Proc. Nati. Acad. Sci. USA Vol. 89, pp. 7732-7736, August 1992 Agricultural Sciences Stable genetic transformation of a beneficial , occidentalis (: ), by a microinjection technique (maternal microiWection/genetic improvement/biological control) JAMES K. PRESNAIL* AND MARJORIE A. HOY*t Department of Entomology, University of California, Berkeley, CA 94720 Communicated by John E. Casida, May 26, 1992 (receivedfor review April 2, 1992)

ABSTRACT A microinjection technique has resulted in future implementation of this vector system in other arthro- stable transformation of the western predatory Metaseiu- pods. Another potential vector system for is under lus occidentalis. Early preblastoderm eggs within gravid fe- development and employs the yeast recombinase FLP and males were microinjected. The needle was inserted through the introduced FRT target sequences (11). cuticle ofgravid females into the egg, or the tissue Immediatel Given the current absence ofefficient vectors forarthropods surrounding the egg. This maternal injection method resulted other than Drosophila, substantial research has been devoted in relatively high levels of survival and transformation. Trans- to developing alternative DNA delivery systems. Investiga- formation was achieved without the aid of any transposase- tors have explored ballistic introduction of DNA-coated mi- producing helper plasmid. The predatory mite was trans- croprojectiles (12, 13), electroporation (unpublished data), and formed with a plasmid containing the Escherwchi cofl -ga- puncturing eggs with DNA-coated silicon carbide filaments lactosidase gene (lacZ) regulated by the DrosophUa hsp7O (31). Another potential DNA delivery system involves binding heat-shock promoter. Putatively transformed lines were iso- heterologous DNA to sperm and introducing it through arti- lated based on P-galactosidase activity in first-generation lar- ficial insemination. This approach may be useful in honeybees vae. Transformation was confirmed in the sixth generation by (14), but it remains to be seen whether the bound DNA can polymerase chain reaction amplification of a region spanning actually be introduced into the embryo by this method. the DrosophikalE. coli sequences. Amplification of a nested Efforts in our laboratory have focused on developing an region, also sp ing the interspecific boundary, provided efficient DNA delivery system to transform the western further evidence for stable transformation. Maternal microin- predatory mite Metaseiulus occidentalis. M. occidentalis is a jection may be adaptable to other beneficial arthropods, par- member of the cosmopolitan family Phytosenidae, a group of ticularly other phytoselid . Genetic transformation of M. mostly predatory mites responsible for controlling spider occidentais may improve its efficiency as a biological control mites and other arthropods in many crops (15). M. occiden- agent as well as provide a method for investigating details of its talis is less than 500 gm long (16) and has five developmental physiology and ecology. stages: egg, larva, protonymph, deutonymph, and adult. At 270C the life cycle from egg to gravid female is completed in The development of germ-line transformation in approximately 6 days. Females lay approximately two eggs a Drosophila day. Only one egg is matured at a time, with the fully mature melanogaster (1, 2) has provided a valuable tool for genetic egg approximately 150 ,um long and occupying approximately analysis of this arthropod. The ability to manipulate other one-third ofthe body cavity. Phytoseiids are thought to have arthropods through efficient transformation would provide a single unpaired ovary. Although M. occidentalis is bisexual researchers with a valuable method to investigate this diverse and eggs must be fertilized, adult males are haploid because phylum. Many arthropod species are disease vectors and a one set of chromosomes is eliminated during embryogenesis far greater number are agriculturally important, including in males (17, 18). both pest and beneficial organisms. M. occidentalis has been the subject of artificial selection Gene transfer in beneficial arthropods has many potential programs aimed at improving its insecticide resistance and its rewards (3). It may be possible to modify a natural enemy's control of spider mites in almonds as well as other crops temperature tolerance, host range, resistance to insecticides, (19-21). Improvement of this species has helped to decrease or sex ratio. Altering these traits could improve the natural the need for chemical control of spider mites, thereby reduc- enemy's performance and result in an overall reduction in the ing production costs (22). Developing an efficient gene trans- use ofinsecticides. Gene transfer may also allow researchers fer method for M. occidentalis, although initially time con- to "tag" a population and, based on this heritable tag, suming, might eventually save time and bypass some of the distinguish them and their progeny from native populations of disadvantages of artificial selection. the same species. In this paper, we report the development of a microinjec- Initial efforts to transform arthropods otherthanDrosophila tion method, expression of the prokaryotic reporter gene focused on employing existingP-element vectors, in particular lacZ under the regulation ofthe Drosophila hsp7O heat-shock pUChsneo (4), and microinjecting these vectors into preblas- promoter, and evidence demonstrating stable transformation toderm eggs (5-7). Transformation was achieved occasionally, of M. occidentalis. but at very low frequencies, and was apparently independent of P-element mediation. The rare transformation events re- ported appear to be the result of random integrations. Present MATERIALS AND METHODS evidence indicates that the utility of P-element vectors is Colony Source and Culture Methods. The strain of M. limited in organisms outside the family Drosophilidae (8), but occidentalis used in these studies (known as Pullman black- characterization of P-element transposition (9, 10) may allow *Present address: Department of Entomology and Nematology, The publication costs of this article were defrayed in part by page charge Building 970, Hull Road, University of Florida, Gainesville, FL payment. This article must therefore be hereby marked "advertisement" 32611. in accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be sent at the present address. 7732 Downloaded by guest on September 28, 2021 Agricultural Sciences: Presnail and Hoy Proc. Natl. Acad. Sci. USA 89 (1992) 7733 berries) was originally collected in 1987 on blackberry bushes produced needles with a fine (1 pkm) short tip. Prior to near Pullman, WA. It is highly susceptible to all pesticides microinjection, the tip of the needle was broken off to tested. The colony is maintained in trays (10 x 10 cm) on produce a jagged tip approximately 3 Am in diameter. Mi- wax-coated paper resting on top of water-saturated cotton. croinjection was performed with the aid of a Zeiss inverted An open water barrier, approximately 1 cm wide, surrounds compound microscope. the cotton. The trays are placed in a plastic box with a 1-cm Mites were prepared for injection in the following manner. petroleum jelly barrier at the top of the box. Predator mites Two layers of Scotch brand double-stick tape were placed were fed a mixture of two-spotted , Tetranychus lengthwise on a microscope slide. The tape was then cut at an urticae, and Pacific spider mite, Tetranychus pacificus. approximately 450 angle to produce a ledge under which the Construction of pJKP2. A 500-base-pair (bp) BamHI- mites could be secured. Adult gravid females were secured EcoRI fragment containing the D. melanogaster hsp7O heat- under the overhanging ledge of the tape in a lengthwise shock promoter (23) was isolated from the plasmid pHSS7/ orientation (Fig. 2). Care must be taken to select females with hsp (24) by restriction endonuclease digestion and gel puri- eggs that are nearly, but not fully, mature. Fully mature eggs fication. This fragment was ligated into the BamHI/EcoRI will burst when probed with the injection needle, leading to polylinker sites of the P-element vector pDM79 (25). The death of the egg and female. Ten females were aligned and resulting plasmid (pJKP2) carries the Drosophila hsp7O pro- covered with distilled water. Females were injected between moter 5' to the adh/lacZfusion gene ofpDM79 (Fig. 1A). The the third and fourth coxae and efforts were made to penetrate plasmid was transformed into Escherichia coli strain DH5a. the egg, but penetration was not always assured and DNA Approximately 100 pg of plasmid was isolated from 50 ml of could have been deposited in close proximity to the egg. After bacterial culture by using a Qiagen-tip 100 column. The injection, the females were carefully lifted from the tape to a plasmid preparation was resuspended in injection buffer (5 clean region of the slide. When fully dry, the females were mM KCl/O.1 mM phosphate, pH 7.8) (26) at a concentration removed from the slide and placed individually on pinto bean of 300 pug/ml. leaf discs (isolated on water-saturated cotton) with spider Mlcroinjection of Gravid M. occidentalis females. Prelimi- mite prey. Females were observed daily, dead females were nary studies indicated that needles used to inject D. mela- removed, and those depositing eggs were maintained on fresh nogaster eggs were too flexible to penetrate M. occidentalis leaf discs. The eggs produced by the injected females are considered the Go generation. G1 progeny were assayed for females. We therefore pulled needles from microcapillary transformation by using a P-galactosidase assay. tubes (1.0-mm external diameter; 0.5-mm internal diameter) Heat Shock and f3-Galactosidase Assay. G1 larvae were with an internal glass filament. An automatic needle puller covered with water and exposed to a standard heat shock of 370C for 30 min. After heat shock, the water was removed and the mites were submersed in 50 mM sodium phosphate (pH 8.6). A solution of 5-bromo-4-chloro-3-indolyl P-D- galactopyranoside (X-Gal) [0.3% X-Gal (in dimethylforma- A mide)/30o Ficoll/50 mM sodium phosphate, pH 8.6/2 mM K4Fe(CN)6] was added and the mites' abdomens were punc- tured. The larvae were kept at 100lo relative humidity and

B JKP2-1 IHsp7O NI Adh/B-gal I ]KP-2i jKP2-2 C JKP2-1. GCTCTTCTCCACTCTGTCACACAGTAAACGGC

JKP2-2. CGCAAAGTAGACACCACGTTGCCCGCGACC

JKP2-21. GGGCGCAATCGTAACCGTGCAATCTGCCAG FIG. 1. Map of the plasmid pJKP2 and associated PCR primers. (A) pJKP2 was produced by introducing a 500-bp BamHI-EcoRI FIG. 2. Gravid females ofM. occidentalis to be microinjected are fragment containing the hsp7O promoter into the polylinker of positioned under two layers ofdouble-stick tape that has been cut to pDM79. (B) Location of the PCR primers in relation to the hsp7O/ produce an overhanging ledge. Microinjection is in the posterior lacZ region of plasmid pJKP2. (C) Sequences of the PCR primers. region and may be directly into the egg. Downloaded by guest on September 28, 2021 7734 Agricultural Sciences: Presnail and Hoy Proc. Nad. Acad Sci. USA 89 (1992) Table 1. Microinjection of adult female M. occidentalis No. No. of No. with No. of No. % laying established G1 expression G6 tested Date injected survival eggs lines of lacZ with PCR 10/18 50 50 12 4 0 10/22 50 24 5 3 1 * 10/23 50 36 12 9 0 10/24 50 66 11 10 1 * 10/26 20 70 6 2 0 10/29 60 48 13 13 4 2 10/30 75 57 16 7 1 0 11/7 50 42 3 0 0 - Total 405 49 78 48 7 2 *Colony was lost before the PCR assay. after 3 hr were scored for the presence or absence of blue acZ Bioassay. Ofthe 48 established lines, 7 displayed lacZ color. expression in the G1 larvae. Two of these putatively trans- The construct pJKP2 was first tested for lacZ expression formed lines were lost, perhaps due to poor rearing condi- in transiently transformed D. melanogaster larvae and ex- tions or decreased fitness. The 5 remaining lines were main- pression was determined to be regulated in a heat-shock- tained and analyzed by PCR. dependent manner. When M. occidentalis larvae were tran- PCR Analysis. In control PCRs, at annealing temperatures siently transformed with pJKP2, the same pattern of heat- below 63°C, amplification of pJKP2 resulted in nonspecific shock-dependent expression was observed (unpublished products (data not shown). At an annealing temperature of data). No activity was detected up to 4 hr after immersion, 66.5°C, only the two predicted bands were produced in but endogenous (3-galactosidase activity could be detected in sufficient quantity to be visible by ethidium bromide/UV uninjected M. occidentalis larvae when larvae were incu- analysis. At temperatures above 53°C, no amplification prod- bated in the X-Gal solution for over 5 hr. ucts were observed from DNA extracted from uninjected M. DNA Extraction. DNA was extracted from individual adult mites by a Chelex method (0. R. Edwards and M.A.H., occidentalis. unpublished data). In brief, specimens were frozen in liquid PCR was first performed in the presence of the hsp7O N2 and ground to powder, 100 ,ud ofa 5% Chelex solution was primer (JKP2-1) and the external lacZ primer (JKP2-2) (Fig. added, and the samples were mixed thoroughly with a vortex 3). The reaction produced the predicted 930-bp product based

mixer. The samples were then heated to 560C for 15 mm r . followed by 4 min at 100TC. The samples were then centri- fuged in a microcentrifuge for 2 min, and the DNA superna- G tant was removed for PCR analysis. Oligonucleotide Primers for PCR. The oligonucleotide primers were produced at the DNA Synthesis Facility, Uni- versity of California, Berkeley. A 30-bp oligomer (JKP2-1) /, . identical to bp 126-155 of the Drosophila hsp7O heat-shock _E ; sr > promoter was used in conjunction with two primers from the .J J o_ E. coli lacZ region of the construct. JKP2-2 (complementary _S to bp 426-475 of lacZ) in conjunction with JKP2-1 amplifies __ a 930-bp fragment. JKP2-21 (complementary to bp 267-294 of -4 > J U lacZ) in conjunction with JKP2-1 amplifies a 775-bp fragment ' > O :iC 1 J internal to the above-mentioned product. The plasmid pJKP2 4 rJ r ;> are illustrated in and associated oligonucleotide primers Fig. _

1 B and C. fF _ D A 10-dul aliquot of DNA extracted from 10 individual G6 _ adult females isolated from lines displaying lacZ expression in the G1 larvae was amplified for 32 cycles under standard reaction conditions (27). The first cycle consisted of 4 min at 950C, 30 sec at 630C, and 90 sec at 750C. The remaining 31 cycles consisted of 30 sec at 950C, 2 min at 630C, and 90 sec . ^. at 750C. PCR products were electrophoresed at 30 mA for 3 hr in a 1% agarose gel (Tris/borate/EDTA buffer), stained with ethidium bromide, and photographed. -,_ RESULTS _ Microinection of Gravid Females. Table 1 provides the results of microinjecting gravid M. occidentalis females. The average survival offemales after 24 hr was 49o, although the FIG. 3. PCR analysis of putatively transformed G6 lines in the number of survivors continued to decline over the next 24 hr presence of primers JKP2-1 and JKP2-2 provides evidence of stable to approximately 30%o. Despite these losses, 78 (191%) of the transformation. DNA samplesfrom uninjected mites (without or with treated females produced between one and eight eggs. From addition of pJKP2), pJKP2, and putative transformed lines were these 78 females, 48 lines were established (established lines amplified. Bgl I/Hinfl-digested pBR328 provided size markers (sizes are defined as those producing G1 larvae). in base pairs at left). Downloaded by guest on September 28, 2021 Agricultural Sciences: Presnail and Hoy Proc. Nadl. Acad. Sci. USA 89 (1992) 7735 gravid female mites is an order of magnitude better than the results of mosquito egg microinjection (5-7). Preliminary efforts to microinject eggs of M. occidentalis Co after they were laid met with marginal success and produced substantial Adult females are easier to 0- mortality. manipulate 0O and line up on the double-stick tape than phytoseiid eggs. Phytoseiids mature a single large egg at a time and the egg (occupying approximately one-third of the mite body) is visible through the translucent cuticle. Thus, one can insert 04 ~ -0 the needle and easily probe the egg; however, it is not always clear whether the needle penetrates the egg. Another advantage to maternal microinjection is that the U C 0. YU' eggs are not subjected to dechorionation treatments. De- chorionation of Drosophila eggs is considered a critical step responsible for much of the mortality incurred during micro- injection. Previous efforts to dechorionate M. occidentalis eggs met with limited success. Response to dechorionation was variable and might depend on the age of the egg (un- published work). A final advantage of maternal microinjec-

21 75 tion is that the eggs are incubated inside the female until 756 oviposition, which may increase the survival rate. Because the efficiency of maternal microinjection is rea- '230 sonably high, this procedure may provide an improved DNA 033 delivery system for other species for which no vector system is available. Furthermore, with improved handling and iso- lation of putatively transformed lines the efficiency of ma- 553 ternal microinjection is likely to improve. In our study, the 454 method was employed without the aid ofany transposase and 394 resulted in a 0.5% rate of transformation. The frequency of transformation may be increased in species in which a vector, FIG. 4. PCR analysis ofputatively transformed G6 lines. Aliquots such as the P element, is available. Further characterization of the PCR mixtures using the external lacZ primer (Fig. 3) were of the transformed lines through Southern blot analysis is reamplified in the presence of primer JKP2-1 and the internal lacZ necessary to characterize the number and location of inser- primer JKP2-21 and confirmed the identity of the sequence. tions, but requires a large number of mites to yield sufficient DNA for analysis. The transformed colonies are not yet on plasmid sequences in the positive control as well as in two sufficiently large to permit this analysis. of the G6 putatively transformed lines. The identity of this These experiments demonstrate the utility of the lacZ amplification product was confirmed by removing an aliquot reporter gene for future genetic transformation of phytosei- of the reaction mixture and conducting a second PCR using ids. The lacZ gene may prove useful in analyzing promoter the hsp7O primer and the internal lacZ primer (JKP2-21) (Fig. sequences in mites and , as it has in Drosophila (25). 4). Again, the product was of the predicted size (775 bp) Finally, another significant result is the demonstration that consistent with amplification of sequences derived from the the Drosophila hsp7O heat-shock promoter functions in this introduced plasmid. Both products span the border between mite. The lacZ gene in this construct is regulated in a the Drosophila hsp7O and the E. coli lacZ genes. It is unlikely heat-shock-dependent manner and was expressed only mar- that products of the specific sizes predicted could be pro- ginally in the absence of heat shock (unpublished data). This duced from endogenous DNA sequences. Further, no am- promoter has been shown to be useful in regulating the in other insects plification was observed from DNA samples ofuninjected M. neomycin phosphotransferase gene (5-7). occidentalis (Figs. 3 and 4). Few native regulatory sequences are known for arthropods and considerable effort may be saved if Drosophila promot- The inheritance of plasmid sequences in the G6 generation ers can be used to regulate genes introduced into diverse argues persuasively for stable transformation in at least two arthropod species. lines. Factors determining the general utility of maternal micro- injection for other arthropod species will relate to their DISCUSSION maternal morphology and egg maturation. Within the family Phytoseiidae, all of which share the basic physical charac- We have developed an approach to transformation of a teristics and behavior of M. occidentalis, the method should beneficial arthropod by microinjecting eggs. This approach, be generally applicable. Because this family contains approx- which we term maternal microinjection, has the following imately 1000 species, a reliable transformation method will be advantages. The method is less laborious than microinjection valuable, even iflimited to this group alone. The potential use of preblastoderm eggs. We can microinject female mites at a of maternal microinjection in transforming other arthropods rate comparable to the rate reported for microinjection of D. will offer opportunities in both basic and applied research on melanogaster eggs (27). Survival of injected M. occidentalis pest and beneficial arthropod species. females is comparable to survival of microinjected Drosoph- ila eggs (28). The efficiency of maternal transformation in our We thank Drs. Caroline Kane and Steven Beckendorf and his tests is approximately 1/10th the efficiency of P-element- laboratory group for their unfailing support and assistance. M.A.H. thanks the Cold Spring Harbor Laboratory and Dr. Gerald Rubin for mediated transformation of Drosophila, but it is comparable assistance in initiating this line of research. This research was to earlier techniques employed for Drosophila (29). Micro- supported by Western Regional Research Project W-84, the Univer- injection of mosquito eggs has typically produced substantial sity of California, and the California Agricultural Experimental mortality and limited transformation (30). Survival of injected Station. Downloaded by guest on September 28, 2021 7736 Agricultural Sciences: Presnail and Hoy Proc. Natl. Acad. Sci. USA 89 (1992) 1. Spradling, A. & Rubin, G. (1982) Science 218, 341-347. 16. Chant, D. A. (1985) in Spider Mites: Their Biology, Natural 2. Rubin, G. & Spradling, A. (1982) Science 218, 348-353. Enemies and Control, eds. Helle, W. & Sabelis, M. W. (Else- 3. Beckendorf, S. K. & Hoy, M. A. (1985) in Biological Control vier, New York), Vol. 1B, pp. 3-32. in AgriculturalIPM Systems, eds. Hoy, M. A. & Herzog, D. C. 17. Hoy, M. A. (1979) Entomol. Exp. Appl. 26, 97-104. (Academic, Orlando, FL), pp. 166-187. 18. Nelson-Rees, W. A., Hoy, M. A. & Roush, R. T. (1980) Chro- 4. Stellar, H. & Pirrotta, V. (1985) EMBO J. 4, 167-171. mosoma 77, 263-276. 5. Miller, L., Sakai, R., Romans, P., Gwadz, R., Kantkoff, P. & 19. Hoy, M. A. (1985) Annu. Rev. Entomol. 30, 345-370. Con, H. (1987) Science 237, 779-781. 20. Hoy, M. A. & Knop, N. F. (1981) Entomol. Exp. Appl. 30, 6. McGrane, V., Carlson, J., Miller, B. & Beaty, B. (1988) Am. J. 10-18. Trop. Med. Hyg. 39, 502-510. 21. Hoy, M. A., Westigard, P. H. & Hoyt, S. C. (1983) J. Econ. 7. Morris, A. C., Eggleston, P. & Crampton, J. M. (1989) Med. Entomol. 76, 383-388. Vet. Entomol. 3, 1-7. 22. Headley, J. C. & Hoy, M. A. (1987) J. Econ. Entomol. 80, 8. O'Brochta, D. A. & Handler, A. M. (1988) Proc. Natd. Acad. 555-559. Sci. USA 85, 6052-6056. 23. Lis, J. T., Simon, J. A. & Sutton, C. A. (1983) Cell 35, 403-410. 9. Rio, D. C., Laski, F. A. & Rubin, G. M. (1986) Cell 44, 21-32. 24. Siefert, H. S., Chen, E. Y., So, M. & Heifron, F. (1986) Proc. 10. Rio, D. C. & Rubin, G. (1988) Proc. NatI. Acad. Sci. USA 85, Natl. Acad. Sci. USA 83, 735-739. 8929-8933. 25. Mismer, D. & Rubin, G. M. (1987) Genetics 116, 565-578. 11. Morris, A. C., Schaub, T. L. & James, A. A. (1991) Nucleic 26. Spradling, A. C. (1986) in Drosophila: A Practical Approach, Acids Res. 19, 5895-5900. ed. Roberts, D. B. (IRL, Oxford), pp. 175-1%. 12. Baldarelli, R. M. & Lengyel, J. A. (1990) Nucleic Acids Res. 27. Saiki, R. K. (1990) in PCR Protocols: A Guide to Methods and 18, 5903-5904. Applications, eds. Innis, M. A., Gelfand, D. H., Sninsky, J. J. 13. Carlson, D. A., Cockburn, A. F. & Tarrent, C. A. (1989) in & White, T. J. (Academic, San Diego), pp. 13-21. Host Regulated Developmental Mechanisms in Vector Arthro- 28. Rubin, G. & Spradling, A. (1983) Nucleic Acids Res. 11, pods, eds. Borovsky, D. & Spielman, A. (Univ. Florida Press, 6341-6351. Vero Beach, FL), pp. 248-252. 29. Germeraad, S. (1976) Nature (London) 262, 229-231. 14. Milne, C. P. (1991) Am. Bee J. 131, 188-189. 30. Walker, V. K. (1990) in Advances in Cell Culture, ed. Maram- 15. Sabelis, M. W. (1985) in Spider Mites: Their Biology, Natural orosch, K. (Academic, New York), Vol. 7, pp. 87-124. Enemies and Control, eds. Helle, W. & Sabelis, M. W. (Else- 31. Kaeppler, H. F., Gu, W., Somers, D. A., Rines, H. W. & vier, New York), Vol. 1B, pp. 35-40. Cockburn, A. F. (1990) Plant Cell Rep. 9, 415-418. Downloaded by guest on September 28, 2021