Genetic variation between strains of the Mediterranean fruit fly, Ceratitis capitata, detected by DNA fingerprinting DAVIDS. HAYMER Department of Genetics and Molecular Biology, University of Hawaii, Honolulu, HI 96822, U. S.A .
DONALD 0. MCINNIS Agricultural Research Service, USDA, P.0. Box 2280, Honolulu, HI 96804, U. S.A.
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
LORETTA ARCANGELI Department of Genetics and Molecular Biology, University of Hawaii, Honolulu, HI 96822, U. S.A . Corresponding Editor: R. Appels Received November 7, 1991 Accepted December 14, 1991 L. 1992. Genetic variation between strains of the Mediterranean HAYMER, D. S., MCINNIS,D. O., and ARCANGELI, fruit fly, Ceratitis capitata, detected by DNA fingerprinting. Genome, 35: 528-533. DNA fingerprinting has been used to detect genetic variation in the Mediterranean fruit fly, Ceratitis capitata. Three different probes have been identified that can be used to detect DNA restriction fragment length polymorphisms between strains of this species. The strains used in this study differ only in terms of their geographic origin or genetic background. One of the probes used is the bacteriophage vector M13, and the other two are repetitive sequences derived from the medfly genome based on a weak homology to M13. Within a strain, each probe produces a consistent restriction fragment profile that is not affected by the method or timing of DNA extraction. Between strains, when M13 is used as a probe, an average of 10% of the observable bands are polymorphic. Use of the medfly genomic sequences as a probe increases the proportion of polymorphic bands between strains up to 30%. The fact that genetic differences between even such closely related strains can be reliably detected by this method holds great promise for studies of insect pests including the ability to monitor the movements of pest species, determining the extent of genetic variation in pest populations, and in making identifications from otherwise unidentifiable material. Key words: genetic variation, Ceratitis capitata, DNA fingerprinting.
'
HAYMER,D. S., MCINNIS,D. O., et ARCANGELI, L. 1992. Genetic variation between strains of the Mediterranean fruit fly, Ceratitis capitata, detected by DNA fingerprinting. Genome, 35 : 528-533. Des cartes molCculaires de 1'ADN ont permis de dCceler des variations gCnCtiques chez le Ceratitis capitata, la mouche a fruit mCditerranCenne. Trois sondes differentes ont CtC identifikes et utiliskes pour dCtecter le polymorphisme des longueurs des fragments de restriction de 1'ADN entre les souches de cette espece. Les souches soumises a I'expCrimentation n'ont diffCrC que par leur origine gkographique et leur bagage gCnCtique. L'une des sondes a CtC le vecteur bactCriophage M13 et les deux autres furent des sCquences rCpCtitives dCrivCes du genome de C. capitata, ayant une faible homologie avec le vecteur M13. A 1'intCrieur des souches, chacune des sondes a produit un profil de fragments de restriction compatibles que ni la mCthode ou le temps d'extraction de I'ADN n'ont affect& Entre les souches, lorsque le M13 a CtC utilisC, une moyenne de 10% des bandes observables furent polymorphes. L'emploi des sCquences gCnomiques de la mouche a fruit mediterrankenne comme sondes s'est traduit par des augmentations des bandes polymorphes entre les souches, atteignant jusqu'a 30%. Que cette mCthode permette de dCtecter des diffkrences gCnCtiques entre de telles souches Ctroitement reliCes devient prometteur pour 1'Ctudes des insectes nuisibles, incluant la capacitC de suivre les mouvements des insectes, de determiner 1'Ctendue des variations gCnCtiques a 1'intCrieur des populations et d'identifier un materiel qui, autrement, ne saurait &re identifik. Mots elks : variation gCnCtique, carte molCculaire de l'ADN, Ceratitis capitata. [Traduit par la rCdaction]
Introduction DNA fingerprinting is a relatively new, powerful method for detecting genetic variation at a variety of levels of biological organization. This technique has been used to detect genetic variation at levels ranging from recombinant gametes within an individual (Arnheim 1989) to differences between populations (Kirby 1990). Beyond simply detecting genetic variation, this technique also offers a method for analyzing the genetic structure of populations, documenting genetic changes that might occur over space or time, and for determining the genetic origin of newly established populations. DNA fingerprinting has also been used in forensic applications, pedigree analysis, and in studies of recombinational processes (Jeffreys et al. 1990). Finally, DNA fingerprinting has also been used to identify molecular markers that may make it possible to establish Printed in Canada / lmprime au Canada
linkage relationships between anonymous DNA sequences and desirable traits in an organism. Once these linkage relationships are established, the polymorphic DNA fragments identified can also be used in efforts to clone the genes of interest (Brilliant et al. 1991). In the DNA fingerprinting method, polymorphic restriction fragment profiles are generated from repetitive sequences present in a genome. The types of restriction fragment length polymorphisms (RFLPs) most commonly identified for this purpose are repetitive sequences known as hypervariable minisatellites or tandemly repeated sequences, sometimes referred to as variable number tandem repeats or VNTRs. These are regions of a genome consisting of a core sequence with variable numbers of repeats that can be both tandemly linked and dispersed at different loci.
HAYMER ET AL.
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
Although the mechanism by which these complex variable repeat sequences arise is not completely understood, the fact that the pattern is specific within an individual genome provides a powerful tool for identifying differences and detecting genetic variation (Jeffreys et al. 1990). In producing a DNA fingerprint, the restriction fragment profiles are visually observed by hybridizing a probe sequence with DNA extracted from a source such as an individual or a strain. The probes used in this type of analysis are often derived from, or have homology to, some of the repetitive sequences mentioned above. Many of the most informative diagnostic repetitive sequences obtained from a variety of species have been identified based on a fortuitous homology with a bacteriophage vector named M13 (Kirby 1990; Maniatis et al. 1982). We have used this fortuitous homology to begin an analysis of the molecular genetic structure of populations of the Mediterranean fruit fly, Ceratitis capitata (Weidemann). At the present time relatively little is known about the molecular structure of medfly populations. This is true despite the fact that the medfly is rated as one of the world's most destructive agricultural pests. These flies are highly vagile, capable of rapidly colonizing new habitats, and responsible for major economic impacts in areas of intensive agriculture where they are found (Saul 1986). Although some surveys of electrophoretic variation have been conducted on medfly populations, the general result produced by these studies has been a paucity of electrophoretic variants that are diagnostic of different strains, even between the most widely separated geographic populations. As shown by Milani et al. (1989) using natural isolates, only the most variable medfly populations have mean per locus heterozygosity values H that approach those reported for Drosophila ( H = 0.15). However, even among these populations, unique alleles were rarely found and with few exceptions, the same alleles tended to reach the highest frequencies in all samples regardless of geographic origin. Moreover, when these strains were put into laboratory culture, even more homogenization was shown to occur. We show here that using DNA fingerprinting techniques, even very closely related geographic isolates and strains of the medfly that have been in laboratory culture can be reliably differentiated. In addition to the fact that these molecular markers are more reliable for discriminating between strains, these same markers can be used to track and monitor the genetic structure, populations dynamics, and movements of such pest species. Finally, these same RFLP markers may be useful in developing a linkage map of a genome such as that of the medfly where only limited amounts of classical genetic analyses have been done.
Materials and methods Strains The Kula and Mauna Loa strains were originally collected from Kula, Maui, and Mauna Loa, Hawaii, respectively, in 1989 and have been maintained in the laboratory since then. The Robinson strains have been in laboratory culture since 1983. The Israel strain is maintained in a USDA certified quarantine facility at the University of Hawaii, and material from this strain was generously provided by Dr. S. Saul. The A5A3 strain used to construct the genomic library is a subline derived from the USDA "Hilab" strain
529
maintained at the USDA Tropical Fruit and Vegetable Laboratory, Manoa, Hawaii. Genomic DNA isolation and library construction Genomic DNA was isolated from the various strains of C. capitata by one of two different methods. Method I . "Lifton" method of DNA extraction This method was used to isolate genomic DNA for library construction and for Southern blots. In this method, DNA was isolated from adult flies by homogenizing approximately 350 mg of freshly etherized flies in 5 mL 0.2 M sucrose, 50 mM EDTA, 100 mM tris pH 9.0, and 0.5% SDS in a Pyrex No. 7725 homogenizer. The homogenate was filtered through autoclaved polyester fiberfil. The filtered homogenate was treated with Proteinase K at a final concentration of 200 pg/mL for 1 h at 65"C, after which potassium acetate was added to a final concentration of 1.2 M. Debris was pelleted out by centrifugation for 15 min at 10 000 rpm, 4°C. DNA was ethanol precipitated from supernatant, resuspended in TE, and extracted with phenol and chloroform - isoamyl alcohol (24:l). DNA was again precipitated with ethanol and resuspended in TE for use. Method 2. "Salt" method of DNA extraction This method was used to isolate genomic DNA for Southern blot analysis. DNA was extracted in this method by homogenizing approximately 350 mg of freshly etherized adult flies in 5 mL 10 mM tris pH 7.4,60 mM NaC1, 10 mM EDTA, 0.15 mM spermine, 0.15 mM spermidine, and 0.5% (v/v) Triton X-100. The homogenate was strained through gauze and centrifuged for 10 min at 7000 rpm at 4°C to pellet nuclei in an SS34 rotor. Nuclei were then resuspended in 5 mL 75 mM NaCl and 25 mM EDTA, and SDS was added to 1%. The sample was treated with Proteinase K at 200 pg/mL for 1 h at 55"C, followed by two chloroform extractions of 30 min each in high salt (1.6 M NaCl). DNA was then precipitated with isopropanol. Each of these procedures were also modified for single fly DNA extractions. This was accomplished by scaling down the procedures approximately 30-fold. Single flies were homogenized in Kontes microhomogenizers. Restriction fragment library This genomic library was constructed by ligating DNA from the A5A3 subline into arms of XgtlO cut with EcoR1. The A5A3 genomic DNA had been completely digested also with EcoR1. This type of construction produces a library of mostly single EcoRl fragments ranging in size from less than 1 kb up to about 7 kb. Probes and hybridizations All probe DNAs were labelled by random priming with the nonradioactive "Genius" labelling and detection kit from Boehringer Mannheim Biochemicals. The plaque hybridizations of the genomic library were conducted as described by Benton and Davis (1977). Probes and hybridizations for plaque lifts and Southern blots using M13 as a probe were conducted under reduced stringency conditions using the Boehringer Mannheim Biochemicals nonradioactive labelling and detection kit as described by the manufacturer. In these cases the entire M13mp18 RF was used to make the probe. Reduced stringency conditions were as follows. Hybridizations were conducted using 35% formamide at 42°C. Reduced stringency washes were conducted at 35 "C in 0.1 % SDS with decreasing concentrations of SSC down to 0.5 x . For Southern blots, genomic DNA was digested at 37°C for 2-3 h using enzymes and buffers from Boehringer Mannheim Biochemicals. DNA was electrophoresed on 0.7% agarose gels with TBE and Southern blotted to Micron Separations Magna Graph nylon membrane as described by Southern (1975). Hybridization, washing, and immunological detection of the Southern blots was conducted as described in the Boehringer Mannheim Biochemicals "Genius" nonradioactive labelling and
GENOME,
-
VOL. 35,
1992
-
0 C C L I
- 2 2 6
2 .O .O
L s s :0z3 m o o
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
uggCb4
FIG. 1. Southern blot of genomic DNA from five different medfly strains digested with EcoRl and probed with M13. Size markers are from a Hind111 digest of lambda DNA. Arrows indicate examples of major RFLP differences consistently observed between strains. detection kit. When either pmedMl3e or pmedMl3f was used as a probe, more stringent conditions were used. Hybridizations were conducted in 43% formamide buffer at 42OC, with washes at 55OC in 0.1 % SDS and decreasing concentrations of SSC down to 0.5 x .
Results Using the bacteriophage vector M13 as a probe, RFLPs can be observed between each of the strains tested here (Fig. 1). Shown here is a Southern blot of EcoRl digests of genomic DNAs. Examples of some of the prominent RFLP differences consistently observed between strains are indicated with arrows. The RFLP differences seen here represent variation in repetitive sequences present in these
FIG. 2. Southern blot of genomic DNAs from the Kula strain digested with EcoRl and probed with pmedM 13f. Lane A contains DNA extracted by the "salt" method, while lanes B and C contain DNA extracted by the "Lifton" method on two different days. genomes, in a manner analogous to hypervariable minisatellite sequences described in the human genome (Jeffreys et al. 1990). Of the strains shown here, the Kula and Mauna Loa strains originated from different geographic localities within the Hawaiian islands. As expected, these strains have many DNA fragments (bands of hybridization) in common but, in addition, there are some notable RFLPs observable between the two strains. The two "Robinson" strains are both derived from a line bearing a translocation of a wildtype gene for pupal color to the Y chromosome (Robinson 1989). They differ only in that the Robinson (Hawaiian) strain has been outcrossed to a strain from the Hawaiian islands (also originally from Kula), while the Robinson (pure) has not been outcrossed. In the same manner discussed above, these two strains are expected to have a great
HAYMER ET AL.
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
Kula
FIG. 3. Southern blot of genomic DNA from individual females of the Kula strain digested with EcoRl and probed with pmedM 13f.
deal of their genetic background in common, although here again some notable polyrnorphisms are consistently detected. Although the M13 probe can be used to elucidate variation between these strains in this manner, these blots are of necessity done under very low stringency conditions (see Materials and methods). To increase the resolving power of the technique and to allow the blots to be done under more stringent conditions, we next isolated some of the M13 homologous sequences directly out of the medfly genome. To accomplish this we screened our Xgt 10 genomic library using M13 as a probe under conditions identical to those described for the Southern blots. We initially isolated five different medfly sequences that, based on hybridization, had some homology to the vector M13. These five sequences were cloned and used individually to probe genomic DNA
FIG. 4. Southern blot of genomic DNA from the five different medfly strains digested with EcoRl and probed with pmedM13f.
from different medfly strains. Two of the five produced informative patterns of RFLP differences between strains and were kept for further analysis. These were designated pmedM 13e and pmedM 13f. These clones contain EcoR 1 inserts 0.9 and 2.0 kb in size, respectively. In addition to examining differences between strains using the DNA fingerprinting methodology, we also wanted to show that we could obtain consistency within a strain or between individuals of a strain. To this end we first compared DNA from the same strains extracted by different methods and on different days. One such result is shown in Fig. 2. Here, the Kula A lane contains genomic DNA extracted from a pool of adult flies by the "salt" method described in the Materials and methods section. Lanes labelled Kula B and C contain genomic DNA extracted from adults by the "Lifton" method (see Materials and methods),
GENOME, VOL. 35, 1992
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
number of DNA polymorphisms can be seen. For the pmedM13f probing (Fig. 4), the majority of the polymorphism~seen are in the range of 3-7 kb in size, although there are some very high molecular weight differences. In the case of pmedM13e (Fig. 5), a large number of polymorphisms can be seen ranging from under 2 kb in size up into a very high molecular weight range.
FIG. 5. Southern blot of genomic DNA from the five different medfly strains digested with EcoRl and probed with pmedMl3e.
each on different days. The probe here is pmedM13f. As can be seen, the restriction fragment profile elucidated here is consistent within a strain, with no detectable differences observable owing to the method or timing of .the DNA extraction. We have obtained similar consistent results for each of the probes discussed above within both the Kula and Mauna Loa strains. Figure 3 shows the results obtained when five individual females of the Kula strain were probed with the same probe, pmedMl3f. Here again, there is consistency within a strain with essentially no qualitative differences in the banding pattern observable. Along this line we tested each of our probes on DNA extracted from individual males and females from both Hawaiian island strains and the results were consistent with those shown here. Each of these probes was also used to probe genomic DNA from each of the different strains tested earlier with M13. These blots were done under much more stringent hybridization conditions. In both of these cases, a large
Discussion We have shown here that DNA fingerprinting can be used to,detect genetic variation between strains of the medfly. Beginning with a fortuitous homology between the bacteriophage vector M13 and repetitive sequences present in the medfly genome, we have shown that DNA RFLPs can be identified that are diagnostic of the genetic makeup of different strains. The strains used here differ only in terms of their geographical origin or their genetic background. Similar approaches have been used to study population differentiation in wildlife conservation studies (Brock and White 1991; Gilbert et al. 1990; Kirby 1990) and to identify cultivar varieties of agriculturally important commodities (Nybom 1990). In our work on the medfly, using M13 as a probe to produce the DNA fingerprint pattern, differences are detected between all of the strains tested here. However, with M13 only a small proportion of the bands are polymorphic. Not counting the Israel strain (which has been included as an out group), only an average of 10% (i.e., 2 of 21) of the observable bands differ between the strains. In addition, the hybridization conditions used with the M13 probe are of necessity very "relaxed," meaning that a fair amount of mismatch is permissible between the probe and the target genomic sequences (Maniatis et al. 1982). On the other hand when one of the medily genomic sequences (pmedM13e or pmedM13f) is used as a probe, much more variation between all of the strains can be detected. In the case of pmedM13f, a smaller number of total bands can be seen (average of 1 9 , and on average 4-5 of these bands (25-30%) are polymorphic between the strains. For pmedMl3e, an average of 6 of the 19 visible bands differ in each case. This means a substantial increase in the information and diagnostic value of this approach when a "native" genomic sequence is used as a probe. In addition, these hybridizations are done under more "stringent" conditions, meaning that a more precise match between the probe and genomic target sequence is required. This can be important as far as the reproducibility of the technique. ' As far as reproducibility is concerned, we have shown that while differences between strains can be reliably detected with all three of the probes described here, there is also consistency within a strain. For example, essentially the same pattern is seen for the Kula strain in different probings of this DNA with pmedl3f in Figs. 2, 3, and 4. Similar consistent results have been obtained using M13 itself and pmedM13e (not shown). We have also shown that the DNA fingerprint pattern obtained within a strain is not affected by the method or timing of the DNA extraction (done on random samples of 20-30 adult flies), nor is the pattern different between single individual males or females from the same strain. One of the great advantages to this approach is that it can be done in a way that requires very little information
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
HAYMER E T AL.
a priori about specific genes and (or) sequences present in a genome. We are now determining the nucleotide sequence of the two clones pmedMl3e and pmedMl3f. Once this information is available, it should be possible to design oligonucleotide primers that can be used in amplification of one or more of these polymorphic sequences by the polymerase chain reaction or PCR method (Kirby 1990). This would allow the identification or assessment of genetic variation to be done on even limited amounts of material such as dessicated or preserved body parts, such as might be recovered from traps in a field collection. The availability of these techniques can be extremely valuable in terms of determining the genetic origin of flies moving into or establishing new habitats and for monitoring genetic changes that might occur in populations over space or time. In conclusion, we have detected genetic variation between strains of the Mediterranean fruit fly using DNA fingerprinting. We have identified probes and described conditions under which considerably more genetic variation can be detected compared to previous methods of population assessment. The variation detected between strains is consistent within individuals or within populations of flies of the same strain, and it is not affected by different methods used to extract DNA. The strains used here differ only in terms of geographic origin (including different islands within Hawaii) or in terms of genetic background.
Acknowledgements This work was supported by California Department of Food and Agriculture grant 90-0486 and USDA grant HAW 00763-G.
533
Arnheim, N. 1989. New approaches to constructing genetic maps: PCR analysis of DNA sequences in individual gametes. In PCR technology. Edited b y H. Erlich. Stockton Press, New York. pp. 119-136. Benton, W., and Davis, R. 1977. Screening XgtlO recombinant clones by hybridization to single plaques-in situ. Science (Washington, D.C.), 196: 180. Brilliant, M.H., Gondo, Y., and Eicher, E. 1991. Direct molecular identification of the mouse pink-eyed unstable mutation by genome scanning. Science (Washington, D.C.), 252: 566-569. Brock, M.K., and White, B.N. 1991. Multifragment alleles in DNA fingerprints of the parrot, Amazona ventralis. J. Hered. 82: 209.-2 12. Gilbert, D.A., Niles, L., O'Brien, S.J., and Wayne, R.K. 1990. Genetic fingerprinting reflects population differentiation in the California channel island fox. Nature (London), 344: 764-766. Jeffreys, A. J., Neumann, R., and Wilson, V. 1990. Repeat unit sequence variation in minisatellites: a novel source of DNA polymorphism for studying variation and mutation by single molecule analysis. Cell, 60: 473-485. Kirby, L.T. 1990. DNA fingerprinting: an introduction. Stockton Press, New York. Maniatis, T., Fritsch, E., and Sambrook, J. 1982. Molecular cloning. Cold Spring Harbor Press, Cold Spring Harbor, New York. Milani, R., Gasperi, G., and Malacrida, A. 1989. Biochemical genetics. In Fruit flies. Vol. 3b . Edited b y A.S. Robinson and G. Hooper. Elsevier, New York. pp. 33-56. Nybom, H. 1990. DNA fingerprints in sports of 'Red Delicious' apples. Hortscience, 25: 1641 - 1642. Robinson, A. 1989. Genetic sexing methods in the Mediterranean fruit fly, Ceratitis capitata. In Fruit flies. Vol. 3b. Edited by A. S. Robinson and G. Hooper. Elsevier, New York. pp. 57-68. Saul, S. 1986. Genetics of the Mediterranean fruit fly, Ceratitis capitata. Agric. Zool. Rev. 1: 73-108. Southern, E. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503.
DONALD 0. MCINNIS Agricultural Research Service, USDA, P.0. Box 2280, Honolulu, HI 96804, U. S.A.
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
LORETTA ARCANGELI Department of Genetics and Molecular Biology, University of Hawaii, Honolulu, HI 96822, U. S.A . Corresponding Editor: R. Appels Received November 7, 1991 Accepted December 14, 1991 L. 1992. Genetic variation between strains of the Mediterranean HAYMER, D. S., MCINNIS,D. O., and ARCANGELI, fruit fly, Ceratitis capitata, detected by DNA fingerprinting. Genome, 35: 528-533. DNA fingerprinting has been used to detect genetic variation in the Mediterranean fruit fly, Ceratitis capitata. Three different probes have been identified that can be used to detect DNA restriction fragment length polymorphisms between strains of this species. The strains used in this study differ only in terms of their geographic origin or genetic background. One of the probes used is the bacteriophage vector M13, and the other two are repetitive sequences derived from the medfly genome based on a weak homology to M13. Within a strain, each probe produces a consistent restriction fragment profile that is not affected by the method or timing of DNA extraction. Between strains, when M13 is used as a probe, an average of 10% of the observable bands are polymorphic. Use of the medfly genomic sequences as a probe increases the proportion of polymorphic bands between strains up to 30%. The fact that genetic differences between even such closely related strains can be reliably detected by this method holds great promise for studies of insect pests including the ability to monitor the movements of pest species, determining the extent of genetic variation in pest populations, and in making identifications from otherwise unidentifiable material. Key words: genetic variation, Ceratitis capitata, DNA fingerprinting.
'
HAYMER,D. S., MCINNIS,D. O., et ARCANGELI, L. 1992. Genetic variation between strains of the Mediterranean fruit fly, Ceratitis capitata, detected by DNA fingerprinting. Genome, 35 : 528-533. Des cartes molCculaires de 1'ADN ont permis de dCceler des variations gCnCtiques chez le Ceratitis capitata, la mouche a fruit mCditerranCenne. Trois sondes differentes ont CtC identifikes et utiliskes pour dCtecter le polymorphisme des longueurs des fragments de restriction de 1'ADN entre les souches de cette espece. Les souches soumises a I'expCrimentation n'ont diffCrC que par leur origine gkographique et leur bagage gCnCtique. L'une des sondes a CtC le vecteur bactCriophage M13 et les deux autres furent des sCquences rCpCtitives dCrivCes du genome de C. capitata, ayant une faible homologie avec le vecteur M13. A 1'intCrieur des souches, chacune des sondes a produit un profil de fragments de restriction compatibles que ni la mCthode ou le temps d'extraction de I'ADN n'ont affect& Entre les souches, lorsque le M13 a CtC utilisC, une moyenne de 10% des bandes observables furent polymorphes. L'emploi des sCquences gCnomiques de la mouche a fruit mediterrankenne comme sondes s'est traduit par des augmentations des bandes polymorphes entre les souches, atteignant jusqu'a 30%. Que cette mCthode permette de dCtecter des diffkrences gCnCtiques entre de telles souches Ctroitement reliCes devient prometteur pour 1'Ctudes des insectes nuisibles, incluant la capacitC de suivre les mouvements des insectes, de determiner 1'Ctendue des variations gCnCtiques a 1'intCrieur des populations et d'identifier un materiel qui, autrement, ne saurait &re identifik. Mots elks : variation gCnCtique, carte molCculaire de l'ADN, Ceratitis capitata. [Traduit par la rCdaction]
Introduction DNA fingerprinting is a relatively new, powerful method for detecting genetic variation at a variety of levels of biological organization. This technique has been used to detect genetic variation at levels ranging from recombinant gametes within an individual (Arnheim 1989) to differences between populations (Kirby 1990). Beyond simply detecting genetic variation, this technique also offers a method for analyzing the genetic structure of populations, documenting genetic changes that might occur over space or time, and for determining the genetic origin of newly established populations. DNA fingerprinting has also been used in forensic applications, pedigree analysis, and in studies of recombinational processes (Jeffreys et al. 1990). Finally, DNA fingerprinting has also been used to identify molecular markers that may make it possible to establish Printed in Canada / lmprime au Canada
linkage relationships between anonymous DNA sequences and desirable traits in an organism. Once these linkage relationships are established, the polymorphic DNA fragments identified can also be used in efforts to clone the genes of interest (Brilliant et al. 1991). In the DNA fingerprinting method, polymorphic restriction fragment profiles are generated from repetitive sequences present in a genome. The types of restriction fragment length polymorphisms (RFLPs) most commonly identified for this purpose are repetitive sequences known as hypervariable minisatellites or tandemly repeated sequences, sometimes referred to as variable number tandem repeats or VNTRs. These are regions of a genome consisting of a core sequence with variable numbers of repeats that can be both tandemly linked and dispersed at different loci.
HAYMER ET AL.
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
Although the mechanism by which these complex variable repeat sequences arise is not completely understood, the fact that the pattern is specific within an individual genome provides a powerful tool for identifying differences and detecting genetic variation (Jeffreys et al. 1990). In producing a DNA fingerprint, the restriction fragment profiles are visually observed by hybridizing a probe sequence with DNA extracted from a source such as an individual or a strain. The probes used in this type of analysis are often derived from, or have homology to, some of the repetitive sequences mentioned above. Many of the most informative diagnostic repetitive sequences obtained from a variety of species have been identified based on a fortuitous homology with a bacteriophage vector named M13 (Kirby 1990; Maniatis et al. 1982). We have used this fortuitous homology to begin an analysis of the molecular genetic structure of populations of the Mediterranean fruit fly, Ceratitis capitata (Weidemann). At the present time relatively little is known about the molecular structure of medfly populations. This is true despite the fact that the medfly is rated as one of the world's most destructive agricultural pests. These flies are highly vagile, capable of rapidly colonizing new habitats, and responsible for major economic impacts in areas of intensive agriculture where they are found (Saul 1986). Although some surveys of electrophoretic variation have been conducted on medfly populations, the general result produced by these studies has been a paucity of electrophoretic variants that are diagnostic of different strains, even between the most widely separated geographic populations. As shown by Milani et al. (1989) using natural isolates, only the most variable medfly populations have mean per locus heterozygosity values H that approach those reported for Drosophila ( H = 0.15). However, even among these populations, unique alleles were rarely found and with few exceptions, the same alleles tended to reach the highest frequencies in all samples regardless of geographic origin. Moreover, when these strains were put into laboratory culture, even more homogenization was shown to occur. We show here that using DNA fingerprinting techniques, even very closely related geographic isolates and strains of the medfly that have been in laboratory culture can be reliably differentiated. In addition to the fact that these molecular markers are more reliable for discriminating between strains, these same markers can be used to track and monitor the genetic structure, populations dynamics, and movements of such pest species. Finally, these same RFLP markers may be useful in developing a linkage map of a genome such as that of the medfly where only limited amounts of classical genetic analyses have been done.
Materials and methods Strains The Kula and Mauna Loa strains were originally collected from Kula, Maui, and Mauna Loa, Hawaii, respectively, in 1989 and have been maintained in the laboratory since then. The Robinson strains have been in laboratory culture since 1983. The Israel strain is maintained in a USDA certified quarantine facility at the University of Hawaii, and material from this strain was generously provided by Dr. S. Saul. The A5A3 strain used to construct the genomic library is a subline derived from the USDA "Hilab" strain
529
maintained at the USDA Tropical Fruit and Vegetable Laboratory, Manoa, Hawaii. Genomic DNA isolation and library construction Genomic DNA was isolated from the various strains of C. capitata by one of two different methods. Method I . "Lifton" method of DNA extraction This method was used to isolate genomic DNA for library construction and for Southern blots. In this method, DNA was isolated from adult flies by homogenizing approximately 350 mg of freshly etherized flies in 5 mL 0.2 M sucrose, 50 mM EDTA, 100 mM tris pH 9.0, and 0.5% SDS in a Pyrex No. 7725 homogenizer. The homogenate was filtered through autoclaved polyester fiberfil. The filtered homogenate was treated with Proteinase K at a final concentration of 200 pg/mL for 1 h at 65"C, after which potassium acetate was added to a final concentration of 1.2 M. Debris was pelleted out by centrifugation for 15 min at 10 000 rpm, 4°C. DNA was ethanol precipitated from supernatant, resuspended in TE, and extracted with phenol and chloroform - isoamyl alcohol (24:l). DNA was again precipitated with ethanol and resuspended in TE for use. Method 2. "Salt" method of DNA extraction This method was used to isolate genomic DNA for Southern blot analysis. DNA was extracted in this method by homogenizing approximately 350 mg of freshly etherized adult flies in 5 mL 10 mM tris pH 7.4,60 mM NaC1, 10 mM EDTA, 0.15 mM spermine, 0.15 mM spermidine, and 0.5% (v/v) Triton X-100. The homogenate was strained through gauze and centrifuged for 10 min at 7000 rpm at 4°C to pellet nuclei in an SS34 rotor. Nuclei were then resuspended in 5 mL 75 mM NaCl and 25 mM EDTA, and SDS was added to 1%. The sample was treated with Proteinase K at 200 pg/mL for 1 h at 55"C, followed by two chloroform extractions of 30 min each in high salt (1.6 M NaCl). DNA was then precipitated with isopropanol. Each of these procedures were also modified for single fly DNA extractions. This was accomplished by scaling down the procedures approximately 30-fold. Single flies were homogenized in Kontes microhomogenizers. Restriction fragment library This genomic library was constructed by ligating DNA from the A5A3 subline into arms of XgtlO cut with EcoR1. The A5A3 genomic DNA had been completely digested also with EcoR1. This type of construction produces a library of mostly single EcoRl fragments ranging in size from less than 1 kb up to about 7 kb. Probes and hybridizations All probe DNAs were labelled by random priming with the nonradioactive "Genius" labelling and detection kit from Boehringer Mannheim Biochemicals. The plaque hybridizations of the genomic library were conducted as described by Benton and Davis (1977). Probes and hybridizations for plaque lifts and Southern blots using M13 as a probe were conducted under reduced stringency conditions using the Boehringer Mannheim Biochemicals nonradioactive labelling and detection kit as described by the manufacturer. In these cases the entire M13mp18 RF was used to make the probe. Reduced stringency conditions were as follows. Hybridizations were conducted using 35% formamide at 42°C. Reduced stringency washes were conducted at 35 "C in 0.1 % SDS with decreasing concentrations of SSC down to 0.5 x . For Southern blots, genomic DNA was digested at 37°C for 2-3 h using enzymes and buffers from Boehringer Mannheim Biochemicals. DNA was electrophoresed on 0.7% agarose gels with TBE and Southern blotted to Micron Separations Magna Graph nylon membrane as described by Southern (1975). Hybridization, washing, and immunological detection of the Southern blots was conducted as described in the Boehringer Mannheim Biochemicals "Genius" nonradioactive labelling and
GENOME,
-
VOL. 35,
1992
-
0 C C L I
- 2 2 6
2 .O .O
L s s :0z3 m o o
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
uggCb4
FIG. 1. Southern blot of genomic DNA from five different medfly strains digested with EcoRl and probed with M13. Size markers are from a Hind111 digest of lambda DNA. Arrows indicate examples of major RFLP differences consistently observed between strains. detection kit. When either pmedMl3e or pmedMl3f was used as a probe, more stringent conditions were used. Hybridizations were conducted in 43% formamide buffer at 42OC, with washes at 55OC in 0.1 % SDS and decreasing concentrations of SSC down to 0.5 x .
Results Using the bacteriophage vector M13 as a probe, RFLPs can be observed between each of the strains tested here (Fig. 1). Shown here is a Southern blot of EcoRl digests of genomic DNAs. Examples of some of the prominent RFLP differences consistently observed between strains are indicated with arrows. The RFLP differences seen here represent variation in repetitive sequences present in these
FIG. 2. Southern blot of genomic DNAs from the Kula strain digested with EcoRl and probed with pmedM 13f. Lane A contains DNA extracted by the "salt" method, while lanes B and C contain DNA extracted by the "Lifton" method on two different days. genomes, in a manner analogous to hypervariable minisatellite sequences described in the human genome (Jeffreys et al. 1990). Of the strains shown here, the Kula and Mauna Loa strains originated from different geographic localities within the Hawaiian islands. As expected, these strains have many DNA fragments (bands of hybridization) in common but, in addition, there are some notable RFLPs observable between the two strains. The two "Robinson" strains are both derived from a line bearing a translocation of a wildtype gene for pupal color to the Y chromosome (Robinson 1989). They differ only in that the Robinson (Hawaiian) strain has been outcrossed to a strain from the Hawaiian islands (also originally from Kula), while the Robinson (pure) has not been outcrossed. In the same manner discussed above, these two strains are expected to have a great
HAYMER ET AL.
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
Kula
FIG. 3. Southern blot of genomic DNA from individual females of the Kula strain digested with EcoRl and probed with pmedM 13f.
deal of their genetic background in common, although here again some notable polyrnorphisms are consistently detected. Although the M13 probe can be used to elucidate variation between these strains in this manner, these blots are of necessity done under very low stringency conditions (see Materials and methods). To increase the resolving power of the technique and to allow the blots to be done under more stringent conditions, we next isolated some of the M13 homologous sequences directly out of the medfly genome. To accomplish this we screened our Xgt 10 genomic library using M13 as a probe under conditions identical to those described for the Southern blots. We initially isolated five different medfly sequences that, based on hybridization, had some homology to the vector M13. These five sequences were cloned and used individually to probe genomic DNA
FIG. 4. Southern blot of genomic DNA from the five different medfly strains digested with EcoRl and probed with pmedM13f.
from different medfly strains. Two of the five produced informative patterns of RFLP differences between strains and were kept for further analysis. These were designated pmedM 13e and pmedM 13f. These clones contain EcoR 1 inserts 0.9 and 2.0 kb in size, respectively. In addition to examining differences between strains using the DNA fingerprinting methodology, we also wanted to show that we could obtain consistency within a strain or between individuals of a strain. To this end we first compared DNA from the same strains extracted by different methods and on different days. One such result is shown in Fig. 2. Here, the Kula A lane contains genomic DNA extracted from a pool of adult flies by the "salt" method described in the Materials and methods section. Lanes labelled Kula B and C contain genomic DNA extracted from adults by the "Lifton" method (see Materials and methods),
GENOME, VOL. 35, 1992
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
number of DNA polymorphisms can be seen. For the pmedM13f probing (Fig. 4), the majority of the polymorphism~seen are in the range of 3-7 kb in size, although there are some very high molecular weight differences. In the case of pmedM13e (Fig. 5), a large number of polymorphisms can be seen ranging from under 2 kb in size up into a very high molecular weight range.
FIG. 5. Southern blot of genomic DNA from the five different medfly strains digested with EcoRl and probed with pmedMl3e.
each on different days. The probe here is pmedM13f. As can be seen, the restriction fragment profile elucidated here is consistent within a strain, with no detectable differences observable owing to the method or timing of .the DNA extraction. We have obtained similar consistent results for each of the probes discussed above within both the Kula and Mauna Loa strains. Figure 3 shows the results obtained when five individual females of the Kula strain were probed with the same probe, pmedMl3f. Here again, there is consistency within a strain with essentially no qualitative differences in the banding pattern observable. Along this line we tested each of our probes on DNA extracted from individual males and females from both Hawaiian island strains and the results were consistent with those shown here. Each of these probes was also used to probe genomic DNA from each of the different strains tested earlier with M13. These blots were done under much more stringent hybridization conditions. In both of these cases, a large
Discussion We have shown here that DNA fingerprinting can be used to,detect genetic variation between strains of the medfly. Beginning with a fortuitous homology between the bacteriophage vector M13 and repetitive sequences present in the medfly genome, we have shown that DNA RFLPs can be identified that are diagnostic of the genetic makeup of different strains. The strains used here differ only in terms of their geographical origin or their genetic background. Similar approaches have been used to study population differentiation in wildlife conservation studies (Brock and White 1991; Gilbert et al. 1990; Kirby 1990) and to identify cultivar varieties of agriculturally important commodities (Nybom 1990). In our work on the medfly, using M13 as a probe to produce the DNA fingerprint pattern, differences are detected between all of the strains tested here. However, with M13 only a small proportion of the bands are polymorphic. Not counting the Israel strain (which has been included as an out group), only an average of 10% (i.e., 2 of 21) of the observable bands differ between the strains. In addition, the hybridization conditions used with the M13 probe are of necessity very "relaxed," meaning that a fair amount of mismatch is permissible between the probe and the target genomic sequences (Maniatis et al. 1982). On the other hand when one of the medily genomic sequences (pmedM13e or pmedM13f) is used as a probe, much more variation between all of the strains can be detected. In the case of pmedM13f, a smaller number of total bands can be seen (average of 1 9 , and on average 4-5 of these bands (25-30%) are polymorphic between the strains. For pmedMl3e, an average of 6 of the 19 visible bands differ in each case. This means a substantial increase in the information and diagnostic value of this approach when a "native" genomic sequence is used as a probe. In addition, these hybridizations are done under more "stringent" conditions, meaning that a more precise match between the probe and genomic target sequence is required. This can be important as far as the reproducibility of the technique. ' As far as reproducibility is concerned, we have shown that while differences between strains can be reliably detected with all three of the probes described here, there is also consistency within a strain. For example, essentially the same pattern is seen for the Kula strain in different probings of this DNA with pmedl3f in Figs. 2, 3, and 4. Similar consistent results have been obtained using M13 itself and pmedM13e (not shown). We have also shown that the DNA fingerprint pattern obtained within a strain is not affected by the method or timing of the DNA extraction (done on random samples of 20-30 adult flies), nor is the pattern different between single individual males or females from the same strain. One of the great advantages to this approach is that it can be done in a way that requires very little information
Genome Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 12/14/14 For personal use only.
HAYMER E T AL.
a priori about specific genes and (or) sequences present in a genome. We are now determining the nucleotide sequence of the two clones pmedMl3e and pmedMl3f. Once this information is available, it should be possible to design oligonucleotide primers that can be used in amplification of one or more of these polymorphic sequences by the polymerase chain reaction or PCR method (Kirby 1990). This would allow the identification or assessment of genetic variation to be done on even limited amounts of material such as dessicated or preserved body parts, such as might be recovered from traps in a field collection. The availability of these techniques can be extremely valuable in terms of determining the genetic origin of flies moving into or establishing new habitats and for monitoring genetic changes that might occur in populations over space or time. In conclusion, we have detected genetic variation between strains of the Mediterranean fruit fly using DNA fingerprinting. We have identified probes and described conditions under which considerably more genetic variation can be detected compared to previous methods of population assessment. The variation detected between strains is consistent within individuals or within populations of flies of the same strain, and it is not affected by different methods used to extract DNA. The strains used here differ only in terms of geographic origin (including different islands within Hawaii) or in terms of genetic background.
Acknowledgements This work was supported by California Department of Food and Agriculture grant 90-0486 and USDA grant HAW 00763-G.
533
Arnheim, N. 1989. New approaches to constructing genetic maps: PCR analysis of DNA sequences in individual gametes. In PCR technology. Edited b y H. Erlich. Stockton Press, New York. pp. 119-136. Benton, W., and Davis, R. 1977. Screening XgtlO recombinant clones by hybridization to single plaques-in situ. Science (Washington, D.C.), 196: 180. Brilliant, M.H., Gondo, Y., and Eicher, E. 1991. Direct molecular identification of the mouse pink-eyed unstable mutation by genome scanning. Science (Washington, D.C.), 252: 566-569. Brock, M.K., and White, B.N. 1991. Multifragment alleles in DNA fingerprints of the parrot, Amazona ventralis. J. Hered. 82: 209.-2 12. Gilbert, D.A., Niles, L., O'Brien, S.J., and Wayne, R.K. 1990. Genetic fingerprinting reflects population differentiation in the California channel island fox. Nature (London), 344: 764-766. Jeffreys, A. J., Neumann, R., and Wilson, V. 1990. Repeat unit sequence variation in minisatellites: a novel source of DNA polymorphism for studying variation and mutation by single molecule analysis. Cell, 60: 473-485. Kirby, L.T. 1990. DNA fingerprinting: an introduction. Stockton Press, New York. Maniatis, T., Fritsch, E., and Sambrook, J. 1982. Molecular cloning. Cold Spring Harbor Press, Cold Spring Harbor, New York. Milani, R., Gasperi, G., and Malacrida, A. 1989. Biochemical genetics. In Fruit flies. Vol. 3b . Edited b y A.S. Robinson and G. Hooper. Elsevier, New York. pp. 33-56. Nybom, H. 1990. DNA fingerprints in sports of 'Red Delicious' apples. Hortscience, 25: 1641 - 1642. Robinson, A. 1989. Genetic sexing methods in the Mediterranean fruit fly, Ceratitis capitata. In Fruit flies. Vol. 3b. Edited by A. S. Robinson and G. Hooper. Elsevier, New York. pp. 57-68. Saul, S. 1986. Genetics of the Mediterranean fruit fly, Ceratitis capitata. Agric. Zool. Rev. 1: 73-108. Southern, E. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503.
