Proc. Nat!. Acad. Sci. USAVol. 91, pp. 10315-10319, October 1994Population Biology

Polymorphisms detected by random PCR distinguish betweendifferent chromosomal forms of Anopheles gambiaeGUIDO FAVIA*t, GEORGE DIMOPOULOS*t, ALESSANDRA DELLA TORRE*t, YEYA T. TOURO§, MARIO COLUZZIt,AND CHRISTOS LoUIs*:*Institute of Molecular Biology and Biotechnology, Foundation of Research and Technology-Hellas, 711 10 Heraklion, Crete, Greece; tIstituto diParassitologia, Fondazione Cenci Bolognetti, Universiti degli Studi, Piazzale Aldo Moro, 00185 Rome, Italy; *Department of Biology, University of Crete,711 10 Heraklion, Crete, Greece; and 1Dpartement d'Epiddmiologie des Affections Parasitaires, Ecole Nationale de Mddecine, Bamako, Mali

Communicated by Fotis C. Kafatos, June 20, 1994

ABSTRACT We have applied PCR amplification usingrandom primers to distinguish between incipient species of thema vector Anopheks gambiae. Individuals belonging tothree chromosomally characterized West African forms of thismosquito, which are important epidemiologically as they differin vectorial capacity, were sampled both from laboratorystocks and from wild populations collected in three localities.The techniques used allowed for the unambiguouscs aof the mosquitoes, providing a tool for rapid and efficientdiaguosis, which previously relied on cytological examinationof polytene chromosomes.

The fact that most Anopheles mosquitoes belong to differentspecies complexes with members that are often hard, or evenimpossible, to distinguish morphologically makes their pop-ulation biology difficult to study and represents a majordifficulty in malaria entomology. Correct analysis of thedistribution of specific malaria vectors is one of the prereq-uisites for meaningful epidemiological studies and for theplanning and monitoring of successful malaria control oreradication programs, particularly those based on antivectormeasures (for review, see ref. 1). Vectorial capacity can varytremendously even between members of the same speciescomplex, as best exemplified by the fact that the anthro-pophilic species Anopheles arabiensis is a very efficientmalaria vector, while its often sympatric sibling speciesAnopheles quadriannulatus does not represent a health prob-lem due to its preference for feeding on animals (2). Thesituation is further complicated by the fact that relevantbiological differences are detected not only between speciesbut also between incipient taxa within the same species. Forexample, three chromosomal forms ofAnopheles gambiae inthe strict sense (s.s.) that carry different strain-specificcombinations ofinversions and differ in vectorial capacity arefound in Mali and are given the names Savanna, Mopti, andBamako (3). These forms appear to be partially or fullyisolated reproductively in nature: in sympatric areas hybridsbetween Savanna and the other two forms have been ob-served at frequencies lower than expected, but no individualscarrying heterozygous complements of the Mopti and Bama-ko inversions are seen in nature, even though the two formsproduce viable progeny under laboratory conditions (4, 5).The larval stages of the Mopti form ofA. gambiae s.s. occurmainly in man-made habitats, such as irrigated areas, even inthe dry season, while the Savanna and Bamako forms havea tendency to breed in more natural sites, exploiting rain-dependent pools for larval development. This affects theirspatial and seasonal distribution, since the Mopti form willbreed throughout the year and can therefore displace theother forms in irrigated areas, leading to changes in the

patterns of malaria entomology (6). The efficient and rapididentification of these three "strains" is therefore of utmostimportance for epidemiological studies in this part of Africa.Since the three forms can, at present, be differentiated onlycytologically, by the combinations of their specific chromo-somal inversions, it becomes obvious that the study of theirbiology and vectorial competence would greatly benefit fromthe development of a more convenient method for distin-guishing between these populations or incipient species.The problems in identifying the members of mosquito

populations have led to extensive efforts to develop tech-niques that would allow for fast and reliable diagnosticprocedures, in particular among species belonging to thesame complex. These techniques include, in addition to thecytological examination of polytene chromosomes, other"traditional" tools such as genetic compatibility (7), isozymeanalysis (8), and immunological procedures (9), and alsomethods based on recombinant DNA technology (10-12).Although the latter tools have proven useful, one majordrawback of techniques based on DNA hybridization orsequence determination of specific genes is that they are timeconsuming. In addition, they require sophisticated equip-ment that is often only found in specialized laboratories.Some ofthese problems have recently been solved by Cramp-ton and his collaborators (13-15), who have improved atechnique based on the hybridization of repetitive oligonu-cleotide probes, specific to the individual sibling species,allowing for the rapid identification of members of the A.gambiae species complex. In another development, mosqui-toes within a species complex have been distinguished byusing PCR-based identification of either rRNA-encoding(rDNA) sequences (16) or anonymous DNA segments byusing random amplified polymorphic DNA (RAPD) markers(for review, see ref. 17). While the former technique can bedeveloped to yield unambiguous information, the latter hasthe advantage that it is extremely fast and it also providesmolecular markers that can be mapped by recombination (18,19). Furthermore, the genomic fingerprints obtained throughRAPD-PCR can also be used in a variety of ways in molec-ular ecological applications (20). In attempts to differentiatebetween mosquito species such fingerprints were shown todistinguish between A. gambiae and A. arabiensis (21) andalso between members of species complexes of anotherdisease vector, Aedes aegypti (22, 23).

In this report we focused our analysis on developing amethodology that would be helpful in differentiating betweenthe Savanna, Mopti, and Bamako forms of A. gambiae. Weused two different techniques that are both based on the PCRamplification of nonspecific sequences. One of them usesRAPD loci to differentiate between the two chromosomalforms, while the second one uses a slightly modified proce-dure based on consensus tRNA gene primers (24). We

Abbreviation: RAPD, random amplified polymorphic DNA.

10315

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

, 202

1

10316 Population Biology: Favia et al.

demonstrate that both methods can be used to distinguishmembers of the A. gambiae complex and, most importantly,also between sympatric chromosomal forms of A. gambiaes.s., including mosquitoes collected in the field.

MATERIALS AND METHODSStrains. The following laboratory strains of A. gambiae

were used in this analysis; their inversion genotype (4) isgiven in parentheses: GASUA (Xag, 2R, 2La, 3R, 3L), astrain extracted in 1989 from a Forest-Savanna 2Rd and 2Lapolymorphic A. gambiae colony, established in 1986 fromadult females collected in the village of Suakoko (Liberia)(this strain is used as the standard reference for A. gambiaes.s.); GASEL-M (Xag, 2Rbc/u, 2La, 3R, 3L), a Mopti strainestablished in 1987 from adult females collected in the villageof Selenkenyi (Mali); GAMOR-B (Xag, 2Rjcu, b/+, 2La, 3R,3L), a Bamako strain ofA. gambiae s.s., established in 1991from adult females collected in the village of Moribabougou(Mali); GAKUI-M (Xag, 2RBC/u, 2La, 3R, 3L), a Moptistrain established in 1991 from adult females collected in thevillage of Kuiti (Burkina Faso); GAMOR-M (Xag, 2Rbc/u,2La, 3R, 3L), a Mopti strain established in 1991 from adultfemales collected in the village of Moribabougou (Mali);ARMAD (Xbcd, 2Rb, 2La, 3R, 3L), an A. arabiensis strainestablished in 1992 from adult females collected in the villageof Alasora (Madagascar); and ARZAG (Xbcd, 2Ra/+, b/+,c/+, d/+, 2La, 3Ra/+, 3L), an A. arabiensis strain estab-lished in 1985 from adult females collected in Zagtouli (Burki-na Faso). Field collections were performed in the villagesSelenkenyi, Balanbani, and Moribabougou in Mali.

Priners. For the tRNA gene amplification the followingprimers were used: T1, 5'-ATCCATAGGTCGCTGGTTC-3'; T2, 5'-CGATAGCTCAGTTGGTAGAG-3'; T3, 5'-TCGTGGCCGAGTGGTTAA-3'; and T4, 5'-TAGCTCAGT-TGGTAGAGC-3'.The decameric RAPD primers were purchased from Op-

eron Technologies (Alameda, CA). Only the results obtainedwith primer L02 (5'-TGGGCGTCAA-3') are shown here.DNA Extraction. Single mosquitoes from laboratory strain

(stored in ethanol) were dried and then soaked in 100 A4 ofextraction buffer (1% SDS/25 mM NaCl/25 mM EDTA) for10 min. Field-collected mosquitoes were stored in Carnoy'sfixative (ethanol/chloroform/glacial acetic acid, 6:3:1 vol/vol) rather than ethanol before further processing. Afterrehydration, all mosquitoes were homogenized with a poly-propylene pestle and placed (after the addition of 200 i1 ofextraction buffer) at 65°C for 10 min. Following the additionof 200 p1 of 3 M potassium acetate (pH 7.2) the homogenateswere placed on ice for 30 min and then centrifuged for 10 miat 15,000rpm in a table-top microcentrifuge (Eppendorf3. Thesupernatants were extracted twice with phenol/chloroformand precipitated with the addition of 2 vol of ethanol, and thepellets were resuspended in 50 p1 of distilled water. Onemicroliter of this DNA solution was used for each PCR-amplification reaction.PCR Amplificaio. Amplfication using the tRNA gene

primers. Reaction mixtures of50 i1 were prepared, using 1.25units of Taq DNA polymerase, 0.2 mM each dNTP, 1 p.Mprimer, and 1 p1 ofDNA solution in the buffer recommendedby Stratagene. The reaction was cycled 40 times through thefollowing scheme: 45 sec at 94°C, 45 sec at 50°C, and 1.2 minat 720C.

Amplification using the RAPD primers. The assays con-sisted of 25 p1 of a solution containing 10mM Tris HCl at pH8.3, 50mM KCI, 1.5 'mM Mg912, 0.01% gelatin, 200 juM eachdATP, dCTP, dGTP, and dTTP, 15 ng of a single RAPDprimer, 0.5 unit of Taq DNA polymerase, and 1 p1 ofDNAsolution. Amplification proceeded through 45 cycles (1 min at

940C, 1 min at 360C, 2 min at 720C). In most experiments TaqDNA polymerase from Stratagene was used.The amplification products were separated electrophoret-

ically on a 1.4% agarose gel and visualized with UV light.

RESULTSTo develop a set of primers suitable to differentiate betweenthe Mopti and Bamako forms ofA. gambiae s.s., we designedoligonucleotides based on a recently developed strategy forfingerprinting bacterial strains (24). This approach uses con-sensus sequences derived from the best alignment of tRNAgenes, and it relies upon the fact that these genes usuallyoccur clustered in multiple copies in the genome of mostorganisms. As no mosquito tRNA sequences were availablein the nucleic acid data bases at the time, we chose to aligndipteran tRNA genes to obtain the desired consensus se-quences. The primers designed, based on the best alignmentof 61 Drosophila tRNA genes, are shown in Materials andMethods. A subsequent search ofthe nucleic acids databaseswith the primer sequences revealed that primers T1, T2, andT4 are colinear with Drosophila genes encoding tyrosinetRNAs, while T3 has a perfect match with serine tRNAs. Ifone allows for mismatches, then more tRNA-encoding genesare found to share extensive similarity with the primers. Allfour primers used in this analysis have the same 5'-to-3'orientation with respect to the corresponding tRNA gene.

In a first series of reactions we amplified genomic DNAfrom individual mosquitoes belonging to different laboratorystrains. These strains had been kept in the laboratory forseveral years under mass mating conditions. All four primers,when used alone (or in combination, results not shown),produce fingerprints consisting of a multitude of bands ofdifferent size and intensity. Fig. 1, as an example, shows thepatterns obtained when primers T2 and T4 are used to amplify

1

A's

B

FIG. 1. Results of the amplification of genomic DNA isolatedfrom three individuals each from the A. gambiae strains GASEL-M,GAMOR-B, and GASUA, and the A. arabiensis strains ARMAD andARZAG (from left to right), using the tRNA gene-based primers T2(A) and T4 (B). The same individuals were used in the two reactions.The symbols x and o mark amplification products that are commonto all five strains but differ quantitatively.

Proc. Natl. Acad Sci. USA 91 (1994)

Ix

Dow

nloa

ded

by g

uest

on

Mar

ch 2

, 202

1

Proc. Natl. Acad. Sci. USA 91 (1994) 10317

the same three individuals from each of the A. gambiae s.s.strains GASUA, GASEL-M, and GAMOR-B and the A.arabiensis strains ARMAD and ARZAG. Clear differencescan be observed among the patterns obtained for the fivedifferent strains, while the fingerprints produced withinstrains are invariant. For reasons that we do not understand,some primers also reproducibly amplified the genomic DNAsof one strain better (or worse) than that of another, althoughthe same DNA extraction protocols were followed through-out. This was also detected for the RAPD-PCR amplifications(see below). The differences observed are both quantitativeand qualitative. Some bands are apparently common to allfive strains, while some of the fragments are specific to onlyone ofthe five. For example, the two bands in Fig. 1A labeledwith a "x" and an "6o" can be detected in all five strains, butthey clearly differ in intensity. In contrast to the three A.gambiae s.s. strains, in the case of the two samples of A.arabiensis, for the most part only quantitative variation canbe observed when the tRNA-based primers are used, sincemost bands are common to both ARZAG and ARMAD. It isalso noteworthy, that in spite of the fact that the T2 and T4primers not only are colinear with Drosophila tyrosine tRNAgenes but also overlap with each other, the fingerprintingpatterns obtained differ markedly between the two primers,demonstrating that they do not necessarily amplify the sameset of tRNA genes.The amplification achieved with primer T3 is shown in Fig.

2. Comparing the patterns obtained, one can detect obviousdifferences between the Forrest-Savanna, Mopti, and Bama-ko laboratory strains, although, again, some bands are com-mon to both strains.

In addition to the tRNA-derived primers, we also testedwhether RAPD-PCR could be used for the purpose of iden-tifying the different strains of A. gambiae, and in particularthe Bamako and Mopti chromosomal forms. Fig. 3 shows theamplification achieved when the decanucleotide L02 is usedto amplify DNA from the same individuals tested with thetRNA-derived primers shown in Fig. 2. Here again, the threeA. gambiae s.s. and the two A. arabiensis strains can readilybe distinguished. No intrastrain variation is apparent whenthis primer is used, while the quantitative and qualitativedifferences in the fingerprints are enough to classify theindividuals as belonging to one or another strain.

GASEL-M GAMOR-B

A, -IA LL -; -) !" ; ;

FIG. 3. RAPD-PCR analysis oflaboratory strains. DNA from thesame individuals shown in Fig. 1 was amplified by using the deca-nucleotide primer L02.

Besides the results presented above, we compared theprofiles obtained from the three additional Mopti laboratorystrains (GASEL-M, GAKUI-M, and GAMOR-M) to oneanother, using both RAPD and tRNA-based primers, and nodifferences could be detected. We should mention, though,that only a limited number of individuals were tested from thelatter two strains, as these died before completion of theanalysis.

Since the mosquitoes initially analyzed were from inbredlaboratory cultures, we decided to test, in a further series ofexperiments, whether strain-specific diagnostic polymor-phisms can also be detected in mosquitoes collected in thefield. For this we concentrated on the examination of mos-quitoes belonging to the two chromosomal forms Mopti andBamako (24 and 23 specimens, respectively). The collectionswere taken in three localities in Mali, located less than 100 kmfrom one another. The mosquitoes were stored in Carnoy'sfixative and, prior to the extraction of the genomic DNA, theovaries were removed and the karyotypes of the mosquitoeswere determined. The results of this analysis are shown inTable 1. While Mopti individuals (with a total offive differentinversion complements) were found only among the Selen-kenyi collection, Bamako mosquitoes homozygous or het-erozygous for two inversion complements were found in all

Table 1. Chromosome arm 2R karyotype determined for the fieldisolates from the three localities in MaliChromosomal

formBamako

2R karyotypejbcu/jbcujbcu/jcu

jcu/jcu

Mopti bc/bcbc/u or bcu/+

bc/+u/+U/u

FiG. 2. Amplification of genomic DNA of three individuals fromthe strains GASEL-M and GAMOR-B (from left to right), using thetRNA gene-based primer T3.

IndividualsSel 50, 58, 60Sel 7, 30, 42, 43Bal 267, 273Mor 105, 175Sel S, 27, 33, 44, 53,

66, 72Bal 244, 259Mor 24, 127, 130Sel 68, 69, 79Sel 2, 9, 23, 29, 45,

47, 48, 55, 56Sel 18, 24, 34, 38Sel 22Sel 1, 15, 17, 32, 39,

61, 70The genotypes were determined by microscopic examination of

the polytene chromosomes prepared from half-gravid females. Sel,Ban, and Mor refer to the collection sites Selenkenyi, Balanbani, andMoribabougou, respectively, while the numbers refer to the code ofthe individual mosquito analyzed.

Population Biology: Favia et al.

L. %.-

Dow

nloa

ded

by g

uest

on

Mar

ch 2

, 202

1

10318 Population Biology: Favia et al.

three localities. Fig. 4 shows the results of the PCR ampli-fications of the individuals with both the T3 tRNA-basedprimer and the L02 decanucleotide primer. Both primers canclearly differentiate between the two chromosomal forms,although the patterns are fairly similar. In both cases diag-nostic amplification products can be observed. Interestingly,no noticeable differences can be observed among individualsfrom each form, although the Mopti mosquitoes shown hererepresent five different genotypes and the Bamako individ-uals are isolates from the three different villages. On the otherhand, it is interesting that there were extensive differencesbetween the patterns ofthe field-collected and the laboratorymosquitoes when the same primers were used (see Figs. 2 and4A for primer T3 and Figs. 3 and 4B for L02).We looked into the potential reasons for the fingerprinting

pattern differences. A summary of this analysis is presentedin Fig. 5. An important reason for irreproducibility may befound in the source of the thermostable DNA polymeraseused in the experiment. Fig. 5A (in which DNA from thelaboratory strain GASEL-M was used for the amplificationwith primer T4) shows that three different polymerases yielddifferent patterns. The second reason for the differences isthat field-collected mosquitoes were stored (for shipmentreasons) in Carnoy's fixative, while theDNA from laboratorystrains was extracted from mosquitoes that had been stored

lM u)p -I lBavnAko

d I- 51 );,S( i 1 1W ili M w M

4\S15Ia,> s i ; ji { |A 2-t|_1 /(-$ |]]'

A

B

I p I - _-----'' - lam ako-}3______" S VCi Vw l$al Hal %:I()h \.rI(-

X, ? 24411 711197

it.

_~~~~~~~~~~~~~~~._

FIG. 4. Amplification of genomic DNA isolated from field-collected mosquitoes, using the tRNA gene-based primer T3 (A) andRAPD-PCR with primer L02 (B). Five Mopti and six Bamakoindividuals are shown. The labels on top of the lanes refer to theindividuals presented in Table 1. M, molecular weight marker.

A\ \ B B 13C BC S > 4( ST

B

FIG. 5. Reproducibility of the technique. Different sources ofcommercially available thermostable DNA polymerases were testedwith primer T4 (A), and the storage of the mosquitoes in Carnoy'sfixative was investigated for the consistency ofthe fingerprint patternproduced by the PCR amplification using primer T4 or T3. (A) DNAwas extracted from GASEL-M mosquitoes and three different ther-mostable polymerases (V, Vent from New England Biolabs; B,Boehringer; and S, Stratagene) were used for the amplification (DNAextracted from two mosquitoes was amplified). In the case of thelanes marked with C, the mosquitoes were stored in Carnoy'sfixative prior to DNA isolation. (B) Mosquitoes were used directly orafter 24-hr storage in Carnoy's fixative before amplification (-C).Lanes Ga and GaC, GASEL-M; lanes SeiC, MorC, and BalC, threefield-collected Mopti mosquitoes from the villages Selenkenyi, Mor-ibabougou, and Balanbani, respectively; lane NT, noDNA template.

in ethanol. In Fig. SA, two ofthe polymerases were tested forthe fingerprinting of the laboratory strain GASEL-M, but themosquitoes had been stored for 24 hr in Carnoy's fixativeprior to the DNA isolation. The differences are marked, bothquantitatively and qualitatively. Fig. 5B shows a similarexperiment using primer T3. Here, in addition, the samebatch of enzyme was used to amplify DNA from threefield-collected Mopti mosquitoes derived from different vil-lages and stored in Carnoy's fixative, as well as two GASEL-M specimens, of which one was also put in Carnoy's fixativefor 24 hr prior to the DNA extraction. It is apparent that thesolvent treatment changes the amplification pattern by pro-ducing one novel strong band, while another band disap-pears. On the other hand, both field-collected samples andthe similarly treated laboratory mosquito produce exactly thesame fingerprint.

DISCUSSIONWe have used PCR amplification with random primers todistinguish between different chromosomal forms ofA. gam-biae and have shown that this approach can be used for theidentification ofmosquitoes collected inthe field. On a purelytechnical basis, this report shows that both kinds of primersused in this analysis can be of value for this purpose.Although it is not readily apparent from the results presentedhere, the protocol based on the primers derived from tRNAgene consensus sequences was, in our hands, more efficientthan the RAPD-PCR method. The reason for this is that allfour primers used in this analysis yielded reproducible resultsand were immediately successful in differentiating betweenthe strains under investigation. In contrast, of 60 decanucle-otide primers analyzed, most were found not to be of use inthe diagnosis ofthe chromosomal forms for different reasons.

Proc. Nad. Acad Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Mar

ch 2

, 202

1

Proc. Natl. Acad. Sci. USA 91 (1994) 10319

These reasons included, among others, the failure of about50% of the primers to detect variation between the strainsand, most importantly, the frequent irreproducibility of theresults (25). Thus, should RAPD-PCR be chosen as themethod for similar investigations, we stress the point thatseveral primers may have to be extensively screened beforeoptimal ones can be identified.

Perhaps the reason that tRNA gene-based primers seem togive better results may be that these are not random, as theywill clearly amplify clustered genes whose sequences wereused to design them. For example, the two marked bands inFig. 1A that are seen in all strains examined could representa tRNA locus that would differ only in the copy number ofindividual genes. Moreover, our results indicate that not onlyrandom primers but also primers recognizing only specificgene arrangements can be useful in the characterization ofdifferent chromosomal forms of A. gambiae, similarly towhat was described for bacteria (24). Indeed, even the twooverlapping primers T2 and T4 yielded different and usefulamplification patterns. Whether these differences are due tothe additional amplification of non-tRNA genes cannot bedetermined.RAPD-PCR has recently been used to distinguish between

sibling species of disease vector mosquitoes (21-23). Weshow that this technique, and, in particular, the relatedtRNA-based primer method, can be extended to differentiatebetween populations ofA. gambiae s.s., for which there areindications that they represent incipient species (4). A recentreport showing that two biotypes of the whitefly Bemisiatabaci represent two different species, and which was based,among others, on the results of PCR amplification withrandom primers (26, 27), raised a controversy as to whetherthe results of such an analysis can be used prima facie foranswering such a question (28-30). Our data clearly supportthe hypothesis of separate taxonomic units, although they donot offer unambiguous clues as to whether the Savanna,Mopti, and Bamako populations do indeed represent differ-ent species. On the other hand, the possibility to distinguishbetween them without the need for specialized cytologicalexamination offers a most important improvement for theepidemiological analysis ofmalaria vectors. In our hands, themain problem that has to be taken care of is that of theirreproducibility of the results in the cases where differentenzyme sources are used for the amplification assays, orwhen the mosquitoes are stored in different media. In spite ofthis, we have seen that when the same conditions are used,all Mopti specimens can be consistently differentiated fromthe Bamako ones.

In conclusion, our results show that the differences ob-served in the comparison ofthe different chromosomal formsare not restricted to the level of the chromosome linearity,but rather they can also be detected on the level of DNAsequences, where polymorphisms between them are reallyabundant.

This work was supported by grants from the John D. and CatherineT. MacArthur Foundation, the European Communities, the WorldHealth Organization, and the Rockefeller Foundation. G.F. was the

recipient of a training grant from the Biotech Programme and A.d.T.is the recipient of a postdoctoral fellowship from the Human Capitaland Mobility Programme of the European Communities.

1. Coluzzi, M. (1992) Parasitol. Today 8, 113-118.2. Coluzzi, M. (1984) Bull. W.H.O. 62, Suppl., 107-113.3. TourE, Y. T., Petrarca, V. & Coluzzi, M. (1983) Parassitologia

(Rome) 25, 367-370.4. Coluzzi, M., Petrarca, V. & Di Deco, M. A. (1985) Boll. Zool.

52, 45-63.5. Persiani, A., Di Deco, M. A. & Petrangeli, G. (1986) Ann. Ist.

Super. Sanitd 22, 221-224.6. Tour6, Y. T., Petrarca, V., Traor6, S. F., Coulibaly, A.,

Maiga, H. M., Sankard, O., Sow, M., Di Deco, M. A. &Coluzzi, M. (1994) Genetica, in press.

7. Fritz, G. N., Narang, S. K., Kline, D. L., Seawright, J. A.,Washino, R. K., Porter, C. H. & Collins, F. H. (1991) J. Am.Mosq. Control Assoc. 7, 198-206.

8. Narang, S. K., Kaiser, P. E. & Seawright, J. A. (1989) J. Am.Mosq. Control Assoc. 5, 317-324.

9. Ma, M., Beier, J. C., Petrarca, V., Gwadz, R. W., Zhang,J.-Z., Song, Q. & Koech, D. K. (1990) J. Med. Entomol. 27,564-569.

10. Collins, F. H., Petrarca, V., Mfopu, S., Brandling-Bennett,A. D., Were, J. B. O., Rasmussen, M. 0. & Finnerty, V.(1988) Am. J. Trop. Med. Hyg. 39, 545-550.

11. Panyim, S., Yasothornsrikul, S., Tungpradubkul, S., Baimai,V., Rosenberg, R., Andre, R. G. & Green, C. A. (1988) Am. J.Trop. Med. Hyg. 38, 47-49.

12. McLain, D. K. & Collins, F. H. (1989) Heredity 62, 233-242.13. Hill, S. M., Urwin, R., Knapp, T. F. & Crampton, J. M. (1991)

Med. Vet. Entomol. 5, 455-463.14. Hill, S. M., Urwin, R. & Crampton, J. M. (1991) Am. J. Trop.

Med. Hyg. 44, 609-622.15. Hill, S. M., Urwin, R. & Crampton, J. M. (1992) Trans. R. Soc.

Trop. Med. Hyg. 86, 213-215.16. Scott, J. A., Gordon, W. G. & Collins, F. H. (1993) Am. J.

Trop. Med. Hyg. 49, 520-529.17. Black, W. C., IV (1993) Insect Mol. Biol. 2, 1-6.18. Williams, J. G. K., Kubelik, A. R., Livak, K. J., Rafalski,

J. A. & Tingey, S. V. (1991) Nucleic Acids Res. 18, 6531-6535.19. Williams, J. G. K., Hanafey, M. K., Rafalski, J. A. & Tingey,

S. V. (1993) Methods Enz4ymol. 218, 704-740.20. Hadrys, H., Balick, M. & Schierwater, B. (1992) Mol. Ecol. 1,

55-63.21. Wilkerson, R. C., Parsons, T. J., Albright, D. G., Klein, T. A.

& Braun, M. J. (1993) Insect Mol. Biol. 1, 205-211.22. Ballinger-Crabtree, M. E., Black, W. C., IV, & Miller, B. B.

(1992) Am. J. Trop. Med. Hyg. 47, 893-901.23. Kambhapati, S., Black, W. C., IV, & Rai, K. S. (1992) J. Med.

Entomol. 29, 939-945.24. Welsh, J. & McClelland, M. (1991) Nucleic Acids Res. 19,

861-866.25. Favia, G., Dimopoulos, G. & Louis, C. (1994) Insect Mol. Biol.

3, 149-159.26. Perring, T. M., Cooper, A. D., Rodriguez, R. J., Farrar, C. A.

& Bellows, T. S. (1993) Science 259, 74-77.27. Gawell, N. J. & Bartlett, A. C. (1993) Insect Mol. Biol. 2,

33-38.28. Campbell, B. C., Duffus, J. E. & Baumann, P. (1993) Science

261, 1333.29. Bartlett, A. C. & Gawel, N. J. (1993) Science 261, 1333-1334.30. Perring, T. M., Farrar, C. A., Cooper, A. D., Bellows, T. S. &

Rodriguez, R. J. (1993) Science 261, 1334-1335.

Population Biology: Favia et A

Dow

nloa

ded

by g

uest

on

Mar

ch 2

, 202

1