b,c; table 1: infection (4) × LPS
ively. Time (in days) was incremented as covariate in the model and infection (7) × LPS). However, this benefit was not
because hazard ratios of the survival functions were not constant constant over time as survival decreased for larvae exposed
across time. All data were analysed using SPSS 10 for Macin- to early fungal infection (table 1: T × infection (4) × LPS).
tosh.

4. DISCUSSION
3. RESULTS
We demonstrated that the immune response following
(a) Antibacterial and PO activities pre-challenge provides resistance against subsequent
MANOVA showed a significant effect of the pre-chal- pathogen insults. LPS-treated T. molitor larvae showed
lenges on the measured immune parameters (these better survival to fungal infection 4 or 7 days after pre-
response variables are zone of inhibition and PO activity challenge (table 1; figure 2b,c). This enhanced resistance
(figure 1; Pillai’s trace: F 4,60 = 9.14, p ⬍ 0.001)). By con- to infection was correlated with elevated antimicrobial
trast, neither time post-injection nor its interaction with activity in the haemolymph of LPS-treated larvae for the
pre-challenge treatments affected the measured immune period of the fungal inoculations (figure 1a). By contrast,
parameters (time post-injection: F 2,29 = 0.34, p = 0.712; PO activity of LPS-treated larvae was lowest when com-

Proc. R. Soc. Lond. B (2003)

 Downloaded from rspb.royalsocietypublishing.org on October 15, 2014

2478 Y. Moret and M. T. Siva-Jothy Responsive-mode prophylaxis in the mealworm beetle

(a) mode prophylaxis, which seems only to be provided by

100
the antimicrobial response. Increased PO activity on
90 wounding in the absence of specific ‘pathogen associated
cumulative survival (%)

80 molecular patterns’ (the ‘Ringer’ treatment) may prevent

70 some low-level opportunistic infections (i.e. as an induced
non-specific response to wounding) and/or may be the
60
result of the mobilization of the wound-healing apparatus,
50 which requires the production of new exocuticle and mel-
40 anin (Ashida & Brey 1995; Sugumaran et al. 2000).
30 Another important result from our study is that the sur-
vival benefit of the LPS pre-challenge is greater for
20
infected larvae, as is shown by the significant interaction
10 terms in table 1: infection (4) or (7) × pre-challenge. The
0 5 10 15 20 25 30 35 40 45 50
analysis shows that the pre-challenge offers increased sur-
vival over the period during which the fungal pathogen
(b) has its negative effects, but bears a slight but significant
100
cost (T × pre-challenges; table 1) once that advantage has
90 passed. By contrast, naive (no pre-challenge) animals
cumulative survival (%)

80 show high initial mortality owing to their lack of resistance

to the infection, which then stabilizes (T × infections at
70
day 4; table 1) in the absence of a cost of the prophylaxis.
60 In our experiment, such a cost, albeit small, was detect-
50 able despite ideal conditions in terms of food and tem-
40 perature that are not commonly found in natural
conditions. The actual mechanism that maintains high
30 levels of antimicrobial activity during long-lasting immune
20 responses in insects is as yet not known. Such a persistent
10 antimicrobial activity may either be due to a continuous
0 5 10 15 20 25 30 35 40 45 50 synthesis of antimicrobial peptides in the haemolymph or
to a slow turnover of these peptides when produced. In
(c) terms of the cost of the immune response, the production
100 mode of these antimicrobial peptides may have different
90 impact, because the first might be more costly than the
cumulative survival (%)

80 second. However, this is as yet not known. Because

70 mounting an immune response is costly (Moret &
60 Schmid-Hempel 2000), a long-lasting immune response
is not expected to be adaptive unless the advantage of pro-
50
ducing it over-compensates for the cost. Variable prob-
40
abilities of reinfection after a primary infection may
30 balance the cost–benefit ratio of such adaptive prophylac-
20 tic investment in immune defence.
10 Homologies exist between vertebrate and invertebrate
innate immunity (Medzhitov & Janeway 1998), but the
0 5 10 15 20 25 30 35 40 45 50 vertebrate acquired immune system has no direct equival-
time post-challenge (d) ent in invertebrates (Hughes 1998). For this reason, it is
generally held that invertebrates are not able to mount any
Figure 2. Survival curves of the mealworm beetle Tenebrio kind of adaptive immune response (Arala-Chaves &
molitor by immune pre-challenge when exposed to (a) no Sequeira 2000). Clearly the definition of ‘adaptive’ is
infection; (b) to fungal infection at day 4 after pre-challenge important in this argument: we argue that if the ultimate
and (c) to fungal infection at day 7 after pre-challenge.
function of vertebrate acquired immunity is to provide
(Circles, naive; triangles, Ringer; squares, LPS.)
better protection against repeated parasite infections then
our data show a similar functional outcome in an invert-
pared with the other treatments, especially when com- ebrate. By looking at the invertebrate immune system
pared with the Ringer pre-challenge where the larvae had from an ecological perspective (i.e. the response to the
an increased enzyme activity (figure 1b). This latter result probability of reinfection), we suggest that the prophylac-
is in accordance with a study in bumble-bees showing a tic benefits of prolonged expression of innate pathways
negative correlation between antibacterial and PO activi- may represent the functional equivalent of the acquired
ties in response to LPS treatment (Moret & Schmid- immune response, and we argue that that kind of regu-
Hempel 2001) and could indicate a trade-off between the lation is an adaptive response. Clearly the resolution and
two immune pathways. The elevated PO activity found in memory of the invertebrate system are not comparable
Ringer-treated larvae did not correlate with a better pro- with that of vertebrates in these mechanistically disparate
tection against fungal infections. This suggests that the PO systems, but both provide prophylactic cover in response
system might not be involved in long-lasting responsive- to the onset of a parasitic insult. The chasm between

Proc. R. Soc. Lond. B (2003)

 Downloaded from rspb.royalsocietypublishing.org on October 15, 2014

Responsive-mode prophylaxis in the mealworm beetle Y. Moret and M. T. Siva-Jothy 2479

Table 1. Results of the time-dependent Cox regression model. The table contains the relevant terms identified by a backward
stepwise procedure.
(The full model is: survival, S(t) = S0(t)p, where p = exp[b(pre-challenges) ⫹ b(infections) ⫹ b(infections × pre-challenges) ⫹
(t/1)(b(T × pre-challenges) ⫹ b(T × infections) ⫹ b(T × infections × pre-challenges)].)

factorsa bb s.e.c Waldd d.f. pe oddsf

pre-challenges 26.25 2 ⬍ 0.001

Ringer ⫺0.39 0.162 5.91 1 0.015 0.67
LPS ⫺0.86 0.169 26.23 1 ⬍ 0.001 0.42
infections 18.15 2 ⬍ 0.001
infection (4) 0.40 0.173 5.43 1 0.02 1.50
infection (7) 0.71 0.168 18.14 1 ⬍ 0.001 2.04
infections × pre-challenges 17.50 4 0.002
infection (4) × Ringer ⫺0.79 0.409 3.72 1 0.054 0.45
infection (7) × Ringer 0.27 0.400 0.46 1 0.496 1.31
infection (4) × LPS ⫺1.43 0.422 11.44 1 0.001 0.24
infection (7) × LPS ⫺0.76 0.406 3.54 1 0.06 0.47
T × pre-challenges 12.20 2 0.002
T × Ringer 0.01 0.007 5.38 1 0.02 1.01
T × LPS 0.02 0.006 11.72 1 0.001 1.02
T × infections 5.02 2 0.081
T × infection (4) ⫺0.01 0.006 4.42 1 0.036 0.99
T × infection (7) ⫺0.01 0.007 2.53 1 0.111 0.99
T × infections × pre-challenges 10.09 4 0.039
T × infection (4) × Ringer 0.03 0.014 4.19 1 0.041 1.03
T × infection (7) × Ringer 0.01 0.017 0.37 1 0.542 1.01
T × infection (4) × LPS 0.04 0.014 9.01 1 0.003 1.04

a
‘Ringer’: (5 µl); ‘LPS’: (5 µl, 0.5 mg ml⫺1); ‘infection (4)’: exposure to fungal inoculums (2.106) at day 4 post immune challenge;
‘infection (7)’: exposure to fungal inoculums (2.106) at day 7 post immune challenge; ‘T’: time covariate (in days).
b
b, regression coefficient of overall survival function for variable.
c
Standard error of regression coefficient.
d
Wald statistic for variable. Values p ⭐ 0.05 are given in bold.
e
Significance level for Wald statistic.
f
Odds ratio of survival for variable relative to control (= exp(b)).

invertebrate and vertebrate immune mechanisms may be REFERENCES

less obvious at this functional level: one system provides
highly specific, responsive-mode prophylaxis in generally Arala-Chaves, M. & Sequeira, T. 2000 Is there any kind of
long-lived complex organisms whereas the other provides adaptive immunity in invertebrates? Aquaculture 191, 247–
generic, responsive-mode prophylaxis in generally short- 258.
lived organisms. Life-history differences between these Ashida, B. & Brey, P. T. 1995 Role of the integument in insect
taxa may be better reflected in the manner in which the defense: pro-phenol oxidase cascade in the cuticular matrix.
available immune responses are coordinated, rather than Proc. Natl Acad. Sci. USA 92, 10 698–10 702.
Barnes, A. I. & Siva-Jothy, M. T. 2000 Density-dependent
the differences in specificity and memory.
prophylaxis in the mealworm beetle Tenebrio molitor L.
Our study demonstrates that invertebrates show long-
(Coleoptera: Tenebrionidae): cuticular melanization is an
lasting (in the context of a semelparous, short-lived indicator of investment in immunity. Proc. R. Soc. Lond.
organism) prophylaxis against secondary parasite insults B 267, 177–182. (DOI 10.1098/rspb.2000.0984.)
after a ‘pre-challenge’. We conclude that such a response Combes, C. 1995 Intéractions durables: écologie et évolution du
is functionally ‘adaptive’ and is not unique to vertebrates. parasitisme. Paris: Masson.
This suggestion is based on the notion that vertebrates Gillespie, J., Kanost, M. R. & Trenczeck, T. 1997 Biological
and invertebrates are subjected to similar parasite- mediators of insect immunity. A. Rev. Entomol. 42, 611–
imposed selection, which has provided a similar ultimate 643.
(functional) response, albeit based on different mech- Goettel, M. S. & Inglis, G. D. 1997 Fungi: Hyphomycetes. In
anisms. Manual of laboratory techniques in insect pathology (ed. L.
Lacey), pp. 213–247. London: Academic.
Gottar, M., Gobert, V., Michel, T., Belvin, M., Duyk, G.,
We thank I. Johnson for technical advice and help. We also Hoffmann, J. A., Ferrandon, D. & Royet, J. 2002 The Droso-
thank K. Reinhardt and J. Rolff for comments on the
phila immune response against Gram-negative bacteria is
manuscript, and two anonymous referees whose comments
greatly improved the manuscript. Y.M. was supported by mediated by a peptidoglycan recognition protein. Nature
a Marie Curie Fellowship. M.T.S.-J. was supported by 416, 640–644.
Natural Environment Research Council (Swindon) grant, Hoffmann, J. A., Reichhart, J. M. & Hetru, C. 1996 Innate
GR3Y12121. immunity in higher insects. Curr. Opin. Immunol. 8, 8–13.

Proc. R. Soc. Lond. B (2003)

 Downloaded from rspb.royalsocietypublishing.org on October 15, 2014

2480 Y. Moret and M. T. Siva-Jothy Responsive-mode prophylaxis in the mealworm beetle

Huang, C. C. & Song, Y. L. 1999 Maternal transmission of Moret, Y. & Schmid-Hempel, P. 2001 Immune defence in
immunity to white spot syndrome associated virus (WSSV) bumble-bee offspring. Nature 414, 506.
in shrimp (Penaeus monodon). Dev. Comp. Immunol. 23, Norusis, M. J. & SPSS Inc. 1989 SPSS advanced statistic user’s
545–552. guide. Chicago: SPSS Inc.
Hughes, A. L. 1998 Protein phylogenies provide evidence of a Plaistow, S., Outreman, Y., Moret, Y. & Rigaud, T. 2003 Vari-
radical discontinuity between arthropod and vertebrate ation in the risk of being wounded: an overlooked factor in
immune systems. Immunogenetics 47, 283–296. studies of invertebrate immune function. Ecol. Lett. 6,
Jomori, T., Kubo, T. & Natori, S. 1990 Purification and 489–494.
characterization of lipopolysaccharide-binding protein from Ratcliffe, N. A., Rowley, A. F., Fitzgerald, S. W. & Rhodes,
hemolymph of American cockroach Periplaneta americana. C. P. 1985 Invertebrate immunity: basic concepts and recent
Eur. J. Biochem. 190, 201–206. advances. Int. Rev. Cytol. 97, 183–350.
Kato, Y., Motoi, Y., Taniai, K., Kadono-Okuda, K., Hiram- Reeson, A. F., Wilson, K., Gunn, A., Hails, R. S. & Goulson,
atsu, M. & Yamakawa, M. 1994 Clearance of lipopolysac- D. 1998 Baculovirus resistance in the noctuid Spodoptera
charide in hemolymph of Bombyx mori: its role in the exempta is phenotypically plastic and responds to population
termination of cecropin mRNA induction. Insect Biochem. density. Proc. R. Soc. Lond. B 265, 1787–1791. (DOI
Molec. Biol. 24, 539–545. 10.1098/rspb.1998.0503.)
Klimovich, V. B. 2000 Actual problems of evolutionary immu- Rinkevich, B. 1999 Invertebrates versus vertebrates innate
nology. J. Evol. Biochem. Physiol. 38, 562–574. immunity: in the light of evolution. Scand. J. Immunol. 50,
Lemaı̂tre, B., Nicolas, E., Michaut, L., Reichhart, J.-M. & 456–460.
Hoffman, J. A. 1996 The dorsoventral regulatory gene cas- Schmid-Hempel, P. 1998 Parasites in social insects. Monograph
sette spaezle/toll/cactus controls the potent antifungal in behavior and ecology. Princeton University Press.
response in Drosophila adults. Cell 88, 973–983. Söderhäll, K. 1982 Prophenoloxidase activating system and
Lemaı̂tre, B., Reichart, J. M. & Hoffmann, J. A. 1997 Droso- melanization: a recognition mechanism of arthropods? A
phila host defence: differential induction of antimicrobial review. Dev. Comp. Immunol. 6, 601–611.
genes after infection by various classes of microorganisms. Söderhäll, K. & Cerenius, L. 1998 Role of the prophenoloxi-
Proc. Natl Acad. Sci. USA 94, 14 614–14 619. dase-activating system in invertebrate immunity. Curr. Opin.
Little, T. J., O’Connor, B., Colegrave, N., Waat, K. & Read, Immunol. 10, 23–28.
A. F. 2003 Maternal transfer of strain-specific immunity in Sugumaran, M., Nellaiappan, K. & Valivittan, K. 2000 A new
an invertebrate. Curr. Biol. 13, 489–492. mechanism for the control of phenoloxidase activity: inhi-
Lui, K. J. 2000 Confidence intervals of the simple difference bition and complex formation with quinone isomerase. Arch.
between the proportions of a primary infection and a second- Biochem. Biophys. 379, 252–260.
ary infection, given the primary infection. Biometrical J. 42, Tzou, P., De Gregorio, E. & Lemaı̂tre, B. 2002 How Droso-
59–69. phila combats microbial infection: a model to study innate
Medzhitov, R. & Janeway, C. A. 1998 An ancient system of immunity and host–pathogen interactions. Curr. Opin.
host defense. Curr. Opin. Immunol. 10, 12–15. Microbiol. 5, 102–110.
Moret, Y. & Schmid-Hempel, P. 2000 Survival for immunity: Wilson, K., Cotter, S. C., Reeson, A. F. & Pell, J. K. 2001
the price of immune system activation for bumble-bee work- Melanism and disease resistance in insects. Ecol. Lett. 4,
ers. Science 290, 1166–1168. 637–649.

Proc. R. Soc. Lond. B (2003)