Abstract
Anopheline mosquito species are obligatory vectors for human malaria, an infectious disease that affects hundreds of millions of people living in tropical and subtropical countries. The lack of a suitable gene transfer technology for these mosquitoes has hampered the molecular genetic analysis of their physiology, including the molecular interactions between the vector and the malaria parasite. Here we show that a transposon, based on the Minos element1 and bearing exogenous DNA, can integrate efficiently and stably into the germ line of the human malaria vector Anopheles stephensi , through a transposase-mediated process.
Genetic and genomic information on malaria vectors has expanded substantially during the last few years. However, further progress is hampered by the lack of germline transformation, a key technology in functional studies. In the fruitfly Drosophila melanogaster the development of gene transfer techniques has promoted, in a short time, a large flow of functional information on genes involved in embryogenesis, tissue modelling, intracellular signalling, neuronal organisation, behaviour and innate immunity. This success has prompted the development of methods for introducing exogenous genes into the genome of insect pests of medical and agricultural importance. Success has only been achieved in the last five years, with the transformation of the mediterranean fruitfly Ceratitis capitata2, the yellow fever mosquito Aedes aegypti3, 4, and the flour beetle Tribolium castaneum 5.
We tested whether the Minos transposable element1 from Drosophila hydei could integrate by active transposition into the germ line of the human malaria vector Anopheles stephensi. Minos was chosen on the basis of the findings that Minos transposase can mediate precise insertions in the genome of Anopheles gambiae cell lines and permits interplasmid transposition in A. stephensi embryos6. We have developed a plasmid vector, pMinEGFP, in which an enhanced version of the green fluorescent protein (EGFP)7, 8 gene was placed within the inverted repeats of Minos under the control of the actin5c promoter from D. melanogaster (Fig. 1a). In five consecutive experiments (I–V) (Table 1), 885 A. stephensi embryos were injected with a mixture of pMinEGFP and the helper plasmid pHSS6hsILMi209. The helper plasmid contains an intron-less version of the Minos transposase gene controlled by the heat shock protein 70 (hsp70) promoter of D. melanogaster and provides the enzyme for Minos transposition. The injections were performed on preblastoderm embryos following a standard procedure10 with an important modification. To prevent the hardening of the chorion, A. stephensi females were allowed to lay eggs in a 0.1-mM solution of p-nitrophenyl p'-guanidinobenzoate (pNpGB), an inhibitor of the enzyme phenoloxidase and thus of the first steps of the melanization process. In this solution, embryos remained soft and transparent for up to 3 h, during which time microinjections were performed.
Figure 1: The Minos element integrates into the A. stephensi genome.
a, Map of the transformation vector pMinEGFP. actinP, D. melanogaster actin5c promoter; hspP, D. melanogaster hsp70 promoter. hspT, D. melanogaster hsp70 terminator sequence. The EGFP, ampicillin-resistance (AmpR) and hygromycin-resistance (HygR) genes are indicated by black arrows and the left (ML) and right (MR) arms of Minos are in green, with the inverted repeats represented by black triangles. H, HindII; E, EcoRI; N, NotI. Black bars represent the three probes (M, EGFP and AMP) used in Southern blot experiments. b, Confocal fluorescence and transmission microphotographs of putative heterozygous (left) and homozygous (middle) transgenic larvae expressing EGFP compared with a wild-type mosquito (right). c, Southern blotting of genomic DNA from transgenic families IV, VD12, VD13, VD14 and VB (lanes 1–5 respectively) digested with either HindII or EcoRI. Lane C, DNA from wild-type A. stephensi mosquitoes, digested with HindII. Upper panel, hybridization with probe M; lower panel, hybridization with probe EGFP (Fig. 1a). d, In situ hybridization on the polytene chromosomes of ovarian nurse cells of transformed lines IV, VB, VD12 and VD14. Arrowheads, distinct chromosomal insertions of Minos in the different lines.
High resolution image and legend (57K)
An average of 29% of injected embryos survived and around 50% of the hatched larvae showed strong transient expression of EGFP, as monitored by fluorescence. Survival to adult stage (G0) averaged 10% and was a good predictor of successful transformation (Table 1). In three early experiments (I–III), marked by low adult survival (average 5%), we detected no fluorescent individuals among 7,672 G1 larvae derived from 22 G0 survivors. In contrast, in two experiments (IV and V) that gave 16% adult survival, the progeny of 69 G0 mosquitoes yielded 92 fluorescent individuals among the 10,539 G1 larvae analysed. In experiment IV, adults were intercrossed, so it was not possible to determine whether the five fluorescent G1 larvae were derived from one or more founders (Table 1). In experiment V, the 42 G 0 survivors were split into four same-sex groups (A–D), which were outcrossed separately (Table 2). Two of these groups (A, C) did not produce fluorescent progeny. The G0 males of group B were tested collectively and it was not determined whether the observed positive G1 larvae had originated from one or more founders. The 14 G0 females of group D were group-mated and laid eggs in isolation. Among these mosquitoes, three different founders (VD12, VD13 and VD14) generated fluorescent progeny. In summary, the 92 fluorescent G1 individuals from experiments IV and V were derived from a minimum of five independent G0 founders, representing a transformation frequency of 7% (5/69 surviving adults). This frequency is higher than that reported in D. melanogaster and C. capitata transformation experiments using Minos marked with the white gene marker2, 11, but is comparable to that achieved using EGFP as a selection marker (C.S., personal observation).
Fluorescent positive and negative G2 larvae generated by all the G1 founders were recorded in nearly equal numbers, in agreement with the 1:1 ratio expected from mendelian segregation of a single transgene insertion. Positive heterozygous G2 adults were then intercrossed to obtain homozygous lines. We attributed an additional phenotype of stronger fluorescence in the G3 generations to the homozygous genotype. The three fluorescent phenotypes (strong, weak and negative, Fig. 1b) were represented in approximately the 1:2:1 ratio expected from mendelian segregation of a single transgene (Table 3 ). To determine the nature of the integration events, we performed Southern blotting of transformed mosquitoes with probes spanning the arms of Minos (probe M), the EGFP sequence (probe EGFP) or the ampicillin-resistance gene present in the backbone of plasmid pMinEGFP (probe AMP) ( Fig. 1c). In both HindII and EcoRI digests from five representative families, two bands of variable size hybridized with probe M, in agreement with the occurrence of a single integration event. The size variation of the bands was consistent with single integrations occurring at different chromosomal sites (Fig. 1c). The integrity of the internal part of the transposon was confirmed by the presence of a single band of constant size that hybridized to the EGFP probe ( Fig. 1c). The AMP probe failed to produce a signal, as expected for Minos-mediated integration (data not shown).
To verify the precision of transposition, we determined the sequences of the junctions between the inserted element and the A. stephensi genome. Genomic DNA libraries from families IV, VD12 and VD14 were screened with probe M. All positive clones were sequenced and the presence of precise Minos-mediated integrations in these three families was confirmed (Table 4). Each of the cloned insertion arms included a complete Minos inverted repeat, a TA dinucleotide and an additional sequence, distinct for each case and presumably derived from a different site of the A. stephensi genome. We then performed in situ hybridizations with plasmid pMinEGFP on polytene chromosome preparations obtained from four representative families. In each case, a single insertion site was detected per chromosome complement, and this site was distinct in each family, confirming the results of the Southern blot analysis ( Fig. 1d).
Our results show that a Minos transposon carrying exogenous DNA can be integrated by a transposase-mediated mechanism into the genome of A. stephensi mosquitoes at high frequency. Furthermore, we have demonstrated that EGFP is a reliable selectable marker for anopheline transformation, as for other insect species5, 12, 13. This is the first reliable system for germline transformation of an anopheline vector of human malaria. We expect this technology to be successfully extended to the most important malarial vector, A. gambiae, as cultured cells of this mosquito species can be transformed in a transposase-mediated manner using a very similar Minos transposon6. The availability of this gene transfer technique may contribute to progress in our understanding of the mechanisms of vector competence, and could open the prospect of engineering anopheline mosquitoes that are refractory to Plasmodium.
Methods
Plasmid construction
The pMinEGFP transformation vector was derived from plasmid pMiNot2 by inserting two DNA fragments: an EcoRI–NotI fragment containing the EGFP gene (Clontech) under the control of the actin5c promoter from D. melanogaster and a NotI–NotI fragment containing the hygromycin B phosphotransferase gene (hyg) controlled by the hsp70 promoter from D. melanogaster . The hyg gene was introduced as an additional selectable marker in the event that selection with EGFP could not be achieved. Attempts to verify whether hyg could function as a selectable marker did not give reproducible results. The helper plasmid pHSS6hsILMi20 has been described9.
Embryo microinjection
Blood-fed A. stephensi
mosquitoes were allowed to lay eggs 48–72 h after a blood meal.
Eggs were laid in a solution of 0.1 mM p-nitrophenyl p'-guanidinobenzoate
(pNpGB) (Sigma) in isotonic buffer (150 mM NaCl, 5 mM KCl; 10 mM
HEPES; 2.5 mM CaCl2; pH 7.2). Eggs were then left in pNpGB
solution until injection, which was carried out 90–120 min after
oviposition. The embryos were microinjected by using glass needles (Eppendorf)
with a mixture of the helper plasmid pHSS6hsILMi20 (100 
g ml
-1) and the pMinEGFP plasmid (400 g ml
-1) in injection buffer (5 mM KCl, 0.1 mM sodium phosphate,
pH 6.8). After injection, the embryos were placed in isotonic buffer and transferred
back to an insectary (80% humidity; 26 °C) to hatch. Injected embryos
were not subjected to heat shock as previous experiments have shown that the
baseline activity of hsp70 promoter provided enough Minos transposase
to mediate integration into the genome of A. gambiae cells6.
Hatched larvae were analysed on an inverted microscope at a wavelength of
490 nm to detect EGFP expression.
Southern blot analyses and sequencing of integration sites
Genomic DNA from transformed adult or fourth instar larvae mosquitoes
(G3 generation) and from wild-type adult mosquitoes was digested
with the restriction endonucleases HindII or EcoRI. Digested
genomic DNA (
4 g per lane) was separated on a 0.8% agarose
gel and transferred onto a nylon membrane. The membrane was hybridized overnight
at 65 °C with three 32P-labelled probes: (1) probe
M, a polymerase chain reaction (PCR) product encompassing sequences of the
left and right arms of Minos; (2) probe EGFP, spanning the EGFP gene;
and (3) probe AMP, encompassing sequences of the ampicillin gene present in
the backbone of plasmid pMinEGFP. Between hybridizing with each probe, the
membrane was washed with a boiling 10-mM EDTA/0.1% SDS solution for 20 min
to dehybridize it. Genomic Lamba Zap EcoRI libraries (Stratagene) were
constructed from the DNA extracted from G3 larvae of transgenic
lines IV, VD12 and VD14 and hybridized with probe M.
The cloned insertion sites were sequenced using primers annealing internally
to the inverted repeats (pMR, 5'-CGAGTTAAATGCGTAATGC-3' and pML,
5'-GCTCTTCTTGAGATTAAGG-3').
In situ hybridization
The integration of the Minos element in the mosquito genome was visualized by in situ hybridization experiments performed on the polytene chromosomes of ovarian nurse cells prepared from four out of the five established lines as described14. The whole plasmid pMinEGFP was used as a template to produce a biotinylated probe.
