"ID

uL SII -/ /1 "

iW F ---- - - -

1111 -2 111111-4 'ii

AD-A206 523

De,, e !cc mmt 3~f a I erc~: cZir i~:~nc the ;!a>

Aropneles gamb iae f cm Ano~phelt-s ai-atmensi 5

DTIC

DJL DI-t L- -

E~ [et r i, Freder~c "P 7

Contract No. DA4iD17-85-C-513L

Emory UniversityAtlanta, GA, 3032?

Approved for public release; di,3ti-iLution is (1nl1imited.

The f indings in this report are nct t~o te cori-;trujed id, cji-

official Departmepnt of the Avtrm-y positian unless so desiq-na ted bV i ther au thor i ed Louin

,te~iRITY CL.4SSIFICATl0N OF 7,61 PAGEForm Approved

REPORT DOCUMENTATION PAGE GAMB No 0704-0188

la. REPORT SECURITY CLASSIFICATION Tlb RESTRICTIVE MARKINGSUnclassifiedI

2a. SECURITY CLASSIFICATION AUTHORITYr 3 O)ISTRIBUTION/AVAILABILiTY OF PTFC;RT

Approved for public release;

2b. DECLASSFCATION / DOWNGRADING SCHEDULE distribution unlimited

4. PERFORMING ORGANIZATION REPORT NUMBER(S) j . MONITORING ORGANIZATION REPORT NUMBER(S)

6a. NAME OF PERFORMING ORGANIZATION '6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATiOrN

Emory University (ifaplcbe I

6c. ADDRESS (City, State, dnat ZIP Code) 7b ADDRESS (Cit), Stare. and ZIP Cocre)

Atlanta, GA 30322

Ba. NAME O PUNDINGSPONSORING Bb OFFICE SYMBOL 9 PROCUREMENT 1NSTRuMENT iDEN71FCALiON NUMBER

ORGA%;ZATION T'S Ar-v Medical (If picable)

Research & Develorment Comm DAMD 17-85-C-5184______8c. ADORESSI.Ly. State. ana ZiP Code) 10 SOURCE OF PUNDNG NVERS

PROGRAM PROjECT TAKWORK UNIT

Ft. Detrick, Frederick, MID 21701-5102 ELMNTNN M1 O CESiN O

11. TITLE (Incluoe Security Classification)Development of a DNA-based method Jor distinguishing the malaria vectors Anooheles carnbiaefrom Anooheles arabiensis

12. PERSONAL AUTHOR(S)Finnerty', Victoria

13a. TYPE OF REPORT 13b. TIME COVERED 14 DATE OF REPORT (Year, Month, Day) 1S. PAGE COUNTAnnual/F"inal IFROM R;-6-15 TO8?1_J__-3 11 1987 November 15 2

16. SUPPLEMENTARY NOTATION 1Annual rep~ort covers period of time 15 June 198 31lOctober 1987

17. COSATi CODES 18. SUjBJECI*TERMS (Continue on reverse if necessary and identify by block numnoer)

FI6L 1RU3 SI-GO Recombinant DNA/Species Identification/Anopheline Malaria06 13 Vectors, RAl

19. ABSTRACT (Continue on rev~erse if necessary arid identify by block number)

The primary African malaria vectors, A. gambiae and A. arabiensis, belonq to a species com-

plex, the members of which are rorphologically indistinguishable. Epidemioloaical studies

to determine the involvement of each in malaria transmission were difficult because two ormore cf the species are commonly sympatric. We have developed a DNA probe (an rDTZP frac-ment from A. gambiae) which reveals RFLPs that distinoruish each member of the comrilex 'I-\Southern analysis. The DNA probe method has been extensivelyr tested with both the existinc7means of distinguishing these species (the isozyme method and the cytogenetic method) and

the results in every case were concordant. The probe can sensitivelv diagnose sincle adul*

mosquitoes of either sex, mosquito parts, larval or pupal. moreover, specimens dessicatei

in the field and stored up to one year can be scored. The test is compatible with ELISA

analyses of dessicated thoraces since the DNA probe can readily diacinose sinale dessicatei

abdomens. Blood Meal analysis can readily utilize the protein pellet ObLained durini DNAextraction, We have extended thEse studies to begin developina a DNA (continued on revers4

20 DISTRIBUTION / AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATIONOUNCLASSIFIED/UNLIMITEO (3 SAME AS RPT 0 DTIC USERS Unclassified

,22a. NAME OF RESPONSIBLE IN.DIVIDUAL 22.bjELEPHONE.lnfude Aea oe OFFICE SYMBOLMary Francis 5 ostian 3U-Ic3-/2 D-RMl -S

.,.J Formn 1473, JUN 86 Previous editions are obsolete. SECURITY CLASSIFICATION OF TRIS PAGE

19. Abstract (continued).probe method which would ut2]iz P riot blnt. And therby eliminate the need fnrrestriction digests, gel separation, and Southern blotting. Therefo -e, we havefocused our efforts on finding and characterizing sp cies specific sequenceswithin our rDNA clones of A. gambiae and A. arabiensis. Thus far tvjo speciesspecific fragments in A. gambiae have been identifieo. The characteriaticn ofthe species specific A. gambiae fragments is expected tQ provide a focus for fu-ture efforts to identify such fragments in A. arabiensis.

e

-1-

FOREWORD

Studies with Recombinant DNA: The investigator has abided by the National Insti-tutes of Health Guidelines for Research Involving Recombinant DNA Molecules(April 1982) and the Administrative Practices Supplements.

Emory University Biosafety File No. 142-85

rT r's c'o;-DiC I Ali L

.. . .. . ... .

By.

A ; 7;

Dist

!~i

-2-

TABLE OF CONTENTS

Page

Foreword.................................

Table of Contents.............................2

List of Tables and Figures........................3

Body of Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41. Statement of Problem Under Study................42. Background............................4

3. Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . 4

4. Experiments and Results......................A. Utility of heterologous conserved sequences.......5B. Mitochondrial DNA probes from A. gambiae ......... 5C. Isolation of a diagnostic cloned rDNA fragment ..... 7D. The diagnostic fragment is useful for single dried

mosquitoes.........................E. The diagnostic probe is sex linked............9F. The DNA probe method is compatible with the sporozoite

assay and is useful for blood meal analysis. ...... 10G. The DNA probe method and ODIA isozyme method give the

same results......................11H. Comparison of DNA probe and cytogenetic methods for

d~stinguishing A. gambiae from A. arab iensis.......15I . Possible use of species-specific rDNA sequences for

dot blot assay......................16

Literature Cited............................1

Bibliography of Publications Supported by DAMD 17-85-C-5194. ......20

Distribution List............................21

-3-

LIST OF TABLES AND FIGURES

Table I (p. 8). Mosquito strains available from the CDC

Table 2 (p. 11). Testing field specimens from A. cambiae complc moscui-toes collected in Asembo, Kenya, October 10R5

Table 3 (p. 14). DNA probe and ODH isozyme analyses of 0. gambiae cox-plex mosquitoes from Kenya

Table 4 (p. 15). Results of comparison of DNA and cytogenetic methods ofidentifying the species of field collected specimens ofA. gambiae complex

Figure 1 (p. 6). Gene map of D. yakuba mt DNA molecule

Figure 2 (p. 7). Lambda AGrI2 rest-ictinn map

Figure 3a (p. 9). Hybridization of pAGr12A to EcoRI digests of singledried female mosquitoes

Figure 3b (p. 9). Hybridization of pAGr12A to EcoRI digests of singlefemale mosquitoes, including four members of the com-plex

Figure 4 (p. 10). Hybridization of pAGrl2A to single dried male and fe-male mosquitoes or mosquito abdomens.

Figure 5 (p. 12). Octanol dehydrogenase electromorphs found in the Kenyafield samples

Figure 6 (p. 13). Hybridization of the pAG1I2A probe to EcoRI digests of

field-collected specimens

Figure 7 (p. 16). Restriction maps of XAGrI2 and XAGr23

-4-

FINAL REPORT

1. Statement of Problem Under Study. Malaria remains the most common disease inthe world today, numbering 250-300 million cases at any given time. Among he 92million new cases each year, there are close to 1 million deaths, nearly all ofwhich are young children. Although malaria is found world-wide, the problem ismost acute in subsaharan Africa where the disease presents an enormous obstacletz social and economic deelopment. Although malaria was eradicated from most ofits original temperate range in the 1960s, its incidence in the tropics continuesto increase, due in part to DDT and chloroquine resistance In mosqtito pcpula-tions. Today. the most common African malaria vectors, Anopheles ganebiae andAnopheles arab ensis (1), belong to the A nopheles gambiae species coplt?>:, ohichcontains six member species. All of the species are mo-phological]v inri'sti,-guishable, and two or three of the species are often sympatric over must of theirrange. The species differ in behavior and preferred habitat. Moreover, there isevidence suggesting that the two major vector species may not be equally involvedin malaria transmission, depending upon the season and location (2). Therefore,one of the requirements for epidemiological studies of these insect vectors is todetermine whether an individual female mosquito is infected with the malaria par-asite, and also, to what species does she belong. The latter consideration ismost pressing for studies of habitat and reproductive behavior which provide in-formation essential for the design of various control strategies. Thus far, theonly reliable means of distinguishing among the members of this complex wer dif-ferences in polytene chromosome banding patterns as observed in either larvalsalivary gland or adult female ovarian nurse cell tissues (3,4). For field-caught specimens a female would have to be blood fed at least once by the experi-menter to insure that ovarian nurse cells could produce the degree of polytenyrequired for examination. Moreover, interpretation of the chromosomal bandingpattern requires considerable skill. Another method, based upon the frequencI Cfcertain isozyme patterns (5,6), is not as reliable because even certain diagnos-tic alleles are found in both species. Moreover, the assay requires that ex-tracts of fresh or frozen specimens be run on polyacrylamide gels. Finally, theA. gambiae and A. arabiqnsis species can be reliably distinguished by virtue ofcharacteristic HPLC patterns from cuticular hydrocarbons 17,8), but the method isnot practical for large numbers of specimens.

2. Backgrpund. Many of the major malaria vectors are members of species com-plexes, for instance, A. culicifacies (9), A. leucosphyrus (10), and the A. far-auti sibling series (11). In these complexes, as well as in the A. qambiae com-plex, reliable species identification of individuals is currently tedious anddifficult. Since malaria continues to represent a major world health problem,epidemiological studies with these species is crucial. The proposal hypothesizedthat the genomic DNA of A. gambiae and A. arabiens.is currently differs in waysthat would permit reliable species identification. In particular, we sought todevelop a species differentiating assay based upon restriction fragment lengthpolymorphism as detected by either heterologous or Aophel.es probes.

3. Rat ionale. Recently, a substantial body of evidence has argued that RFLPs ex-ist between members of closely related species as well as within a single species(12,13,14). We expected to find such differences between A. gambiae and A. ara-bilesi.s, and such differences should provide an excellent epidemiological tool.The major advantages of a DNA-based assay are (1) the great sensitivity of South-ern analysis so that single mosquitoes (or parts thereof) could be examined, and(2) the ability to use dried material so that field specimens could readily beassayed. In order to find a diagnostic RFLP, we suggested using heterologous

-5-

probes such as the highly conserved actin sequence. A second possibility wnuld

be to use mosquito mitochondrial DNA or rDNA for RFLP analysis. Since the rDNAproved most useful, its properties are discussed in more detail below.

The rDNA genes of most eukaryotic organisms contain spacer regions which, al-

though transcribed, do not appear in the final gene product; these were formerly

called NTS (non-transcribed spacer) regions. Such regions have been shown to di-verge rapidly compared to the rDNA coding sequences and even single copy genes.

Insect rDNA genes also contain intervening sequences (IVS) and/or various other

moderately repeated sequences (15). These introns as well as other repeated se-

quences are similarly much more free to diverge than are coding regions. Quite

possibly these IVS could diverge so much that they could lose all homology to

those of other members of a species complex.

The rDNA genes have another advantage because thus far, dipterans appear to havethese genes present in at least 200 copies per genome arranged in a few large

tandem assays (15). Therefore, the rDNA genes possess the ability to yield use-

ful RFLPs as well as species-specific sequences, both of which would be the basis

for a diagnostic assay.

4. Experiments and-Results. The object of our work was to develop a fast and re-

liable single mosquito assay to distinguish A. gambiae from A. arabiensis. The

first strategy was to determine whether certain heterologous probes would reveala diagnostic RFLP. The second strategy was to search for diagnostic RFLP by us-ing mosquito mitochondrial DNA. The final strategy was to obtain molecular

clones of the rDNA genes of A. gambiae and then to quickly identify non-codino

portions of the genes (spacers and IVS).

(A). Utility of heterologous conserved sequences. Our proposal suggested an

initial survey of heterologous probes known to be highly conserved throughoLt

various phyla. One choice was the Drosophila actin sequence (16,17) because inaddition to its highly conserved sequence, it is a member of a multigene family,

thus giving more copies per genome for detection of single mosquito patterns by

Southern analysis. However, in our preliminary studies, hybridization with the

DPro.5ohil actin probe required such low stringency conditions that the signal tonoise ratio made the blots impossible to interpret. We quickly decided to exam-

ine appropriate mosquito sequences to use as probes for distinguishing RFLPs.

Moreover, since our long term goals include isolation of species-specific mos-

quito sequences, we did not examine any other heterologous probes.

(B). Mitochondria 1 DNA probes from A. gambiae. Studies of mitochondrial RFLPS

have been used to construct the phylogeny of closely related species. We there-

fore decided to utilize this approach for A. gambiae and A. arabiensis. We ob-

tained a clone pDyHB from David Wolstenhome, bearing a 4.Skb HindIII fragmentfrom Drosophila yakuba which includes the cytochrome oxidase I and II genes.

Since these genes are highly conserved, we sought to use this fragment to iden-tify a homologous sequence in A. gambiae. When pDyHB (shown below in red) was

used to probe A. gambiae DNA restricted with various enzyems, Southern analysis

showed a small number of hybridizing bands.

0

IAE A

2 /

A\\ Dic '8-.

AA 8

1.0

3,,

Gene 'ao o :'e lo,'1iL nucleotide palr rosoo iI a V3KtDa mtD'iA molecuie. The

locations of tne '-f-ricn reqicn, :rii orl rn or reuiication ), ind :hedirect oh or nedilcalon r R) relative to tcnRl and Hino[i l restriction si-es(inner drcties) -.ere detemineo oy elect-cr qnicroscoe studies (Fiurn 3no'olste no he t 76 ) 'oc . 'atl . Acad. Sci 73: )2 3-3o2?; (.96,J) Nuci. c-;sRes. d:24 9-2,52; oc dard and 4ostennoime ( 1980) Nuc1. Acids Res. l:?-7 ). acn tRNA cene (natcn e area) is iientified oy thne one lette, 3riino

acid code, an i ndividual serne and leucine t NRA genes are identitied y :hecovon raily (in -areicneses) nc er e ?rinscrortiO) orocucts recn:nid.Arrcws .Itnin and outside tne 7olecule indicate the direc:!on or ahrdn : :rdhof each :ene. The nunoers of apparently noncod ng nucleoctr es nicn 3c jrtetween tre oifferent genes are shown at the gene nounari-es on the innpr sice7r -ne gene -ap. iegative nuoers indicate overlaDoing Cuc)eotides orrd4acent 2enes. An asterisi indicates an incomplete termination Cool - )rTA). Tne location of all EcoRl, qin{dil , Clal , 3o_ ll and Inol restrictionsites., incn were used to 5otain tne clones of se.nents of the 0. 3kOejA Mzu!4Amolecule from ,nicn nucleotide sequences were ootalned are shown on :7e innermap. The letters within the three concentric circles identify tie frajmentsof the restriction enzymes Indicated (references attached).

The combined molecular size of these bands was in no case greater than that(14-17kb) expected for the mitochondrial genome. One of the hybridizing bands isa 5.5kb EcoRi fragment which is a convenient size for cloning into the lambda-gtlO vector. We therefore electroeluted 5-6kb DNA from a preparative EcoRI di-gest, ligated the fragments with lambda-gtlO, and packaged the phage using thecommercially available "Gigapack." The resulting minilibrary was screened usinglowered stringency (hybridization and washing at 420 C) with the pDYHB probe.

One of the selected clones, MR7, had a 5.5kb insert which was used for furtherstudy. Of the several enzymes tried, the most interesting differences betweengambiae and arabiensis were seen with PstJ digests probed with MR7. In additionto the interspecies difference, there was also a suggestion that this probe may

be useful for diagnosing various geographical isolates of a given species, suchas those listed in table I. However, we did not investigate this interestingpossibility any further, because we questioned whether it actually represents mi-tochondrial DNA. One possibility would be that it is actually a nuclear DNA se-

-7-

quence containini a tRiNA also present ( nee hatched ar-eas aboe' ~n *ni tccondr ', .Since tRN'A sequences tend to be highly conserved, he D. yaVur3 n-rcbe c;uid hri..ij

selected either oucloar or initochondciai senlutncv. t-;e founo tlat *, ,(-cod -ccent fraciment, nDvHC (,:hc,n in iri:en), also hyrdzsto NP7. . hich irdicdtr-=

that MR7 most likely is i mitochondrial seoiuence. Htai-ver. -?~ J,' rnot ch~rc:,--ize MR7 any further and therefore vie cannot mak definite statements concerr'lnqits orioin. Regard less of its oriQin. MR7 izz fcraiil K eqll'iale: to a initoc-

drial probe since it hybridizes to mitochondrial DAg.

In summarv. oje Found interestino differences 1et-.ieten a. o A.no+r an-i

ensis usi-c NR- H it, -,,Fr . t1h - P stI catf:e rn di1 e' n.-'ud _- ? f: u t

torpret 'rnth confide-ice if one were using si-ple aootes -- fr'c~

would be inaciorooriate in any case for a general dia~nostic test. .Je msjh't vthen cloned othe:- portions. of the mosquito aiiticcndrial uenorne to - e if a -oreuseful orone could be -Fourd, had we not aiready ontairved interescirQ prelirmir-f-data from mosoulto -DNP coes. Thus. we necan VC exatrir!e r- " r~e 4h1t

might be mo-e usE~ul -2s pot ential diaqncstic tooje.

(C). Isolationi uf diagqnostic cloned ~~ fra _"'nnt.^ nt gi O3 i 1C o eno in clrary was screened with a Sciora ccprophila rDNA clone (!E) vhicf- contain-, -r~e

cunoilete cistron. Thirtv/-tv.o A. gamoiae rDtlP-containiro ohace 4er- isolated -2,d

selecte tor frevealys Tiheenese ontes to rces. cThe 4 clt a s erlr , tr-~and subjected to Sou ther T anajysis, in ordJer tV_ finld nOOOE2O'~ve' rs2ions that

fore oronein w~ith Sc i -ra i- NA w-h i ch t. - ro t e - :e t d o ) u r i -e t 6 tr aome nts rotnie nonconserv.ed r-coi ons . RE-tr,.ct:0)- fro ,nt rr such recions ithoise Tocttridizing to the Sciara probe) vjtere 'hen ic ol~itod frrom _jpls :,rOc USEd to Dr:icE- Cle-nomic Soutnern blots of A. gamoire and 4. at-abieisi- Coe s.' 2 hc.N, :fi

figure 1, was found to contain a3 Q.E-coRi-:E-lI restriction' fracoent whicOconsistently showed a different pattern of hybridization to H. -amtiae Versus A.arabiensis Qenomic D,'4.

18S

BHBH BH H H LONG

AR SE E E S EAM

1KB28S 28S

FIGURE2. XAgr 12 restriction map. The approximate locations of the I SS and :8S regions were determinedby hybridization with heterologous Siaral rDNA (p[3C2). and Calihora" rDNA (pKB-42 and pKB-33).)AAgrl2 contains slightly more than I rDNA ciisron. including the NTS. The dashed line indicates weakhybridization to the heterologous probes. The 0.59 Kb EcoRi-Sal I restriction fragment which reveals a diagnosticrestriction fragment length polymorphism between .4. gamnhar and .A. arabi'nsis is shown as a darkened bar.

The 0.59kb EcoRI-Sall fragment is very close to the 3' terminus of the 28S re-gion of the mosquito rDNA cistron. Hybridization of the Sciara and Call-iphora(15) probes is very weak in this region, suggesting a 'low degree of conservationi,yet this fragment is highly conserved among different geographic isolates of thethree member species in the A. qarnbiae complex first examined. EcoRI-SalI geno-mic digests invariably show the 0.59kb fragment, and there is no evidence for de-tectable levels of inter-cistronic variation in either of these two restriction)sites.

In summary, the nrobe oHGrI2 Bhow- an unambiauous difference between :4. sr,-

biae and A. arabiensls as weil as 2. nelas.

ID) The di gnoicic -fraoment is -isefu1 :?r sia.le dried C',,'U 1 i s T, -

0.59kb fraGient was sulicloned inWu te iluestriLe Th3-iaisJii jS-:: re

ing Systems), nd this COnstruLt, 43rl2A, na-s bj-en ubed to prftje ir.2 ,

cf A. gambise ccmole -olonies atno Field isolates. The A. lsn:iae :mmle-nies available are shown in table .

Table 1. List of -osouto strains ava ladie from t C .

Cesuigat cn Cr nin

Mosquito gercmic DNA was isolaltec from indi¢ia1 ai :" s231itccs nec .r e S! -i0 ,

dessicated at room temperature in the presence of anhydrous calcium sulfate. FAfrom individual mosquitoes or mosquito abdomens was isolated by a simple protcal

(19), digested with EcoRI, separated on a O.7 acarose gel, transferred to Gene-Screen Plus (NEll Research Products), and then probed with nick-translated

pAGrl2A. A. gamoiae mosquitoes from both east and v,est Africa shco a single,

consistent pattern of hybridization to a l.4 b fragment, and this is illustrated

in figure 3a.

]Z1 2 3 4 5 6 7 8 9 10 - ,

236-

95- 4 -

4 66- I"I 23.6 -

43- 9.5 -- 5.5 - - ""

23- 4.3 --2 0--

2.3-

0 56- 2.0-

FtcuRFA. Hybndizaton ofpA rl2A to EcoR[ di-

gtsts oi sinie dned lemaie rtmsouiocs. Secies andL-cn.raon:c Oricin Ol specim-ns are as lof]oVs': 11) A.

P, ( Yhe Gambia). )2) .1 arr Pi tsis (Sudan. SEN- 0.56 -

NAR coilonvL. (3) .-. arar ncst, ISudan. (,NI.-\L :oi-

ofn). (4) 4 ruwnsis (Kcn'aL 5) ..1 arOtnsts (Bur- F6 RFF3B. H hnduation otpA (rl2A to :R d-

kma Fasol. (0) .4 amria ( ran ania., d

ttit

aeI i c',[)\Aesraclcd romsInC emale rosdul 1cs(7znzinarl. t8) .A i'rrrria' e, ' ). ) .4. CjPriO iase and Leoviapbic orcien ol SDeimnens arc ir

(N eri. , and I 0I .4 1'1;: !he (jamnia. 63. - hios, , 'u' es The Gamma: In oani ai .Z :,'ar

ons, ,! r.1':e''sl Sudan (( \ L -. ion', and [ ,"' -

'P.AlIn M ;i::,~.: Io-qicw

This fraonent corresoonds to the .4kb se-ment of AGrl oe iini ted by the

Ec2RI site in the nrce and the ne,.t EcoRI site dovinstream of the EE2 3 codirg

region in the spacer region. None of the more than 1000 individual A. gmbIae

examined from colonies representing 9 geographically different field isolates re-

vealed any variation in this region. All the A. arabiensis examined to date,representing isolates from four different geographic areas, showed a cluster of

EcoRI fragments in the 6-9kb range which contain segments homologous to pAGrl2A.We also examined one available A. melas colony which showed a cluster of bands

centered around 3.5kb. No individuals from the A. arabiensis and A. melas coo-ries have been found with the 1.4kb A. gambiae type fragment. Multiple bands of

hybridization in A. arabiensis and A. melas are probably due to inter-cistronicvariation in the spacer region. In addition, a fourth member of the complex, A.

quadriannulatus, is clearly distinguished by the probe, giving a consistent 2.3kb

EcoRI band, as illustrated in figure 3b. Therefore, the probe to pAGrl2A can beused to examine populations where 2, 3, or 4 of the species are found.

In summary, the diagnostic difference revealed by the probe pAGrl2A was found

without exceptiwi i, individual mosquitoes. Further,, specimens dessicated by a

,,ery simple method -_,Ov. ao ev tden-:e of :)NA denradation even when stored at rooattc-roea tUre fo, 1-1 1,rI q s one year. Moreover- in other preli.T inaiy expe :,,tO

6, r-u, d that oth r lire r ,t.ges such :- secood ,.st:;, dO -rd pup~e (and c''.,i-zJsly' .Zth .:ees ar e-ai'i -5 ,_red ire the P,'A Dro-e.

E) Th' tiog, o t ,_- ag-,be is . ii ed C a, .itlJ() ." '.'e ,rN ci.t.: z,-

pears to be the same in both males and females, as judged by Southern blots of

male and female DNA. However, the intensity of hvbridiztion of 2;AGn-i2A to ca-nomic Southern blots, as shown in fiqure 4, nonicates that nal,.s have a smaji,number of total cc)ios , ich is ei:oected if the ,FL A .enes res ide cn the Xmosome. A. amb 1 ae-, arabiensi s hybrid female ossmi toes mra1 ,i too lob,:,tory contain bcth of the parental tvoes of NA it:-,n .Fig. 4. h? yb

on tne other hand, show the cistron structure of the Female pii.t, Ii:-t- 1 at Itnot the rRNA 7e _es are located on the X Lchro,:o.,c:re. -,,is fir i ' _ t 1-ciates the diacooscic probe with that oart cf the snoscuito qnome cK chr:.- -some) currently Iisec as the basis for cvoceet ic s-eciation.

1 2 3 4 5 6 7 8 9 10 I112

9,5-

66-

4.3-

23-2.0-

0.56-

Fi( ' H' br:diRa4io it pi 2, Ti siniwi dried male and feimalc mosquitoes or Mosquito abdomens.Ian. J , I '| I ' Z."t 2 i ii:''& " ri " ! "! i"[' ic. ) I irrlo,' te a i ando eri oil' 1, (41

$ '.1 . niare'r y,:', hr t c ilc, e r m l t- I t.'' r m alt ' h ih r lemaN. m) I iir maw. ')II0'i~ [Ji ieiirau. '( 'o'>' -, 1r'r~ hthrid mnale. i 1 ia''ti~ ejoi1

il'brid Itmai . anrd I I ha , r I eadii'"ri'tr rid male. female parent is isted first Ir all hsbrtds. D \irom a sr nIcc aindoni en is clearh imoe that suliclen to ni ake a species identilicat on. Furthernore. the presence

ota bloudniwai in the abdomcn dces not sivnl.icanilv reduce DN-\ ield. Dessicated indiv idual pupae and larxae(all inslars except the ttrst) can also be readil> specated (data not sho.,n). DNA extraction'- and h.bnd:zation'" are as described for Fig. 2. Lanes I-8 ,cre exposed to nim, [9 hr. lanes 9-12, C hr.

(F). The DNA probe method is compatible with_ the sporozoite assay and is use-ful for blood meal analysis. In order to determine ,hether the probe could beused to assay single mosquitoes which are also assayed for the presence of themalaria parasite, we obtained a number of field specimens which had been dessi-cated for at least 14 months. The mosquitoes were cut so that Dr. Collins re-taired the head and thorax ;or the sporozoite assay (20) and we tested the abdo-mens. The results, shown in Table 2, indicate that the diagnostic probe canreadily distinguish species using only part of a dried specimen.

Table 2. Testing field sectrens from A. gamoia complex mosquitoes

collected in lisembo, Kenya, October 1985.

Species I-NArOt

Abdomens from: .. g bi A.aae-is nraatile

Plas.,odlum falcicarum

infected mosquiitoes 47 (757.) 17 7/.8

Urrinfected mosquitoes 78 (491.) 80 (51,') 19

Note: Percentaces are based On specimen% which were identified as tosoecies. The soorazoite assay and DNA prboe assay were perfor.-d in

December 1986.

The proport ions of gamb'lae and 6rabiensis in ~Aemno wrticn- fc Tojnd 5re ajswj'jar

to those found by other worker-,. F-i nce the se so ecimen_ we re cu ite clz, an ci i;these experiments at a time w~hen our DNA ex:tractmon. proceourie nad not bee- :o:-mized, there are an unaccsiptatle_ rumoer of unreadasle indi1. _A: corI: T..zi1 . Since then, however , we have had fewi i f at),ura~r a 2 -~r o rspecimens so treated.

A second imoortan, consideration for -i diainostic Dcotie ~S ~ it is C:s-

patible "i th blood (real 6nalysas. The DIN e-trac:ticn protnicu viE. :urje'--lyuse allows blood meal analysis: A S I IL'i710SQui t o (oer p o r tic t E? -eo f I s h --Fr

oenized in a 1.5ml plastic Eppendorf tube (usina a ccinica. iac-;sle 1

50PIl .08M NaCI, .16M sucrose, .OoM EDTfg, .5'. EDS, IlM Tris-CI, _-H G7.6. Tre hcmo-aenate is incubated, 651C, 30 min; 8M potassium acetate is added to a final con-centration of IM. After incubatior, 4~0C, :30 min. and centrifugation at room ten-perature, 12,000xg, 10 min, supernata~it is removed to a fresh tube and the pelletisaved. 100H]1 95%. ethanol is added to the supernatant and centrifuced aga.In.

is00xg 10 min. After discarding the supernatant, the DNrA-containicto pelle-t iLvwashed with 701' ethanol, dried and resuspended in 16H1l 10mM- Tris, 1mTM EDT",. oHe.0. The first (potassium acetate) pellet contains most of the protein. Dr.Collins has examined this protein pellet from a number of the infecteo soeci fenslisted in Table I for blood meal IgG and it can be readily scc-eo fc- rL boc~jmeal source, using commercially available methods (21). Moreover, D:-. CJ.0llinshas also found that the Plasmodium circumsporozoite antigens can also be detected

i n this protein pellet with no apparent loss of sensitivity.

(G). The DiNAprobe method and ODH isozyme method give_ the samne reuil-ts. TheCimportant strategy we employed in comparing the methods was tu test isofenialefamilies. For these experiments female mrosquitoes collected b. Dr. Collins inwestern Kenya (Ahero, Asembo, and Gombe) and one coastal arEa (Sabaki) vieiE used.These are heavily infested areas where the two species are known to be syrnpatr mc.Aherio is an irrigated, rice-growing area on the Kano plain approxsimately 10,.meast of the city of Kisumu. Previous studies of the mosquito fauna of this loca-tion have shown that most of the A. gambiae complex mosquitoes breeding in therice fields are Ai. arabiensis (22). Asembo is a farming community located on the

north shore of Lake Victoria, approximately 45km west of Kisumu. Gombe is a rn-ilar community located at a slightly higher elevation, roughly 20km north of

Asembo. Th2 upland Gombe terrain is considerably more hilly and has a longerrainy season than either Asembo or Ahero. Both A. gambiae and A. arabiensis hsve

been reported from these areas (22). Sabaki is a small coastal village apprcxt-

mately 170km north of Mombassa. Using chromosomal methods of species determina-

tion, Mosha and Subra (23) reported only A. gambiae and A. arabiensis in thisarea. All sites fall in the Sudan-Savannah ecological zone. The western envacollections were made during mid-May 1986, approximately 6 weeks after the onsetof the major rainy season; mosquitoes were collected in Sabaki during mid-Au:ust

1986, a considerably drier season. Gravid specimeis were placed individually incotton-covered, 7-dram vials. After oviposition, the vials were capped andshipped back to the CDC insectaries for rearing as isofemale families. TheSabaki larvae were subjected to salinity testing to screen for the presence of

the brackish water breeding member of the complex, A. merus, but none were found.Some mosquitoes from each family were analyzed for ODH isozymes (EC 1.1.1.73) by

polyacrvlamide gel electrophoresis (PAGE) in Dr. Collins' laboratory. Five dJf-ferent ODH alleles were found among the specimens examined: most of the fre-

quently encountered allele combinations are illustrated in figure 5.

1 2 3 4 5 6 7

95-

100- -

105-

Fig. 5. Octanol dehvdrocenase electromorphs foundin the Kenya held samples. Lane 1. ,A. arabfensts from

the G;NIAL coionv; lanes 2-3. A. ar blensts from .hero,lane 4. an A. araoiensis iGNIAL) x ,A. gam ae kG.3)hybrid produced in the laboratorv lanes 5-7_-. a nl'aefrom the GO-66 colony established with specimens col-lected in Gombe.

The number system used to identify the alleles follows Miles (23). The A.gambiae reference colony 16C/SS used by Miles was employed to identify allele

100. Identities of other alleles were inferred from relative mobilities and put-lished allele frequencies. ODH-98, which we found only in specimens from westernKenya, has not previously been described. This number is assigned on the basisof its mobility relative to ODH-95 and ODH-100. The allele was present in 15 of168 families examined (one family from Asembo was homozygous for the allele).thus it is clearly not rare. Because of the small differences in their relative

mobilities, ODH-98 may not have been differentiated from ODH-100 in previousl

punlished work that used either starch gel or nonstacking PAG-E gel systems.

The most frequent ODH allele in A. gambiae populations is ODH-100, although

ODH-105 and ODH-95 have been observed in relatively low frequencies. A. arabieri-sis populations are typically characterized by the ODH-95 allele, with low fre-quencies of ODH-90 and ODH-100. Thus, specific OH alleles do not absolutel'identify soecies (A. gambiae versus A. arabierisis), but genotypes that do notcontain the shared alleles can be used to separate the species with a very highdegree of reliability.

-13-

The DNA probe method of distinguiuhinqo Species provides a Vt'-vs a,.Fi(quou-sult. All A. gambiae soecimens consistently p,-oduce the same cistinct patt-rr Dfhybridization. Pigure 6 shows nine individual specimens from the Ahero and 3jmoesamoles tested by the NA robe; individuals 2, 7, and 3 are scored as A. oa.-biae, while the other six are . arabiensis.

1 2 3 4 5 6 7 8 9

23.4 -

9.5 - ,-

6-6 - CJC

43 -

2.3-

2.0-

0.56 -

Fi,. b. 1 1briduatoon of the p.-\ cii 2.i prohe to Fco L) !di.ests tiel.-coilecte( speciTnens.

1.djII s I--) .i Ii lai

MosqUimes tromn dilferer)t Ahern Iam~lies, jafns fl-U. ]I)-dividuais frorn Gumue tafimlhes,

The results for 'iDecim.ens and families analyzed v both DNA cobe and CDH iso-zyme are oresented in Table 3. The individual mosquitoes from Sabaki, vihich w eresplit into two portions and analyzed by both methods, clearly show that the DNAprobe and enzymatic methods divide the sample in exactly the same way. All indi-viduals with ODH-1O0 or ODH-105 alleles test as A. gambiae by DNA probe; those

with ODH-95 have the A. arabiensis DNA pattern.

-14-

Table-3. DNA-prohe antd dh ..s.,me analyses o1 .4.gambae complex mosquitoes Iront Kenya

Results for m aterial frotn hero. A em bn and .,n ib , re .resent .ii s r t sl$ t at least t n r osit t nsitiirises t roo t ac, tairurc i, rD N A ~ pe and an tduition -l t ino q uitoes for th lh 151 s t sle uhts tr material tort, 51lak represent N. A-iroie and (dh

isozs me arial% ses on since MStIIr toes tile .h omen i 15 tued 6irOdh anar1sis arid the head-tto a x portion ias Ued Ior .NAts pingl

Probe- nt ec ke a l l .Iiie - nLoca- famijies OiD t alls rriirtion r

ihero 1 .1, gan'ilite

3 A, aratr ria

I A airat.iensisI A, uroarens s

\Xser i A, iamiiae

AI amtae1 .4. gj :mtiaeI A arorneniiS - - -

I . arat,;erir s

1 A arabrii -4 A. irat' iss3 A aratnnsu

.,imre 10 A ramiutae

I A aroinien ris

I A. ra ie nt ,s

[ -V aratenrils

1), kl '6 A g'a mtnaei Ia ga ma e

-J3 A. drtrtren$ is

Specimens from the western Kenya locations were treated so that individudlsfrom each isofemale family were analyzed either by DNA orobe (two individuais/family) or ODH isozyme (two or three individuals/family). Of the 41 differentfamilies so analyzed, none showed any within-family variation in the DNA probehybridization pattern. Furthermore, only the expected 1.4kb or 6-Bkb bands ofhybridization were observed.

All the previously cited studies of ODH alleles in field specimens of A. gam-biae and A. arabiensis indicate that, with near certainty, families with only theODH-100 or ODH-105 alleles can be classified as A. gambiae and families with onlythe ODH-90 or ODH-95 alleles are A. arabiensis. Indeed, the 20 families fromAhero, Asembo, and Gombe with the A. gambiae isozyme types show the DNA probepattern diagnostic of A. gambiae. Also, the six families with only alleles ODH-90 or ODH-95 are identified by the probe as A. arabiensis. The fifteen familieswith other combination of ODH alleles cannot be assigned to species on the basisof their isozyme pattern. However, the DNA probe test of these families indi-cates that 14 of the 15 are A. arabiensis, a finding that is consistent withthose of Miles (24) and others who have reported considerably higher frequenciesof ODH isozyme polymorphism in populations of A. arabiensis than A. gambi-ae.None of the more than 200 individuals examined in this study gave a DNA probe re-sult that would suggest an interspecies hybrid.

In summary, in the 112 individual specimens from Sabaki and the 26 familiesfrom the western Kenya sites where isozyme results permit a reliable species di-agnosis to be made, the results are in agreement with those provided by the DNA

-15-

probe. While this consistent pattern does not represent an absolutely unambiou-

ous validation of the DNA probe method of species determination, it does provide

very strong support. This is especially obvious in the mosquitoes from Sabaki,

where the two methods subdivide the sample identically.

(H). Comparison of DNA.probe..andcy.togenetic methods for dist

inguis .pn -

oambiae from A. arabiensis. Since the chromosome method provides the primary

means for unambiguous differentiation of species, Dr. Collins arranged to collab-

orate with Dr. Vincenzo Petrarca. at the University of Rome, who is expert at the

chromosomal diagnosis of these species. Individual gravid females collected in

Kenya and Zimbabwe by Dr. Collins were dissected so that ovaries were placed in

Carnoy's solution for chromosome squashes. Dr. Petrarca prepared squashes in his

laboratory and scored them. The carcasses were dried and brought to Emory for

testing with the DNA probe.

The Kenya specimens were collected from Asembo and Ahero, where prior studies

(22,23) indicate A. gambiae and A. arabiensis are present. The other specitrens

were collected from the shore of the Lundi River, from a single site in southern

Zimbabwe, where A. arabiensis and A. quadriannulatus have been identifieo.

The results are shown in Table 4:

T nie 4. Jesuin, of couoartoan ot bo *000to byoCCenIba O. of

identif l rit u garcie. of fild coie~te *nceea*~n 01 til A'tvrir

0r- -eroe0 in Co -- 'edCrnoyS (nantmnt far su Ooefutnnifni. offo--rn

frnure iit Coil' Lne cflro.ono. es na- thor.. norti ont aere OnenCed in

the Pe-sence of nnydrou .l. hm - l -fnte for iter OIA eotrsctioh n

telrln a. cmiy rspri~en.I ithl rena. Di orotytene oroolool .- re tested Of

theh tia-- ...t1-1

0.0 PHlUct kthLLT

CYTOCEiOTIC SAML A UtiitLEaT NOT

SITE RESULT NO. CI(iOlUSO iE FRUM qHMOM.bho 1O CIOCILJ

Z mnb .&0. a 9nn~oi ii 0 1)

K.'r. All A.nCtennJ, [us 104

All of the specimens from Kenya matched excent for two. which were mosquitoes

numbered 132 and 133. We believe that these exceptions are due to a mix-up of

samoles because it is highly unlikely that adjacent tubes would contain the only

two exceptions, and these two were therefore dropped from the study. Three of

-16-

the chromosomally scored A. gambiae showed bands diagngstic for both species (6-8kb and 1 .4kb), but we believe these resulted from partial digestion rather thanthat they are indicative of hybrids. Approximately 97% of the cvtogeneticallv

identified specimens were also identified with the DNA probe. In 9 of the 257samples, the DNA had been degraded, possibly due to entry of moisture into themicrofuge tube containing the carcass. Such degradation is evident in the ds-tribution of DNA fragaments in the agarose gels.

The mosquito chromosome squashes were also categorized as to the type of vari-ous polvmorphic chromosomal inversions, which are characteristic of each speciesand are reported to exist in certain frequencies. The data from this aspect ofthe study showed that there was general agreement between observed and expectedfrequencies of these rearrangements. These data will be reported in detail else-where (see bibliography of publications supported by this contract) and are not

treated here.

In summary, 97% of the cytogenetically identified specimens could also beidentified witi the DNA probe, and in every case the DNA probe and cytooeneticmethods of species identification produced concordant results.

(I). Possible use of so ec ies -_soecif ic_ r DNA _seque nces for dot blot ass -ay. Aswas detailed earlier, a dot blot assay would eliminate the need for runnino andblotting gels, the use of restriction enzymes, and possibly also obviate the needfor a radioactive probe. During the latter portion of the contract ,eriod. eexplored the possibility that some of the A. gambiae comple>: rDNA oenes couldcontain species-specific seQuences. In particular, certain dipterans are ,nownto have some rDNA genes which contain seQuences (IVSI that interrupt the BBS ccc-ing region. These IVS should be urder fewer selective pressures than other oor-tions of the rDNA genes and could therefore contain seOuencec tjhich, under normalDNA hybridization conditions, behave in a species-specific mrnner. We theref'-esearched for a possible IVS-containing rDNA clone in A. aambiae. Aooroximatelv15 of the previously identified rDNA clones were subjected to Southern anaLsisand probed with genomic DNA from A. a-rabiensis. We therefore identified a cone.AGr23, which contained two EcoRi. Sail fragments which were not recognized by thearabiensis DNA probe. This clone (and AGr12) was more carefully mapoed and theresults are shown in figure 7.

pAGr i2ASSE B G X X G E B XX xx X SE BAGP1I I I I I I ! I J1 I 'I

CH H HH HH

pAGr23B PAGr23ASE S SE S S E B GX X G

>AGr23 I KBH H HHC C

Fioure 7. Restriction maos of lambda-Atrl2 and lambda-Agr23. Both clonesare shown to illustrate the (probable) position of the IVS. As inother dipterans the IVS interrupts the 26S codinc sequerce. Anotherfeature of the rDNA genes is the internal transcribed sopcer (ITS).The species-specific fragments pAGr23A and pAGr23B, as expected, do

not contain coding region.

-17-

The putative species-specific fragments, shown as pAGr23A and pAGr23B, weresubcloned into Bluescript Ml3 (Stratagene Cloning Systems) and were then usedseparately to probe Southern blots containing single mosquito digests. All ofthese mosquitoes were diagnosed as to species using the pAGrl2A probe. About S00individuals were tested in this way, and we found that only the A. gambiae indi-viduals were positive with pAGr23A or pAGr23B; i.e., some gambiae individuals re-acted with only one of the two IVS probes. The DNA of A. melas, A. merus, and A.qadr annul atus did not react with either probe. We found that there is indivi-dual variation in the intensity of hybridization. This is expected since otherdipterans are known to possess variable numbers of such IVS-containing genes(25).

In summary, the two A. gambiae sequences (in pAGr23A and pAGr23B) behave asspecies-specific probes under the conditions normally used for DNA hybridization.These sequences could be used to design specific probes for dot blot diagnosis ofthe proportion of A. gambiae individuals in a population. Similarly, these ex-periments strongly suggest that analogous arabiensis-specific sequences could beobtained, and thereby allow diagnosis of both species from a single dot blotanalysis.

Literature Cited

1. Coluzzi M, A Sabatini, V Petrarca, and MA DiDeco. 1979. Chromosomal differen-

tiation and adaptation to human environments in the Anopheles gambiae complex.

Trans R Soc Trap Med Hyg 73:483-497.2. White GB. 1974. Anopheles gambiae complex and disease and transmission in

Africa. Trans R Soc Trap Med Hyg 68:278-301.

3. Coluzzi M. 196e. Chromosomi politenici delle cellule nutrici ovariche nel

complesso gambiae del genera Anopheles. Parasitol l0:179-183.

4. Coluzzi M and A Sabatini. 1967. Cytogenetic observations on species A and Bof the Anopheles gambiae complex. Parasitol 9:73-8.5. Miles SJ. 1979. A biochemical key to adult members of the Aropheles gambiae

group of species. J Med Entomol 15:297-299.

6. Marchand RP and AEP Mnzava. 1985. A field test of a biochemical key to iden-tify members of the A. gambiae group of species in northeast Tanzania. J Trap MedHyQ BB:205-210.

7. Carlson DA and MW Service. 1q0. Identification of mosquitoes of Anopheles

gambiae species complex A and B by analysis of cuticular components. Science 207:

1089-1091.8. Hamilton RJ and MW Service. 19B3. Value of cuticular and internal hydrocar-

bons for the identification of larvae of Anopheles gambiae Giles, Anephel .r-

biensis Patton and Anopheles melas Theobald. Ann Trap Med PlSitol 77:203-21 .9. Miles SJ. 1981. Unidirectional hybrid sterility Trom crosses between speciesA and species B of the taxon Annpheiles (Cellia) culicifacies Giles. J Trap MedHyg 84:13-16.

10. Baimai V, BA Harrison, and L Somchit. 1981. Karvotype differentiation of

three anopheline taxa in the Balabacensis complex of Southeast Asia (Diptera:

Culicidae). Genetica 57:81-96.11. Mahon RJ and PM Miethke. 1982. Anopheles farauti No. 3, a hitherto unrecog-nized biological species of mosquito within the taxon A. farauti Laveran (Dip-tera: Culicidae). Trans R Soc Trop Med Hyg 76:8-12.12. Jeffreys AJ. 1981. Recent studies of gene evolution using recombinant DNA.In: Genetic Engineering 2 (R Williamson, ed.), NY, Academic Press, pp. I-4e.

13. Avise J, RA Lansman, and RO Shade. 1979. Use of restriction endonucleases tomeasure mtDNA sequence relatedness in natural populations. I. Peromyscus. Gene-

tics 92:279-295.

14. Langley CH, E Montgomery, and W Quattlebaum. 1982. Restriction map variationin the AdH region of D.r..osop.h.i..1. PNAS/USA 79:5631-5635.15. Beckingham K. 1980. The ribosomal DNA of Calliphora erythrocephala: An analy-sis of hybrid plasmids containing ribosomal DNA. J Mol Biol 137:349-373.

16. Ng E and J Abelson. 1980. Isolation and sequence of the gene for actin in S.cerevisiae. PNAS/USA 77:3912-3916.

17. Sanchez F, SL Tobin, U Rdest, E Zulauf, and BJ McCarthy. 1983. Two Drosophila

actin genes in detail: Gene structure, protein structure, and transcription dur-ing development. J Mol Biol 163: 533-551.

18. Gerbi S and R Renkowitz. 1979. Sciara rDNA has a small homogenous repeat. Mol

Gen Genet 173:1-13.

19. Livak K. 1984. Organization and mapping of a sequence on the Drosophila mela-

nogaster X and Y chromosomes that is transcribed during spermatogenesis. Genetics

107:611-634.20. Collins FH, F Zavala, PM Graves, AH Cochrane, RW Gwadz, J Akoh, and RSNussenzweig. 1984. First field trial of an immunoradiometric assay for the detec-

tion of malaria sporozoites in mosquitoes. Am J Trap Med Hyg 33:227-230.21. Washino RK and CH Tempelis. 1983. Blood meal analysis. Ann Rev Entomol 28:

179-202.

-19-

22. White GB. 1972. The Anopheles gambiae complex and malaria transmission around

Kisumu, Kenya. Trans R Soc Trop Med Hyg 66:572-58l.

23. Mosha FW and R Subra. 1982. Ecological studies on Anopheles gambiae complex

sibling species in Kenya. I. Preliminary obseriations on their geographical dis-

tribution and chromosomal polymorphic inversions. WHO/VBC/82.867.

24. Miles SJ. 1978. Enzymatic variation in the Anopheles gambiae Giles group of

species (Diptera: Culicidae). Bull Entomol Res 68:85-96.

25. Gerbi S. 1985. Evolution of ribosomal DNA. In: Molecular Evolutionary Gene-

tics (RJ MacIntyre, ed.), NY, Plenum Press, chapter 7.

-20-

Bibliography of Publications Supported by DAMD 17-85-C-5184

Collins FH, AM Mendez, MO Rasmussen, PC Mehaffey, NJ Besansky & V Finnert,.1987. A ribosomal RNA gene probe differentiates member species of the Ano-pheles gamrbiae complex. Amer 3 Trop Med Hyg 37:37-41.

Collins FH, PC Mehaffey, MO Rasmussen, AD Brandling-Bennett, 3 Odera, & V Fin-nerty. 1988. Comparison of DNA probe and isozyme methods for differentiat-ing Anopheles gambiae and Anopheles arabiensis. I Med Entomol 25:l16-120.

Collins FH, V Finnerty & V Petrarca. 1988. Ribosomal DNA probes differertlaefive cryptic species in the An. gambiae complex. Parasitol (in press).

Collins FH, V Petrarca, S Mpofu, AD Brandling-Bennett, JBO Were, MO Rasmusser,& V Finnerty. 1988. Comparison of DNA probe and isozyme methods for differ-entiating Anopheles gambiae and Anopheles arabiensis. J Med Entomol 'inpress).

Finnerty V and FH Collins. 1988. Ribosomal DNA probes for identification of mem-ber species of the Anopheles gambiae complex. Symp Fla Ent Soc (Daytona Eeach.August 1987) (in press).

*

DISTRIBUTION LIST

5 copies DirectorWalter Reed Army Institute of ResearchWalter Reed Army Medical CenterATTN: SGRD-UWZ-CWashington, DC 20307-5100

1 Copy CommanderUS Army Medical Research and Development CommandATTN: SGRD-RMI-SFort Detrick, Frederick, Maryland 21701-5012

2 copies Defense Technical Information Center (DTIC)ATTN: DTIC-DDACCameron StationAlexandria, VA 22304-6145

1 copy Dean

School of MedicineUniformed Services University of the

Health Sciences4301 Jones Bridge RoadBethesda, MD 20814-47)9

I copy CommandantAcademy of Health Sciences, US ArmyATTN: AHS-CDMFort Sam Houston, TX 78234-6100