WOLBACHIA-MEDIATED POPULATION REPLACEMENT IN DENGUE MOSQUITO
VECTORS
By
Shawna Ryan

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Entomology - Master of Science
2013

ABSTRACT
WOLBACHIA-MEDIATED POPULATION REPLACEMENT IN DENGUE MOSQUITO
VECTORS
By
Shawna Ryan
There are many vector-borne diseases, including dengue, that lack vaccines or effective
treatment options, resulting in vector control being the primary disease control strategy.
Wolbachia, an intracellular bacterium that can spread itself through a population via cytoplasmic
incompatibility (CI), has been shown to inhibit the transmission of a number of the deadly
human pathogens, like dengue and Plasmodium, in mosquitoes. In order to utilize Wolbachia to
make mosquitoes inhospitable to the pathogens, we have to create a more efficient population
replacement strategy such that disease transmission can be interrupted completely and rapidly. In
this work, we performed Aedes aegypti laboratory cage studies in which the Wolbachia-infected
females were released once, at the beginning, followed by continued inundative infected male
release at every generation. We found that this inundative male release could accelerate the
process of population replacement. We also designed a new mathematical model that is capable
of accurately predicting the generation in which population replacement will occur. To develop a
population replacement strategy for Aedes albopictus we introduced the third type of Wolbachia,
wPip, into this mosquito species to create a transinfected line carrying a triple Wolbachia
infection. We characterized the pattern of CI induced by this novel artificial infection through
crossing assays. We found that the triply infected Ae. albopictus induces unidirectional CI when
crossed with the wild type doubly infected mosquitoes, supporting its potential to be used in
population replacement study.

Copyright by
SHAWNA RYAN
2013

ACKNOWLEDGEMENTS

I would like to begin by acknowledging my advisor, Dr. Zhiyong Xi, for providing the
opportunity to work on this project and for all of his help and support. In the same way, I want to
thank my graduate committee members, Drs. Michael Kaufman and Moxun Tang, for all of their
guidance, support, and assistance. I would like to give a special thank you to Dr. Tang for all of
his help in the design of the mathematical model. Thank you to all of the lab members for their
help, support, and for being willing to assist me in furthering my laboratory skills. Thank you to
the Entomology department staff for helping to keep me on track to graduate and for answering
any questions I may have had. I greatly appreciate the help provided by CSTAT consultant
Wenjuan Ma. Finally, thank you to all of my friends and family for their editing, support, and
advice while I was writing this thesis.

iv

TABLE OF CONTENTS

LIST OF TABLES…..…………………………………………………………………………viii
LIST OF FIGURES.....………………………………………………………………………….ix
Chapter 1 An introduction to Wolbachia, dengue, cytoplasmic incompatibility, and
population replacement …...…………………………………….……..………………………..1
Chapter 2 Increasing the rate of population replacement in Ae. aegypti mosquitoes ..........10
Introduction..……………………………………………………………………………10
Materials and Methods..………………………………………………………………..11
Rearing Practices………………………………………………………………...11
Blind Tests ………………………………………………………………………12
Population Cage Setup…………………………………………………………..13
Population Replacement experimental Design …………………………………14
Polymerase Chain Reactions ...………………………………………………….15
DNA Extraction and Incubation ………………………………………...15
PCR Assay ………………………………………………………………15
Mathematical Modeling …………………………………………………………16
Results ...………………………………………………………………………………...18
Blind Tests ………………………………………………………………………18
Population Replacement 1 ………………………………………………………20
Mathematical Modeling …………………………………………………………22
Discussion ...…………………………………………………………………………….23
Blind Tests ………………………………………………………………………23
Population Replacement 1 ………………………………………………………25
Mathematical Modeling………………………………………………………….26
Chapter 3 Reducing the initial female release frequency threshold for population
replacement …………………………………………………………………………………….27
Introduction ...…………………………………………………………………………..27
Materials and Methods..………………………………………………………………..28
Population Replacement experimental design…………………………………28
Mathematical Modeling …………………………………………………………29
Results ...………………………………………………………………………………...31
Discussion……………………………………………………………………………….33
Chapter 4 Characterization of Cytoplasmic incompatibility Phenotypes with Triple
infected Aedes albopictus .……………………………………………….……………………..35
Introduction...…………………………………………………………………………...35
Materials and Methods………………………………………………………………...36
Results...…………………………………………………………………………………39
Discussion...……………………………………………………………………………..40

v

Chapter 5 Mosquitoes, Wolbachia, and cytoplasmic incompatibility: Conclusions and
future directions ……..…………………………………………………………………………43
Conclusions ……………………………………………………………………………..43
Future Directions ..……………………………………………………………………..44
APPENDIX...……………………………………………………………………………………48
BIBLIOGRAPHY……...…...…………………………………………………………………..50

vi

LIST OF TABLES

Table 2.1 Experimental design of population replacement in the laboratory cages. The Waco
line represents the uninfected line and the WB1 line represents the infected line ……………...14
Table 2.2 The results of the first blind bias assessment. The first column denotes the cage
number. The second and third columns represent the infection rate estimate achieved by taking
10 and 25 samples, respectively. The final column is the actual infection percentage for each of
the three cages………………...………………………………………………………………….19
Table 2.3 The results of the second blind bias assessment. This blind test involves taking 25
samples of female ovaries over the course of 3 days. Ten samples were taken the first day and
five samples were taken on subsequent days………...…………………………………………..19
Table 2.4 The results of the PCR assessment blind test. This blind was run by dissecting all of
the live females in each cage, extracting the DNA, and running PCR. Cages one and three had
15/15 samples dissected and cage two had 14/15 samples dissected (one female died before
dissection) ……………………………………………………………………………………….20
Table 2.5 The percentages represent the number of eggs for each set of mosquitoes that hatched
at each generation. ………………………………………………………………………………21
Table 2.6 The percentages represent the percentage of each sample set that were found to be
infected. The results were obtained using PCR to image a Wolbachia surface protein (WSP)…22
Table 2.7 Table of predicted replacement generations, based on the mathematical model, and
the actual generations during which replacement took place……………………………………23
Table 2.8 The equivalence test for the comparison of each group of samples. For the purpose of
this table, each cage has been assigned a different number (1-9). Numbers 1, 4, and 7 designate
the groups with a sample size of 10. Numbers 2, 5, and 8 designate the groups where an
additional 15 samples were taken after the sample size of 10. Numbers 3, 6, and 9 designate the
final group where 25 samples were taken over the course of several days and as many females as
possible were collected and sub-sampled for each set (extra females were released back into their
cages to be sampled again the next day)…………………………………………………………24
Table 3.1 The experimental design of Population Replacement 2: Reducing the threshold
number of Wolbachia-infected females necessary for the bacterium to achieve fixation within the
population. This table describes the numbers of infected and uninfected mosquitoes released into
each of the experimental cages…………………………………………………………………..29
Table 3.2 The expected generations when the infection frequency has reached 99.9%. The
predicted generations were calculated using the mathematical model (Eq. 3.1 through 3.3)…...32

vii

Table 3.3 Predicted and actual replacement generations achieved using inundative release of
Wolbachia-infected male Anopheles stephensi mosquitoes to reduce the number of infected
female mosquitoes necessary for complete population replacement. …………………………...32
Table 4.1 Crossing patterns and expected CI responses for all possible CI crossings of HT1,
Hou, and HC Ae. albopictus mosquitoes………………………………………………………...38
Table 4.2 The analysis of variance for the Ae. albopictus CI crossing hatch data. This test shows
that there is no significant difference between the average hatch rates produced by each of the CI
crosses……………………………………………………………………………………………40

viii

LIST OF FIGURES

Figure 1.1 A Unidirectional Cytoplasmic Incompatibility occurs when Wolbachia-infected
individuals (Solid symbols) cross with the uninfected individuals (Empty symbols) (Brelsfoard
& Dobson 2009)……………………………………………………………………..…………….5
Figure 1.2 Bidirectional CI occurs when a population is infected by two types of Wolbachia.
Grey: type 1, Black: type 2 (Brelsfoard & Dobson 2009).………………………………………..6
Figure 1.3 An additive CI occurs in the crosses between superinfected and single infected lines.
Solid black: single infection, Solid white: no infection, 1/2 solid and 1/2 patterned: 2 types of
infection (Brelsfoard & Dobson 2009; Dobson, Additive
Incompatibility)….………………………………………………………………………………...7
Figure 4.1 The average hatch rates for all crosses. The *# represents how many replicates of
each cross were run. The HC line was in its eighth generation during this experiment…………39
Figure 4.2 The one way analysis of the asin proportion of hatchings by crossing showing that
none of the hatchings are significantly different from each other……………………………….40

ix

Chapter 1
An introduction to Wolbachia, dengue, cytoplasmic incompatibility, and population
replacement
Insect-borne diseases are the leading causes of illness and death in the world today
(Iturbe-Ormaetxe et al 2011). The case numbers increase with the expansion of globalization and
human travel. Even with inspection protocols in place, it is impossible to catch every single nonnative and possible pathogen carrying insect as it enters the country. This leaves every country
that imports and exports goods, or has an international travel industry vulnerable to exotic
vector-borne diseases. Particularly dangerous are those diseases, such as dengue virus, that have
no vaccine and no treatment options.
Approximately 2.5 billion people are living in areas at risk for dengue infection and there
are at least 50-100 million cases of dengue fever and 500,000 cases of dengue hemorrhagic fever
(DHF) reported each year (Gubler and Clark 1995; WHO 2009). However, even these numbers
are conservative due to the probability of a large number of cases remaining unreported and the
limited availability of data for those that are reported (Gubler 1998). Reports of DHF have
increased by an average of 5-fold on a global scale, with the majority in Asia (Gubler 2002). Due
to a lack of vaccines and treatment options, vector control is a primary dengue control strategy.
Dengue virus, with four serotypes that are all transmittable to humans, can infect all ages
and races of people. The common symptoms of dengue fever are mostly similar to the influenza
virus, including a high fever accompanied by some of the following: chills, pain, nausea,
vomiting, headache, and a rash (Cobra et al 1995, WHO 2012). The symptoms usually last for
approximately one week and begin between 4-10 days after bitten by the infected mosquito
1

(World Health Organization 2012). The disease becomes fatal when it develops into dengue
hemorrhagic fever. This form of the disease has symptoms of nausea, vomiting, headache,
abdominal pain, a fever that lasts 2-7 day, easily bruising, bleeding from the gums and nostrils,
skin hemorrhaging, and possible internal bleeding (CDC 2012). Dengue Shock Syndrome (DSS),
in addition to the symptoms of dengue fever and dengue hemorrhagic fever, causes circulatory
failure and death (WHO 2009).
Ae. aegypti and Ae. albopictus are two important disease vectors for dengue virus. Ae.
aegypti, also known as the Yellow Fever Mosquito, is the primary vector of dengue fever and
dengue hemorrhagic fever. This mosquito originated in Africa and likely began to spread through
the world via the slave trade and it continued to spread through various trading routes and wars
(Mousson et al. 2005). Ae. aegypti is very well adapted to living in urban areas. They prefer
stagnant water sources for their larvae, so man-made containers (like buckets, tires, etc.) are
perfect habitats for them. As of 2001, Ae. aegypti has spread across parts of Asia, Africa, South
America and North America (Gubler 2002). It is now found in tropical areas around the world
(Mousson et al. 2005).
Ae. albopictus, also known as the Tiger Mosquito, is another mosquito species that has
readily adapted to urban habitats. This species can breed in both man-made and natural
containers (Knudsen 1995). It originated in South-East Asia and spread through the world via
trade routes and human travel (Mousson 2005). Until now, this mosquito species has been found
on six of the seven continents and is rapidly spreading; Antarctica is the only continent lacking
the species due to climate conditions (Paupy et al 2009).

2

Both societal and environmental stumbling blocks have caused issues with the
implementation of many vector control strategies. In the case of pesticides to control the vector
population, there are increasing problems in development of resistance (Lenormand et al. 1999;
Chareonviriyahpap et al. 1999; Rose 2001), possible negative effects on non-targeted species,
and potential toxicity to vertebrates. These have been observed for the insecticides including
Temephos, Methoprene, Permethrin, DDT, etc. (Rose 2001). There are several problems with the
sterile male release strategy. It is not self-sustainable and it is work-intensive, as the males must
be released every generation or the natural population will be able to build up again. In addition,
it is not always effective. It is possible for some males that are released to be fertile or to be too
weak to properly mate with the females. Transgenic mosquito technology is facing challenges of
containment, risk management, regulation approval, fitness costs (Irvin et al. 2004; Sperança and
Capurro 2007) and the capability of the transgene to spread itself through a population
(Crampton et al. 1990; Scott et al. 2002). These issues cause many of these options to be
impractical for long-term use (Kyle and Harris 2008).
Despite the problems associated with many of the control strategies, each approach has
its unique advantage compared to the others, resulting in a possibility to develop an integrated
vector control strategy. Pesticides, for example, are generally fast acting and, while slower, with
proper application sterile males can be an effective method of control for disease vectors (Esteva
and Yang 2005). Transgenic mosquitoes that are altered to be resistant to the diseases they
transmit have a great deal of potential use in disease control projects (Marshall and Taylor 2009).
Release of incompatible males for population suppression has been shown to be effective in
eliminating a population (Laven 1967). Wolbachia–mediated population replacement, due to its

3

self-sustaining nature, use of naturally occurring bacteria, and use of the native species, can be
considered a far more appealing option for vector control strategies.
Wolbachia is a genus of obligate reproductive endosymbiotic bacterium related to
Rickettsia (Werren and Windsor 2000). Found in all of the major insect orders (Werren and
Windsor 2000), Wolbachia is estimated to be present in approximately 66% of insect species
(Hilgenboecker et al. 2008). Wolbachia can be transmitted vertically, from mother to offspring.
The bacterium is associated with cytoplasmic incompatibility (CI), parthenogenesis, feminization
of males, decrease in adult life span, and male killing (Iturbe-Ormaetxe et al 2011, Werren and
Windsor 2000; Werren 1997). Recently, Wolbachia was found to inhibit the transmission of
several types of human pathogens within the vector (Moreira et al. 2009; Bian et al. 2010)
Wolbachia has great potential in the control of dengue because it can spread into
mosquito population through CI and cause the mosquitoes to be refractory against dengue virus.
In the CI cross, a Wolbachia infected male mates with a female that is uninfected or one lacking
the same strain of Wolbachia, resulting in few to no viable offspring (Jansen et al. 2008;
Hoffmann et al. 1990). We do not fully understand the mechanism by which CI occurs, but one
model to explain it is that Wolbachia function through a modify/rescue system. In this model, the
sperms are modified by Wolbachia in the infected males such that they cannot fertilize the eggs,
unless the same type of Wolbachia presents in the eggs to rescue this sperm modification
(Zabalou et al. 2008, Dobson 2004). In a simple unidirectional CI, Wolbachia provides
advantage to infected females such that it can invade into the uninfected population (Figure 1.1).
An example of this is seen in the single infected Ae. aegypti line generated via microinjection (Xi
et al. 2005b).

4

Figure 1.1: A Unidirectional Cytoplasmic Incompatibility occurs when Wolbachiainfected individuals (Solid symbols) cross with the uninfected individuals (Empty symbols)
(Brelsfoard & Dobson 2009).
It gets more complex when the host is infected with two or more Wolbachia types. This
can lead to bidirectional CI and additive unidirectional CI. Bidirectional CI and additive CI are
types of CI involving at least two different Wolbachia infections. In bidirectional CI the
population has some individuals infected with one type of Wolbachia and some individuals that
are infected with another type of Wolbachia (Figure 1.2). This type of CI results in
approximately a fifty percent chance of compatible matings and a fifty percent chance of
incompatible matings in the population. Additive CI occurs when it is possible to have
individuals that are infected with two or more Wolbachia types at the same time and within the

5

same mosquito (Figure 1.3). (Dobson Lab Web site (citation))

Figure 1.2: Bidirectional CI occurs when a population is infected by two types of Wolbachia.
Grey: type 1, Black: type 2 (Brelsfoard & Dobson 2009).

6

Figure 1.3: An additive CI occurs in the crosses between superinfected and single infected lines.
Solid black: single infection, Solid white: no infection, 1/2 solid and 1/2 patterned: 2 types of
infection (Brelsfoard & Dobson 2009; Dobson, Additive Incompatibility).
Using Wolbachia infected males to induce CI is a possible alternative to the sterile insect
technique (SIT), also referred to as population suppression or Incompatible Inset Technique
(IIT). SIT is often induced by irradiating males to kill the sperm which results in a lack of viable
offspring (White, Rohani, and Sait 2010; Weeks, Turelli, Harcombe, Reynolds, Hoffmann 2007).
However, irradiation can have serious fitness cost for the treated insects. This includes increased
mortality and decreased longevity, greatly lowering the probability of mating success. SIT can
also result in other fitness costs and incomplete sterility (Helinski, Parker, and Knols 2009). As

7

another advantage over the sterile male technique, the Wolbachia infection persists throughout
generations, and is thus a self-sustaining alternative to repeated use strategies.
Wolbachia can convey a resistance to several different diseases, including dengue virus
in its mosquito hosts (Hedges 2008; Mousson et al. 2012; Teixeira et al. 2008; Moreira et al.
2009; Bian et al. 2010). By conveying this resistance it is possible to block dengue transmission
to humans. Thus, the various types of Wolbachia are viable options for disease control if they
can be established within a population. The bacterium, despite not wiping out or greatly reducing
the mosquito population in the release area like the sterile male technique or IIT, still reduces
disease transmission. A method for using CI to drive Wolbachia into a population is referred to
as population replacement. Population replacement is where a naturally disease susceptible
population is replaced by a population that is resistant to pathogens. Population replacement
studies in Ae. aegypti benefit from the fact that the wild population is not infected with
Wolbachia and thus avoid a need of complicated system. Establishment of a stable single
Wolbachia infection allows CI to take place in the pattern shown in Figure 1.1. There is
approximately a 50% chance that the offspring will be infected with Wolbachia, a 25% chance
that the offspring will be uninfected, and a 25% chance of CI occurring and causing the embryo
to fail to form. In the case of Ae. albopictus, the system is much more complicated because the
two types of Wolbachia, wAlbA and wAlbB, are present in the natural populations. For CI or
population replacement to occur this line requires either a third Wolbachia type, such as wPip
which is derived from Culex pipiens mosquitoes (Hertig 1936), to be added or the current
Wolbachia types to be cleaned from the mosquito line and replaced with a novel infection type.
Population replacement has been shown to be possible in both laboratory studies (Xi et al.
2005b) and in field studies (Turelli and Hoffmann 1991 and Hoffmann et al. 2011). This is
8

extremely important due to the disease limiting capabilities of Wolbachia. As mentioned earlier,
the transmission of dengue virus can be limited and, potentially stopped, by Wolbachia. The key
is how to effectively drive the bacterium into a target population and establish a permanent and
stable infection.
Typically embryonic microinjection is used to establish a novel Wolbachia-host
association (Xi et al. 2005a). At an early embryonic development stage, Wolbachia is transferred
from the infected donors into the posterior area of recipients, followed by intensive screening by
PCR to identify a line with a stable symbiosis formed. A number of stable infected lines carrying
different infection types have been generated using this method, including Ae. aegypti with
wAlbB (Xi et al. 2005b), wMelPop (McMeniman et al. 2009), and wMel (Walker et al 2011),
Anopheles stephensi with wAlbB (Bian et al. 2013), Ae. albopictus with wRi (Fu et al. 2010 and
Xi et al. 2006), wPip (Calvitti et al. 2010), and wMel (Blagrove et al. 2012). All these
transinfected lines maintain perfect or nearly perfect maternal transmission, and induce strong CI
when crossed with individuals that are either uninfected or infected by a different type of
Wolbachia (Xi et al. 2005b; McMeniman et al. 2009; Walker et al 2011; Bian et al. 2013; Fu et
al. 2010; Xi et al. 2006; Calvitti et al. 2010; Blagrove et al. 2012).
There were two main focuses in my research. The first focus of my thesis studies was to
determine the optimum conditions necessary for an effective population replacement in Ae.
aegypti. The second focus was to determine the potential of a novel Wolbachia infection to
invade into Ae. albopictus.

9

Chapter 2
Increasing the rate of population replacement in Ae. aegypti mosquitoes
Introduction
Ae. aegypti (L.) (Diptera: Culicidae) mosquito is a primary vector of dengue virus in
tropical regions. There have been some moderately successful attempts at controlling Ae. aegypti
and innovative approaches continue to be developed. One method to control Ae. aegypti
populations is the elimination of its habitat, particularly in urban areas. Eggs often hatch in
stagnant water where there are many artificial water sources (e.g. buckets, tarps with water pools,
etc.), a reduction or elimination of these habitats can significantly decrease Ae. aegypti
density(Chan et al. 1971). Another common approach to control this mosquito species is ultralow volume (ULV) application of insecticides (Gubler and Clark 1996). It is well known that,
while providing partial control of these pests, insecticides can cause the target organisms to
development resistance. This makes future control of the pest far more difficult. The use of
microorganisms, such as reproductively disruptive bacteria, has become a way to assist in
preventing the mosquito from transmitting the disease. Specifically, large efforts have made to
develop bacteria within the genus Wolbachia to control vector-borne diseases through population
replacement, such that disease susceptible mosquitoes could be modified to become disease
resistant individuals (Xi et al. 2005 a & b; Iturbe-Ormaetxe et al 2011).
In previous studies the replacement strategies require the release of a number of infected
females that equates to at least 20% of the entire population (Xi et al. 2005b). How long it takes
for Wolbachia to spread into the populations partially dependent on the size of the overall
population (Jansen et al. 2008) and the number of infected individuals that are released. For
example, with a 20% release of Wolbachia infected females (calculated based on the number
10

infected females released compared to the total number of females released into a cage),
population replacement takes seven generations (Xi et al. 2005b). However, when the initial
female release frequency is below a certain threshold, population replacement is likely to fail (Xi
et al. 2005b).
The objective of the first study was to increase the rate of population replacement. If the
replacement can be accelerated, then the use of Wolbachia as a method to control the spread of
certain vector-borne diseases will be more effective in field applications. The hypothesis is that
additional release of Wolbachia-infected males during the spread of Wolbachia into a population
will accelerate population replacement to occur.
Materials and Methods
Rearing Practices
We used the Waco line of mosquitoes (Ae. aegypti), originating from Texas, for our
experiments. The Wolbachia infected line of mosquitoes (referred to hereafter as WB1) was
generated at the University of Kentucky in 2005 (Xi et al. 2005b). The cages used in this study
are made of a light weight aluminum frame (Size 12x12x12” W:H:L Bioquip, Rancho
Dominguez, CA, USA) and with hard mesh covering four sides, while the front of the cage is
covered in a cloth netting that can be used to seal the cage. The top of the cage can be opened
and closed, and is held closed with two metal latches.
All mosquito eggs were stored at a constant temperature and moisture before hatching.
The wild type (Waco) and infected type (WB1) eggs were stored for no more than six and three
months, respectively because storage for a longer amount of time may result in a dramatic
decline in the hatch rate. The eggs were laid on egg paper over the course of three days. The egg
papers are pieces of water absorbent brown paper that was cut, to fit the egg cups, is from a

11

9x11” sheet (Norton, USA). Three egg papers were placed into a 5 ½ oz plastic soufflé cup (Solo
Cup co, Lake Forest, IL, USA) and then moistened with water.

The eggs were hatched in a 6% liver powder (Now Foods, Bloomingdale, IL, USA) and
water mixture that had been fermented for at least one week to create deoxygenated water and
then was diluted to between 30% and 50% with purified water. The water mixture and the egg
papers were left in the water overnight in a plastic hatch tray (8.2”x8.34”x2”, 49oz, Pactiv
Foodservice/Food Packaging, Lake Forest, IL, USA) and the emerging larvae were split into new
hatch trays into groups of approximately 100 larvae per tray. Fresh water was added to dilute the
deoxygenated water further. The larvae were fed a 6% liver powder water mixture each day.
After 5-6 days the larvae pupated and were collected from the pans for use in the experiment.

Blind tests
Two blind tests were conducted to assess possible technical errors. The first blind test
was used to check for and reduce sampling biases in the study. In this experiment, a number of
infected and uninfected females were released into a series of cages, each with a different
infection frequency. The releasing was done by one other member in the lab. Ten samples were
collected in one day in the first test. In the second test, the sample size was increased to 20
samples collected over the course of three days.
The second blind test was run to assess efficacy of PCR assay used to diagnosis
Wolbachia in this study, (Table 2.3). Fifteen Ae. aegypti were released into each of three cages.
All of the mosquitoes were released by another lab member. All fifteen mosquitoes were
dissected and then assessed for infection using PCR.

12

Population cages setup
Once the larvae pupated, they were separated by sex and placed into individual test tubes
until they emerged. In the parental generation (F0), approximately 350 male Waco mosquitoes,
350 female Waco mosquitoes, 90 female WB1 mosquitoes, and 500 male WB1 mosquitoes were
placed into individual test tubes. At the onset of the experiment six cages were used, one of
which was used for mating between WB1 males and WB1 females. In the mating cage, 70-80
WB1 females and males were released and left alone to mate for two days. The cage was bloodfed with live mice and thirteen blood-fed satiated females (N=52) were selected randomly for
release into each of four of the five experimental cages.
In five experimental cages, male mosquitoes were released first to allow both Waco and
WB1 males an equal opportunity to mate with the females. Once the males were released, both
the infected and uninfected females were released into the cages. They were allowed to mate
until they reached 7-10 d post emergence, and they were subsequently blood-fed on live mice.
After the blood meal the mosquitoes were given two days to rest for oogenesis, then two 5 ½ oz
cups were placed into each cage to collect eggs. Once a sufficient number of eggs
(approximately 1000) were laid, which occurred five days after the blood meal, the egg cups
were removed and incubated for four days in an environmental chamber in which all of the live
mosquito cages were maintained at an average temperature of 26.9C and humidity of 70%. After
incubation, the eggs were stored in the environmental chamber in a plastic bag with a moist
paper towel.
Once the next generation began, the same hatching procedures as previously described
were followed. After the hatching was completed, the number of hatched, un-hatched, and
broken eggs on the egg papers for each experimental generation and WB1 were counted. The egg

13

numbers were then used to calculate the hatch rate for each set of eggs and this procedure was
repeated for each generation.
Population Replacement experimental design
There were a total five cages used in this experiment. The cages were set up with varying
numbers of Wolbachia infected males, with the exception of negative control (Table 2.1). This
experiment used a system of single Wolbachia infected female release in generation zero (F0)
and continued inundative release of infected males. For each generation, once the eggs were
collected ten adult females were selected at random, dissected, and their ovaries were collected
for use in PCR.

Table 2.1: Experimental design of population replacement in the laboratory cages. The Waco
line represents the uninfected line and the WB1 line represents the infected line.
Waco♂: WB1♂

No. of Waco

No. of

No. of

No. of

Predicted

♀

Waco ♂

WB1 ♀

WB1 ♂

replacement
generation

1:0

50

50

13

0

F7

1:1

50

50

13

50

F5

1:2

50

50

13

100

F4

1:4

50

50

13

200

F3

Negative control

50

50

0

0

NA

Total

250

250

52

350

14

Polymerase Chain Reactions
DNA Extraction and Incubation
Prior to dissection, 10-20 1.5mL lidded tubes for each experimental set were prepared.
Each tube contained 50µL Sodium Chloride-Tris-EDTA (STE) Buffer. One pair of ovaries was
placed into each tube. Each sample was then ground, using either a battery operated mortar and
pestle, for 90 seconds, or an electrically operated homogenizer (if the homogenizer was used,
100 µL STE buffer and homogenization beads were placed in the bottom of each tube instead of
50 µL). After grinding, 2µL of proteinase K (Roche, USA) was added to each sample and the
sample was vortexed and centrifuged. If a pellet had formed after centrifugation, the pellet was
re-suspended and the samples were placed in a 55 degree Celsius incubation plate for 1h
followed by 15min in a 97 degree Celsius incubation plate to deactivate the proteinase K.
PCR assay
Once the DNA extraction was complete, the samples were kept on ice while 8-strip PCR
tubes were prepared for use in the thermo-cycler. Into each tube I placed 22.5 µL of Platinum
PCR Supermix (Invitrogen Inc., Grand Island, NY, USA), 0.2 µL of the 81F (forward) and 691R
(reverse) Wolbachia Surface Protein (WSP) primers (Zhou et al. 1998) and 2.1µL of the DNA
sample. The tubes were placed in the thermocycler and a program that heats the machine to 95°C
for five minutes, then there were 40 repetitions of spending 30s at 95 °C (melting), 30s at 55 °C
(annealing), and 45s at 72 °C (elongation), followed by 7min at 82°C, and then idling at 4 °C.
The program ran through 40 cycles to produce the DNA copies.
Once the PCR products obtained, they were mixed with 3µL of loading dye (Invitrogen,
CA, USA). Each well of the 2% agrose gel (lifetech, USA) was loaded with 5-10 µL of the
sample. Electricity was run through the gel for 25min at 180V and 1A in the case of larger gels

15

and 12-15min at 150V in the case of smaller sized gels. Once the gel was run it was stained for
15min with an ethidium bromide (10mg/ ml, Sigma, USA) and water mixture. The gel was
imaged with UV light using Kodak Molecular Imaging System software (standard edition) in
combination with a specialized camera (EL Logic 100 Imaging System) and the results were
compared to the 100Kb ladder (Invitrogen) that was prepared by mixing 1µL of the ladder, 7µL
of DNAse/RNAse free water and 2µL of loading dye and loaded at the same time as the sample.
The WSP shows up as a band that is approximately 610 base pairs long.
Mathematical modeling
In our research, we employ a method of increasing the rate of population replacement in
such a way that it can take place in two to six generations as opposed to the current minimum of
seven (Xi et al. 2005b). The generations for Population replacement 1 were predicted using the
equation that follows:
a n = 50(1 − p n ) , An = 50 p n + 50 m , bn = 50(1 − p n ) , Bn = 50 p n . When n>0. (Eq 2.1)

Where an is the number of uninfected males, An is the infected males, bn is the uninfected
females, and Bn infected females. The number of uninfected females in the next generation
G

n +1

corresponds to the matching frequency between the number of uninfected male and

female mosquitoes, and therefore corresponds to an bn = [50(1 − pn )]2 . The number of infected
females also corresponds to the matching frequency of infected male and female mosquitoes and
then is multiplied by the infection cost, ξ (a n + An ) × Bn = 2500 (1 + m )p n ξ . Together, these
numbers define the total infection frequency for G

16

n +1

, by calculating:

p

n +1

=

2500 (1 + m ) p n ξ
2500 (1 + m ) p n ξ + [50 (1 − pn )] 2

=

(1 + m ) p n ξ
(1 + m ) pn ξ + (1 − pn ) 2

for all generations

n > 0. (Eq 2.2)
To calculate the infection frequency for generation 0 to generation 1, it is necessary insert
the actual number of males and females into each of the equations: a n = 50(1 − p n ) ,
An = 50 p n + 50 m , bn = 50(1 − p n ) , B n = 50 p n , meaning a n = 50 , An = 50 m ,
bn = 50 , and B n = 13 . This results in the equation:

13(1 + m )ξ
p =
(Eq 2.3)
1 13(1 + m )ξ + 50
For each given fitness cost ξ , and ratio number of males m the infection frequency
p

n +1

is determined by the infection frequency p n in the previous generation. This equation

follows the assumptions that 1) there is an equal chance that any female will mate with any male,
she will not discriminate based on infection presence or absence, 2) there is equal sex
determination for each experimental generation, the probability that any one egg will be male or
female is .5, 3) there is complete CI sterility, all incompatible matings will result in no offspring,
and 4) 100% maternal transmission, all offspring of infected females will have the same
infection type as the female. This model is designed to take into account a varying number of
infected males being released into the population.
The equation that was designed follows the properties of uniqueness and monotonicity.
So, for each given fitness cost ξ , the infection frequency p

n+1

is uniquely determined by the

infection frequency pn in the preceding generation; and the sequence pn is increasing (or

17

decreasing) for all n ≥ 0 . In other words, p

n+1

> pn ( p

n+1

< pn ) for all n ≥ 0 . To verify the

monotonicity, the value of p * as p * = 1 (1+ m )ξ
From Eq 2.2 it is found that

p

n+1

- pn =

pn (1- pn )( pn - p*)
. (Eq 2.4)
(1+ m)pn ξ + (1- pn )2

Now, if p* = 1 (1+ m )ξ ≤ 0 , i.e., ξ ≥ 1/(1+ m ) , then p
p* = 1 (1+ m )ξ > 0 , i.e., ξ < 1/(1+ m ) , then p

n+1

n+1

> pn for all n ≥ 0 . If

> pn for all n ≥ 0 if and only if p > p * .
0

In this case, p = p * is the unique unstable equilibrium within the interval (0,1) . The predicted
generations are shown in Table 2.7.

Results
Blind tests
The results of the blind tests (Table 2.2 & Table 2.3) show that the sampling 20 samples
over the course of 3 days produced a more accurate result. The results of the third blind test show
that the PCR system does result in an accurate display of both positive (Wolbachia infected) and
negative (Wolbachia uninfected) results. The sampling size was increased from ten samples to
twenty and the sampling style altered to allow for multiple sampling sessions based upon these
results. Additionally, the sampling style was changed so that s many females as possible were
collected, for each sampling set, and then sub-sampled. The extra females from the set were
released back into their cages to be collected and sub-sampled again for the next set (Table 2.3).
The test of the validity of the PCR system results shows that the system is accurate (Table 2.4).

18

Table 2.2: The results of the first blind bias assessment. The first column denotes the cage
number. The second and third columns represent the infection rate estimate achieved by taking
10 and 25 samples, respectively. The final column is the actual infection percentage for each of
the three cages.

Result

Result

Actual

(10 samples)

(25 samples)

1

40%

28%

50%

2

70%

40%

30%

3

30%

36%

20%

Table 2.3: The results of the second blind bias assessment. This blind test involves taking 25
samples of female ovaries over the course of 3 days. Ten samples were taken the first day and
five samples were taken on subsequent days.
Result

Actual

1

28%

37.5%

2

28%

28.6%

3

16%

15.3%

19

Table 2.4: The results of the PCR assessment blind test. This blind was run by dissecting all of
the live females in each cage, extracting the DNA, and running PCR. Cages one and three had
15/15 samples dissected and cage two had 14/15 samples dissected (one female died before
dissection)

Cage

Result

Actual

1

100%

100%

2

71.4%

66.7%

3

33.3%

33.3%

Population Replacement 1
Two strains of a single species, Ae. aegypti, were used for this experiment. No less than
ten female mosquitoes were randomly selected each generation and tested using PCR targeting
the WSP gene. Of the five experimental sets, the 1:4 (Waco male: WB1 male) cage reached full
population replacement the quickest at two generations (Table 2.6). The percentage of infection
in the other three cages (1:0, 1:1, and 1:2) progressed at a relatively stable rate, culminating in
full population replacement during various generations (Table 2.6). The hatch rates of each of
the generations were counted and calculated. At each generation at least 1000 eggs were counted

20

and the number that hatched was divided by the total number of mosquito eggs to obtain an
approximate hatch rate (Table 2.5).

Table 2.5: The percentages represent the number of eggs for each set of mosquitoes that hatched
at each generation.
Negative
Generation

Control

1:0

1:1

1:2

1:4 WB1

1

84%

64%

50%

44%

43%

79%

2

82%

56%

53%

34%

74%

60%

3

68%

48%

65%

62%

61%

62%

4

68%

47%

26%

19%

18%

28%

5

80%

47%

53%

57%

56%

58%

6

60%

50%

47%

52%

54%

63%

7

53%

46%

46%

52%

55%

48%

8

51%

43%

58%

62%

63%

44%

21

Table 2.6: The percentages represent the percentage of each sample set that were found to be
infected. The results were obtained using PCR to image a Wolbachia surface protein (WSP).
Negative
Generation

Control

1: 0

1:1

1:2

1:4

1

0%

50%

70%

80%

100%

2

0%

50%

30%

70%

100%

3

0%

80%

90%

90%

100%

4

0%

100%

100%

100%

100%

5

0%

100%

100%

90%

100%

6

0%

95%

100%

80%

95%

7

0%

100%

100%

100%

95%

8

0%

100%

100%

96%

100%

Mathematical modeling
Based upon the results obtained through use of the mathematical model and a comparison
of them to the actual results obtained via completion of the experiment (Table 2.7) it is possible
to say that the mathematical model is accurate. There is a maximum of a three generation
difference between the actual and the predicted generation, with a difference of two generations
or less being the norm.

22

Table 2.7: Table of predicted replacement generations, based on the mathematical model, and
the actual generations during which replacement took place.
Waco♂ : WB1 ♂

Predicted Replacement

Actual Replacement

Generation (mathematical)

Generation (experimental)

1:0

F7

F7

1:1

F5

F4

1:2

F4

F7

1:4

F3

F2

Negative Control

N/A

N/A

Discussion
Blind Tests
The necessity of an increased sample size from 10 to 20 samples is shown through the
use of a chi-squared (observed v. expected) statistical analysis. The chi-square observed vs.
expected analysis showed that the actual and observed infection frequency was significantly
different for the first set of blind assays (Table 2.2) that was used to assess the validity of a
2

sampling size of ten (χ (df=2)=6.033 p=.049) (α = .05 used for all statistical tests). The result of
the chi-squared analysis shows that there is a significant difference between the observed and
expected values when only assaying 10 samples that are all taken the same day. By analyzing a
further set of 15 samples it was possible to show that taking twenty samples on the same day also
2

displayed a significant difference between the observed and expected values (χ (df=2)=6.453
p=.0397). The second sampling size assessment shows that the process of sampling multiple sets,
23

10 on the first day and five on the subsequent days, is an adequate method for accurately
2

assessing an approximate infection frequency (χ (df=2)=0.6173 p=.7344). This supports the use
of the second sampling method over the first method.
After assessing that the final sampling method of collecting as many mosquitoes as
possible from the cage and sub sampling from that group results in a there being no significant
difference between the observed and expected values, an equivalence test was run to ensure that
the results of the blind tests were not statistically the same. The epsilon value (calculated by
taking 20% of the variance of each group) is considered to be equal for each group, due to all
values being very similar. Epsilon is defined as .042. The null hypothesis is that the difference
between the groups is greater than the epsilon value, meaning that the groups are not the same.
All groups were found to be statistically different from each other (Table 2.8).

Table 2.8: The equivalence test for the comparison of each group of samples. For the purpose of
this table, each cage has been assigned a different number (1-9). Numbers 1, 4, and 7 designate
the groups with a sample size of 10. Numbers 2, 5, and 8 designate the groups where an
additional 15 samples were taken after the sample size of 10. Numbers 3, 6, and 9 designate the
final group where 25 samples were taken over the course of several days and as many females as
possible were collected and sub-sampled for each set (extra females were released back into their
cages to be sampled again the next day).
Group Group Infection
X

Y

Infection

Percentage Percentage
for group
X

Value

Confidence

(difference Interval (for

for group Y between
means)

24

value)

P

Null
reject
(Yes/No)

Table 2.8 cont’d
1

2

50

50

.12

[-0.1806674,

.6682519

No

0.4206674]
1

3

50

37.5

.12

[-0.1806674,

0.6682519 No

0.4206674]
4

5

30

30

0.3

[-0.01372013,

0.91335

No

0.61372013]
4

6

30

28.6

.42

[0.1254568,

0.9814239 No

0.7145432]
7

8

20

20

-.06

[-0.3690375 ,

0.5389628 No

0.2490375]
7

9

20

15.3

.14

[-0.1175705,

0.7379567 No

0.3975705]
2

3

50

37.5

0

[-0.2173931,

0.3736589 No

0.2173931]
5

6

30

28.6

.12

[-0.1075098,

0.7160195 No

0.3475098]
8

9

20

15.3

0.2

[-0.006782086, 0.8969214 No
0.406782086]

Population Replacement 1
Based upon this experiment we conclude that the release of additional males that are
infected with a strain of Wolbachia bacteria can increase the rate at which a population of
25

uninfected Ae. aegypti is replaced by a population of infected individuals. This is shown by the
release of WB1 males into cages that contained only a low base number of infected female
mosquitoes and an equal number of male and female uninfected mosquitoes. The cage that
contained a ratio of 1:4 (uninfected males : infected males) became fully infected at a higher rate
than the cage with no additional release of Wolbachia infected male mosquitoes(Table 2.6).

Mathematical modeling
The mathematical model proved to be relatively accurate. It predicted the replacement
generation number, but it needs further refinement in order to predict an accurate infection rate at
each other generation. With future refinements the mathematical model designed for this project
could be possibly be used for field studies.
The issue with this sample set comes from the fact that there were a minimum of ten
females sampled of the 50 collected. The fifty were already a randomly selected subset from the
total number of hatched mosquitoes for each cage at each generation. The number of mosquitoes
selected for the first subset of 50 may be less than 10% of the total hatched population. However,
the sample set of ten was a good starting point, by beginning there, a minimum number of
samples to begin the assessment of population replacement rate was established. The issue of the
selection size may have been one of the causes in the discrepancy between the number of
generations that population replacement with no additional males took place (four to five
generations) and previous studies where it took seven generations (Xi et al. 2005b).

26

CHAPTER 3
Reducing the initial female release frequency threshold for population replacement
Introduction
As mentioned in chapters 1 and 2, Dengue is a very prevalent mosquito-borne disease.
Due to its vector it has both a high incidence of initial infection and a risk of re-infection. Over
the years, many attempts have been made towards controlling this disease and its mosquito
vectors (Ae. aegypti and Ae. albopictus) and have been met with varied levels of success. Many
of the currently used mosquito control techniques have an assortment of objections and
difficulties attached to them. These problems range from ethical dilemmas concerning GMOs
(Resnik 2012) to the development of resistance to insecticides (Lenormand et al. 1999;
Chareonviriyahpap et al. 1999; Rose 2001).
With the objections, hindrances, and dilemmas a different type of control program is
needed. That is not to say that the new program would be without its own share of any or all of
the afore mentioned types of stumbling blocks, but that each group of troubles would be easier to
manage or of a reduced negative impact. One such program is the application of inundative
release techniques using Wolbachia to replace a Dengue vulnerable population with a disease
refractory population (as in chapter 2). While Wolbachia-mediated disease control techniques
may be difficult to classify under current regulations, as in the case of the release of Wolbachia
infected mosquitoes in Australia (DeBarro et al. 2011), the lack of genetic modification avoids
the ethics struggle associated GMOs and the lack of actual chemicals dodges the potential issues
of both insecticide resistance and toxicity brought on by the use of pesticides. In addition, the
mosquito release group, while containing some infected female mosquitoes at the first
generation, would be entirely made up of males after the initial release. This may reduce

27

concerns, as the males are not the transmitters of disease. All of these factors combined make the
use of Wolbachia-mediated disease control rather attractive and the inundative release of male
mosquitoes, as opposed to females, can potentially make it more so.
The goal of this experiment was to decrease the threshold number of Wolbachia infected
female mosquitoes that must be released into a population to achieve fixation of the infection
within the population. The hypothesis is that adding male release to the population replacement
strategy is a viable way to decrease the number of females required for the infection to reach
fixation within the population. This strategy requires the inundative release of Wolbachia
infected male mosquitoes at each generation.
Materials and Methods
The general procedure is similar to what has been described in the chapter 2 except the
below:
Population Replacement design
This experiment entails the use of an extra number of wAlbB infected males being
released into the population in order to reduce the necessary number of infected females. The
idea behind this is that the extra infected males would cause more incidences of CI by mating
with the uninfected females. This would reduce the number of uninfected offspring. The infected
females are able to mate with both infected and uninfected males successfully, increasing the
chance of infected offspring being produced. Depending on the number of infected males
released, this would cause a rapid decrease in the uninfected population and, by comparison, an
increase in the early generation infection frequency, allowing the infection too persist more
readily within a population.

28

Table 3.1: The experimental design of Population Replacement 2: Reducing the threshold
number of Wolbachia-infected females necessary for the bacterium to achieve fixation within the
population. This table describes the numbers of infected and uninfected mosquitoes released into
each of the experimental cages.
WB1 ♀ release %

No. of Waco ♀

No. of Waco ♂ No. of WB1
♀

No. of WB1
♂

20%

50

50

13

100

10%

50

50

6

100

5%

50

50

3

100

2%

50

50

1

100

Negative control

50

50

0

0

Total

250

250

23

400

There were a total five cages used in this experiment. The cages were set up with varying
numbers of Wolbachia infected females, with the exception of negative control (Table 3.1). This
experiment used a system of single Wolbachia infected female release in generation zero (F0)
and continued inundative release of infected males. For each generation, once the eggs were
collected ten adult females were selected at random, dissected, and their ovaries were collected
for use in PCR.

Mathematical modeling
The equation for predicting the generations at which the replacement would take place,
depending on the size of the initial female seeding population, is a modification of the equation

29

used for the prediction of the generations of population replacement in increasing the rate of
population replacement. These generations were predicted using the equation that follows:
a n = 50(1 − p n ) , An = 50 pn + 100 , bn = 50(1 − pn ) , B n = 50 p n . When n>0. (Eq 3.1)

Where an is the number of uninfected males, An is the infected males, bn is the uninfected
females, and Bn infected females. The number of uninfected females in the next generation
G

n +1

corresponds to the matching frequency between the number of uninfected male and

female mosquitoes, and therefore corresponds to an bn = [50(1 − pn )]2 . The number of infected
females also corresponds to the matching frequency of infected male and female mosquitoes and
then is multiplied by the infection cost, ξ (an + An ) × Bn = 7500 pn ξ . Together, these numbers
define the total infection frequency for G

p

n+1

=

3 pn ξ
3 p n ξ + (1 pn ) 2

n +1

, by calculating:

for all generations n > 0. (Eq 3.2)

To calculate the infection frequency for generation 0 to generation 1, it is necessary insert
the actual number of males and females into each of the equations: a n = 50(1 − p n ) ,
An = 50 pn + 100 , bn = 50(1 − pn ) , B n = 50 p n , meaning a n = 50 , An = 100 , bn = 50

, and Bn = f . This results in the equation:

3fξ
p1 =
(Eq 3.3)
3fξ + 50
For each given fitness cost ξ and seeding number of infected females f the infection
frequency p

n +1

is determined by the infection frequency p n in the previous generation. This

equation follows the assumptions that 1) there is an equal chance that any female will mate with

30

any male, she will not discriminate based on infection presence or absence, 2) there is equal sex
determination for each experimental generation, the probability that any one egg will be male or
female is .5, 3) there is complete CI sterility, all incompatible matings will result in no offspring,
and 4) 100% maternal transmission, all offspring of infected females will have the same
infection type as the female. This model is designed to take into account a varying number of
infected females being released into the population at only generation 0 and the fixed number of
infected males that are released at every generation.
The same concepts of uniqueness and monotonicity that were applied to the equation for
population replacement 1 were applied to equation 3.2, resulting in
p* = 1 - 3ξ .

Then from (3.2) we find

p

n+1

pn =

pn (1 - p n )( p n - p*)
.
3 pn ξ + (1 - pn ) 2

> pn for all n ≥ 0 . If p* = 1 - 3ξ > 0 , i.e.,
Now, if p* = 1 - 3ξ ≤ 0 , i.e., ξ ≥ 1/ 3 , then p
n+1
> pn for all n ≥ 0 if and only if p > p * . In this case, p = p * is the
ξ < 1/ 3 , then p
n+1
0
unique unstable equilibrium within the interval (0,1) .

Results
Based on mathematical modeling it is possible to show that the introduction of

Wolbachia infected males will lower the threshold number of infected females necessary for
population replacement (Table 3.2). The model shows that a female release that is as low as 2%
of the total female population will still result in complete population replacement when
combined with continued inundative Wolbachia infected male release.
31

The use of inundative infected male release as a method to lower the threshold number of
necessary infected females needed to achieve population replacement is demonstrated using the

Anopheles stephensi mosquito (Table 3.3) (Bian et al. 2013)

Table 3.2: The expected generations when the infection frequency has reached 99.9%. The
predicted generations were calculated using the mathematical model (Eq. 3.1 through 3.3)
Female infection Frequency

Predicted generation

20%

F4

10%

F5

5%

F6

2%

F7

Negative Control

No population replacement

Table 3.3: Predicted and actual replacement generations achieved using inundative release of
Wolbachia-infected male Anopheles stephensi mosquitoes to reduce the number of infected
female mosquitoes necessary for complete population replacement.
Infected percentage of ♀

Mathematically predicted

Experimentally achieved

replacement generation

replacement generation

5%

F7

G8

10%

F6

G8

20%

F5

G7

32

Discussion
Infected male inundative release can potentially decrease the number of infected females
necessary for a seeding population. At present, a minimum of a 20% Wolbachia infected female,
calculated using the number of infected females divided by the total number of females in the
entire population, release is necessary for full population replacement. The equation designed for
this projected indicates that the number of Wolbachia infected females can be reduced by
inundative release of infected males at every generation. This reduction in the number of
required females results in a more socially acceptable program. The decreased number of
infected females being released does bring with it the potential for a decreased rate at which
population replacement will take place. Applying the same concept used in Population
Replacement 1, it may be possible to limit the rate decrease by altering the number of infected
males that are released into the population.
In support of the utility of inundative male release and the mathematical model the
comparison between the generation of replacement generated through mathematical modeling
and the generation of replacement that was experimentally derived has been provided (Table
3.3). This table shows that the model is relatively accurate, but still might need some fine tuning.
This may be due to a need to further assess fitness costs and population dynamics associated with
the novel infection of An. stephensi with wAlbB. So, while not 100% accurate when comparing
predicted and actual replacement generations, the model still supports both the ability of this
inundative release technique to reduce the necessary number of females and the applicability of
the model to different genii of mosquito. It is likely that the model will need to be fine tuned for
each new genus of mosquito it is applied to. The usefulness of the model for more than one
genus of mosquito can makes it a very practical research tool in both laboratory and field studies.

33

In the lab it could be used to assess viability of and created a plan for a replacement strategy. In
field studies it could be used to help predict generations of replacement, optimize the
replacement rate, and help decide an approximate number of mosquitoes to begin the
replacement with.

34

CHAPTER 4
Characterization of Cytoplasmic incompatibility Phenotypes with Triple infected Aedes
albopictus
Introduction
Aedes albopictus (Skuse) is the secondary vector of dengue virus. It is a mosquito that
has adapted to the urban environment provided by human habitation, as well as the sub-urban
and semi-rural areas (Knudsen 1995). Due to the lack of vaccines for dengue and the spread of
its vectors, it is imperative to come up with effective disease control strategies. Many methods
have been explored for the control of disease vectoring species like Ae. albopictus. These control
techniques focus on both larval and adult stages of the mosquitoes, such as larvicides and adult
insecticides, sterile male release techniques, biological control agents, and genetically modified
mosquito release (Paupy et al 2009). One strategy with potential is the use of Wolbachia to
control the spread of dengue. In studies with Aedes aegypti it has been shown that Wolbachia
infection inhibits the transmission of dengue (Bian et al. 2010) and that it takes seven generations
in Ae. aegypti to drive an infection into the population when a specific minimum initial infection
threshold is met (Xi et al. 2005b). There are hopes of a similar response in Aedes albopictus.

Ae. albopictus naturally carries two types of Wolbachia, wAlbA and wAlbB
(Kittayapong et al. 2002) and studies have shown that it is possible to transfer infection types
within Ae. albopictus via microinjection of the embryo (Xi et al 2005a). There have been
different types of Wolbachia used to generate a triple infected line (Fu et al. 2010). However, the
focus thus far has been in using single infected males that contain an incompatible infection type
for population suppression (Calvitti et al. 2010; Laven 1967).

35

Ae. albopictus exhibits a unidirectional type of CI called additive CI (Figure 1.3).
Additive CI occurs when there are at least two Wolbachia infection types within individuals of
the population. This type of crossing displays an increasing percentage of total crossings that
result in CI directly related to the increasing number of infection types. Due to being a type of
unidirectional CI, the mod/resc system holds true. The Wolbachia modify the sperm and they
make it so that an egg with the same infection type can rescue it (Werren 1997), as discussed in
chapter 1.
Despite the evidence for the ability of Wolbachia to inhibit dengue, it is still a secondary
disease vector for dengue virus. There is hope that a new, third, type of Wolbachia infection may
prevent the transmission of dengue by Ae. albopictus. In order to effectively design an
experiment to drive this bacterium into the population and assure that there are three infection
types within the population, phenotypic characterization of the CI must occur.

Materials and Methods
The Houston line of Ae. albopictus mosquitoes was used for this experiment. Through
microinjection techniques a line (HC) of Ae. albopictus mosquitoes was developed that
contained three types of Wolbachia: wAlbA and wAlbB (from Ae. albopictus) and wPip (from

Culex pipiens (Hertig 1936). It was generated via infected cytoplasm microinjections to the
embryo of the Ae. albopictus, using cytoplasm from the Culex pipiens, similar to how the WB1
line was created (Xi et al 2005b).
Once the lines were established and enough eggs were generated to maintain the lines,
crossing experiments commenced. This type of crossing experiment used three lines of Ae.

albopictus mosquitoes, Houston Tetracycline treated (HT), Houston wild type (Hou), and
36

Houston triple infected (HC). The HT line was the uninfected line. It was generated by feeding
Hou mosquitoes tetracycline to wipe out the natural Wolbachia infection types (Dobson and
Rattanadechakul 2001). The HC line was created via microinjection and the experiment took
place during the eighth generation of the HC line. The Hou line is the wild type and originated in
Houston, Texas (1986).
The cages used for this study were medium sized Tupperware bowls modified so that
there was a hole in the side of the bowl and a tube of cloth netting was attached to cover the hole
and allow the cage to seal. The top was cut so that only the rim of the lid remained and was used
to hold plastic mesh netting in place to allow for blood feeding and sucrose feeding.
All mosquito eggs were stored at a constant temperature and moisture before hatching
(average 26.9C) and moisture (70% relative humidity). The eggs were stored for no more than
three months because storage for a longer amount of time may result in a sharp decline in the
hatch rate. The eggs were laid on egg paper over the course of two to three days. The egg papers
are pieces of water absorbent brown paper that was cut to fit the egg cups from a 9x11” sheet
(University Stores, Michigan State University, East Lansing, MI, USA). Four egg papers were
placed into a 5 ½ oz plastic soufflé cup (Solo Cup co, Lake Forest, IL, USA) and then moistened
with water.

The eggs were hatched in a 6% liver powder (Now Foods, Bloomingdale, IL, USA) and
water mixture that had been fermented for at least one week to create deoxygenated water and
then was diluted to between 30% and 50% with purified water. The water mixture and the egg
papers were left in the water overnight in a plastic hatch tray (8.2”x8.34”x2”, 49oz, Pactiv
Foodservice/Food Packaging, Lake Forest, IL, USA) and the emerging larvae were split into new
hatch trays into groups of approximately 100 larvae per tray. Fresh water was added to dilute the
37

deoxygenated water further. The larvae were fed a 6% liver powder water mixture each day.
After 5-6 days the larvae pupated and were collected from the pans for use in the experiment.
The pupae were placed into individual test tubes, by sex, to provide enough virgin males
and virgin females for each cage. The pupae were sexed via size comparison. The males tend to
be noticeably smaller than the females. The mosquitoes were checked for sex after emergence
and before release in the cages. This was a simple matter of comparing antennae and proboscis
structures. After emergence, the males and females were released into the cages in preset crosses.
Each cage received ten males and ten females. The crossing types and expected pattern of CI are
shown in Table 4.1.

Table 4.1: Crossing patterns and expected CI responses for all possible CI crossings of HT1,
Hou, and HC Ae. albopictus mosquitoes.

Female

HT

Hou

HC

HT

Compatible

Compatible

Compatible

Hou

CI

Compatible

Compatible

HC

CI

CI

Compatible

Male

After release into the cages, they were allowed to mate until they reached 7-10 d post emergence,
and they were subsequently blood-fed on live mice. After the blood meal the mosquitoes were
given two days to rest for ovigenesis, and one 5 ½ oz cup was placed into each cage to collect
eggs. Approximately 2 days after the egg cups were placed in the cages they were removed and
placed inside of a hatch tray under a damp paper towel for 2 days. The eggs were then dried for
around 7 days and then hatched. After 2 days the egg papers were pulled out of the water and the
hatch rate was counted. The blood feeding was performed a second time for the same cages and
38

egg collection process was repeated. The second set of eggs was then hatched and the hatch rate
was counted (Figure 4.1).

Results
The crossing results were as expected as shown by the comparison of the table (Table
4.1) and figure (Figure 4.1). The results show that wPip is incompatible with a double infection
of wAlbA and wAlbB. The triple infected line induced unidirectional CI and the wPip infection is
incompatible with a double infection of wAlbA and wAlbB. The expected incompatible cross
types resulted in a hatch rate of zero. There were some lower hatch rates associated (Figure 4.1)
with the Hou and HC lines, but these rates were not significantly different when compared to the
hatch rates of the other crosses (Table 4.2 and Figure 4.2).

Figure 4.1: The average hatch rates for all crosses. The *# represents how many replicates of
each cross were run. The HC line was in its eighth generation during this experiment.

Hatch Rates
70
HTXHT

*3

50

*3

*3

HTXHou
*3 *3 *3

40

HTXHC

30

HouXHou

20

HouXHT
*3

39

HCx HT

HCx Hou

HCx HC

Houx HT

HCXHC
Houx Hou

HTx Hou
HT x HC

0

HouXHC

*5

*3

Houx HC

10
HT x HT

% Hatched

60

HCXHou
HCXHT

Figure 4.2 The one way analysis of the asin prop
proportion of hatchings by crossing showing that
none of the hatchings are significantly different from each other.

Table 4.2: The analysis of variance for the Ae. albopictus CI crossing hatch data. This test shows
that there is no significant difference between the average hatch rates produced by each of the CI
produced
crosses.

Source

DF

Sum of Squares

Mean Square

F Ratio

Prob > F

5

0.12012592

0.024025

0.9045

0.5093

Error

12

0.31875256

0.026563

C. Total

17

0.43887848

Crossings

Discussion

40

It is important to note that this third type of infection within the Ae. albopictus system
using wPip is novel. There has been at least one other triple infection type used for Ae.

albopictus mosquitoes, but the type of Wolbachia was wRi from Drosophila simulans (Fu et al
2010). This infection characterization showed the possibility of using triple infected Ae.

albopictus to act as the CI mechanism for a population replacement experiment. There have been
other studies involving the introduction of wPip into a population of Ae. albopictus, however
these studies introduced the infection into a Wolbachia negative line of the mosquito to create a
single infected line of the species (Calvitti et al. 2010). That study assessed the maternal
inheritance of the infection, the fitness effects, and the CI expression within the new
transinfected line. They found that there was a significant fecundity cost for the new single
infected line; however continued selection may reduce the fitness costs. There was complete CI
sterility, and 100% maternal transmission (Calvitti et al 2010).
It is also important to note that there are no known systems that contain four types of

Wolbachia within the same individual. This may be due to a combination of stress upon the host
caused by supporting multiple infection types or that not all Wolbachia types have an additive
effect and those that do not often compete with each other within the host and one or more
infections are lost.
There is great disease control potential for this triple Wolbachia infected line based on the
results of this experiment. It has been shown that the wPip infection type is resistant to dengue
virus in both the midgut and salivary gland (Xi et al. unpublished data). Due to this resistance it
is possible that this third infection type within the Ae. albopictus system could be used to clean
the mosquitoes of the dengue virus and halt the disease transmission via the secondary vector.

41

The selection of an infection type that matches well with desired characteristics for using
a Wolbachia driven system is extremely important. In order to make this type of selection, the
infection and the fitness costs, maternal transmission rate, CI sterility rate, and the additive
effects within the population must be observed. All of these factors can influence the
effectiveness of both population replacement and population suppression experiments. If the
fitness costs are too high, the maternal transmission to low, or if the CI sterility is incomplete it
can reduce the effectiveness of the vector control strategy and, could cause the project to fail.

42

CHAPTER 5
Mosquitoes, Wolbachia, and cytoplasmic incompatibility: Conclusions and future
directions
Conclusions
Wolbachia has been proven to inhibit the transmission of dengue virus in mosquitoes (Lu
et al 2012; Walker et al 2011; Hoffmann et al 2011; Iturbe-Ormaetxe et al 2011; Blagrove et al.
2011). Thus, it is important to find an effective way to drive this bacterium into disease vector
populations. We observed that the release of additional infected males into the population can
accelerate the population replacement. These additional males cause an increase in the frequency
of incompatible matings and result in the decline of the uninfected population. There is a direct
correlation between an increased ratio of infected males and an increased rate at which
population replacement will occur, as shown by equation 1. Therefore, it is possible to conclude
that an increased number of Wolbachia infected males released into a population will cause an
increased rate of population replacement. In addition, we have made mathematical modeling to
predict in which generation population replacement will occur with this additional male release
strategy.
We also found that wPip is incompatible with wAlbA and wAlbB infection types. Due to

wPip being incompatible with wAlbA and wAlbB, the triple infected Ae. albopictus line
generated for the studies within the Xi lab causes unidirectional CI when crossed with the superinfected wild type mosquito and the tetracycline treated, Wolbachia-negative, Ae. albopictus
line. Together with PCR diagnosis result, the CI cross data confirm that the HC triple infected

Ae. albopictus line does actually contain three types of Wolbachia.

43

Overall, both population replacements in Ae. aegypti and characterization of CI hatch
phenotypes in triple infected Ae. albopictus are important steps toward disease control. Both of
these vectors are now wide spread invasive species in the majority of tropical and subtropical
areas around the world. There are still many research directions that need to be explored to
further this style of disease control.

Future Directions
The next step for the first population replacement experiment is to repeat the experiment
to gather more data and ascertain the most efficient ratio of male release to achieve population
replacement and to increase the experiment size to get a better view of a larger population, such
as a field population. In laboratory cage tests are often designed to assess the feasibility of a
study within a controlled environment and they are designed to be used a model to base fieldcage and field studies upon. Once a field-cage study, is completed and the results are confirmed
the data can be used for projects like increasing the predictive power of the mathematical model,
modification of the model to account for a different species, or population replacements of other
species of vector containing Wolbachia,.
These results can be used to build several models that can be applied to other disease vectors
that either do not naturally carry Wolbachia or that do not carry the particular type of Wolbachia
being introduced into the population. Considering that some types of the bacteria can shorten the
life span of the vector in question, as well as be used to drive desirable traits into the population
(Werren et al. 2008), it is a small leap to extend this idea to other types of disease vector.
Research delving into the possibility of controlling diseases that have a specific incubation
period within the insect before they are able to be transmitted has begun. This research relies on
the alteration of the age population structure by shortening the life-span of the insect with

44

Wolbachia (Cook et al. 2008). The data collected in this experiment may make it easier for such
studies to predict how many Wolbachia infected insects must be released to complete the
replacement and determine ways to reduce the number of blood-meal requiring females that must
be released.
There are several future directions associated with the triple infected Ae. albopictus line.
The first project should be fitness costs studies, such as fecundity, survivorship, etc similar to
what was done by Calvitti et al. (2010) for the single wPip infected Ae. albopictus line and
studies that look at the stability of the three infection types over time. Once these studies have
been completed it will be possible to set up several cage population replacement experiments.
Due to wPip being resistant to dengue virus (Xi et al. unpublished) it is a good candidate for
future field replacement studies. Another future project to be completed, after the cage studies, is
a field population replacement, such as what was done using wMel infected Ae. aegypti in
Australia, wMelPop was also investigated, but the fitness costs for infected mosquitoes were too
high and the infection had trouble spreading in the population (Hoffmann et al. 2011).
The single infected or the triple infected lines of Ae. albopictus can be used for population
suppression similar to the suppression of Culex pipiens fatigans (Laven 1967), so long as the
infection type makes the released mosquitoes incompatible with the current population. The
suppression strategies (Dobson et al. 2002) can be used in tandem with population replacement
to create a more effective control strategy or the suppression strategies can be used to create an
artificial unstable equilibrium that continually drives the population size down over several
generations (Dobson et al. 2002).
Population replacement studies are applicable in other species besides Ae. aegypti. There
have already been studies in population replacement using Anopheles stephensi. A line of this

45

species has been artificially infected with Wolbachia wAlbB and population replacement studies
have been worked on to assess the lowest number of infected females that can be released into
the population and have the infection reach fixation (Bian et al. 2013). This work was undertaken
due to the ability of wAlbB to cause the An. stephensi mosquito to be resistant to Plasmodium

falciparum (Hughes et al. 2011).
Another option that is currently undergoing study is the use of transgenic mosquitoes. In
various studies there have been modifications to the genome of certain mosquito vectors. These
modifications were designed to elucidate the purpose of a variety of genes within the mosquito
system and find possible methods for the control of the spread of disease. An example of one of
these types of studies involves an alteration of the REL1, which is related to the Toll immune
pathway of insects. The over expression of which causes an up-regulation of Spatzle and
immune response (Bian et al 2005), meaning an up-regulation of the Toll pathway. The Toll
pathway plays a significant role in the regulation of the Ae. aegypti mosquito’s resistance to
dengue infection and its ability to limit the viral infection (Xi et al. 2008). Should the fitness
costs associated with transgenic mosquitoes, not just those that were modified using the REL1
gene, (Irvin et al 2003; Catteruccia et al. 2003; Moreira 2003) be mitigated or alternate genes
with a similar effect but lacking the fitness cost be discovered this could easily become a valued
disease control strategy. The other issues with this strategy involve mitigating the biosafety risk,
regulation approval for field releases, public approval, and environmental hazards (Sperança and
Capurro 2007).
Due to the prevalence of diseases like dengue, it is important to find an effective control
strategy. Every year thousands died from dengue due to the lack of vaccines and treatments to
combat this disease. Vector control is the only real option for controlling the spread of dengue

46

and dengue hemorrhagic fever. The development of resistance to various pesticides in
mosquitoes and the lack of a self-sustaining solution make the Wolbachia-mediated disease
control strategy very promising. Wolbachia-based population replacement is unique in that it is
naturally occurring, self-sustaining, and host specific, thus it will not affect non-target species.
All of the above make the use of Wolbachia for disease control a very attractive alternative and
future studies, as described above, are expected to further improve the efficiency and efficacy of

Wolbachia-based population replacement strategies.

47

APPENDIX

48

APPENDIX
RECORD OF DEPOSITION OF VOUCHER SPECIMENS
The specimens listed below have been deposited in the named museum as samples of those
species or other taxa, which were used in this research. Voucher recognition labels bearing the
voucher number have been attached or included in fluid preserved specimens.

Voucher Number: ___2013-06________________

Author and Title of thesis:

Wolbachia-Mediated Population Replacement in Dengue Mosquito Vectors By Shawna Ryan

Museum(s) where deposited:
Albert J. Cook Arthropod Research Collection, Michigan State University (MSU)

Specimens:
Family

Genus-Species

Culicidae

Aedes aegypti

adult

10

pinned

Culicidae

Aedes aegypti

adult

10

alcohol

Culicidae

Aedes albopictus

adult

10

pinned

Culicidae

Aedes albopictus

adult

10

alcohol

Life Stage

49

Quantity

Preservation

BIBLIOGRAPHY

50

BIBLIOGRAPHY
Bian, G., Joshi, D., Dong, Y., Lu, P., Zhou, G., Pan, X., Xu, Y., Dimopoulos, G., & Xi, Z.
(2013). Wolbachia invades Anopheles stephensi populations and induces refractoriness to
Plasmodium infection. Science, 340(6133), 748-751.
Bian, G., Shin, S. W., Cheon, H. M., Kokoza, V., & Raikhel, A. S. (2005). Transgenic alteration
of Toll immune pathway in the female mosquito Aedes aegypti. Proceedings of the
National Academy of Sciences of the United States of America, 102(38), 13568-13573.
Bian, G., Xu, Y., Lu, P., Xie, Y., & Xi, Z. (2010). The endosymbiotic bacterium Wolbachia
induces resistance to dengue virus in Aedes aegypti. PLoS pathogens, 6(4), e1000833.
Blagrove, M. S., Arias-Goeta, C., Failloux, A. B., & Sinkins, S. P. (2012). Wolbachia strain
wMel induces cytoplasmic incompatibility and blocks dengue transmission in Aedes
albopictus. Proceedings of the National Academy of Sciences, 109(1), 255-260.
Brelsfoard, C. L., & Dobson, S. L. (2009). Wolbachia-based strategies to control insect pests and
disease vectors. Asia Pac J Mol Biol Biotechnol, 17(3), 55-63.
Calvitti, M., Moretti, R., Lampazzi, E., Bellini, R., & Dobson, S. L. (2010). Characterization of a
new Aedes albopictus (Diptera: Culicidae)-Wolbachia pipientis (Rickettsiales:
Rickettsiaceae) symbiotic association generated by artificial transfer of the w Pip strain
from Culex pipiens (Diptera: Culicidae). Journal of medical entomology, 47(2), 179-187.
Catteruccia, F., Godfray, H. C. J., & Crisanti, A. (2003). Impact of genetic manipulation on the
fitness of Anopheles stephensi mosquitoes. Science, 299(5610), 1225-1227.
CDC. (2012). Dengue frequently asked questions. Fact Sheet.
Chan, K. L., Ho, B. C., & Chan, Y. C. (1971). Aedes aegypti (l.) and Aedes albopictus (skuse) in
singapore city. Bulletin of the World Health, 44(5), 629-633.
Chareonviriyahpap, T., Aum-Aung, B., & Ratanatham, S. (1999). Current insecticide resistance
patterns in mosquito vectors in Thailand.
Cobra, C., Rigau-Pérez, J. G., Kuno, G., & Vomdam, V. (1995). Symptoms of dengue fever in
relation to host immunologic response and virus serotype, Puerto Rico, 1990–1991.
American journal of epidemiology, 142(11), 1204-1211.
Cook, P. E., McMeniman, C. J., & O'Neill, S. I. (2008). Modifying insect population age
structure to control vector-borne disease. Transgenesis and the Management of VectorBorne Disease, 627, 126-140.
Crampton, J., Morris, A., Lycett, G., Warren, A., & Eggleston, P. (1990). Transgenic
mosquitoes: a future vector control strategy?. Parasitology Today, 6(2), 31-36.

51

De Barro, P. J., Murphy, B., Jansen, C. C., & Murray, J. (2011). The proposed release of the
yellow fever mosquito, Aedes aegypti containing a naturally occurring strain of
Wolbachia pipientis, a question of regulatory responsibility. Journal für
Verbraucherschutz und Lebensmittelsicherheit, 6(1), 33-40.
Dobson, S. L. (Designer). Additive Incompatibility [Web Graphic]. Retrieved from
http://dobsonserv.ca.uky.edu/DobsonSite/Super2.html
Dobson, S. L. (2004). Evolution of Wolbachia cytoplasmic incompatibility types. Evolution,
58(10), 2156-2166.
Dobson, S. L., Fox, C. W., & Jiggins, F. M. (2002). The effect of Wolbachia-induced
cytoplasmic incompatibility on host population size in natural and manipulated systems.
Proceedings of the Royal Society of London. Series B: Biological Sciences, 269(1490),
437-445.
Dobson, S. L., & Rattanadechakul, W. (2001). A novel technique for removing Wolbachia
infections from Aedes albopictus (Diptera: Culicidae). Journal of medical entomology,
38(6), 844-849.
Fan, W., Yu, S., & Cosgriff, T. M. (1989). The reemergence of dengue in China. Review of
Infectious Diseases, 11(Supplement 4), S847-S853.
Fu, Y., Gavotte, L., Mercer, D. R., & Dobson, S. L. (2010). Artificial triple Wolbachia infection
in Aedes albopictus yields a new pattern of unidirectional cytoplasmic incompatibility.
Applied and environmental microbiology, 76(17), 5887-5891.
Gubler, D. J. (1998). Dengue and dengue hemorrhagic fever. Clinical microbiology reviews,
11(3), 480-496.
Gubler, D. J. (2002). Epidemic dengue/dengue hemorrhagic fever as a public health, social and
economic problem in the 21st century. Trends in microbiology, 10(2), 100-102.
Gubler, D. J., & Clark, G. G. (1995). Dengue/dengue hemorrhagic fever: the emergence of a
global health problem. Emerging infectious diseases, 1(2), 55.
Gubler, D. J., & Clark, G. G. (1996). Community involvement in the control of Aedes aegypti.
Acta Tropica, 61(2), 169-179.
Hedges, L. M., J.C. Brownlie , S.L. O'Neill, & K.N. Johnson. 2008. Wolbachia and virus
protection in insects. Science, 322: 702.
Helinski, M. E., Parker, A. G., & Knols, B. G. (2009). Radiation biology of mosquitoes. Malar J,
8(Suppl 2), S6.
Hertig, M. (1936). The rickettsia, Wolbachia pipientis (gen. et sp. n.) and associated inclusions
of the mosquito, Culex pipiens. Parasitology, 28(04), 453-486.

52

Hilgenboecker, K., Hammerstein, P., Schlattmann, P., Telschow, A., & Werren, J. H. (2008).
How many species are infected with Wolbachia?–a statistical analysis of current data.
FEMS Microbiology Letters, 281(2), 215-220.
Hoffmann, A. A., Montgomery, B. L., Popovici, J., Iturbe-Ormaetxe, I., Johnson, P. H., Muzzi,
F., Greenfield, M., Durkan M., Leong, Y. S., Dong, Y., Cook, H., Axford, J., Callahan,
A. G., Kenny, N., Omodei, C., McGraw, E. A., Ryan, P. A., Ritchie, S. A., Turelli, M., &
O’Neill, S. L. (2011). Successful establishment of Wolbachia in Aedes populations to
suppress dengue transmission. Nature, 476(7361), 454-457.
Hughes, G. L., Koga, R., Xue, P., Fukatsu, T., & Rasgon, J. L. (2011). Wolbachia infections are
virulent and inhibit the human malaria parasite Plasmodium falciparum in Anopheles
gambiae. PLoS pathogens, 7(5), e1002043.
Irvin, N., Hoddle, M. S., O'Brochta, D. A., Carey, B., & Atkinson, P. W. (2004). Assessing
fitness costs for transgenic Aedes aegypti expressing the GFP marker and transposase
genes. Proceedings of the National Academy of Sciences of the United States of America,
101(3), 891-896.
Iturbe-Ormaetxe, I., Walker, T., & LO'Neill, S. (2011). Wolbachia and the biological control of
mosquito-borne disease. EMBO reports, 12(6), 508-518.
Jansen, V. A. A., Turelli, M., & Godfray, C. J. (2008). Stochastic spread of Wolbachia.
Proceedings of the Royal Society Biological Sciences, 275, 2769-2776.
Kittayapong, P., Baimai, V., & O'Neill, S. L. (2002). Field prevalence of Wolbachia in the
mosquito vector Aedes albopictus. American Journal of Tropical Medicine and Hygiene,
66(1), 108-111.
Knudsen, A. B. (1995). Global distribution and continuing spread of Aedes albopictus.
Parassitologia, 37(2-3), 91.
Kyle, J. L., & Harris, E. (2008). Global spread and persistence of dengue. Annu. Rev. Microbiol.,
62, 71-92.
Laven, H. (1967). Eradication of Culex pipiens fatigans through cytoplasmic incompatibility.
Nature, 216, 383-384.
Lenormand, T., Bourguet, D., Guillemaud, T., & Raymond, M. (1999). Tracking the evolution of
insecticide resistance in the mosquito Culex pipiens. Nature, 400(6747), 861-864.
Esteva, L., & Mo Yang, H. (2005). Mathematical model to assess the control of Aedes aegypti
mosquitoes by the sterile insect technique. Mathematical biosciences, 198(2), 132-147.
Lu, P., Bian, G., Pan, X., & Xi, Z. (2012). Wolbachia induces density-dependent inhibition to
dengue virus in mosquito cells. PLoS neglected tropical diseases, 6(7), e1754.

53

Marshall, J. M., & Taylor, C. E. (2009). Malaria control with transgenic mosquitoes. PLoS
medicine, 6(2), e1000020.
McMeniman, C. J., Lane, R. V., Cass, B. N., Fong, A. W., Sidhu, M., Wang, Y. F., & O'Neill, S.
L. (2009). Stable introduction of a life-shortening Wolbachia infection into the mosquito
Aedes aegypti. Science, 323(5910), 141-144.
Moreira, L. A., Iturbe-Ormaetxe, I., Jeffery, J. A., Lu, G., Pyke, A. T., Hedges, L. M, Rocha, B.
C., Hall-Mendelin, S., Day, A., Riegler, M., Hugo, L. E., Johnson, K. N., Kay, B. H.,
McGraw, E. A., van den Hurk, A. F., Ryan, p. A., & O'Neill, S. L. (2009). A Wolbachia
Symbiont in Aedes aegypti Limits Infection with Dengue, Chikungunya, and
Plasmodium. Cell, 139(7), 1268-1278.
Mousson, L., Dauga, C., Garrigues, T., Schaffner, F., Vazeille, M., & Failloux, A. B. (2005).
Phylogeography of Aedes (Stegomyia) aegypti (L.) and Aedes (Stegomyia) albopictus
(Skuse)(Diptera: Culicidae) based on mitochondrial DNA variations. Genetical research,
86(1), 1-12.
Mousson L, Zouache K, Arias-Goeta C, Raquin V, Mavingui P, et al. (2012) The Native
Wolbachia Symbionts Limit Transmission of Dengue Virus in Aedes albopictus. PLoS
Negl Trop Dis 6(12): e1989. doi:10.1371/journal.pntd.0001989
Paupy, C., Delatte, H., Bagny, L., Corbel, V., & Fontenille, D. (2009). Aedes albopictus, an
arbovirus vector: From the darkness to the light. Microbes and Infection, 11(14), 11771185.
Rose, R. I. (2001). Pesticides and public health: integrated methods of mosquito management.
Emerging Infectious Diseases, 7(1), 17.
Scott, T. W., Takken, W., Knols, B. G., & Boëte, C. (2002). The ecology of genetically modified
mosquitoes. Science, 298(5591), 117-119.
Sperança, M. A., & Capurro, M. L. (2007). Perspectives in the control of infectious diseases by
transgenic mosquitoes in the post-genomic era: a review. Memórias do Instituto Oswaldo
Cruz, 102(4), 425-433.
Teixeira, L., Ferreira, Á., & Ashburner, M. (2008). The bacterial symbiont Wolbachia induces
resistance to RNA viral infections in Drosophila melanogaster. PLoS biology, 6(12),
e1000002.
Turelli, M., & Hoffmann, A. A. (1991). Rapid spread of an inherited incompatibility factor in
California Drosophila. Nature, 353(6343), 440-442.
Walker, T., Johnson, P. H., Moreira, L. A., Iturbe-Ormaetxe, I., Frentiu, F. D., McMeniman, C.
J., Leong, Y.S., Dong, Y., Axford, J., Kriesner, P., Lloyd, A. L., Ritchie, S. A., O’Neill,
S. S. L., & Hoffmann, A. A. (2011). The wMel Wolbachia strain blocks dengue and
invades caged Aedes aegypti populations. Nature, 476(7361), 450-453.

54

Weeks, A. R., Turelli, M., Harcombe, W. R., Reynolds, K. T., & Hoffmann, A. A. (2007). From
parasite to mutualist: rapid evolution of Wolbachia in natural populations of Drosophila.
PLoS biology, 5(5), e114.
Werren, J. H. (1997). Biology of Wolbachia. Annual Review of Entomology, 42, 587-609.
Werren, J. H., Baldo, L., & Clark, M. E. (2008). Wolbachia: Master manipulators of invertebrate
biology. Nature Reviews Microbiology, 6, 741-751.
Werren, J. H., & Windsor, D. M. (2000). Wolbachia infection frequencies in insects: evidence of
a global equilibrium?. Proceedings of the Royal Society of London. Series B: Biological
Sciences, 267(1450), 1277-1285.
White, S. M., Rohani, P., & Sait, S. M. (2010). Modelling pulsed releases for sterile insect
techniques: fitness costs of sterile and transgenic males and the effects on mosquito
dynamics. Journal of Applied Ecology, 47(6), 1329-1339.
WHO. (2009). Dengue: Guidelines for diagnosis, treatment, prevention and control. Fact sheet,
(117).
WHO. (2012). Dengue and severe dengue. Fact sheet, (117).
Xi, Z. Wolbachia pipientis and dengue transmission. Unpublished data.
a.Xi, Z., Khoo, C. C. H., & Dobson, S. L. (2005). Generation of a novel Wolbachia infection in
Aedes albopictus (Asian tiger mosquito) via embryonic microinjection. Insect
Biochemistry and Molecular Biology 35 (2005) 903–910
b. Xi, Z., Khoo, C. C. H., & Dobson, S. L. (2005). Wolbachia establishment an invasion in an
Aedes aegypti laboratory population. Science, 310, 326-328.
Xi, Z., Khoo, C. C., & Dobson, S. L. (2006). Interspecific transfer of Wolbachia into the
mosquito disease vector Aedes albopictus. Proceedings of the Royal Society B:
Biological Sciences, 273(1592), 1317-1322.
Xi, Z., Ramirez, J. L., & Dimopoulos, G. (2008). The Aedes aegypti toll pathway controls
dengue virus infection. PLoS pathogens, 4(7), e1000098.
Zabalou, S., Apostolaki, A., Pattas, S., Veneti, Z., Paraskevopoulos, C., Livadaras, I., Markakis,
G., & Brissac, T., Mercot, H., Bourtzis, K. (2008). Multiple rescue factors within a
Wolbachia strain. Genetics, 178(4), 2145-2160.

55