Content uploaded by Ahmad Abu Hassan
Author content
All content in this area was uploaded by Ahmad Abu Hassan
Content may be subject to copyright.
Journal of Parasitology and Vector Biology Vol. 1 (2) pp. 013-018, August, 2009
Available online at http://www.academicjournals.org/jpvb
Academic Journals
Full Length Research Paper
Fluorescence can be used to trace the fate of
exogenous micro-organisms inside the alimentary tract
of mosquitoes
Tomomitsu Satho1, Hamady Dieng1*, Tetsuya Mizutani2, Yuki Eshita3, Takeshi Miyata1, Parimal
Talukder1, Nobuhiro Kashige1, Abu Hassan Ahmad4 and Fumio Miake1
1Faculty of Pharmaceutical Sciences, Fukuoka University, 8-19-1 Nanakuma, Jonan-ku, Fukuoka 814-0180, Japan
2Virology 1, National Institute of Infectious Diseases, 4-7-1 Gakuen, Musashimurayama, Tokyo 208-0011, Japan.
3Department of Infectious Diseases, Faculty of Medicine, Oita University, 1-1 Idaigaoka, Hasama-machi, Yufu, Oita 879-
5593, Japan.
4School of Biological Sciences, University Sains Malaysia, 11800, Pulau Pinang Malaysia.
Accepted 28 August, 2009
There is a great deal of current research interest in utilising bacteria for the control of intractable
arthropod-borne diseases such as dengue. Although there is accumulating evidence that bacterial
infection is a promising control strategy, most studies on bacteria-insect interactions lacked useful
markers for detecting pathogenesis. This provided the impetus to investigate bacterial infection in the
dengue vector Aedes albopictus. The infection persistence patterns in key organs of the alimentary
canal of females were examined using a GFP-expressing strain of Escherichia coli (Migula). Just after
feeding with sugar meal containing the bacteria, the crop and midgut as well as parts of the Malpighian
tubules showed fluorescence. From 1 h onwards, bacterial populations declined sharply in both the
midgut and crop, with complete elimination in the former but persistence of bacteria at 7 h post-feeding
in the latter. After 24 h, neither organ retained the fluorescent marker. However, culture of homogenates
of these organs in Luria-Bertani medium revealed the presence of a bacterial population in the crop, but
not in the midgut. These observations suggest a difference in the potential physiological actions
expressible by the two organs. In fact, both are storage sites for ingested fluids, but the midgut has
greater physiological activity. Presumably, one of these activities contributed to eliminating GFP-
expressing E. coli from the A. albopictus midgut after 24 h. The results of the present study using a
fluorescent marker to detect infection may be useful for developing strategies to fully characterise the
main steps involved in the bacterial infection process in insects.
Key words: Bacteria infection, fluorescent marker, crop, midgut, persistence.
INTRODUCTION
There is growing concern regarding the eventual impact
of global climate change on the evolution of arbovirus
infections (Chastel et al., 2002), particularly dengue.
Despite tremendous effort to control this disease, its
transmission is increasing due to the combined effect of
changing human demographics and the spread of
pathogens and vectors (Gubler, 2004). Dengue is caused
*Corresponding author. E-mail: hamachan1@yahoo.com.
Tel: +81-92-871-6631 (Ext: 6613). Fax: +81-92-863-0389.
by a flavivirus transmitted by the mosquitoes Aedes
aegypti (L.) and A. albopictus. The later mosquito, which
is native to Southeast Asia, is ranked fourth on the list of
the most invasive organisms worldwide (http://www.is-
sg.org/data base/welcome). It is an important vector of
several arboviruses, including those responsible for
yellow fever and various types of encephalitis as well as
a competent laboratory vector of at least 23 arboviruses
(Mitchell, 1995). Its larvae emerge from the eggs in
containers and ingest some of the microbial fauna (Sota
et al., 1992) associated with organic detritus, their major
carbon source (Clements, 1992). As with most insects,
014 J. Parasitol. Vector Biol.
the adult stage has a relatively restricted diet and thus
harbours a community of microbiota, in particular
bacterial fauna, which participates in many types of
interaction ranging from pathogenesis to obligate
mutualism (Dillon and Dillon, 2004; Dharne et al., 2006).
Many types of bacteria have been isolated from the
midgut of field-collected mosquitoes (Straif et al., 1998).
The principal route of the interactions between bacteria
and the insect interior milieu is through the ingestion of
contaminated food (Valley-Gely et al., 2008). In
mosquitoes, there is evi-dence that bacteria can be
transmitted transstadially from larvae to adults and in
nectar/sugar-feeding at the adult stage (Pumpuni et al.,
1996). On feeding, ingested food or blood meal and
associated microorganisms and/or pathogens are
transferred passively through the foregut and stored in the
midgut, the site of digestion and absorption of nutrients
(Clement, 1992) as well as infection (Wang et al., 2001). In
nature, both sexes feed on nectar, which is stored
temporarily in the crop and then gradually released into
the midgut (Fisk and Shambaugh, 1954).
Ingested bacteria encounter the insect immune system
and can often counteract host defences (Valley-Gely et
al., 2008). There is evidence that most ingested bacteria
are eliminated by mechanisms such as reactive oxygen
species (ROS) and peristalsis, and persistence within the
host requires ingested bacteria to be able to survive the
conditions of the alimentary tract (Valley-Gely et al.,
2008). As more bacteria survive in the midgut, their inte-
ractions with the epithelium molecules may increase and
perturb gut physiology (Jackson et al., 2001). The degree
to which these phenomena can occur may be dependent
on bacterial persistence.
Recently, there has been renewed interest in
characterising insect gut microorganisms because they
are a potential source of novel bioactive compounds,
such as antimalarial, antiviral, and antitumour peptides
(Chernysh et al., 2002) as well enzymes (Zhang and
Brune, 2004). In addition, studies on insect–bacteria
interactions have attracted interest because manipulating
microbial symbionts is thought to be an effective strategy
for controlling the spread of pathogens that use insects
as hosts (Dillon et al., 2005). Although bacterial infection
has been reported to suppress vector competence in
mosquitoes (Pumpuni et al., 1993; 1996; Lowenberger et
al., 1996), there are still no satisfactory candidate
vaccines or environmentally safe insecticides for
controlling dengue (Chaturvedi et al., 2005). This
situation in combination with the observation that
bacterial infection can potentially suppress pathogen
transmission provided an impetus to explore bacterial
infection in A. albopictus. However, one problem faced by
most studies on insect–pathogen interactions is the lack
of means to identify pathogenesis. Here, the location and
persistence of an orally ingested bacterium were in-
vestigated in key parts of the female digestive tract using
fluorescent bacteria because monitoring of fluorescence
allows detection of infection and its intensity can be used
to determine persistence of infection.
MATERIALS AND METHODS
Rearing and producing sterile adult mosquitoes
A. albopictus used here originated from a colony maintained at the
insectarium of Nagasaki University. To establish a new colony at the
Microbiology Laboratory of Fukuoka University for the purpose of
this study, eggs from the Nagasaki colony were hatched in cool
boiled water and newly hatched larvae were raised at a density of
100 per plastic tray (27 × 36 × 6 cm) filled with 2.5 L of sterilised
and dechlorinated water. They were fed daily with a mixture of
powdered mouse pellet diet (CLEA Japan, Inc., Tokyo, Japan) and
dry yeast (1:1) (Tanabe Seiyaku Co. Ltd., Osaka, Japan). Pupae
were transferred into sterilised plastic dishes, and emerging adults
placed in cages (20 x 20 x 30 cm) were given access to a 3%
sucrose solution. Three-day-old females were blood-fed on
immobilised mice. Eggs were dried under rearing laboratory
conditions (27 ± 2.0°C, 70 ± 10% RH and 12:12 h photoperiod) and
were kept at room temperature. For aseptic production of adults for
the experi-ments, all instruments used for rearing were sterilised
and wiped frequently with isopropanol prior to use. 100 newly
hatched larvae were reared in covered plastic trays. Sterilised food
was supplied to larvae at 2 g per day. The pupae were transferred
to sterilised plastic dishes filled with sterilised and dechlorinated tap
water. Emerging adults were placed in cages without food prior to
experimental use.
Construction of GFP-expressing E. coli
The plasmid vector pGFPuv (Clontech Laboratories, Inc., Mountain
View, CA) was transformed into E. coli DH5 (Takara Bio Inc. Shiga,
Japan) by electroporation using a Gene Pulser (Bio-Rad Laborato-
ries, Inc., Hercules, CA). The transformants were cultured in Luria-
Bertani rich nutrient (LB) medium overnight at 37°C. All reagents for
bacterial culture were purchased from Wako Pure Chemical
Industries Ltd. (Osaka, Japan).
Feeding A. albopictus with GFP-expressing E. coli-infected
meal
Recombinant E. coli carrying the plasmid pGFPuv were grown in LB
medium supplemented with isopropyl -D-1-thiogalactopyranoside
(IPTG) at 37°C until the optical density at 660 nm (OD660) reached
4.0. Aliquots of 5 mL of the bacterial cultures were centrifuged at
5000 rpm for 10 min at 4°C. The resulting pellets suspended in 5 ml
of sterilised 3% sucrose solution served as the experimental meal.
A total of 50 three-day old A. albopictus females were transferred
into the cages and allowed free access to the infected meal. Adults
were further sampled and feeding status was confirmed under a
dissecting microscope based on expansion of the abdomen. Fed
individuals were maintained without food in sterilised glass test
tubes until dissection.
Dissection of crop and midgut
Dissection was performed under laminar air flow to avoid potential
contamination. The fed mosquitoes were divided into two groups:
those in the first group were dissected immediately after feeding (0
h), while those in the second group were dissected 24 h after the
infected meal. Following the method reported by Pumpuni et al.
(1996), individuals were anaesthetised with CO2 and wiped with
70% ethanol before dissection under a binocular microscope (Nikon
SMZ-1B; Nikon, Tokyo, Japan). Only successful dissections, de-
fined as leaving the crop and midgut intact, were considered in the
analyses. Body carcasses containing an undamaged crop and
midgut were irradiated with UV (366 nm) on a UV transilluminator
(Funakoshi Co. Ltd., Tokyo, Japan) and photographed. A similar
group of mosquitoes were used to confirm the presence of bacteria
at different time’s post-ingestion of the infected meal. A group of fed
individuals were dissected at 1, 3, 5, 7 and 24 h. Those dissected
at t = 24 h will have had longer to interact with the meal than their
counterparts dissected at t=1 h.
The crop and midgut were then placed in 1.5 mL plastic tubes
with 200 l of physiological saline solution and homogenised with
an additional 800 l of the same solution. The homogenates were
diluted 1×, 10×, and 100× and aliquots of 100 l from each dilution
were spread on LB agar plates and incubated at 37°C overnight.
Data collection and analysis
Photographs of carcasses showing the intact crop and midgut were
taken with a digital camera fixed to a binocular microscope with a
UV lamp (Model UVL/Blak-Ray UV Lamp, long wavelength UV 366
nm; UVP, San Gabriel, CA) as the light source. At least 4 body car-
casses were photographed at 0 and 24 h post-feeding, and if
visualised by UV were considered present if observed in all photos.
Bacteria were scored as present if consistent images were obtained
in all photos. The homogenates were diluted as described above
and three replicates of each dilution were spread on Luria-Bertani
medium (LB). To estimate the number of GFP-expressing E. coli
present, colonies that showed luminescence under UV illumination
were counted 48 h post-feeding. For each homogenate, the mean
value of the number of bacteria was calculated and expressed as
colony forming units per ml (cfu/mL). The midgut homogenates
were treated in a similar manner. This procedure revealed the
number of bacteria for both the crop and migdut throughout the time
course of the experiment. The mean bacterial numbers were
calculated and expressed as the mean ± SE (standard error). The
proportion of bacteria disappearing from each organ at each time
point was calculated using the formula: mean number of bacteria at
a given time point/mean number of bacteria at 1 h post-feeding ×
100. Statistical analyses were performed using Analysis of Variance
and Tukey’s HSD multiple comparisons with the SYSTAT®11
software package (SYSTAT®11 DATA, 2004).
RESULTS
Analysis of organ photographs
Visual analysis of photographs of the carcasses of A.
albopictus females just after feeding (0 h) with the infec-
ted meal indicated the presence of the green fluorescent
marker in the crop and midgut in most cases (Figure 1A).
In contrast, no luminescence of GFP was observed in
the midgut or crop on photographs from those dissected
24 h post-feeding (Figure 1B). In some cases, the
Malpighian tubules also showed fluorescent labelling
(Figure 1A).
Bacterial density
The time post-feeding, but not the organ type,
significantly affected the number of bacteria (Table 1).
Satho et al. 015
The bacterial population present in the midgut showed a
shorter period of persistence than that in the crop (Figure
2). In both organs, the size of the bacterial population
tended to decrease over time. In the midgut, the number
of bacteria differed significantly between 1, 3, and 5 h (F
= 123.23, df = 4, p < 0.001), while a pairwise comparison
probability matrix did not reveal any differences between
population sizes at 5, 7 and 24 h. In the crop, the
bacterial popula-tion size decreased significantly between
1 and 3 h post-feeding. Thereafter, there was a steady
decrease of that population about ten times less cfu/ml
are found after 7 hours than at after 3 hours (F = 9.30, df
= 4, p < 0.001). The bacterial population size was similar
between the two organs at 1 (F = 0.38, df = 1, p = 0.540)
and 3 h (F = 0.28, df = 1, p = 0.601) post-feeding. However,
at 5 (F = 9.92, df = 1, p = 0.004), 7 H (F = 12.81, DF = 1, p =
0.001), AND 24 H (F = 4.19, DF = 1, p = 0.05) after
ingesting the infected meal, the size of the bacterial
population in the midgut was significantly smaller than
that in the crop. The poor persis-tence of transgenic E.
coli in the midgut may be related to the noxious contents
of physiological activity in the lumen.
The data shown in Table 2 indicate variable patterns of
bacterial disappearance from the crop and the midgut.
Both organs showed a marked decrease in size of the E.
coli population between 1 and 3 h post-ingestion. Further,
the sizes of the bacterial populations decreased slowly, and
at 5 h post-feeding both the crop and the midgut still
harboured bacteria. At 7 h post-feeding, the crop still har-
boured a residual bacterial population, whereas this was
completely absent in the midgut.
DISCUSSION
Preliminary trials on the transformation of E. coli using E.
coli DH5 and the expression plasmid vector pGFPuv
produced viable bacteria characterised by the presence
of green fluorescence under UV illumination. These
bacteria were maintained at -20°C and sub-colonies still
showing luminescence after months. We assume that the
fluorescence actually measures the GFP. Such sub-
colonies were mixed with sucrose solution to prepare the
experimental meal for A. albopictus.
Most females showed fluorescence in the midgut and
crop following meal intake, and the marker had dis-
appeared completely from the epithelium in both organs
after 24 h. There have been few studies of time-related
bacterial persistence in mosquitoes. St. John et al. (1930)
reported that ingested Staphylococcus aureus, Bacillus
prodigiosus, Bacillus leprae and Cytoryctes variolae
mixed with blood remained in the A. aegypti midgut for
more than 24 h. In a related study, Pumpuni et al. (1996)
fed a population of A. stephensi a 1% sucrose solution
containing Xanthomonas maltophilia, Serratia marce-
scens or Pseudomonas aeruginosa. They reported reten-
tion of the two first bacteria in the midgut after 24 h in
75% of cases, whereas only 19% showed persistence of
016 J. Parasitol. Vector Biol.
Figure 1. Carcasses of A. albopictus females showing the presence of the green fluorescent marker in
the crop and midgut just after (A) feeding with the sugared meal contaminated with the transgenic E.
coli and its absence 24 h later (B).
Table 1. Analysis of Variance of the effects of organ type and
time on bacterial population size of GFP E. coli within the
digestive tract of A. albopictus
Source variables df F-ratio P
Organ
1 0.20 0.65
Time post-feeding (TPF)
1 40.81 0.000
Organ × TPF 1 11.67 0.74
Crop and Midgut;
0 h and 24 h; x=Interaction.
Pseudomonas aeruginosa. These reports suggested that
the period of bacterial retention in the midgut is depen-
dent on the species of bacteria and/or mosquito, but this
may not be the sole reason for disappearance of bacteria
from the insect alimentary canal.
The results obtained using the homogenates in the
present study indicated that the period of persistence of
ingested bacteria was much shorter in the midgut, in
which the number of bacteria disappeared almost
completely after 5 h post-ingestion of the infected meal.
In the crop, however, although the size of the bacterial
population decreased in the initial 3 h after meal ingestion,
it remained at a relatively constant level thereafter. This
difference in retention between the crop and midgut may
be related to the difference in physiological roles of the
epithelium between the two organs. As in any epithelial
tissue, the physiological roles are derived from cell func-
tion and the degree at which they vary depends largely
on the number and proportion of different cell types.
Indeed, an epithelium with a limited number of cells will
tend to have reduced physiological roles as compared to
another with a larger cell number. The crop epithelium is
composed of unspecialised cells, which therefore have
reduced potential to impact the ingested meal. Although
this epithelial tissue has been reported as a site of partial
hydrolysis of disaccharides, it is important to realise that
this action was due to a salivary enzyme (Clements,
1992). In contrast, the midgut contains rege-nerative,
endocrine, and columnar cells, which account for much of
its epithelium. The functions of this organ are well
established: it is the principal organ of digestion and thus
the site of synthesis of digestive enzymes, including
trypsin, proteinases, peptidases, esterases, and glucosi-
dases. In addition to its digestive function, the midgut is
the site of nutrient absorption, hormonal synthesis
(Clements, 1992), infection processes and defensive
responses (Wang et al., 2001). Although morphologically
different, the midgut and the crop share a similar
physiological function, that is, storage of sugar-containing
Satho et al. 017
A B C D E
Time post-feeding (hours)
1
10
100
1000
Log bacterial density (cfu/ml)
a
b
c
c
c
1
2
2
2
2
Crop
Midgut
A = 1 h
B = 3 h
C = 5 h
D = 7 h
E = 24 h
Figure 2. E. coli populations in homogenates of the crop and midgut of A.
albopictus females fed the sugared meal contaminated with transgenic E. coli
and allowed to interact with the meal at different time points. The vertical axis
of this figure shows the mean bacteria density on a log10-based scale.
Table 2. Proportions of E. coli cells that disappeared from the crop and midgut
of A. albopictus. The meal was taken only once, and that the fate of the
bacteria inside the crop and midgut was examined at different time points after
feeding. By ANOVA, bars of the color and with the same letter do not show a
significant difference (P < 0.05).
% disappearance Mean ±SE bacterial cells
at 1 hour post-feeding 3 h 5 h 7 h 24 h
Crop 1389.5 ± 390.83 81.2 97.8 99.1 99.75
Midgut 1692.2 ± 1 69.64 87.6 99.96 99.99 100
meals. These two organs have been shown to interact
during digestion processes in A. aegypti, a species
closely related and ecologically similar to A. albopictus
studied here (Hawley, 1988). Sugar solu-tion ingested by
this mosquito has been detected in the midgut 30 min
after filling the crop (Jones and Brandt, 1981). The
difference in bacterial persistence between the crop and
midgut seen in the present study was likely due to
discrepancies in their potential actions on the ingested
meal. In addition to its suggested hydrolytic role, the crop
is also known to release its contents in small amounts
into the midgut (Clements, 1992). Here, ingested meals
and any associated micro-organism may be exposed to
lumen conditions, such as pH, digestive enzymes, ionic
strength, ROS, and the immune system (Valley-Gely et
al., 2008), which may alter the bacterial population in the
midgut.
In addition to the severe midgut conditions, the bacte-
rial population associated with an ingested meal may be
affected by post-midgut physiological processes as food
residue undigested in the midgut moves to the hindgut
and is excreted (Clements, 1992; Chapman, 1998).
These phenomena may have occurred in the present
study, but no experiments were performed to determine
whether bacteria were present in the excreta of the
mosquitoes. Therefore, further studies should include
monitoring of the post-midgut bacterial populations.
The results of the present study demonstrated that the
crop and midgut differ fundamentally in their interactions
with sugar-containing meals, despite their similar status
as storage organs. Another interesting result of the
present study was the ease of detection of bacterial infec-
tion, which was made possible by the use of a fluorescent
marker.
018 J. Parasitol. Vector Biol.
There are, however, three possible sources of error in
our approach. First, we did not control for loss of plasmid
by the E. coli, without a selective preasure the E. coli
might have lost the gfp-expressing plasmid and still
undetected in the mosquitoes after 24 h. Secondly, the
bacteria culture used for the experimental meal was
grown to OD660. In reference to the OD600 used in a pre-
vious study (Riehle et al., 2007), one may consider that
our OD is overgrown and may result in many dead and
ruptured bacteria. The cell material from these dead
bacteria might trigger immune defences in the mosquito
that would not happen in the intact bacteria cells intact.
This in turn might have lead to a faster clearence of the
bacteria from the midgut. Finally, after spreading the
diluted homogenates on LB plates, no deliberate efforts
were performed to count the colonies that did not show
any luminescence. According to Riehle et al. (2007), the
number of fluorescent colonies on each plate represents
one half of the total E. coli population for an individual
mosquito midgut.
Further studies of bacteria-Aedes interactions may
facilitate the use of such microorganisms to control
disease transmission by arthropod vectors. Therefore,
further studies involving one or prefereably several
bacteria isolated from A. albopictus or other Aedes mos-
quitoes, if a known insect pathogen, are required.
ACKNOWLEDGMENTS
The authors thank Dr. Takagi (Department of Vector Eco-
logy and Environment, Nagasaki University, Japan), and
the staff of the Department of Microbiology, Faculty of
Pharmaceutical Sciences, Fukuoka University, Japan.
REFERENCES
Chapman RF (1998) The Insects: Structure and Function. Cambridge
University Press, Cambridge.
Chastel C (2002) Impact of global climate changes on arboviruses
transmitted to humans by mosquitoes and ticks. Bull. Nat. Acad. Med.
186: 89-100.
Chaturvedi UC, Shrivastava R, Nagar R (2005) Dengue vaccines:
problems and prospects. Ind. J. Med. Res. I2l: 639-652.
Chernysh S, Kim SI, Bekker G, Pleskach VA, Filatova NA, Anikin VB,
Platonov VG, Bulet P (2002) Antiviral and antitumor peptides from
insects. Proceedings of the Nat. Acad. Sci. USA 99: 12628-12632.
Clements AN (1992) The biology of mosquitoes. Chapman and Hall,
London.
Dharne M, Patole M, Shouche YS (2006) Microbiology of the insect gut:
tales from mosquitoes and bees. J. Biosci. 31: 293-295.
Dillon RJ, Dillon VM (2004) The gut bacteria of insects: nonpathogenic
interactions. Ann. Rev. Entomol. 49: 71-92.
Dillon RJ, Vennard CT, Buckling A, Charnley AK (2005) Diversity of
locust gut bacteria protects against pathogen invasion. Ecol. Lett. 8:
1291-1298.
Fisk FW, Shambaugh GF (1954) Invertase activity in Aedes aegypti
mosquitoes. The Ohio J. Sci. 54: 237-239.
Gubler DJ (2004) The changing epidemiology of yellow fever and
dengue, 1900 to 2003: full circle? Comparative Immunol. Microbiol.
Infect. Dis. 27: 319-330.
Hawley WA (1988) The biology of Aedes albopictus. J. Am. Mosquito
Control Assoc. 4(1): l-39.
Jackson TA, Boucias DG, Thaler JO (2001) Pathobiology of amber
disease, caused by Serratia spp., in the New Zealand grass grub,
Costelytra zealandica. J. Invertebrate Pathol. 78: 232-243.
Jones JC, Brandt E (1981) Fluid excretion by adult Aedes aegypti
mosquitoes. J. Insect Physiol. 27: 545-549.
Lowenberger CA, Ferdig MT, Bulet P, Khalili S, Hoffmann JA,
Christensen BM (1996) Aedes aegypti: induced antibacterial proteins
reduce the establishment and development of Brugia malayi. Exp.
Parasitol. 83: 191-201.
Mitchell CJ (1995) The role of Aedes albopictus as an arbovirus vector.
Proceedings of the Workshop on the Geographic Spread of Aedes
albopictus in Europe and the Concern among Public Health
Authorities; 1990 Dec 19-20; Rome, Italy. Parassitol. 37: 109-113.
Pumpuni CB, Beie, MS, Nataro JP, Guers LD, Davis JR (1993)
Plasmodium falciparum: inhibition of sporogonic development in
Anopheles stephensi by gram-negative bacteria. Exp. Parasitol. 77:
195-199.
Pumpuni CB, Demaio J, Kent M, Davis JR, Beier JC (1996) Bacterial
population dynamics in three anopheline species: The impact on
Plasmodium sporogonic development. Am. J. Trop. Med. Hyg. 54:
214-218.
Riehle MA, Moreira CK, Lamp D, Lauzon C, Jacob-Lorena M (2007).
Using bacteria to express and display anti-Plasmodium molecules in
the mosquito midgut. Int. J. Parasitol. 37: 595-603.
Sota T, Mogi M, Hayamizu E (1992) Seasonal distribution and habitat
selection by Aedes albopictus and Ae. riversi (Diptera: Culicidae) in
northern Kyushu, Japan. J. Med. Entomol. 29: 296-304.
St. John JH, Simmons JS, Reynolds FHK (1930) The survival of various
microorganisms within the gastro-intestinal tract of Aedes aegypti.
Am. J. Trop. Med. Hyg. 10: 237-241.
Straif SC, Mbogo CN, Touré A M, Walker ED, Kaufman M, Touré YT,
Beier JC (1998) Midgut Bacteria in Anopheles gambiae and An.
funestus (Diptera: Culicidae) from Kenya and Mali. J. Med. Entomol.
35: 222-226.
Systat®11 Data (2004) Systat for Windows: Statistics, Systat Software,
Inc., San Jose p.265.
Valley-Gely I, Lemaitre B, Boccard F (2008) Bacterial strategies to
overcome insect defences. Nat. Rev. Microbiol. 6: 302-313.
Wang P, Conrad JT, Shahabuddin M (2001) Localization of midgut-
specific protein antigens from Aedes aegypti (Diptera: Culicidae)
using monoclonal antibodies. J. Med. Entomol. 38: 223-230.
Zhang H, Brune A (2004) Characterization of partial purifi cation of
proteinases from highly alkaline midgut of the humivorous larvae of
Pachnoda ephippiata (Coleoptera: Scarabaeidae). Soil Biol. Biochem.
36: 435-442.
