WO2005042751A1 - Controlling the spread of infective agents - Google Patents

Abstract

Parasite-specific factors present in a host organism are activated in the presence of the parasite to kill the host.

Description

CONTROLLING THE SPREAD OF INFECTIVE AGENTS

FIELD OF THE INVENTION

The present invention relates to methods for controlling the spread of infective agents, especially parasites, by reducing compatibility with a vector or host therefor.

BACKGROUND OF THE INVENTION

Insect-borne diseases such as malaria, dengue fever, West Nile virus, trypanosomiasis (sleeping sickness), and many others, are major causes of morbidity and mortality in humans and animals around the world, and particularly in developing countries. One approach to this problem is genetically to manipulate the insect vectors such that they are no longer capable of transmitting the disease. Such insects are generally described either as "refractory to disease transmission", or as "refractory insects" for short. Such a strategy has two essential steps:

1. create a strain of insect in the lab which is refractory to a relevant disease, for example a mosquito which is incapable of transmitting malaria; and

2. cause this refractory gene or gene system to spread throughout the target population.

Many laboratories are working on the first step and significant progress has been reported (Adelman et al, 2002; Ghosh et al., 2002; Ito et al., 2002; James, 2002). However, although the molecular basis differs, each of these approaches is essentially a variation on a single approach, which is to attempt to make the mosquito less compatible with, or "resistant" to, the parasite. This may take the form of modifying the mosquito so that its cells fail to bind or take up the parasite, or to prevent the parasite from expressing its genes, etc., and may be likened to a vaccination program.

L. Alphey, M. Andreasen (Molecular & Biochemical Parasitology 121 (2002) 173 - 178) suggest that insects can be killed under any particular condition if a suitable promoter can be found that is responsive to infection. Using such a promoter, it might be possible to engineer dominant susceptibility to parasite infection, such that insects are killed by infection. The drawback to such a scenario is that insects do not recognise specific parasites but, rather, respond to a type of infection. Thus, a promoter that is activated in response to the malaria parasite would also be activated by other infections, so that insects which died when the promoter was activated would suffer a large fitness penalty in the wild, which would make it difficult to introduce the phenotype into the population.

Rather than make use of an insect mechanism for detecting any particular type of infection, it has now, surprisingly, been found that it is possible to use parasite functionality to trigger death of the insect.

SUMMARY OF THE INVENTION

Thus, in a first aspect, the present invention provides a genetic system for a host, said system being responsive to the presence of a parasite for said host, wherein functionality present in the parasite, but not in the uninfected form of the host, is effective to activate the system.

The parasite functionality suitable to trigger the genetic system in the host may take various forms. For example, the parasite may encode a protease which cleaves at a site not recognised by any of the host proteases. Such a site may then be engineered into a fusion protein, for example, wherein the fusion protein itself is either inactive or sequestered or securely compartmentalised, and wherein cleavage by the parasite protease serves to release an active protein.

Another suitable form of parasite functionality is a promoter which is active only in the presence of parasite products, such as a medium or late viral promoter. Such a promoter can be linked to a suitable genetic sequence to provide a response to parasite infection. Other forms of parasite functionality may include binding proteins and metabolites characteristic of the parasite which may be recognised by an otherwise silent host receptor which, on activation, provides a suitable response.

The genetic systems of the present invention generally comprise sensor and effector functions such that, when the sensor is in the presence of the parasite, the effector is triggered.

As indicated above, the sensor may take any suitable form, and effectively provides a link between the presence of the parasite and the effector function. It is generally preferred that systems of the present invention be kept as simple as possible, in order to avoid disruption by recombination events, for example.

The sensor may, for instance, be expressed primarily in cells, or in some part of a subset of cells, which come into contact with the parasite or which come into contact with products resulting from the presence of the parasite in the host. If sufficiently stable in the absence of the parasite, the sensor may be expressed in a wider range of cells, for instance all cells, and may even be expressed at all developmental stages.

However, if expression of the sensor, or a sensor-effector cassette, is deleterious to the cells which express it, the overall fitness cost may be minimised by restricting the expression to cells which contact the parasite or which come into contact with products resulting from the presence of the parasite in the host. For example, in the case of mosquito -borne diseases such as malaria or dengue fever, only the adult females are typically exposed to the parasite. Furthermore, not all tissues in said females are necessarily exposed, even in adult females. Therefore, the fitness cost associated with expression of the sensor may be minimised by restricting expression, for instance, to specific tissues or to adult females only.

The present invention envisages cascades, whether simple or complex, connecting the sensor and effector functions. A simple cascade might, for example, be where a parasite protease cleaves a previously inactive, or otherwise ineffective, unique activator or transcriptase for a promoter which, in turn, allows the transcription of RNAi which blocks the translation, for example, of a critical mRNA.

It is preferred that cleavage is of a protein and, furthermore, that the cleavage product of the protein is a toxin, or is a transcriptase for a toxin operon.

In one preferred form, the sensor may simply be a late promoter for a virus, for example, which is linked to lethal DNA, such as that encoding ricin, such that when the viral replication cycle reaches a certain stage, the promoter is activated, and ricin is produced, killing the cell and, depending on the effectiveness of the promoter, the whole organism.

Preferably, the promoter is part of an operon encoding an effector function.

It is not critical to the present invention what parasite functionality is used to generate the response, provided that the functionality is either not present in the host, or is dormant or in some other way incapacitated. In the case of a viral promoter, for example, this maybe specifically incorporated into the host cell line or cells in association with an effector, but is inactive in the absence of a transcription factor from the target parasite.

Proteases have been discussed above and are further discussed below. In addition, single chain antibodies linked to different fluorescent markers, for example, may recognise a parasite surface epitope, or epitopes, so that both will bind the parasite, enabling the presence of the parasite to be determined by Fluorescence Resonant Energy Transfer (FRET). Further applications of FRET are discussed below.

As will be apparent from the above, while it is generally preferred that the effector function be a lethal function, in order to eliminate pests infected with parasites from the wild, it is also possible to use the system of the present invention in order to track parasite infestation in a host species. Thus, the effector function may be a lethal function, or may provide any other functionality desired, such as a visible or behavioural marker. While it is generally preferred to provide an effector not generally found in the host species, this need not necessarily be the case, and it may be that the effector serves either to up-regulate or down-regulate an important host function, for example, thereby either serving to kill the host, or otherwise to incapacitate the host or provide some indication of parasite infection.

The term "lethal" is used herein to indicate a high fitness cost. In the case of insects, for example, this may include semi-lethal, or sterile, or semi-sterile, or flightless, or blind, or generally any condition which imposes a sufficiently high fitness cost on the target. It is preferred that the effector function leads to an inability to transmit the disease.

With regard to insects, effector functions leading to changes in host-seeking behaviour, for example a reduced ability to find a blood meal, are preferred. Furthermore, the effector function preferably leads to a reduced tendency to select humans as hosts.

Even where an effector is potentially 100% fatal to the organism, it may be expressed in low quantities after contact with the parasite such as simply to slow down, or incapacitate, the host through localised apoptosis, for example. Toxic compounds may also have a similar effect, and inhibiting crucial enzymes may likewise have such an effect, the only requirement being a substantially increased fitness cost.

Preferred parasites include Plasmodium sp., preferably Plasmodium falciparum, nematodes or viruses, preferably Baculo viruses or Flaviviruses.

Preferred hosts are fish, mammals, arthropods and plants. In the case of arthropods, it is preferred that the arthropod is a shrimp, for instance. More preferably, however, the arthropod is an insect, preferably an insect such as a bee, and most preferably a mosquito. Preferably, the mosquito is the Malarial mosquito Anopheles sp, preferably Anopheles stephensi. Alternatively, it is also preferred that the mosquito is the Dengue or Yellow Fever mosquito Aedes sp., and most preferably Aedes aegypti. Preferred fish and mammals include those commonly farmed by humans. In the case offish the host is preferably a bass, trout, cod, mackerel or salmon species. With regard to mammals, the host is preferably a cow, sheep or pig.

Preferred plants include crop plants, preferably maize, corn, wheat, barley, oats, rape, soya or corn.

Therefore, while the hosts of the present invention do not include humans, they may include anything else from the animal and plant kingdoms, all the way down to single-celled organisms.

In the case of insects, for example, it will often be the case that the systems of the present invention are most suitable simply to kill the insects in order to prevent transmission of disease, although the invention is not so limited.

In the case of plants, however, it may be that it is simply necessary to initiate the release of a dye, for example, in order to indicate infected plants, or to initiate localised tissue necrosis to prevent spread of the disease throughout the plant. It will be appreciated that whole-scale killing of the plant on simple infection is also contemplated by the present invention.

Preferably the system may be used to control parasitic diseases that threaten captive, domestic or farmed host populations of animals or plants. Examples include, but are not limited to, the production of silk moths, which are susceptible to viral infections (Watanabe 2002), or the protection of aquaculture species, such as penaeid shrimp (Loh et al 1997) and farmed fish (Goldberg et al 2003), from a variety of viral pathogens. Indeed, the management of honeybee populations infected by a range of micro- organisms (Williams 2000) is also envisaged. In such captive, farmed or domestic populations, it will be possible to ensure the proportion of individuals carrying the engineered genetic system, a, is maintained at a high value, preferably 1. The genetic system may be used to provide sentinel individuals used as an early warning surveillance system. For example, a visible rather than a lethal effector may be engineered, as discussed above, which maybe used to detect the presence of a specific pathogen, or a range of pathogens.

In crop plants, for example, genetically modified sentinel crops may be planted around larger stands of non-GM crops. Early detection of a pathogen in the sentinel crops may then trigger a series of pathogen and/or vector control methods, such as chemical spraying.

The effect of "Death on infection," discussed below, on the control of directly transmitted pathogens in captive, domestic or farmed populations may be illustrated using a simple mathematical model (Anderson & May 1991, pages 122-123), see Example 2.

The present invention further provides a host organism transformed with a system of the present invention. Preferred hosts are, as mentioned above, fish, mammals, arthropods and plants, especially arthropods and, most preferably, insects.

Therefore, although it is particularly preferred that the host is an insect, the present invention may be applied to a range of hosts. Furthermore, although the present application discusses exiamples of selector and effector functions, often in regard to insects, it is to be understood that this is by way of example and systems according to the present invention may be used in other hosts, as would be apparent to the skilled person.

Further provided are vectors comprising systems of the present invention, capable of transforming a host. Therefore, the present invention provides vectors capable of integrating genetic information into a host genome such as those based on transposons, or retrovirαses, or capable of inducing integration by homologous, site- specific or non-homologous recombination. Whilst retroviral vectors, such as adenovirus, are envisaged, it is preferred, in the case of insects, that the vectors comprise transposons, whilst Agrobacterium or vectors suitable for homologous recombination are preferred for plants. A plasmid-based system, which does not integrate but is maintained episomally, is preferred for microbes, although not for multi-cellular organisms, where such systems tend not to be stable. Also envisaged are paratransgenesis systems, where the system according to the present invention is used in a gut bacterium, Wolbachia, or other commensal or symbiotic bacterium, for instance.

The present invention further provides methods for transforming a host with a system of the present invention, and especially a method for transforming the germ line of the host. In this respect, it is often preferred to use a selectable marker in association with the system of the invention, in order to be able to select individuals transformed with the system of the invention. Suitably selected individuals may then be established as breeding colonies which may then be released into the wild, preferably in combination with an introgression program (Braig and Yan (2001), (Gould and Schliekelman, 2003)).

DECRTPTION OF THE DRAWINGS

Figure 1. Illustration of the dynamic and equilibrium properties of the simple model, as well as the impact of introgressing the "Death on Infection" trait into the population. At start of the simulation there is an introduction of an infectious host into the otherwise susceptible population. The proportion of infected hosts and vectors increases up to some stable level. After 5 years, the intervention is introduced such that 50% of the population is introgressed with the "Death on Infection" trait that reduces the life expectancy of the vector, following infection, by 50%. The parameter values, taken to represent dengue fever in a low transmission setting (e.g. Singapore), are m=20, l///=30 days, l/α=3 days, l/r=5 days, b=l and c=0.005, giving an ?₀=1.667.

Figure 2. The impact of an early introduction of intervention following an initial outbreak. Parameters as Figure 1 except β=0.85, although intervention is earlier. Figure 3. The impact of introducing the intervention at 0.25 years, 0.5 years and 1 year following the initial outbreak, relative to no intervention. The parameter values are as in Figure 2.

Figure 4. The effect of Ro on transmission. The line represents the isocline of R₀=l. Above this line the combinations of a and β result in the interruption of transmission (i.e. R₀<1) while below the line transmission persists (i.e. Ro>l). The parameter values are as in Figure 1 with an Ro— 1.667.

Figure 5. The effect of incubation on Ro and, thereby, on transmission. As in Figure 4 the lines represent the isoclines of i?o—l for two scenarios of no incubation period in the vector and a _T=10 day incubation period. In both scenarios the value of Ro was maintained at 1.667 by adjusting the value of w=30 in the incubation scenario. All other parameter values remain the same as in Figure 1. Note the incremental benefits of the intervention under the incubation period, i.e. the greater amount of a and β parameter space under which transmission cannot occur.

Figure 6. The dependence of the critical value of ? required to halt transmission on the basic reproductive number Ro- Here, 1/μ = 200 days, 1/r = 150 days, N= 1000 and λ is allowed to vary according to the relationship R₀/N(r + β) i.e. the situation in a wild-type captive population in which β=0.

Figure 7. An example of a sensor/effector construct is shown.

DETAILED DESCRIPTION OF THE INVENTION

Essentially, a benefit of the present invention is that it makes the presence of the parasite deleterious or even fatal to the host, such as a mosquito, in either a cell specific manner, or to the whole organism. These are broadly exemplified, as follows: (i) kill the cell fast enough that it prevents replication of the parasite, or (ii) kill the mosquito before it becomes infectious (for dengue or malaria this is 10 or so days from taking an infected blood meal). The former is particularly applicable to intracellular parasites such as viruses, but may also be applicable to extracellular parasites, while the latter may be used, for example, for dengue or malaria, where the incubation period is 10 days or so.

The first is reminiscent of programmed cell death, or apoptosis, where a cell commits suicide to save the rest of the organism. There is substantial literature on this. For example, Takaoka et al. (Nature, 424:516), discusses interferons, p53, tumour suppression and (natural) anti-virus defence in mice. They conclude that p53 has an important role in complex innate antiviral host defences, and that prompt induction of apoptosis of virus-infected cells via p53 activation is beneficial to the host in limiting virus replication. Apoptotic cells are swiftly engulfed by phagocytic cells in vivo, further inhibiting the spread of virus.

The second can be understood in terms of elimination. A dead mosquito will not infect anyone. If every dengue-infected mosquito throughout an area is killed, and this is kept up, dengue infection can be prevented in that area.

Indeed, not every mosquito needs to be killed, only a sufficient proportion. Without being bound by theory, according to models of vector borne disease, if a sufficient proportion of mosquitoes are killed quickly enough, it is thought that Ro (the basic reproductive rate) would be forced to a value less than 1. If Ro falls below 1, then the population is not replacing itself and so tends to decline. More generally, killing any dengue-infected mosquitoes will tend to reduce infection, and so reduce Ro and, indeed, the burden of disease. If enough mosquitoes can be killed, and this threshold will vary from one location to another, then Ro may be reduced locally below 1 and so break the transmission cycle in that area. It will be appreciated that the threshold proportion and time of killing may need to be varied between environments and diseases.

However, since only about 1% of Aedes aegypti mosquitoes, which is the main dengue vector, in an area are infected, even in highly endemic areas where this can occasionally reach as high as 2.5%, such a program would have a minimal impact on the overall mosquito population. While genetic systems conferring lethality are not normally easy to drive through a population, if only -1% of the population are in fact killed, then this is not a large fitness penalty, especially if only females are affected.

A further benefit of the present invention is that it may be used as a method of identifying or tracking infection in a host, as mentioned above.

Either version of the system requires a parasite sensor, and an effector, as illustrated above. Ideally, the sensor and effector are closely connected. It is preferred that the sensor has high sensitivity and accuracy. It is also preferred that both the sensor and the effector have a high signal-to-noise activity ratio.

In a preferred embodiment, the parasite sensor is a parasite-encoded function that the host does not have. This is particularly suitable for viruses, but may also be used for cellular and multicellular parasites, such as Plasmodium or nematodes. Plasmodium, for example, is known to secrete a number of proteins including chitinases, as well as expressing a variety of proteins on its surface. A parasite sensor may also be constructed based on a specific parasite-binding protein, for example a single-chain antibody, where the altered conformation of the protein is used as the basis for a detection system.

Baculoviruses, for example OpMNPV and ylcMNP V, have a cascade of gene expression following infection, such that the "immediate early" genes are transcribed immediately, then "early", "intermediate" and so on. Part of the regulation of this system is that later gene expression depends on the protein products of earlier genes. For example, the iel and ie2 genes of both OpMNPV and^cMNPV are transcription factors. Therefore, a suitable later promoter from a baculovirus, operably linked to a sequence encoding a reporter or effector molecule, forms a baculovirus sensor. The promoter is silent in the absence of baculovirus-encoded factors, but on infection these are provided and the sensor is activated. It will be appreciated that this arrangement provides a tight connection between the virus sensor and the effector. A suitable effector, in this case, might be, for example, a pro-apoptotic gene, or ricin, or any other strongly deleterious gene, or an anti-viral protein, such as are illustrated below. Alternatively, a reporter such as EGFP allows easy tracking of the progress of virus infection through the host, or of viral spread through a population.

Flaviviruses are a group of viruses which include dengue fever, yellow fever, West Nile and various encephalitis viruses, and these have a much simpler genome than baculoviruses, without the cascade of gene expression. Dengue and all other flaviviruses encode only a single polyprotein, which is then processed into about 10 proteins (C preM, E, NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5, arranged in that order in the genome). Some processing is done by host-encoded signal peptidase, but some is done by NS3, a multi-functional protein of which the NH -terminal domain is a trypsin-like serine protease. Full activity of NS3 requires NS2B as a cofactor. NS3 can act in trans, though it may also act in cis. For some of the polyprotein junctions, for example NS4B/5, there is no host enzyme that will cleave this site. NS3 is active against short peptides corresponding to the junction sequence, as well as longer proteins (Leung et al, 2001; Yusof et al., 2O00). Accordingly, the NS2B/NS3 protease provides a virus-encoded function useful in the present invention. A suitable form of such a sensor/effector construct is shown in Figure 7:

The presence of a signal peptide and transmembrane domain, in this particular construct, serves to anchor the sensor protein in the membrane. This restricts the subcellular localisation of the sensor. Other anchoring systems are envisaged, for example heat shock proteins through a binding protein (such as GeneSwitch), to mitochondria, nuclei (nuclear localisation signal) etc. On exposure to NS3, for example following infection of the cell by the virus, the linker will be cut and the effector/reporter released from the membrane.

In a preferred embodiment, the linker region contains a plurality of protease cleavage sites. This allows a single molecule to detect multiple viruses, for example dengue viruses 1-4, yellow fever, etc., releasing the same effector molecule in each case. In this way, a mosquito strain simultaneously refractory to multiple viruses could be constructed relatively easily, and with a very compact, single construct. It will be appreciated that such a construct may also be used to include multiple cleavage sites derived from a single virus type, thereby to reduce the possibility of resistance by a slight change in the recognition site of the protease.

A suitable non-lethal effector may use FRET (Fluorescence Resonant Energy Transfer). FRET involves the transfer of energy from one fluorochrome to another, and requires them to be very close together. FRET between ECFP and EYFP results in absorption of short-wave (for example 405nm) radiation by the ECFP, transfer of energy to EYFP (rather than emission in cyan) and emission by EYFP (yellow). This can be observed by connecting the two with a protease-sensitive linker. FRET will break down in the presence of the protease, giving increase emission in cyan and decreased in yellow.

Such systems, and various other sensor configurations, may be used in vitro or in tissue culture cells, allowing the presence of parasite to be detected and potentially quantified.

As indicated above, the nature of the effector is not critical to the present invention. In the case of a transcriptional response system, as described above for baculovirus, any desired gene may be expressed, for example a hairpin RNA to silence virus gene expression, or a pro-apoptotic gene such as hid, grim, reaper, activated caspase or Bax, for example. Alternatively, ricin, the anti-ribosome action of which would help suppress virus protein production as well as tending to kill the cell, or diphtheria toxin. These examples are substantially cell autonomous in effect, while others are not, such as activated Ras or secreted proteins, such as peptide hormones, signalling molecules, or toxins. Cell-autonomous effectors may be rendered non cell- specific by causing them to be secreted, such as by the addition of a signal sequence, and optionally adding a cell-penetrating domain such as that of Antp.

A protease sensor may be used to provide a transcriptional response by using a transcriptional activator as the effector molecule, for example GAL4 or GAL4-VP16, which, when released from the membrane, will activate expression of a sequence operably linked to its UAS_GALΛ binding site (Brand et al., 1994; Brand and Perrimon, 1993). A nuclear localisation signal may be added to GAL4, although it works well without one.

The protease sensor can be used directly, by cleaving an effector molecule such as ricin or Bax. This has the advantage of not requiring host cell transcription, which can be advantageous where the virus suppresses host cell transcription. As some viruses suppress the RNAi response, and/or suppress apoptosis, the effector should be selected appropriately. Thus, Bax may be preferable over reaper, for example, as Bax is several steps further down the pathway, and less sensitive to repression. Another suitable effector includes a member of the caspases, as these are normally inactive and are activated by a proteolytic cascade. This proteolytic activation can be directly mimicked by a protease sensor, releasing active caspase. This has the advantage that the signal is amplified by the catalytic action of the released protease. The use of a transcriptional activator, as described above, also has this advantage of amplification of the signal. Another variation with amplification is to use a caspase, for example, as the effector, and include a cleavage site for that caspase in the linker, so that activating a few molecules will set off a positive feedback amplification. A possible drawback in this case is the extreme sensitivity to low levels of background activation.

In the case of proteolytic activation of a substance such as caspase, it will be appreciated that membrane association is not necessary. Caspase is inactive before proteolytic activation, so a caspase variant with an activating cleavage site sensitive to a parasite-encoded protease would not need to be sequestered at a location where it cannot function. Such sequestering may be useful, however, as this adds an additional level of noise-reduction.

An alternative signal amplification mechanism would be to use a site-specific recombinase, such as ere, as the effector. Released from the membrane, it would enter the nucleus and promote recombination between suitable sites, for example lox sites. A strong expression cassette with a stuffer fragment flanked by these sites could be activated, leading to high level expression of a reporter or effector. Silencing of ere by GeneSwitch-type anchoring has been demonstrated in Drosophϊla (Heidmann and Lehner, 2001). Other suitable effectors include inhibitors of signal protease, or any other function that the virus needs for its replication. Although this would kill the cell more slowly, virus replication is prevented in the mean time. Such an inhibitor may be a protein inhibitor or a transcriptional response, such as RNAi.

Multi-functional effectors may be constructed, either multifunctional proteins or self processing ones, for example caspase and some other functional domains separated by caspase cleavage sites.

One embodiment of the present invention is provided by a protease sensor for dengue 3 NS2B/NS3 protease, where a signal peptide and at least the transmembrane domain from CD8a are linked to protease linkers of various sizes, and a reporter molecule such as DsRed2 or EGFP, linked to a nuclear localisation signal. Tissue culture cells transfected with such a construct show fluorescence at the cell membrane. On exposure to the virus, the linker is cleaved, leading to an accumulation of fluorescence in the nucleus and, where the cleavage is very efficient, a reduction in fluorescence at the cell membrane.

Preferably, the parasite sensor rapidly and sensitively detects infection and kills the cell or organism with 100% reliability. However, in most cases, this is not necessary for effectiveness as a disease control tool. Slow or ineffective killing of a cell may permit parasite replication, but as the parasite spreads through the host, many cells will be affected and the fitness or viability of the host compromised. For mosquito- borne diseases, even a modest reduction in viability of the infected mosquito can have a dramatic effect on disease transmission, indeed this is a parameter highlighted as one to which transmission is particularly sensitive (Anderson and May, 1991; Macdonald, 1957; Ross, 1911).

Other refractory systems, previously described, aim at reducing the ability of the parasite to infect the host, such an insect, and/or to replicate within it. These are entirely compatible with the system described herein. For example, an RNAi system against dengue may be designed to destroy the dengue genome and/or transcripts and so prevent virus replication. This prevents virus gene expression, and so stops the virus sensor triggering. As the cell or organism will then live, the RNAi system reduces the fitness cost of the system of the invention. Should the RNAi system fail, the system of the invention is present as a fall-back. Even a partial effect of the RNAi system, inadequate to prevent virus replication, may be beneficial if it slows virus replication and thereby provides more time for the system of the invention to operate, or reduces the amount of virus released before the cell becomes incompetent for virus replication. Such considerations also apply to single-chain antibodies, which inhibit parasite function by steric interference, or by marking the parasite for attack by complement or other components of the mammalian immune system contained within the blood meal.

Systems of the present invention will be associated with a fitness penalty, although it will often be a small one, and will tend not to spread through a population if introduced at a relatively low frequency. While inundative release, particularly with insects carrying multiple copies of the system, or versions of the system, could result in a high proportion of insects carrying at least one copy of the system, this is unlikely to lead to fixation of any one copy of the system, and will typically lead to loss of the system from the population when the inundative release is stopped. Such an approach is useful, however, and may be used for a trial program, for example, or to protect a limited area.

It is preferred that large-scale use involves additional genetic elements to promote the spread of the system through a population and/or its maintenance at high frequency. Such systems are known as gene drivers, as they are intended to drive genes into populations, and introgress the genes into populations (Braig and Yan, 2001; Gould and Schliekelman, 2003).

The systems of the present invention are compatible with any gene driver systems known in the art. It is preferable to provide a tight linkage of the drivers with the cargo, otherwise, if this linkage is broken, the gene driver will tend to spread itself through the population without the cargo. One method for introgressing a gene into a population comprises a method for introgressing a genetic sequence into a target population, the method comprising the use of at least two different lethal-suppressor pairs, wherein at least one lethal from a first pair and at least one suppressor from a second pair are linked to the genetic sequence, at least one lethal-suppressor pair being effective through RNAi.

Another method for introgressing a system of the invention into a host population, comprises the use of an inheritable lethal factor and an inheritable suppressor therefor, wherein the system is linked to at least one of the lethal factor and the suppressor.

Both of the above methods provide the possibility of tight linkage between the driver system and the cargo. This may be contrived by arranging that the cargo is an integral part of the gene driver; preferably the part that is under positive selection. Each of these systems uses one or more lethals with a corresponding suppressor.

A system of the present invention may be integrated with these or other gene drivers in various ways. A preferred example of a Dengue sensor, as described above is for a so-called "killer" element, in gene driver system, where RNA is used against an essential transmembrane protein.

The essential transmembrane protein can be suppressed by expressing a full- length, functional transmembrane protein with silent substitutions that make it insensitive to the RNAi. The virus sensor can then be constructed by using this transmembrane protein with a fusion to the protease-sensitive linker and killer. Embedding the linker and effector into the transmembrane protein, such that they are flanked by essential domains of the transmembrane protein, reduces the risk of inactivation by deletion.

In one preferred embodiment, cleavage of the original transmembrane protein is made lethal, i.e. have a domain of the transmembrane protein act as the effector. Notch is though to work by proteolytic cleavage leading to a fragment entering the nucleus. The sequence-modified version of the transrαembrane protein, with plasmid sensor, provides the suppressor in this system, and hence is the element under positive selection, as required.

A preferred variant of this is to use a substance normally activated by proteolysis, such as caspase or elements of the Notch and other signalling pathways, and substitute a parasite-protease sensitive site for the normal cleavage site, or add it to the normal site. The killer in the gene driver system would then be RNAi against the endogenous caspase, the suppressor and refractory gene would both be a sequence- modified version of the caspase (insensitive to the RNAi) with the additional parasite- sensitive cleavage site.

A further preferred example is the case of a gene driver with a lethal protein and an RNAi suppressor directed against some part of the transcript expressing the lethal protein, the target region for the RNAi in the transcript encoding the lethal protein could be flanked by lox sites, and ere used as the effector. This links the system of the invention to an RNAi-based refractory system itself integrated with a gene driver.

The invention is not to be limited by what has been particularly shown and described, except as indicated by the appended claims. Indeed, while the invention will now be illustrated in connection with the following Examples, it will be understood that it is not intended to limit the invention to these particular embodiments. On the contrary, it is intended to cover all alternatives modifications and equivalents, as may be included within the scope of the invention as defined by the appended claims.

Example 1 Transmission model of "Death on Infection"

The basic framework

The Ross-Macdonald model (Anderson and May, 1991; Macdonald, 1957; Ross, 1911) captures the essentials of vector-bome disease transmission but excludes many of the complicating details.

The model consists of two linked differential equations describing changes in the proportion of infected (≡ infectious) humans, x, and the proportion of infected (≡ infectious) vectors, y.

— = [ab(l - )my(l - x)]- rx dt

Here, m is the number of female mosquitoes per human host, a is the biting rate on humans (so that 1/α is the average interval between blood meals), b is the proportion of infectious bites on susceptible humans that produces a patent infection, r is the per capita rate of human recovery from infection (such that 1/r is the average duration of infection), c is the proportion of susceptible mosquitoes feeding on infected humans that develop a patent infection, and μ is the per capita mosquito mortality rate (1/μ is the average mosquito life expectancy).

The model assumes stable human and vector population sizes, so the dynamical variables are the proportion infected in each population (as shown in Figs 1-3).

The basic reproduction number for this simple model is given as ma²bc

K₀ = rμ and the equilibrium prevalence of infection in the human host, x^*, and vector, v , populations are

x = ^■ Ro - 1 R₀ +(ac/μ)

and

respectively,

Including "Death on Infection" as a public health intervention

The basic model can be simply extended to include the effects of introgressing an engineered "Death on Infection" trait through a proportion, α, of the vector population. Following infection with the vector borne pathogen, the average life expectancy of the vector carrying the genetically engineered trait is reduced by a given proportion, β, relative to the life expectancy of the wild vector. In all respects other than life expectancy following infection, it is assumed that the wild and genetically engineered vectors are similar. It is also assumed that the intervention does not alter the total vector and human population sizes (i.e. immediate density-dependent replacement of dead vectors).

The model is extended to include three equations, describing temporal changes in the proportion of human hosts, x, wild vectors, y_w, and vectors carrying the genetically engineered trait, y_g. dx

— = [ab(l - )my_w(l - x)]+ [abamy_m (l - x)]- rx

acx(l-y_w) -μy_H dt d g = _ acg(l-y_w) - μ dt (ι-^β)

An example of the impact of the intervention on the prevalence dynamics is shown in Figure 1.

The basic reproduction number in this modified version of the simple vector transmission model is the sum of the component Ros contributed by the wild and engineered vector populations, and is defined as

which simplifies to

Defining transmission interruption

Setting Ro to 1 and solving for a gives the critical proportion of the vector population that must express the genetically modified trait, , for any given value of β, so that transmission is interrupted. Following this rationale gives

ma bc The above expression succinctly defines the parameter space under which the "Death on Infection" intervention is capable of driving Ro<l, thereby interrupting transmission, as shown in Figure 4. Importantly, if the proportional reduction in life expectancy is high, then it may be possible to halt disease transmission by introgressing the trait to only a fraction (i.e. a<l) of the population. But it is equally important to note that the absolute minimum level of introgression occurs when , i.e. engineered vectors die as soon as they pick up the pathogen. Under these optimal conditions the critical level of introgression into the population necessary for transmission interruption is at a minimum and is given by =l-l/Ro. [As (ma²bc-ru)lma²bc = l-(ru/ma²bc) = 1-1/Ro].

For example in Figure 4, the Ro for the pathogen in the absence of any control is 1.667. If the 30-day average life expectancy of a wild vector was reduced to only 6 days in an engineered insect following infection (i.e. a proportional reduction of β=0.S), then the "death on infection" genetic trait would have to be expressed in a =0.5 of the population in order to halt transmission. If death were instantaneous following infection (i.e. β=l), then =0.4 (=1-1/1.667).

Obviously the critical threshold for β and a required to drive Ro<l will vary between settings and diseases.

How does this differ from "classical" refractoriness?

Model refractoriness can be simplified by saying a proportion of the vector population are refractory to infection (i.e. are not capable of transmitting the pathogen) and so the effective reproductive number, R, is simply

(l — a)ma²bc

R = (l -α)R₀ rμ

and the critical proportion of the population that must be refractory to interrupt transmission, a , is simply found by setting the above expression to 1 and solving with respect to a, to give a = 1

So, under the simplest set of assumptions, "Death on Infection" approximates to a classical refractoriness model when the mortality effects are instantaneous upon infection. When β<l then the "Death on Infection" traits requires a higher level of population introgression to result in transmission interruption relative to a fully refractory trait.

Transmission reduction

This analysis has only looked at transmission interruption (i.e. an all or nothing outcome). However, there are potentially huge public health benefits to be gained from reducing Ro from a high to a lower level, even if it still persists above 1. The model set out above can also be used to quantify these benefits, for example, in terms of percentage reduction in pre-intervention equilibrium prevalence levels in the human population.

Beyond the simple model - incubation period

The first step in making the model more biological realistic is to include an extrinsic incubation period, T, in the vector.

Following a similar rational as above, the R₀ for the infection in the absence of any intervention is

_D ma be __uT

R₀ = exp rμ

The introduction of genetically engineered vectors into the wild population results in Ro being modified to

Again, following the same reasoning as above, we can set Ro equal to 1 and solve for a to give an analytical expression for the critical threshold combinations of a and β that result in Ro falling below 1 and so transmission interruption. The resulting expression is

Again, however, when the "Death on Infection" trait is optimal (i.e. β=\) the critical level of introgression again is still a =l-l/Ro, which is also the value required for a fully refractory trait to block transmission.

Importantly, the intervention is more effective, relative to the no incubation period model, for values of β<\. This is because the term for mortality also appears as an exponential in the definition of Ro because an infected vector has to live long enough to develop infectiousness. The impact of the effect of incubation period is illustrated in Figure 5, where the overall Ro under the no incubation and incubation periods are the same (both set at 1.667).

Example 2 The application of "Death on Infection" in captive/domestic populations

The effect of "Death on infection" on the control of directly transmitted pathogens in captive, domestic or farmed populations may be illustrated using a simple mathematical model (Anderson & May 1991, pages 122-123) that describes the total number of susceptible (S), infectious (1) and recovered immune (R) host animals over time (t). If we assume Type II survival (an accepted assumption for most arthropod populations), with a constant mortality rate μ, and that all animals carry the "Death on infection" trait (i.e. α=l) then the following set of linked, first-order, differential equations maybe used to describe host dynamics

~ = μN-λSI-μS dt ^

*L _{= λsI}__rI_ μ τ dt (1-β)

^ T v dt

Here we assume that all births are into the susceptible class, and that births directly balance all deaths, so that the total population size is constant and given by:

N= S + /+R.

The rate of recovery from infectious to immune status is r, (such that 1/r is the average duration of infectiousness), and λ is the transmission co-efficient. As before, prefers to the proportional reduction in the life expectancy (\lμ) of the animal carrying the genetically engineered trait following infection.

The basic reproductive number, Ro, may then be shown to be

The critical level of β required in order to force Ro<l and so interrupt pathogen transmission in the captive/domestic population may be found by setting the above expression equal to 1 and solving with respect to /?to give

_ r+μ-λN critical r — λ ΛNI T

Figure 6 shows how this critical value of β increases with increasing magnitude of Ro.

In this example, 1/μ = 200 days, 1/r = 25 days, N= 1O00 and λ is allowed to vary according to the relationship Ro/N(r + β) i.e. the situation in a wild-type captive population in which β=0

The basic model described above may also be extended to include an incubation period and also natural pathogen-associated mortality (see Anderson & May 1992, Chapter 4).

References

Adelman, Z., Sanchez- Vargas, I., Travanty, E., Carlson, J., Beaty, B., Blair, C, and Olson, K. (2002). RNA silencing of dengue virus type 2 replication in transformed C6/36 mosquito cells transcribing an inverted-repeat RNA derived from the virus genome. J Virol 76, 12925-12933.

Anderson, R., and May, R. (1991). Infectious diseases of humans (London, Oxford University Press).

Braig, H., and Yan, G. (2001). The spread of genetic constructs in natural insect populations. In Genetically Engineered Organisms: Assessing Environmental and Human Health Effects (Boca Raton, CRC Press).

Brand, A., Manoukian, A., and Perrimon, N. (1994). Ectopic expression in Drosophila. Meth Cell Biol 44, 635-654.

Brand, A., and Perrimon, N. (1993). Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401-415.

Ghosh, A., Moreira, L., and Jacobs-Lorena, M. (2002). Plasmodium-moscpxito interactions, phage display libraries and transgenic mosquitoes impaired for malaria transmission. Insect Biochem Mol Biol 32, 1325-1331.

Goldberg, T.L., Coleman, D.A., Grant, E.G., Inendino, K.R., Philipp, D.P. (2003) Strain variation in an emerging iridovirus of warm-water fishes. J Virol 77, 8812-8818.

Gould, F., and Schliekelman, P. (2003). Population genetics of autocidal control and strain replacement. Annu Rev Entomol 49, RiA: doi:10.1146/annurev.ento. 1149.061802.123344.

Heidmann, D., and Lehner, C. (2001). Reduction of Cre recombinase toxicity in proliferating Drosophila cells by estrogen-dependent activity regulation. Dev Genes Evol 211, 458-465. Ito, J., Ghosh, A., Moreira, L., Wimmer, E., and Jacobs-Lorena, M. (2002). Transgenic anopheline mosquitoes impaired in transmission of a malaria parasite. Nature 417, 452- 455.

James, A. (2002). Engineering mosquito resistance to malaria parasites: the avian malaria model. Insect Biochem Mol Biol 32, 1317-1323.

Leung, D., Schroder, K., White, H., Fang, N.-X., Stoermer, M., Abbenante, G, Martin, J., Young, P., and Fairlie, D. (2001). Activity of recombinant dengue 2 virus NS3 protease in the presence of a truncated NS2B co-factor, small peptide substrates and inhibitors. J Biol Chem 276, 45762-45771.

Loh, P.C., Tapay, L.M., Lu, Y. and Nadala, E.C. Jr. (1997) Viral pathogens of the penaeid shrimp. Adv Virus Res 48, 263-312.

Macdonald, G. (1957). The epidemiology and control of malaria (London, Oxford University Press).

Ross, R. (1911). The prevention of malaria (2nd edition), 2nd edn (London, Murray).

Yusof, R., Clum, S., Wetzel, M., Krishna Murphy, H., and Padmanabhan, R. (2000). Purified NS2B/NS3 serine protease of dengue virus type 2 exhibits cofactor NS2B dependence for cleavage of substrates with dibasic amino acids in vitro. J Biol Chem 275, 9963-9969.

Watanabe, H. (2002) Genetic resistance of the silk worm, Bombyx mori to viral diseases. Current Science 83, 439-446.

Williams, D.L. (2000) A veterinary approach to the European honey bee (Apis mellifera) Vet J 160, 61-73.

Claims

CLAIMS:

1. A genetic system for a non-human host, said system being responsive to the presence of a parasite for said host, wherein functionality present in the parasite, but not in the uninfected form of the host, is effective to activate the system.

2. A system according to claim 1, comprising a parasite sensor and an effector.

3. A system according to claim 1 or 2, wherein the functionality is responsive to the presence of the parasite and triggers an effector function.

4. A system according to claim 3, wherein the effector is detrimental to the cell or the host organism.

5. A system according to claim 4, wherein the effector is lethal.

6. A system according to claim 3, wherein the effector provides a marker for the presence of the parasite.

7. A system according to any preceding claim, wherein the functionality is a parasite-specific enzyme.

8. A system according to any preceding claim, wherein the functionality is a protease whose cleavage site is not shared with that of the selected host.

9. A system according to claim 7 or 8, wherein the action of the enzyme or protease releases or activates an effector function, or triggers a cascade to provide the effector function.

10. A system according to claim 9, wherein the protease cleaves a protein, preferably a fusion protein, anchored in a cell membrane.

11. A system according to claim 10, wherein the protein has a plurality of cleavage sites, either for variants of one protease, or for different proteases, or both.

12. A system according to claim 10 or 11, wherein the cleavage product of the protein is a toxin, or is a transcriptase for a toxin operon.

13. A system according to any of claims 1 to 6, wherein the functionality is a promoter only transcribed by the host in the presence of the parasite.

14. A system according to claim 13, wherein the functionality is a parasite-specific promoter.

15. A system according to claim 13 or 14, wherein the promoter is part of an operon encoding an effector function.

16. A system according to claim 15, wherein the effector is a marker.

17. A system according to claim 15, wherein the effector is a toxin.

18. A system according to claim 15 or 17, wherein the effector is RNAi.

19. A system according to any preceding claim, in association with a second refractory system against the selected parasite.

20. A system according to any preceding claim, wherein the parasite is Plasmodium sp.

21. A system according to any of claims 1-19, wherein the parasite is a virus.

22. A system according to claim 21, wherein the virus is a Baculovirus or a Flavivirus.

23. A system according to any of claims 1-19, wherein the parasite is a nematode.

24. A host organism transformed with a system according to any preceding claim.

25. A host according to claim 24, which is an arthropod.

26. A host according to claim 25, wherein the arthropod is an insect.

27. A host according to claim 26, wherein the insect is a mosquito.

28. A host according to claim 24, which is a plant.

29. A host according to claim 24, wherein the plant is a crop plant.

30. A host according to claim 24, which is a farmed mammal or fish.

31. A vector comprising a system according to any of claims 1 to 19.

32. A method for transforming a host with a system according to any of claims 1 to 19, comprising transforming the germline of the host.

33. A method for introgressing a system according to any of claims 1 to 19 into a host population, the method comprising the use of at least two different lethal- suppressor pairs, wherein at least one lethal from a first pair and at least one suppressor from a second pair are linked to the genetic sequence, at least one lethal-suppressor pair being effective through RNAi.

34. A method for introgressing a system according to any of claims 1 to 19 into a host population, comprising the use of an inheritable lethal factor and an inheritable suppressor therefor, wherein the system is linked to at least one of the lethal factor and the suppressor.