Content uploaded by Fabio Manfredini
Author content
All content in this area was uploaded by Fabio Manfredini on Oct 12, 2017
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Original article
Parasitic infection protects wasp larvae against a bacterial challenge
Fabio Manfredini
a,
*, Laura Beani
b
, Mauro Taormina
a
, Laura Vannini
a
a
Department of Evolutionary Biology, University of Siena, via Aldo Moro 2, 53100 Siena, Italy
b
Department of Evolutionary Biology, University of Florence, via Romana 17, 50125 Florence, Italy
Received 7 February 2010; accepted 5 May 2010
Available online 2 June 2010
Abstract
Host antibacterial defense after Strepsiptera parasitization is a complex and rather unexplored topic. The way how these parasites interact
with bacteria invading into the host insect during an infection is completely unknown. In the present study we demonstrate that larvae of the
paper wasp Polistes dominulus are more efficient at eliminating bacteria when they are parasitized by the strepsipteran insect Xenos vesparum.
We looked at the expression levels of the antimicrobial peptide defensin and we screened for the activity of other hemolymph components by
using a zone of inhibition assay. Transcription of defensin is triggered by parasitization, but also by mechanical injury (aseptic injection).
Inhibitory activity in vitro against the Gram positive bacterium Staphylococcus aureus is not influenced by the presence of the parasite in the
wasp or by a previous immune challenge, suggesting a constitutive power of killing this bacterium by wasp hemolymph. Our results suggest
either direct involvement of the parasite or that defensin and further immune components not investigated in this paper, for example other
antimicrobial peptides, could play a role in fighting off bacterial infections in Polistes.
Ó2010 Elsevier Masson SAS. All rights reserved.
Keywords: Host-parasite-bacteria interactions; Innate immunity; Wasps; Defensin; Inhibition zone assay
1. Introduction
An emerging field in immunology is the interaction
between immunity and other important physiological func-
tions, such as feeding behavior, ageing, energy use, circadian
rhythm and reproduction: all these aspects can noticeably
impact the outcome of an immune response [1]. In insects like
bees, ants, wasps and termites, “social immunity” is an addi-
tional variable [2]. Colonies of social insects have been
described as “factory fortresses” by Wilson [3], since multiple
barriers are interposed between the colony and the external
“abrasive environment” [4], a continuous source of predators,
parasites and pathogens. Both individual immunity and
cooperative hygienic behavior contribute to maintain the nest
safe and stable [5]. Colony parasites need to progressively
overwhelm multiple defenses, including colony-level
defenses, if they want to parasitize a member of the colony
inside the nest [6]. Endoparasites must accomplish an addi-
tional step to perform successful infection, which consists in
overcoming host individual immunity, since they have to settle
within the hemocoel, another small homeostatic fortress [7].
An insect society with numerous prey items in a single
sheltered environment may be attractive for many parasites
and/or pathogens at the same moment. Given the comfortable
environmental temperature, the crowded living conditions and
the high genetic relatedness of colony members, insect soci-
eties are highly susceptible to microbial infections [8] espe-
cially for young individuals of the colony, which have soft
cuticles and limited movements. Data from the literature claim
that immatures from social insects rely more than adults on
molecular defence mechanisms against parasites and patho-
gens, since they are practically devoid of behavioural
responses, being confined within a comb cell with limited
possibility of movement and necessarily defended by the other
colony members [9].
* Corresponding author at: Department of Entomology, Centre for Chemical
Ecology, Huck Institutes of the Life Sciences, Pennsylvania State University,
University Park, PA, USA. Tel.: þ1 814 863 4432; fax: þ1 814 862 4489.
E-mail address: manfred_f@libero.it (F. Manfredini).
1286-4579/$ - see front matter Ó2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.micinf.2010.05.001
Microbes and Infection 12 (2010) 727e735
www.elsevier.com/locate/micinf
Author's personal copy
The scenario depicted so far corresponds exactly to what
we expect to find in annual colonies of the primitively eusocial
wasp Polistes dominulus (Hymenoptera: Vespidae) infected by
the strepsipteran endoparasite Xenos vesparum (Strepsiptera:
Stylopidae). Infective stages of the parasite (1st instars, also
called triungulins) manage to reach Polistes nests carried by
foraging wasps (for infection modalities different from phor-
esy see Refs. [10,11]). Due to their non-selective host-seeking
behaviour (unpublished), they quickly penetrate into imma-
tures ethus escaping possible aggressions by adult nestmates
eand reach the hemocoel, where they perform all the steps of
their endoparasitic development, rendering the host unable to
mount an effective immune response [12]. The cost of para-
sitism for wasp larvae has been shown to be irrelevant in terms
of mortality and mass loss [13]. This is a good example of
virulence trade-off: the host is valuable to the parasite if it
lives a long life [14]. As a testament to this, coexistence
between X. vesparum and P. dominulus is quite long (around
one year for female parasites) compared to the limited life
span of a worker wasps (usually a few weeks, although they
may reproduce and/or overwinter, due to their caste flexibility
[15]). Xenos achieves this goal by castrating the host and
manipulating its behaviour in such a manner that parasitized
wasps are no longer social and active on the nest but turn into
simple vectors that facilitate the development, the reproduc-
tion and the spread of the parasite [16].
In a system where host and parasite are so deeply con-
nected, it is evident that pathogens which are potentially
dangerous for the host represent a secondary risk for the
parasites that would clearly get important benefits by killing
them straight or pre-activating the immune defense of the host.
In the present study, we tested the antibacterial response of
P. dominulus larvae against pathogenic bacteria injected in
their hemocoel before and after parasitization by X. vesparum.
From previous analyses we knew that wasp immatures are
equipped with a functional immune system provided with both
cellular and humoral responses [12,17,18]. We performed
different measures of immunocompetence (see Ref. [19]):
bacterial clearance, expression levels of the antimicrobial
peptide defensin and the inhibition zone assay. This is the first
time that an antimicrobial peptide ei.e. defensin eis
sequenced and characterized in P. dominulus, a “model
organism” for studies on social insects [20].
2. Material and methods
2.1. Study animals
Insect colonies for laboratory experiments were obtained
from hibernating clusters of P. dominulus (Christ) females (i.e.
future queens), which were collected at the end of the winter in
three different sites in Tuscany (Italy): Impruneta (Florence),
Renaccio and San Gimignanello (Siena). Specimens were split
in pools of three females inside 20 20 20 cm Plexiglas cages
under standard conditions (15L/9D and 28 2C, with paper,
sugar, water and Sarcophaga sp. larvae ad libitum) to allow
polygynous colony foundation (30 large nests after 4e6 weeks).
To obtain parasitized larvae in the lab, we performed arti-
ficial (i.e. laboratory) infections. Our sources of X.vesparum
(Rossi) triungulins were 15 wasps coming from the same
hibernating clusters and parasitized by a single (rarely two)
Xenos female. For each trial, we matched nesting and para-
sitized wasps coming from the same hibernation site. Parasite
females, extruding their cephalothorax (top portion of the
body) through a wasp’s abdomen, started to release batches of
triungulins after four weeks at 15L/9D and 28 C, i.e. when
wasp larvae began to develop inside nests. The procedure was
the same as described in a previous work [12]: briefly, we used
a thin needle to transfer Xenos triungulins from mothers’
cephalothorax to wasp larvae at their 3rde4th developmental
stage. To simulate a natural infection we used a pool of tri-
ungulins (around 5) for each wasp larva and then we painted
the respective nest cell in order to be able to find parasitized
larvae at the moment of dissection. The measures of immu-
nocompetence were performed by collecting hemolymph at 4
days post-infection. During post hoc dissections, we sought for
X. vesparum exuvia and/or 3rd instars, all clues for the para-
site’s successful development. Infections were carried out
from mid-June to mid-July, thus we presumably infected
larvae of early and late workers, i.e. the main target for tri-
ungulins in the field [16]. For each experiment we clumped
together pools of 3rde4th instars larvae eparasitized and
unparasitized efrom multiple nests (15 total, randomly
chosen among 30 laboratory colonies), in order to avoid any
pseudo-replication due to the colony of origin. The mixed
pools of our relatively limited sample prevented from doing
any nested analysis, nevertheless we controlled for random
effects of colony (see Section 2.7).
2.2. Injection of bacteria and bacterial clearance
As immune elicitors we used Gram positive and Gram
negative bacteria, i.e. Staphylococcus aureus and Escherichia
coli (strain ATCC 23739 and ATCC 25923, respectively) that
are common model pathogens for insects normally absent
from the wasp’s hemolymph [18]: therefore, this was likely to
be a novel immune challenge for P. dominulus larvae. Bacte-
rial cultures were grown overnight at 37 C in Luria-Bertani
Broth (LB) to an optical density of OD
600
¼2. After centri-
fugation, bacteria were washed twice, resuspended in phos-
phate-buffered saline (PBS) and diluted to the desired
concentration with PBS. Then the two solutions were mixed
together for inoculation in wasp larvae, in order to test the
effect of both Gram positive and Gram negative bacteria in our
relatively limited set of samplings. Preliminary trials were
conducted to determine bacterial dosage. A higher mortality
rate was registered at injections of more than 10
5
cells in both
parasitized and unparasitized larvae, therefore a dose of 10
5
bacteria was chosen as our experimental challenge; a similar
amount has been used previously with honey bees [21]. Under
these optimized conditions, the survival rate of wasp larvae at
24 h post-injection was about 70% for both groups.
During three independent trials, Polistes larvae were
divided into three groups and treated without being removed
728 F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
from their nest: one group was kept as a non-injected control
(C), the second group was injected with PBS (PBS-i) and the
third group was injected with bacteria (Bac). This procedure
was followed for each of two animal pools: unparasitized
wasps (C¼3; PBS-i ¼12; Bac ¼17, from eight colonies) and
parasitized wasps at 3 days after infection with X. vesparum
(C¼7; PBS-i ¼12; Bac ¼22, from seven colonies). Polistes
larvae were challenged with 1 ml of the bacterial mixture
(total ¼10
5
cells) or 1 ml PBS using a microsyringe (Hamilton
Microliter, series 700, ⌀33 gauge); before injection, the
needle of the syringe was washed with 75% ethanol and then
in sterile distilled water. Nests were left in standard condition
for 24 h to allow recovering of larvae. The next day, 10 ml
hemolymph was collected from each larva, diluted in 90 ml
PBS 1and then serially diluted as described elsewhere [18].
We plated 20 ml of the original solution and 20 mlofthe
1000dilutions on LB agar and we evaluated bacterial
clearance by optically recording the total colony forming units
(CFU) for both S. aureus and E. coli.
2.3. DNA and RNA isolation and cDNA synthesis
Total DNA was extracted from a P. dominulus naı
¨ve 4th
instar larva using the Wizard SV Genomic DNA purification
system (Promega Corporation, Madison, WI, USA). For RNA
extraction, sample larvae were frozen in liquid nitrogen and
subsequently homogenized using a Polytron homogenizer
(Kinematica AG). From each sample total RNA was extracted
following a standard “TRI REAGENTÔ” (SIGMA-Aldrich,
St. Louis, MO, USA) procedure. The quantity of the extracted
RNA was assessed with a Nanodrop ND-1000 UVevis spec-
trophotometer (NanoDrop Technologies, Wilmington, DE,
USA) and absorbance ratio at 260/280 nm and 260/230 nm
was used to assess purity of the RNA samples. For each
sample, 1 mg of total RNA was used to make Oligo(dT) cDNA
by means of ‘‘RevertAid H Minus First Strand cDNA
Synthesis Kit” (Fermentas AB, Vilnius, LT) according to the
manufacturer’s specifications.
2.4. Sequencing of defensin
Full-length defensin (def ) cDNA of P. dominulus was
obtained by RT-PCR and rapid amplification of cDNA ends
(RACE) method. A multiple alignment of Hymenoptera def
sequences (Genbank source) was performed using CLUSTAL
X[22] to identify highly conserved nucleotide sequence
portions used to design a pair of degenerate primers
(Hym_deg_FW vs Hym_deg_RV) that allowed to amplify and
sequence the first portion (158 bp long) of P. dominulus def.
Gene-specific primers (Def_RACE_50and Def_RACE_30) and
nested primers (Def_NEST_50and Def_NEST_30) were
designed from previously determined DNA sequence to get the
full-length def cDNA by means of 50- and 30-RACE analyses
performed using the SMART RACE cDNA amplification Kit
(Clontech, Mountain View, CA, USA) and following manu-
facturer’s instructions. For all primer sequences see Table 1.
Both fragments obtained (50- and 30-regions) were cloned in
the pGEM-T Easy vector (Promega Corporation, Madison,
WI, USA) and three clones for each fragment were randomly
chosen and sequenced in both strands, using M13 forward and
reverse primers. All sequences obtained were corrected
manually and assembled using Sequencer 4.2.2 (Gene Codes,
Ann Arbor, MI, USA) in a complete def transcript, which
shows a 3X coverage, at least, for each nucleotide position.
Protein analysis was performed using Prosite database [24] to
eventually identify conserved domains and motifs. By SignalP
3.0 software [25] the sequence was screened for signal peptide
presence. A search for similarities within known genes was
performed using BLAST (http://blast.ncbi.nlm.nih.gov/Blast.
cgi). A pair of species-specific primers was designed on the
untranslated regions, which were identified at the 50and 30
(50-UTR and 30-UTR) of the transcript obtained to get intron
sequences using DNA as a template in a PCR. The sequence of
the isolated gene was submitted to GenBank with accession
number GU327374.
2.5. PCR analysis of defensin
For both unparasitized and parasitized wasps (see Section
2.2) we screened the following samples to establish def tran-
scription level by means of PCR: 3 C, 3 PBS-i and 9 Bac
specimens (total 15 specimens for each group). The relative
transcription of cytoplasmic actin (act) gene was also investi-
gated to normalized def expression data (internal control). One
microliter of cDNA was used as a template in a 25 ml PCR
reaction combining two different primer pairs: Def_FW vs
Def_RV and Act_FW vs Act_RV (Table 1). After an initial
denaturation at 95 C for 5 min, the PCR conditions were:
95 C60s,54C60s,72C 45 s, for 32 cycles. PCR mixture
was: 1 mM of each primers, 1 mM of each dNTPs, 1X PCR
reaction buffer (Promega Corporation, Madison, WI, USA),
2.5 mM Mg
2þ
and 1U of recombinant Taq DNA polymerase
(Promega Corporation, Madison, WI, USA). This allowed to
amplify a 365 bp portion of the def gene and a 399 bp portion of
the act gene. PCR products were run in a 1.5% agarose gel and
for each sample we performed 3 experimental replicas.
Densitometric quantification of unsaturated images was per-
formed using ImageQuant 5.2 software (Molecular Dynamics).
Table 1
Primer sequences.
Primer name Sequence 50/30
Hym_deg_FW GAACGTGCCGAYAGACAWAGAAGA
Hym_deg_RV TTCTCGCARYGACCTCCAGCTTT
Def_RACE_50TTTCTCGCAATGACCTCCAGCTT
Def_RACE_30AACGTGCCGATAGACAAAGAAGA
Def_NEST_50GTCTTCTTCTTTGTCTATCGGCACGT
Def_NEST_30AAGCTGGAGGTCATTGCGAGAAA
Def_FW CGTCATAGTTGCGGTCAATATGGC
Def_RV CGCAAATACCACTGCTGCAATATCC
Def_Gen_FW GACTTCGATAATTTTATCTAATAAC
Def_Gen_RV GTGATACATTAAATTATCAAAATGTTG
Act_FW
a
AGCAGGAGATGGCCACC
Act_RV
a
TCCACATCTGCTGGAAGG
a
Sequence obtained from Ref. [23].
729F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
The optical density of each band was established based on an
area (with the same size for all samples) including the band;
pixel density for each area was quantified and adjusted by
means of a background correction.
2.6. Inhibition zone assay
As a third measure of immunocompetence we performed an
inhibition zone assay. This is a test to explore the ability of
antibacterial substances present in the hemolymph to inhibit
bacterial growth in vitro: the result is an approximately
circular, clear zone around the hemolymph (where bacteria
were unable to grow) with a diameter proportional to the
strength of the inhibition. The procedure was the same
described in Ref. [26], with a few minor modifications and as
follows. Antibacterial test plates were prepared separately for
S. aureus and E. coli. Live bacteria from an overnight culture
were added to sterile LB containing 1% agar and maintained at
45 C: the final concentration was 10
6
cells/ml for each
bacterial strain. One percent Mead’s Anticoagulant Buffer
(NaOH 98 mM, NaCl 145 mM, EDTA 17 mM, citric acid
41 mM; pH 4.5) was added as melanization inhibitor [27]. The
inoculated agar broths were poured onto 9 cm Petri dishes.
Wells (⌀2 mm) were made by puncturing the agar with a glass
capillary. Hemolymph samples (2 ml) were pipetted directly
into wells and plates were then incubated for 24 h at 37 C. We
evaluated the antibacterial activity as the mean of the min and
max diameters (mm) of inhibition zone per each specimen
[26]. Three replicates per specimen for both bacterial strains
were conducted. Antibiotics (10,000 U penicillin and 10 mg
Streptomycin/ml, 2 ml per well) were added to each plate as
a control sample and had an inhibition zone of approximately
36 mm diameter.
We conducted a preliminary test of the antibacterial activity
of hemolymph samples from the previous year collected from
17 unparasitized Polistes larvae and 26 larvae parasitized by
X. vesparum, which were diluted in a 4X mixture of Grace’s
Insect Medium (SigmaeAldrich) and Mead’s Anticoagulant
Buffer (1:1) and immediately stored at 80 C. Thereafter, we
measured the antibacterial activity of hemolymph samples from
16 controls (C: unparasitized ¼4; parasitized ¼12) and 26
wasp larvae challenged with bacteria (Bac: unparasitized ¼8;
parasitized ¼18), from the same sample groups as above (see
Section 2.2). In this case, we used wasp larvae stored at 80 C,
which were placed into microcentrifuge tubes and spinned at
500gfor 5 min at 4 C, as described in Ref. [28].
2.7. Statistical data analysis
Descriptive statistics were computed for the central
tendency and variability of each dependent variable. Barplots
were used to visualize differences between parasitized and
unparasitized larvae and among treatments. Datasets of
bacterial clearance, expression of defensin and inhibition zone
assay (further trials) had a hierarchical structure with speci-
mens nested within colony. In order to test the effect of
treatments and/or parasitization we used mixed effects models
[29] to account for random effects of colony and specimen.
Normality of residuals was evaluated using the ShapiroeWilk
test. When the assumption of homogeneity was not met, the
final equation was weighted by a power variance function to
correct for heteroskedasticity.
For the quantitative analysis of defensin expression,
a measure of the repeatability of the same measurement taken
three times was computed using an intraclass correlation
coefficient. All analyses were performed with R version 2.9.2
[30]. Mixed effects models were conducted using the lme
function in library nlme.
3. Results
3.1. Bacterial clearance
Non-injected larvae did not harbour any bacterial strain
capable of growing on LB agar, as detected in three unpara-
sitized wasps and in seven wasps parasitized by X. vesparum
(data not shown): this is consistent with previous data [18].
Analogously, LB agar plates from PBS-i larvae did not show
any evidence of bacterial presence in both pools of animals
(12 specimens each, data not shown), thus we may assume that
injection itself was performed in sterile conditions and no
bacteria were accidentally introduced during the experimental
procedure.
The pattern was different in Bac: two bacterial morpho-
types were clearly distinguishable on LB agar plate 24 h after
the inoculation of 10
5
bacterial cells. In 10 ml hemolymph per
larva, on average 1.03 10
4
total CFU were recorded in
unparasitized wasps and 1.67 10
2
in wasps parasitized by
X. vesparum (Fig. 1A): the difference between the two pools
was highly significant (F
(1,24)
¼13.66, P<0.001). The result
is in line with previous observations on the increased bacterial
clearance in 24-h-parasitized wasps [18]. Focusing on unpar-
asitized specimens, animals were able to eliminate around
90,000 CFU within 24 h, which corresponded to a reduction of
98% of injected bacteria. This finding is consistent with data
reported for mosquitoes and fruit flies [31,32].
When we performed separate statistical analyses for the two
bacterial strains, we noticed that the pattern was slightly
different (Fig. 1B). E. coli was significantly reduced in para-
sitized larvae (F
(1,24)
¼9.23, P<0.01; total CFU ¼1.88 10
3
and 0.34 10
2
, unparasitized and parasitized wasps, respec-
tively) while S. aureus did not change significantly
(F
(1,24)
¼1.07, P¼0.31; total CFU ¼1.12 10
3
and
1.05 10
2
, unparasitized and parasitized wasps, respectively).
3.2. Wasp defensin
The complete cDNA sequence of P. dominulus def (Fig. 2)
is 482 bp long and shows a poly(A)
26
sequence at the 30-end. A
predicted coding region of 104 amino acids, as well as 50
(60 bp) and 30(110 bp) untranslated portions are present in this
sequence. Comparison with genomic sequence reveals the
presence of three exons of 73 bp (E1), 202 bp (E2) and 37 bp
(E3), and two introns of 135 bp (among E1 and E2) and 141 bp
730 F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
(among E2 and E3). Prosite database search allowed the
identification of several hypothetical functional sites through
the deduced amino acid sequence: two cAMP- and cGMP-
dependent kinase phosphorylation sites (51-RRVT-54 and
91-RKDS-94), four N-myristoylation sites (62-GMIGSS-67,
65-GSSACA-70, 80-GGYCSS-85 and 86-GICVCR-91), one
protein kinase C phosphorylation site (94-SFK-96) and one
casein kinase II phosphorylation site (94-SFKD-97). Through
exon 2, six cysteines were identified as involved in intra-chain
disulfide bonds (positions: 55, 69, 73, 83, 88 and 90). A most
likely cleavage site between positions 20 and 21 of the
deduced protein (NMA-AP) was identified by SignalP 3.0
software revealing the presence of a signal peptide. Also
a mature peptide cleavage site between position 52 and 53
(RR-VT) and a C-terminal amidation site between position
101 and 104 (KRF-G) were recognized, comparing the
sequence with Apis mellifera defensin 1 protein [33]. A pol-
yadenilation signal sequence AATAAA was identified at the
C-terminal of the sequence.
BLAST analysis of the complete deduced amino acid
sequence showed an amino acid identity with A. mellifera def
1 isoform (GenBank accession number: NP_001011616) of
58%. There is a gap of seven amino acids between Polistes and
Apis sequences located at the position 38e44 in the wasp
protein (FREDMVE), this portion seems to be absent only in
Apoidea sequences recorded in GenBank. Regarding non-
Apoidea sequences the best match (58% identity) was
obtained with Nasonia vitripennis (Hymenoptera, Chalcidoi-
dea) def (GenBank accession number: ACX54960).
The analysis of P. dominulus defensin sequence revealed
several features that are typical of this gene. First, defensin
shows three exons as the A. mellifera defensin1 isoform [33],
moreover these are similar in position and length. Second, the
presence of a signal peptide on the N-terminal of the deduced
defensin protein indicates that this protein is secreted in the
hemolymph. Third, the existence of a mature peptide cleavage
site and a C-terminal amidation site suggests that the native
protein is composed by 51 amino acids, in line with data from
the literature showing defensin ranges in length from 32 to 51
amino acids [34]. Fourth, the six cysteines involved in hypo-
thetical disulfide bonds are a distinctive character of inverte-
brate defensins, which are a family of cysteines-rich
antimicrobial peptides that show six conserved cysteines, all
involved in intra-chain disulfide bonds.
3.3. Expression of defensin
Defensin was constitutively expressed in C unparasitized
wasp larvae (Fig. 3, left side), which showed a significant
increase in transcriptional levels after the injection of PBS
(F
(2,5)
¼51.12, P<0.001, ICC ¼0.55). The relative
expression of the gene remained at similar levels at 24 h
following a bacterial challenge. On the other hand, the pres-
ence of Xenos (Fig. 3, right side) increased the expression of
def in C parasitized individuals (F
(1,2)
¼17.39, P¼0.05,
ICC ¼0.77), thus neither PBS injection nor a bacterial
inoculum were able to trigger a significant boost in transcript
levels, (F
(2,4)
¼0.27, P¼0.77, ICC ¼0.83).
3.4. Inhibition zone assay
No inhibitory activity was recorded against the Gram
negative test bacterium E. coli in Polistes hemolymph (data
not shown), whereas evident antimicrobial activity occurred
against the Gram positive bacterium S. aureus. The inhibitory
activity was strong in preliminary analyses with frozen
hemolymph from untreated animals (mean diameter of the
inhibition zones ¼17.65 mm). No significant difference
(Fig. 4A) was detectable between unparasitized and parasit-
ized wasps (F
(1,41)
¼0.85, P¼0.36). Further trials with
hemolymph collected from frozen larvae (Fig. 4B) revealed
Fig. 1. Bacterial clearance. (A) Total CFU (log
10
transformed) in 10 mlof
hemolymph sampled from 17 unparasitized wasp larvae challenged with
bacteria and 22 wasps challenged with bacteria at 3 days post-infection with
Xenos. (B) Separate analysis for the two bacterial components (E. coli and
S. aureus). Means SE. **P<0.01; ***P<0.001. Arrows denote the
injection dose (10
5
bacterial cells). No CFU grew in control wasps (non-
injected individuals) or in wasps injected with PBS.
731F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
a smaller diameter of the inhibition zones (2.69 mm on
average), whereas the inhibitory activity was not affected by
X. vesparum, nor challenge with bacteria or combination of
both (F
(1,21)
¼0.52, P¼0.47).
4. Discussion
Recent studies on host-parasite interactions have suggested
that parasites do not always irreversibly damage their host [35],
thus the scenario may be different compared to insects affected
by hymenopteran parasitoids (for a review of the subject, see
Ref. [36]). In this perspective, parasites are organisms that
redirect the physiological pathways of the host to gain the
energy necessary for their development and dispersal. In
particular, parasitic castrators are able to exploit resources that
were allocated for reproduction of the host [14]. These two
aspects apply to the X. vesparum-P. dominulus system: Xenos is
a permanent parasitic castrator and, as a consequence of that,
parasitized hosts represent a complex “extended phenotype”
(sensu Dawkins) and become members of the parasite
populations.
In this playground, it may be easier to interpret our main result:
P. dominulus larvae parasitized by X. vesparum are not compro-
mised in their antibacterial response. On the contrary, they are
better equipped than unparasitized specimens to face a huge
bacterial challenge, as reported in previous studies [18].The
lower number of bacterial colonies in parasitized wasps (Fig. 1)
may have two possible justifications: a more effective antibac-
terial response due to the infected wasp’s innate immune system,
or a direct involvement of the parasite in killing bacteria. In this
study we focus on the first possibility in order to point out which
pathway/s is/are specifically targeted by the parasite. However, it
is not always easy to distinguish between pathways, which may
be mechanistically linked via a sophisticated network of signaling
cascades (namely Toll, Imd, Jak/Stat and JNK pathways) which
regulate the effectors of innate immunity [37,38].
Fig. 3. Transcriptional levels for the antimicrobial peptide defensin. Relative
expression of defensin in 15 unparasitized and 15 parasitized wasp larvae
parted into 3 groups: non-injected controls (C, n¼3), injected with phosphate
buffer saline (PBS-i, n¼3), injected with bacteria (Bac, n¼9). All of the
values shown are mean þSE; the bars with different letters are significantly
different (ANOVA, P<0.001).
Fig. 2. Nucleotide sequence and deduced amino acid sequence of P. dominulus defensin. Open reading frames are in bold. Amino acid deduced sequences of each
exon are in grey boxes. Vertical arrows mark a cleavage site for signal peptide processing, a mature peptide cleavage site and a C-terminal amidation site. In square
are marked cytosines involved in disulfide bonds. Splicing signals and the polyadenylation signal sequences are underlined.
732 F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
As a critical component of Polistes humoral immunity, we
analyzed the def transcriptional rate, in combination with
a bacterial challenge and infection by X. vesparum. A similar
approach is frequently used to investigate immune mecha-
nisms in insects which are important disease vectors [39].
Based on our results, defensin seems to play a marginal role in
clearing bacteria. First, the injection itself triggers a significant
increase in def expression (Fig. 3), as demonstrated by higher
levels of transcripts during aseptic injury in unparasitized
wasps (with no further increase after injection of bacteria): this
is in line with dynamics reported in Apis [5,6,40] and Bombus
[41]. Second, visible levels of def expression in unparasitized
non-injected individuals suggest that this gene is constitutively
expressed in wasp larvae, similarly to genes for several anti-
microbial peptides in other organisms [4]. Concerning
a possible effect of Xenos on def transcription levels, parasit-
ized controls show a higher transcription rate than unparasit-
ized ones, but the expression levels of the gene, 24 h after
bacterial challenge, are not different in the two animal pools.
The most obvious explanation (that the wasp increases the
production of defensin to fight Xenos parasites) does not
appear fully appropriate for this case, since antimicrobial
peptides are normally released to selectively kill unicellular
invaders, such as bacteria and fungi [34] or in the case of
protozoan infections [42]. Nevertheless, whether a macro-
parasite can induce the production of antimicrobial peptides
remains an open question. For example parasitoid insects
during their development seem to maintain the host free of
opportunistic infections [43] while antimicrobial peptides in
the hemolymph of a mollusk affected by Schistosoma
contribute to produce lesions to the sporocysts of the parasite
[44]. An alternative explanation is that the expression of def is
just a side effect of a broad immune response activated by the
parasite, although in our experience the defense reaction of the
wasp is apparently silent during the first three days of para-
sitization. Within this time span, in fact, hemocyte numbers
slightly increase [18], phagocytosis in vivo is not significantly
affected by the parasite (unpublished), the encapsulation
process is just starting and the melanization response is absent
[12]. On the other hand, the simple entry of a foreign body into
the larval host may be an elicitor of def expression, a sort of
priming that produces an immune memory and enables para-
sitized wasps to respond more promptly to subsequent bacte-
rial challenges: a similar mechanism has been tested in other
insects for several stimuli [45,46]. Though in the past we
described this process as “soft entry” and a “non-traumatic
event” [12], nevertheless X. vesparum must interrupt the
epidermal layer to reach the hemocoel. The expression of
antimicrobial peptides is persistent for several days at the level
of many epithelia [37,47].
Through the inhibition zone assay we tested the in vitro
activity of Polistes hemolymph. One unexpected result is the
absence of antibacterial activity towards the Gram negative
bacterium E. coli but not against the Gram positive S. aureus,
unlike what observed in bacterial clearance assay. This
discrepancy could be due to specific antimicrobial responses.
In the yellow fever mosquito Aedes aegypti, the clearance of
E. coli relies principally on phagocytosis (a typical cellular
mechanism), while S. aureus on both cellular and humoral
processes such as melanization and lysis [48]. Since hemo-
lymph aliquots used in our assay were previously frozen, we
probably inhibited hemocyte activity, thus we lost the cellular
component of the process. Also the smaller inhibition zone in
S. aureus trials with frozen larvae may be linked to a different
modality of hemolymph storage (see Methods and Fig. 4). On
the whole, these results suggest a constitutive power of killing
bacteria which is not influenced by septic injury or by the
presence of Xenos parasites.
In conclusion, parasitization by X. vesparum evokes
a better response in wasp larvae against a subsequent micro-
bial challenge. This effect could be partially due to a higher
level of defensin, here sequenced for the first time in a paper
wasp. In this study we focused on this antimicrobial peptide
and performed functional analysis following the concomitant
attack of a macroparasite and bacteria. So far only two anti-
microbial peptides, Dominulin A and B, have been described
Fig. 4. Inhibition zone assay. (A) Preliminary analysis of the antibacterial
activity in the hemolymph of 17 unparasitized wasp larvae and 26 individuals
parasitized by Xenos. (B) Comparison of the antibacterial activity in the
hemolymph obtained from 26 wasp larvae challenged with bacteria (Bac:
unparasitized ¼8; parasitized ¼18) and 16 untreated controls (C:
unparasitized ¼4; parasitized ¼12). The antibacterial activity is expressed as
the zone of inhibition (mm diameter) around a drop of hemolymph on
a bacterial test plate inoculated with S. aureus. Data are shown as means SE.
No significant differences were found.
733F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
on the cuticle and in the venom of P. dominulus and represent
a social protection against infection for the nest [49]. Further
antimicrobial peptides are likely to be present in wasp
hemolymph. Since each peptide may have individual patterns
of expression [50], it would be incorrect to extend what we
found for def to other genes. Moreover, the speed of the
response, rather than the total amount of transcripts, could be
the core of a successful defense strategy. Finally, transcript
abundance does not always reflect actual protein levels [50]
which are the real weapon against bacterial cells. Quanti-
fying protein levels in the hemolymph would be the best way
to link the activity of def gene with the bacterial clearance: but
this is a different approach, which is beyond the scope of the
current study. Further investigations on the expression
patterns of other antimicrobial peptides as well as additional
immune measures (for example the phenoloxidase system)
will help elucidating whether bacteria are better eliminated in
parasitized wasps due to a more effective response of their
immune system or to a direct antibacterial activity of Xenos
parasites.
Acknowledgements
The authors are thankful to Romano Dallai and Davide
Malagoli for fruitful discussions and support during the
experimental design and to Alessio Bianciardi for his help in
semi-quantitative analysis of defensin expression. A special
thank to Amy L. Toth for her contribution in revising and
improving the manuscript with valuable advices and nice
suggestions.
References
[1] D.S. Schneider, Physiological integration of innate immunity. in: J. Rolff,
S.E. Reynolds (Eds.), Insect Infection and Immunity. Oxford Univerity
Press, Oxford, 2009, pp. 106e116.
[2] S. Cremer, S.A.O. Armitage, P. Schmid-Hempel, Social immunity. Curr.
Biol. 17 (2007) 693e702.
[3] P. Schmid-Hempel, Parasites in Social Insects. Princeton University
Press, Princeton, NJ, 1998.
[4] P. Bulet, R. Stocklin, L. Menin, Anti-microbial peptides: from inverte-
brates to vertebrates. Immunol. Rev. 198 (2004) 169e184.
[5] J.D. Evans, Transcriptional immune responses by honey bee larvae during
invasion by the bacterial pathogen, Paenibacillus larvae. J. Invertebr.
Pathol. 85 (2004) 105e111.
[6] K. Randolt, O. Gimple, J. Geissendorfer, J. Reinders, C. Prusko,
M.J. Mueller, S. Albert, J. Tautz, H. Beier, Immune-Related proteins
induced in the hemolymph after aseptic and septic injury differ in honey
bee worker larvae and adults. Arch. Insect Biochem. Physiol. 69 (2008)
155e167.
[7] M.R. Strand, L.L. Pech, Immunological Basis for Compatibility in
parasitoid host Relationships. Ann. Rev. Entomol. 40 (1995) 31e56.
[8] A. Stow, A. Beattie, Chemical and genetic defenses against disease in
insect societies. Brain Behav. Immun. 22 (2008) 1009e1013.
[9] N. Wilson-Rich, S.T. Dres, P.T. Starks, The ontogeny of immunity:
development of innate immune strength in the honey bee (Apis melli-
fera). J. Insect Physiol. 54 (2008) 1392e1399.
[10] D.P. Hughes, J. Kathirithamby, S. Turillazzi, L. Beani, Social wasps
desert the colony and aggregate outside if parasitized: parasite manipu-
lation? Behav. Ecol. 15 (2004) 1037e1043.
[11] L. Vannini, A. Carapelli, F. Frati, L. Beani, Non-sibling parasites
(Strepsiptera) develop together in the same paper wasp. Parasitology 135
(2008) 705e713.
[12] F. Manfredini, F. Giusti, L. Beani, R. Dallai, Developmental strategy of
the endoparasite Xenos vesparum (Strepsiptera, Insecta): host invasion
and elusion of its defence reactions. J. Morphol. 268 (2007) 588e601.
[13] D.P. Hughes, J. Kathirithamby, Cost of strepsipteran macroparasitism for
immature wasps: does sociality modulate virulence? Oikos 110 (2005)
428e434.
[14] K.D. Lafferty, A.M. Kuris, Parasitic castration: the evolution and ecology
of body snatchers. Trends Parasitol. 25 (2009) 564e572.
[15] H.K. Reeve, J.M. Peters, P. Nonacs, P.T. Starks, Dispersal of first
“workers” in social wasps: causes and implications of an alternative
reproductive strategy. PNAS 95 (1998) 13737e13742.
[16] L. Beani, Crazy wasps: when parasites manipulate the Polistes pheno-
type. Ann. Zool. Fenn. 43 (2006) 564e574.
[17] F. Manfredini, R. Dallai, E. Ottaviani, Circulating hemocytes from larvae
of the paper wasp Polistes dominulus (Hymenoptera, Vespidae). Tissue
Cell. 40 (2008) 103e112.
[18] F. Manfredini, D. Benati, L. Beani, The strepsipteran endoparasite Xenos
vesparum alters the immunocompetence of its host, the paper wasp
Polistes dominulus. J. Insect Physiol. 56 (2010) 253e259.
[19] B.M. Sadd, P. Schmid-Hempel, Principles of ecological immunology.
Evol. Appl. 2 (2009) 113e121.
[20] M.J. West-Eberhard, Polistine passion. Ann. Zool. Fennici 43 (2006)
387e389.
[21] X.L. Yang, D.L. Cox-Foster, Impact of an ectoparasite on the immunity
and pathology of an invertebrate: evidence for host immunosuppression
and viral amplification. Proc. Natl. Acad. Sci. U.S.A. 102 (2005)
7470e7475.
[22] J.D. Thompson, T.J. Gibson, F. Plewniak, F. Jeanmougin, D.G. Higgins,
The CLUSTAL_X windows interface: flexible strategies for multiple
sequence alignment aided by quality analysis tools. Nucleic Acids Res.
25 (1997) 4876e4882.
[23] D. Malagoli, S. Sacchi, E. Ottaviani, Unpaired (upd)-3 expression and
other immune-related functions are stimulated by interleukin-8 in
Drosophila melanogaster SL2 cell line. Cytokine 44 (2008) 269e274.
[24] N. Hulo, A. Bairoch, V. Bulliard, L. Cerutti, B.A. Cuche, E. de Castro,
C. Lachaize, P.S. Langendijk-Genevaux, C.J.A. Sigrist, The 20 years of
PROSITE. Nucleic Acids Res. 36 (2008) 245e249.
[25] J.D. Bendtsen, H. Nielsen, G. von Heijne, S. Brunak, Improved predic-
tion of signal peptides: signalP 3.0. J. Mol. Biol. 340 (2004) 783e795.
[26] Y. Moret, P. Schmid-Hempel, Survival for immunity: the price of
immune system activation for bumblebee workers. Science 290 (2000)
1166e1168.
[27] G.P. Mead, N.A. Ratcliffe, L.R. Renwrantz, The separation of insect
hemocyte types on Percoll gradients emethodology and problems. J.
Insect Physiol. 32 (1986) 167e177.
[28] K. Meylaers, D. Freitak, L. Schoofs, Immunocompetence of Galleria
mellonella: sex- and stage-specific differences and the physiological
during cost of mounting an immune response metamorphosis. J. Insect
Physiol. 53 (2007) 146e156.
[29] J.C. Pinheiro, D.M. Bates, Mixed-effects Models in: S and S-PLUS.
Springer Verlag, New York, 2000.
[30] R Development Core Team, R: a Language and Environment for
Statistical Computing. R foundation for statistical computing, Vienna,
Austria, 2009.http://www.R-project.org URL.
[31] J.F. Hillyer, S.L. Schmidt, J.F. Fuchs, J.P. Boyle, B.M. Christensen, Age-
associated mortality in immune challenged mosquitoes (Aedes aegypti)
correlates with a decrease in haemocyte numbers. Cell Microbiol.
7 (2005) 39e51.
[32] C.A. Brennan, J.R. Delaney, D.S. Schneider, K.V. Anderson, Psidin is
required in Drosophila blood cells for both phagocytic degradation and
immune activation of the fat body. Curr. Biol. 17 (2007) 67e72.
[33] J. Klaudiny, T. Albert, K. Bachanova, J. Kopernick, J. Simuth, Two
structurally different defensin genes, one of them encoding a novel
defensin isoform, are expressed in honeybee Apis mellifera. Insect.
Biochem. Mol. Biol. 35 (2005) 11e22.
734 F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
Author's personal copy
[34] P. Bulet, C. Hetru, J.L. Dimarcq, D. Hoffmann, Antimicrobial peptides in
insects; structure and function. Dev. Comp. Immunol. 23 (1999)
329e344.
[35] J. Rolff, M.T. Siva-Jothy, Invertebrate ecological immunology. Science
301 (2003) 472e475.
[36] O. Schmidt, U. Theopold, M. Strand, Innate immunity and its evasion
and suppression by hymenopteran endoparasitoids. Bioessays 23 (2001)
344e351.
[37] N.A. Broderick, D.P. Welchman, B. Lemaitre, Recognition and response
to microbial infection in Drosophila. in: J. Rolff, S.E. Reynolds (Eds.),
Insect Infection and Immunity. Oxford University Press, Oxford, 2009,
pp. 13e33.
[38] M. Elrod-Erickson, S. Mishra, D. Schneider, Interactions between the
cellular and humoral immune responses in Drosophila. Curr. Biol.
10 (2000) 781e784.
[39] Y.M. Dong, F. Manfredini, G. Dimopoulos, Implication of the mosquito
midgut microbiota in the defense against malaria parasites. PLoS Pathog.
5 (2009) e1000423.
[40] P.G. Gregory, J.D. Evans, T. Rinderer, L. de Guzman, Conditional immune-
gene suppression of honeybees parasitized by Varroa mites. J. Insect Sci.
5 (2005) insectscience.org/5.7.
[41] Y.S. Choi, Y.M. Choo, K.S. Lee, H.J. Yoon, I. Kim, Y.H. Je, H.D. Sohn,
B.R. Jin, Cloning and expression profiling of four antibacterial peptide
genes from the bumblebee Bombus ignitus. Comp. Biochem. Physiol. B.
Biochem. Mol. Biol. 150 (2008) 141e146.
[42] C.A. Lowenberger, S. Kamal, J. Chiles, S. Paskewitz, P. Bulet, J.A.
Hoffmann, B.M. Christensen, Mosquito ePlasmodium interactions in
response to immune activation of the vector. Exp. Parasitol. 91 (1999)
59e69.
[43] Y. Carton, M. Poirie, A.J. Nappi, Insect immune resistance to parasitoids.
Insect Sci. 15 (2008) 67e87.
[44] C.A.J. Pereira, R.L. Martins-Souza, A. Correa, P.M.Z. Coelho,
D. Negrao-Correa, Participation of cell-free haemolymph of Bio-
mphalaria tenagophila in the defence mechanism against Schistosoma
mansoni sporocysts. Parasite Immunol. 30 (2008) 610e619.
[45] P. Mowlds, A. Barron, K. Kavanagh, Physical stress primes the immune
response of Galleria mellonella larvae to infection by Candida albicans.
Microbes Infect. 10 (2008) 628e634.
[46] B.M. Sadd, P. Schmid-Hempel, Ecological and evolutionary implica-
tions of specific immune responses. in: J. Rolff, S.E. Reynolds (Eds.),
Insect Infection and Immunity. Oxford University Press, Oxford, 2009,
pp. 225e240.
[47] V. Leclerc, J.M. Reichhart, The immune response of Drosophila mela-
nogaster. Immunol. Rev. 198 (2004) 59e71.
[48] J.F. Hillyer, S.L. Schmidt, B.M. Christensen, The antibacterial innate
immune response by the mosquito Aedes aegypti is mediated by hemo-
cytes and independent of Gram type and pathogenicity. Microbes Infect.
6 (2004) 448e459.
[49] S. Turillazzi, G. Mastrobuoni, F.R. Dani, G. Moneti, G. Pieraccini, G.
Lamarca, G. Bartolucci, B. Perito, D. Lambardi, V. Cavallini, L. Dap-
porto, Dominulin A and B: two new antibacterial peptides identified on
the cuticle and in the venom of the social paper wasp Polistes dominulus
using MALDI-Tof, MALDI-TOF/TOF, and ESI-Ion Trap. J. Am. Soc.
Mass Spectrom. 17 (2006) 376e383.
[50] M.D. Lavine, G. Chen, M.R. Strand, Immune challenge differentially
affects transcript abundance of three antimicrobial peptides in hemocytes
from the moth Pseudoplusia includens. Insect. Biochem. Mol. Biol.
35 (2005) 1335e1346.
735F. Manfredini et al. / Microbes and Infection 12 (2010) 727e735
