5

Genetic approaches for malaria control

Marcelo Jacobs-Lorena#

Abstract
The already unacceptable large burden of malaria continues to increase, indicating
that the available means to fight the disease are insufficient. The genetic manipulation
of the mosquito vectorial competence is a potential new promising weapon for the
control of malaria. Considerable progress has been made in recent years towards this
goal. It is now possible to introduce synthetic genes into the mosquito germ line,
promoters have been identified that effectively drive gene expression in tissues and at
times appropriate to target the parasite, and effector genes that impair parasite
development in the mosquito have been identified. With these tools, proof-of-concept
experiments have already demonstrated that it is possible to interfere genetically with
the vectorial competence of the mosquito. At least some of the transgenic mosquito
lines that have been created appear to be as fit as their wild-type counterparts.
Presently, the major unresolved challenge is the development of methods to drive
effector genes into mosquito populations in the field. While several approaches are
under consideration, such as transposable elements, Wolbachia, meiotic drive and
paratransgenesis, their relative feasibility remains to be demonstrated. Additional
challenges are the resolution of safety concerns and satisfactorily addressing social,
ethical and political considerations. Hopes remain high that the remaining challenges
will be solved and that we shall be able to deploy this new genetic weapon in the
foreseeable future.
Keywords: mosquitoes; malaria; transgenesis; effector genes; driving mechanisms;
refractoriness

Introduction
Insect-transmitted diseases impose an enormous burden on the world population in
terms of loss of life (millions of deaths per year) and morbidity. These diseases
impose huge economic losses both in terms of health-care costs and lost productivity,
mostly in countries that can least afford it. Three basic approaches have been
attempted to contain these diseases: 1) treat infected people with drugs that kill the
pathogen; 2) control insect vector populations; and 3) develop vaccines that prevent
infection.

Drugs
Drugs have been at the forefront of the fight against many arthropod-transmitted
diseases. For decades, chloroquine was successfully used against malaria. However,
#
Johns Hopkins School of Public Health, Malaria Research Institute, Dept. Molecular Microbiology
and Immunology, 615 N. Wolfe St.,Baltimore, MD 21205. E-mail: [email protected]

53
Chapter 5

with time pathogens are selected for drug resistance, forcing the development of new
drugs. Unfortunately, this cycle of drug discovery followed by resistance becomes
more difficult to perpetuate with the passage of time. In principle, combination drug
therapy should greatly alleviate this problem. In practice, other factors (mostly
economic) make the implementation of this strategy difficult. While drugs are
extremely useful in containing and treating diseases, they are not sufficient on their
own for disease eradication. Clearly, a combination of strategies is needed.

Control of insect populations

On the contrary, reduction of vector insect populations will reduce disease
transmission. This can be accomplished in various ways, for instance, with
insecticides, by managing the environment (elimination of breeding sites) or by
interfering with reproduction (sterile-insect releases). Recent technological advances
suggest an alternative approach, namely genetic modification of the competence of
the vector arthropod to transmit pathogens (vectorial competence), which is the main
subject of this article.

Vaccines
The only successful vaccine in existence against an arthropod-transmitted pathogen
is the yellow-fever vaccine. Even in this case, the disease has not yet been eradicated.
Decades of intense research aimed at the development of other vaccines, notably for
malaria and dengue fever, but this has yet to yield a viable product. At the heart of the
problem, at least for malaria, is that during thousands of years of association with
humans pathogens have been selected that efficiently evade the host immune system.
High genetic diversity, variability of potential target molecules (e.g., Plasmodium var.
genes), and intracellular sequestration are strategies frequently used by pathogens that
allow them to elude immune attack. While the search for effective vaccines should
continue, this has been an uphill battle.

Insecticides
Under appropriate circumstances, insecticides are powerful weapons to fight
vector-borne diseases. For instance, they have been crucial in the eradication of
malaria in Europe and of An. gambiae in Brazil, and they are important in controlling
disease epidemics (e.g., dengue, West Nile) in urban areas. The introduction of DDT
in the mid 1940s heightened the hopes of disease eradication. A case in point is the
WHO campaign to eradicate malaria, which was successful at its inception (for
instance, malaria was almost eliminated from the entire Indian subcontinent).
However, problems such as development of insecticide resistance by mosquitoes, the
discovery of the harmful effects of DDT to the environment and to non-target
organisms, and the ‘letting down of the guard’ when disease was almost under
control, led to the reversal of most initial successes. While judicious use of
insecticides is still a powerful weapon to fight disease, one aspect of its use is
frequently overlooked: insecticides usually leave intact the biological niche where the
target insects reproduce. Therefore, insect populations rapidly return to pre-treatment
levels as soon as application is halted, which is a very serious problem. For instance,
one can hardly hope for large-scale mosquito-population reduction in Africa,
especially if one considers that management of breeding sites (e.g. widely distributed
small pools of water) would be required. While residual spraying of house interiors or
use of bednets will lower transmission rates and reduce prevalence and incidence of

54
 Jacobs-Lorena

infections, the breeding sites will continue to generate mosquitoes and this cycle of
breeding and killing is likely to hasten the development of insecticide resistance.
Thus, insecticides are useful to bring temporary relief but cannot be considered as
solutions by themselves. In future, insecticides are likely to become key weapons
when used in combination with other approaches such as vaccines, drugs or when
used for population replacement (see below).

Sterile-insect technique (SIT)

Insect populations can be controlled by the release of large numbers of sterile
males. Thus, if a female mates with a male that has no sperm or whose sperm was
rendered unviable, this female will have fewer or no progeny. When many sterile
males are released, the local population tends to decline or become extinct. There are
a number of cases of the successful local application of this technique, for example, in
the control of the Mediterranean fruit fly in Latin America, the New World
screwworm in the Americas and Libya, and for tsetse in Zanzibar, Africa. SIT also
has been applied, on a limited scale, to Culex in India and Anopheles albimanus in El
Salvador (see Curtis, Chapter 3).
For population control, the crucial parameter is the ratio of the number of released
sterile males to the number of males in the local population, which ideally should be
around 10:1. Therefore, sterile-insect control is only effective when the resident
population to be controlled is small relative to the number of sterile males that can be
mass-produced for release or when it can be reduced to very low levels with
conventional control tools before the start of releases. It is highly desirable that only
males be released for two reasons: 1) In most cases only females bite and transmit
disease while, moreover, sterile females can also transmit; 2) Males would court and
mate with the released sterile females (instead of local females), thus reducing the
efficacy of the programme (Alphey and Andreasen 2002). Large-scale production in
the laboratory of a pure male population by non-genetic means may be problematic. It
may rely on sex-specific differences of pupal size (culicine mosquitoes) or adult
eclosion times (tsetse), but these protocols rarely yield a 100% male population.
Clearly, genetic sexing methods (see below) are far superior. The most commonly
used technique for male sterilization is exposure to high doses of radiation, a
procedure that damages chromosomes and results in unviable sperm. Sterilization by
chemical means also has been employed. Because of the large numbers of insects that
need to be released, it is crucial that the effectiveness of the sterilization procedure
approaches 100%. However, the large doses of radiation and chemicals needed to
achieve this effectiveness may reduce insect fitness, survival and mating
competitiveness. These strategies can fail if the laboratory-reared males do not mate
as effectively as their field counterparts.
The advent of germ-line transformation for a number of different insects has led to
the development of genetic alternatives for production of sterile insects (Heinrich and
Scott 2000; Thomas et al. 2000). In one version of this approach (Release of Insects
carrying a Dominant Lethal or RIDL, Thomas et al. 2000), a conditional dominant
lethal gene is introduced into the target insect genome. This gene has two important
properties: 1) it is expressed only in females (or it kills only females); and 2) the gene
is effectively repressed by a compound that does not occur normally in nature (e.g.
tetracycline). Large insect populations are maintained by rearing them in the presence
of tetracycline, which represses the dominant lethal gene and allows the survival of
equal numbers of males and females. Prior to release, the insects are reared in the

55
Chapter 5

absence of tetracycline, a condition that allows the expression of the dominant lethal
gene and the death of all females. The resulting males can be released without further
manipulation or treatment. Males carry two copies (homozygous) of the dominant
lethal gene. When these males mate in nature, all female progeny will be killed and
only males will be produced. Since these surviving males are heterozygous for the
dominant lethal gene, the population-reducing effect is still manifest in the second
generation.
It should be emphasized that the effectiveness of the SIT is dependent on
population structure and dynamics. Furthermore, this technique leaves intact the
biological niche in which the target insect is found. SIT is most likely to succeed in
cases where target populations are small, the number of target insects is low, and the
target area is sufficiently isolated, thereby reducing the likelihood of re-invasion. It is
unlikely to be effective for controlling mosquito populations in highly endemic areas
of Africa where the mosquito population consists of several vector species in high
densities, where access to breeding sites is difficult and where poorly interbreeding
mosquito populations co-exist.

Genetic manipulation of vectorial competence

Germ-line transformation
Drosophila melanogaster was the first multicellular organism to be stably
transformed (Spradling and Rubin 1982). The same general principles that were used
in this pioneering work are still employed today for all germ-line transformation work
in insects (Atkinson and James 2002). Embryos are injected with two DNA
constructs. One construct contains a gene encoding a dominant selectable marker
(e.g., eye color, a fluorescent protein) and the gene of interest, each driven by a
separate promoter, and both sequences are together flanked by the inverted repeats of
a transposable element. The second construct encodes a transposase, which is an
enzyme that recognizes the inverted repeats and catalyses the insertion of the
intervening sequences into the genome of the host insect. It took from 1982 until the
mid 1990s to develop two crucial technologies: an appropriate transposable-element
system (at first scientists did not realize that the P transposable element is not active
in non-Drosophila organisms) and a suitable transformation marker (e.g., GFP). Since
then, germ-line transformation of many insects has been accomplished but mosquitoes
(Aedes, Anopheles, Culex) are the only insects of medical importance in this list.
Importantly, both An. stephensi and An. gambiae can be transformed, though the
success rate in the latter case is still low. Improvement of the transformation
efficiency of An. gambiae is a high-priority topic for future research. It would also be
desirable to develop germ-line transformation procedures for other medically
important insects such as sand flies and black flies. Current technology cannot be
applied to germ-line transformation of tsetse because these do not lay eggs (that
would need to be injected), only fully formed larvae. However, genetic modification
of tsetse vectorial capacity could be achieved via genetic modification of one of its
symbionts.
The net result of germ-line transformation is the integration into the genome of the
host organism of a relatively large DNA sequence, flanked by inverted repeats of the
transposable element. The inserted DNA contains at least two genes, the gene to be
investigated and a transformation marker gene (e.g., eye color, GFP) that allows
transformed individuals to be identified. The integrated DNA is usually stable and
transmitted in a Mendelian manner from one generation to the next. In the following

56
 Jacobs-Lorena

text, we will consider transformation of insects with genes that affect their ability to
transmit pathogens. We will start by considering promoters to be used for driving
gene expression and then effector genes capable of interfering with parasite
development.

Promoters
Promoters that drive the expression of effector genes in transgenic insects should
be strong, that is, they should result in abundant transcription. This is because the
effectiveness of the gene products is expected to increase with increased abundance.
Two general types of promoters can be considered: ubiquitous and tissue-specific.
Ubiquitous promoters are less desirable because general expression of a foreign gene
product in all tissues of the insect and at all times is likely to impose a fitness load.
Tissue-specific promoters have the advantage of being restricted to a tissue and are
often developmentally and/or physiologically regulated (not constitutive). Strong
tissue-specific promoters that have been characterized in mosquitoes include gut
carboxypeptidase (Moreira et al. 2000), fat body vitellogenin (Kokoza et al. 2000) and
gut peritrophic matrix protein 1 (PM1; Jacobs-Lorena laboratory, manuscript in
preparation). The carboxypeptidase promoter has the advantage of being induced by
blood intake, and thus its activation coincides with parasite arrival in the gut. The
carboxypeptidase signal sequence effectively promotes secretion into the midgut
lumen, which is the compartment where the parasite initially resides. The PM1
promoter and the protein’s signal sequence direct synthesis and storage of the protein
in vesicles of the midgut-epithelial cells prior to a blood meal. The vesicle contents
are released into the midgut lumen immediately after blood ingestion. The
vitellogenin promoter is also induced by the blood meal (expression peaks at ~24 h)
and the peptide signal sequence promotes secretion into the mosquito body cavity
(haemocoel), where the parasite later develops. This promoter is ideally suited for
expression of effector molecules that target the ookinete soon after its crossing of the
midgut epithelium and emergence into the haemocoel. Once it reaches the haemocoel,
the ookinete transforms into an oocyst that is difficult to target because it is covered
by a thick protective layer. Upon maturation (at about 10 days after the infective
blood meal), the oocyst releases sporozoites that disperse though the haemocoel until
they come in contact with, and invade, the salivary gland. It would be desirable to find
promoters that drive synthesis of proteins secreted into the haemolymph at the time of
sporozoite release. While sequences that direct secretion into the salivary-gland lumen
have been identified, their expression levels of the corresponding promoters are low
(Coates et al. 1999). It would be desirable to identify a strong salivary-gland promoter
to target pathogens stored in the salivary glands. The pathogen usually is stored in the
salivary-gland lumen for extended periods of time, and this increased period of
contact may favour parasite inactivation by the effector-protein product. One should
keep in mind though, that mosquito saliva (and any effector protein secreted into it) is
transferred to its vertebrate (human) host, raising safety and ethical questions.

Effector genes
The term effector gene is used here for genes whose products interfere with the
development of a pathogen. At least four classes of effector genes can be identified: 1)
Genes whose products interact with insect host tissues crucial for parasite
development: Examples of this class are SM1, a peptide that occupies putative
salivary-gland and midgut receptors for the malaria parasite (Ghosh, Ribolla and
Jacobs-Lorena 2001) and phospholipase A2 (PLA2), which is a protein that interferes

57
Chapter 5

with the malaria ookinete invasion of the midgut (Zieler et al. 2001); 2) Genes whose
products interact with the pathogen: Examples of this class are genes encoding single-
chain monoclonal antibodies that bind to the parasite’s outer surface thus blocking
their development (Yoshida et al. 1999; De Lara Capurro et al. 2000); 3) Genes whose
products kill the pathogen: Examples are peptides from the insect’s innate immune
system such as defensins and cecropins, and peptides from other sources that act as
selective toxins to parasites but do not affect the host insect, such as magainins,
Shiva-1, Shiva-3 and gomesin (Kim et al. 2004). Most published work on effector
genes deals with effects on the malaria parasite and little is known about such genes
for other pathogens. In particular, it is not clear what class of effector genes would be
useful for nematodes (filaria). Since these may be encapsulated in certain mosquito
strains, genes that activate encapsulation could be considered as possible effector
genes. For viruses, genes of the first class (interference of host-tissue invasion) or
genes that interfere with virus replication (Olson et al. 1996) are possible candidates.
4) Another possible strategy is to reduce vector competence by manipulation of its
immune genes, for instance by using RNA interference or ‘smart sprays’
(Christophides, Vlachou and Kafatos 2004).
Another important strategic consideration is the stage of malaria parasite
development to target. When a mosquito ingests an infected blood meal, it acquires
thousands of gametocytes of which only few (usually less than ten) manage to cross
the midgut and form oocysts. Later, each oocyst produces thousands of sporozoites, a
significant proportion of which invade the salivary gland. Because the strong
bottleneck at the level of midgut invasion, this stage of parasite development
constitutes a prime target for intervention. Midgut invasion is also a strong bottleneck
in the process of arboviral transmission.

Genetically modified mosquitoes

Successful development of the technology described above (transgenesis, promoter
characterization and effector-gene identification), permitted the creation of genetically
modified mosquitoes impaired in their ability to transmit the malaria parasite. An
early example was the creation of an Ae. aegypti expressing defensin in the
haemolymph (Kokoza et al. 2000). However, the effect of defensin on malaria
parasite development has not been reported. At about the same time, the James
laboratory reported that a single-chain monoclonal antibody that recognizes a
sporozoite surface protein inhibits invasion of the salivary gland (De Lara Capurro et
al. 2000). In this instance, the effector gene was transiently expressed from a viral
vector that is not inherited by the mosquito progeny. The Jacobs-Lorena laboratory
showed that a stably integrated gene encoding SM1 strongly inhibits parasite
development in transgenic mosquitoes (Ito et al. 2002). In another example,
transgenic mosquitoes expressing PLA2 also had much reduced vectorial competence
(Moreira et al. 2002). Recently, it was demonstrated that the capacity to transmit the
malaria parasite is reduced by about 60% in transgenic An. gambiae expressing
cecropin from a carboxypeptidase promoter (Kim et al. 2004). Thus, it is clear that
mosquitoes can be genetically modified to reduce their vectorial competence. To date,
most reported experiments have been done with non-human malaria parasites. An
important next step is the transfer of this technology to human pathogens.

Insect fitness
For the introduction of an effector gene into populations, it is important that it
confers the least possible detrimental effect on mosquito survival or reproduction

58
 Jacobs-Lorena

(fitness load). This parameter can be initially tested in the laboratory by use of
population cages. For instance, SM1-transgenic mosquitoes do not seem to have any
load, while PLA2-transgenic mosquitoes lay significantly fewer eggs and therefore
carry a significant fitness load (Moreira et al. 2004). In contrast to the apparent lack of
fitness load of SM1-transgenic mosquitoes, Cateruccia, Godfray and Crisanti (2003)
reported that transgenic mosquitoes expressing GFP from an actin promoter may have
a fitness disadvantage. It appears however that in these experiments, loss of fitness
was mainly due to inbreeding (the experiments were conducted with homozygous
transgenic mosquitoes that may have been subject to the ‘founder effect’) and perhaps
to generalized foreign gene expression from a ubiquitous promoter (see above).
Moreover, Irvin et al. (2004) have detected a fitness load in transgenic Ae. aegypti that
express an eGFP marker gene. However, as for the experiments by Cateruccia et al.,
they used homozygous transgenic mosquitoes to measure fitness and these
experiments cannot determine whether the fitness load originates from nearby
recessive genes that were homozygosed with the transgene (‘hitchhiking effect’) or
from a true fitness load imposed by the transgene itself.
Another consideration is that the malaria parasite itself reportedly imposes a fitness
load on the mosquito (Hogg and Hurd 1997). In agreement with this observation, the
Jacobs-Lorena laboratory has preliminary results indicating that in cage experiments,
transgenic mosquitoes expressing SM1 out-compete wild-type mosquitoes with the
same genetic background when fed on P. berghei-infected mice, presumably because
the transgenics have a lower parasite load (unpublished observations).
Eventually tests will have to be devised that measure insect fitness in the field (as
opposed to laboratory cages) because other factors may come into play. Moreover,
laboratory mosquitoes may not compete well with their field counterparts (important
for release studies). One possible solution to this issue may be to cross the effector
genes into wild-caught local mosquito populations prior to release.

Between now and field release

While genetic modification of mosquitoes to resistance to the malaria parasite is
clearly feasible in a laboratory setting, many issues remain to be addressed before
implementation of this approach in the field can be envisioned. Examples of
unresolved issues follow.
i) Parasite resistance and multiple effector genes
Parasites tend to have a heterogeneous genome that favours selection of individuals
able to overcome barriers such as drugs or possibly effector gene products. It will
therefore be crucial that transgenic mosquitoes incorporate more than one (ideally
several) effector genes, each of which blocks parasite development by a different
mechanism.
ii) Driving effector genes into field populations
This is undoubtedly the major unresolved technical issue. Several approaches have
been suggested.
Population replacement by inundatory release. A possible scenario would be to
start with an isolated area (e.g., an island) where malaria is prevalent, and reduce to
the maximum extent possible the mosquito population by use of insecticides. As
discussed above, this would leave an empty biological niche. The next step would be
the release of transgenic mosquitoes to occupy this niche. New transgenic releases
could follow periodically with the expectation that the original mosquito population
will eventually be replaced by the transgenic one. While waiting for the development
of an effective driving mechanism (see below), comparison of malaria transmission

59
Chapter 5

before and after population replacement should provide valuable data to assess
effectiveness of the transgenic approach. It should be noted however that while
population replacement is conceivable for research purposes in small physical or
ecological islands, its implementation on a large country- or continent-wide scale is
not feasible.
Transposable elements. There is an excellent example in Drosophila of how an
element can spread through wild populations. In a matter of a few decades, the P
element spread through virtually all D. melanogaster in the world. Presumably, this
happened because the transposase causes the element to multiply in the genome,
resulting in non-Mendelian transmission. Unfortunately, P elements are not active in
non-drosophilid insects and more importantly, to date no transposable element with
similar properties has been identified in mosquitoes. Even when such elements are
identified, there will be major issues to be addressed. One relates to ‘cargo’ size. As
mentioned in the preceding text, it will be critical to use at least two different effector
genes. In addition, a gene encoding the transformation marker (e.g., GFP) and a gene
encoding the ‘driver’ (transposase) will be needed. Note that it is crucial for cargo and
driver to be tightly linked. Thus, a minimum of four genes are needed representing a
substantial cargo of 12~16 kb (calculated at 3~4 kb/[gene + regulatory sequences]). In
nature, transposable elements are known to become truncated as they hop from one
position to another, and the probability of truncation can be expected to increase with
the size of the cargo (for comparison, a typical transposable element has about 3 kb).
Thus, cargo damage and consequent inactivation of the genes carried by the element
is a major concern. Another consideration is that in some instances, insects carrying a
transposable element accumulate a repressor of the transposase, precluding
introduction of a different gene with the same transposable element and into the same
population. In other words, this could be a ‘one-shot’ proposition: should one discover
that the wrong transgenes were used, that resistance developed or that gene
inactivation occurred, there would not be a second chance to spread another set of
genes with the same element.
Wolbachia. Wolbachia are intracellular bacteria that inhabit the germ line of a
number of insects and distort reproduction by killing progeny that do not contain it, by
a phenomenon known as cytoplasmic incompatibility (CI). Compelling evidence in
favour of Wolbachia as a drive mechanism comes from Drosophila. Turelli and
Hoffmann (1991) observed that Wolbachia swept through the D. simulans population
in California at the rate of 100 km per year. In principle, Wolbachia could provide a
powerful driving mechanism. However, no Wolbachia have yet been identified in
anopheline mosquitoes (these are the exclusive vectors for human malaria), although
they have been observed in culicine mosquitoes. A major limitation of Wolbachia is
that it inhabits the germ line while the pathogen develops in the soma. Thus, it is
difficult to target parasites with genes introduced into Wolbachia. A possible solution
to this problem is the identification of genes that cause CI. Currently little is known at
the molecular level about how CI functions or how many genes are involved. When
identified, such gene(s) could conceivably be used to create a driving mechanism via
their insertion into the mosquito genome.
Meiotic drive. Population replacement can be driven by certain genes, such as the
Drosophila segregation distorter gene, that favour its inheritance over individuals not
containing the gene. Unfortunately, very little is known about such genes in insects of
medical importance. One complication is that at least in model systems (Drosophila,
mouse), the drive mechanism depends on multiple genes (e.g., distorter and
responder) and this could complicate the implementation of this system in

60
 Jacobs-Lorena

mosquitoes. Moreover, if such genes were to be employed to drive effector genes into
populations, all meiotic drive and effector genes would have to be tightly linked to
avoid loss of effectiveness due to recombination.
Paratransgenesis. An alternate approach to spread effector genes through
mosquito populations is to introduce effector genes into bacteria that inhabit the
mosquito gut, rather than introducing them into the mosquitoes themselves. This
approach (known as paratransgenesis) has been successfully tested in another
vector/parasite system to disrupt the transmission of Trypanosoma cruzi, the causative
agent of Chagas disease, by the triatomid bug Rhodnius prolixus. R. prolixus harbours
an obligate bacterial gut symbiont that lives in close proximity to the T. cruzi parasite.
When this symbiotic bacterium was transformed with an effector gene encoding
cecropin A, and fed to naive R. prolixus nymphs, T. cruzi’s ability to survive in the
nymphs was significantly reduced (Durvasula et al. 1997). Preliminary experiments
from the Jacobs-Lorena laboratory suggest that the same principle can be used for
reducing the capacity of mosquitoes to vector malaria parasites. When E. coli
displaying the inhibitory SM1 peptide (Ghosh, Ribolla and Jacobs-Lorena 2001) or
PLA2 (Zieler et al. 2001) on its surface was fed to An. stephensi followed by an
infectious blood meal, significantly fewer P. berghei parasites developed as compared
to control mosquitoes fed on wild-type bacteria. Several considerations argue in
favour of the paratransgenic approach: 1) Bacteria live in the same compartment
where the initial stages of Plasmodium development occur; 2) In principle, bacteria
should be easier to introduce into mosquito populations than transgenes. One possible
scenario is that baits containing the modified bacteria, a source of sugar and mosquito
attractant(s) would be placed in strategic locations (e.g. huts) around a village where
malaria transmission occurs. Moreover, genetic modification of bacteria is
straightforward and efficient. Bacteria can be easily and cheaply grown in large
quantities. This facilitates the introduction of multiple effector genes into mosquito
populations, each bacterium being transformed with a different transgene. Unlike
mosquito transgenes, inactivation of bacterial transgenes after many generations in the
field is not a major concern because of the likely easier logistics of introducing freshly
transformed bacteria. Moreover, if an effector gene fails to perform as promised,
introduction of alternate transgenes is relatively simple. 3) A paratransgenic approach
also poses fewer risks compared with a large-scale release of genetically modified
vectors. Not only does a large-scale mosquito release cause increased nuisance to the
local population, but there may be an increased health risk if the mosquitoes are
capable of vectoring other pathogens. However, the paratransgenic approach is not
without concerns: 1) It is not known how adult mosquitoes acquire their bacterial
flora. Thus, we do not yet know how to implement a plan to control Plasmodium
transmission with genetically modified bacteria; 2) Introduction of a genetically
modified organism of any kind into the field needs to be approached with much
caution because of the unknown consequences. In particular, bacteria are known to be
able to spread genes by horizontal transfer (especially if present in plasmids). It is not
known whether the modified bacteria could populate the guts of non-target organisms.
If so, the consequences need to be assessed.
iii) Population structure
It is quite clear that at least for anopheline mosquitoes, population structure is
complex. This is because a number of morphologically identical but chromosomally
different (cytotypes) mosquito populations can co-exist in any given area.
Importantly, these populations may not freely interbreed and this may seriously affect

61
Chapter 5

efforts to spread effector genes through populations. A better understanding of

population structure and mosquito ecology should be given a high priority.
iv) Safety concerns
While there is no reason to believe that any of the effector genes identified to date
have any effects on non-target organisms, concerns are being raised by the scientific
and lay communities regarding the safety of transgenic mosquitoes. For instance, it
has been suggested that these mosquitoes might be better vectors for other (non-
malaria) pathogens. While there is no evidence to suggest that this is the case, caution
should be used and these possibilities should be tested. Another concern is that of
horizontal gene transfer from the transgenic mosquitoes to other organisms, even to
humans. Again, this possibility is remote. Moreover, the possibility of horizontal
transfer of the effector gene to the germ cells of another organism is even more
remote. Finally, most (if not all) effector genes being considered should be innocuous
to higher organisms.
v) Political, social and ethical considerations
In addition to scientific concerns, it is important to address concerns of public
perception. The issues and debates over genetically modified crops and other attempts
of insect releases provide a useful precedent from which we should learn. It will be
important to inform the public at large about the benefits and risks of the transgenic
technology. The time to start is now, because full public awareness of these issues
may take an entire generation to accomplish. The targets of such a campaign must
include residents of the affected countries and their government officials. The results
of safety tests should be openly and broadly divulgated. It is important to emphasize
that no approach will be entirely risk-free and that the balance between potential
benefit and risk should prevail in deciding about implementation of a new genetic
strategy. It is also crucial to emphasize that any single approach is unlikely to be
completely effective on its own, and that a final solution will have to incorporate a
combination of several weapons, such as drugs, insecticides, bed nets, and hopefully
vaccines and genetic modification of vector competence.

Priorities
Following are some current research priorities.

- First and foremost, a method to drive effector genes into field mosquito populations
needs to be devised.
- The efficiency of An. gambiae transformation needs to be improved.
- Anopheline mosquitoes cannot be stored frozen or desiccated. Establishment of
repository centres for transgenic lines is desirable.
- Whenever possible, work should be conducted with the organisms that are most
relevant to human disease (An. gambiae, P. falciparum).

62
 Jacobs-Lorena

Figure 1. Laboratory and field scientists have mostly worked independently of each other in
the past. However, the task of implementing malaria control via genetic manipulation of
mosquitoes is so large that only close interactions and collaboration among investigators in
the different lab and field disciplines will allow the final goal to be attained.

Prospects
During an historical 1991 meeting sponsored by the MacArthur Foundation and the
WHO/TDR in Tucson, Arizona, a consensus was reached that the genetic
manipulation of mosquito vectorial competence is an important and attainable goal.
Hard work by an emerging community of mosquito molecular biologists delivered the
first germ-line transformation in 1998, followed in rapid succession by the
development of the necessary tools and by the proof-of-principle demonstration that
the approach is feasible. At the same time, field researchers have made fundamental
discoveries concerning population structure and vector ecology. Clearly,
implementation of the genetic modification of vector competence that was considered
in this article will depend on close collaboration among many groups of scientists
with a broad range of lab- and field-based expertise (Figure 1). The task ahead of us is
so big that it is inconceivable to proceed otherwise. While major lab and field issues
remain to be solved, the rapid progress made to date gives one reason to be optimistic
that this important new weapon (genetic modification of vector competence) will be
ready for field testing within the next decade or so.

Acknowledgments
The author is grateful for comments by Anthony James on an early version of this
manuscript. Work from this laboratory was supported by the National Institutes of
Health, U.S.A. and by the WHO/TDR.

63
Chapter 5

References
Alphey, L. and Andreasen, M., 2002. Dominant lethality and insect population
control. Molecular and Biochemical Parasitology, 121 (2), 173-178.
Atkinson, P.W. and James, A.A., 2002. Germline transformants spreading out to
many insect species. Advances in Genetics, 47, 49-86.
Catteruccia, F., Godfray, H.C.J. and Crisanti, A., 2003. Impact of genetic
manipulation on the fitness of Anopheles stephensi mosquitoes. Science, 299
(5610), 1225-1227.
Christophides, G.K., Vlachou, D. and Kafatos, F.C., 2004. Comparative and
functional genomics of the innate immune system in the malaria vector
Anopheles gambiae. Immunological Reviews, 198, 127-148.
Coates, C.J., Jasinskiene, N., Pott, G.B., et al., 1999. Promoter-directed expression of
recombinant fire-fly luciferase in the salivary glands of Hermes-transformed
Aedes aegypti. Gene, 226 (2), 317-325.
De Lara Capurro, M., Coleman, J., Beerntsen, B. T., et al., 2000. Virus-expressed,
recombinant single-chain antibody blocks sporozoite infection of salivary
glands in Plasmodium gallinaceum-infected Aedes aegypti. American Journal
of Tropical Medicine and Hygiene, 62 (4), 427-433.
Durvasula, R.V., Gumbs, A., Panackal, A., et al., 1997. Prevention of insect-borne
disease: an approach using transgenic symbiotic bacteria. Proceedings of the
National Academy of Sciences of the United States of America, 94 (7), 3274-
3278.
Ghosh, A.K., Ribolla, P.E.M. and Jacobs-Lorena, M., 2001. Targeting Plasmodium
ligands on mosquito salivary glands and midgut with a phage display peptide
library. Proceedings of the National Academy of Sciences of the United States
of America, 98 (23), 13278-13281.
Heinrich, J.C. and Scott, M.J., 2000. A repressible female-specific lethal genetic
system for making transgenic insect strains suitable for a sterile-release
program. Proceedings of the National Academy of Sciences of the United
States of America, 97 (15), 8229-8232.
Hogg, J.C. and Hurd, H., 1997. The effects of natural Plasmodium falciparum
infection on the fecundity and mortality of Anopheles gambiae s.l. in north
east Tanzania. Parasitology, 114 Part 4, 325-331.
Irvin, N., Hoddle, M.S., O'Brochta, D.A., et al., 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.
Ito, J., Ghosh, A., Moreira, L.A., et al., 2002. Transgenic anopheline mosquitoes
impaired in transmission of a malaria parasite. Nature, 417 (6887), 452-455.
Kim, W., Koo, H., Richman, A.M., et al., 2004. Ectopic expression of a cecropin
transgene in the human malaria vector mosquito Anopheles gambiae (Diptera:
Culicidae): effects on susceptibility to Plasmodium. Journal of Medical
Entomology, 41 (3), 447-455.
Kokoza, V., Ahmed, A., Cho, W.L., et al., 2000. Engineering blood meal-activated
systemic immunity in the yellow fever mosquito, Aedes aegypti. Proceedings
of the National Academy of Sciences of the United States of America, 97 (16),
9144-9149.

64
 Jacobs-Lorena

Moreira, L.A., Edwards, M.J., Adhami, F., et al., 2000. Robust gut-specific gene
expression in transgenic Aedes aegypti mosquitoes. Proceedings of the
National Academy of Sciences of the United States of America, 97 (20),
10895-10898.
Moreira, L.A., Ito, J., Ghosh, A., et al., 2002. Bee venom phospholipase inhibits
malaria parasite development in transgenic mosquitoes. Journal of Biological
Chemistry, 277 (43), 40839-40843.
Moreira, L.A., Wang, J., Collins, F.H., et al., 2004. Fitness of anopheline mosquitoes
expressing transgenes that inhibit Plasmodium development. Genetics, 166
(3), 1337-1341.
Olson, K.E., Higgs, S., Gaines, P.J., et al., 1996. Genetically engineered resistance to
dengue-2 virus transmission in mosquitoes. Science, 272 (5263), 884-886.
Spradling, A.C. and Rubin, G.M., 1982. Transposition of cloned P elements into
Drosophila germ line chromosomes. Science, 218 (4570), 341-347.
Thomas, D.D., Donnelly, C.A., Wood, R.J., et al., 2000. Insect population control
using a dominant, repressible, lethal genetic system. Science, 287 (5462),
2474-2476.
Turelli, M. and Hoffmann, A.A., 1991. Rapid spread of an inherited incompatibility
factor in California Drosophila. Nature, 353 (6343), 440-442.
Yoshida, S., Matsuoka, H., Luo, E., et al., 1999. A single-chain antibody fragment
specific for the Plasmodium berghei ookinete protein Pbs21 confers
transmission blockade in the mosquito midgut. Molecular and Biochemical
Parasitology, 104 (2), 195-204.
Zieler, H., Keister, D.B., Dvorak, J.A., et al., 2001. A snake venom phospholipase
A(2) blocks malaria parasite development in the mosquito midgut by
inhibiting ookinete association with the midgut surface. Journal of
Experimental Biology, 204 (23), 4157-4167.

65