Biolostic transformation of a procaryote, bacillus megaterium

Vol. 57, No. 2APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 1991, p. 480-4850099-2240/91/020480-06$02.00/0Copyright ©) 1991, American Society for Microbiology

Biolistic Transformation of a Procaryote, Bacillus megateriumKATHERINE B. SHARK,' FRANZINE D. SMITH,2* PETER R. HARPENDING,' JEANETTE L. RASMUSSEN,3

AND JOHN C. SANFORD2Department of Horticultural Sciences, Cornell University, Geneva, New York 144562; School of Biological Sciences,

University of Nebraska, Lincoln, Nebraska 68588-0118'; and Department of Biological Sciences,Plattsburgh State University College, Plattsburgh, New York 129013

Received 16 July 1990/Accepted 25 November 1990

We present a simple and rapid method for introducing exogenous DNA into a bacterium, Bacillusmegaterium, utilizing the recently developed biolistic process. A suspension of B. megaterium was spread ontothe surface of nonselective medium. Plasmid pUB110 DNA, which contains a gene that confers kanamycinresistance, was precipitated onto tungsten particles. Using a biolistic propulsion system, the coated particleswere accelerated at high velocities into the B. megaterium recipient cells. Selection was done by use of an agaroverlay containing 50 ,Ig of kanamycin per ml. Antibiotic-resistant transformants were recovered from themedium interface after 72 h of incubation, and the recipient strain was shown to contain the delivered plasmidby agarose gel electrophoresis of isolated plasmid DNA. All strains of B. megaterium tested were successfullytransformed by this method, although transformation efficiency varied among strains. Physical variables of thebiolistic process and biological variables associated with the target cells were optimized, yielding >104transformants per treated plate. This is the first report of the biolistic transformation of a procaryote.

A great deal of effort has gone into the development oftransformation technology, resulting in a large body ofknowledge and diverse methodologies. Some methods arevery simple but are not widely applicable. Other methodsmay be more widely useful but are more complex or difficult.Many gram-negative bacteria, such as Escherichia coli, aretransformed easily by pretreatment with divalent metal ionsbefore the addition of plasmid DNA (6). Conversely, manygram-positive bacteria are notoriously difficult to transform.Natural transformation of gram-positive bacteria is onlyknown to occur among Bacillus subtilis and a few species ofStreptococcus. With considerable effort, protoplasting, con-jugation, electroporation, and injured-cell techniques havemade possible the genetic manipulation of various gram-positive genera, including Bacillus, Lactobacillus, Listeria,Staphylococcus, Streptococcus, Streptomyces, and coryne-form bacteria (15, 16, 23, 24). However, some recalcitrantspecies remain untransformable even after exhaustive test-ing of currently available techniques. Also, many of theprotocols now in use must be painstakingly adjusted forindividual strains to achieve satisfactory results. It is appar-ent that a widely applicable, rapid method of plasmid trans-fer for procaryotic cells is still highly desirable.The genus Bacillus contains species that are both scientif-

ically and commercially important (18). Strains of B. mega-terium developed for industry are used in the production ofvitamin B12, penicillin amidase, nucleotides, and otherchemicals. Some strains of B. megaterium are also signifi-cant as plant pathogens. The technique currently employedfor plasmid transformation in this species is protoplasting (4,25, 26). Although this method has resulted in successfulplasmid transfer, preparation and regeneration of protoplastsis an onerous process, often resulting in low efficiency andinability to achieve reproducible results. In this report, wedemonstrate the use of biolistic technology (formerly used asa method of gene transfer only in eucaryotic cells) as anexpeditious means for high-efficiency plasmid transforma-

* Corresponding author.

tion of B. megaterium. Some biological and physical param-eters affecting the efficacy of this technique are elucidatedwith respect to this species.

MATERIALS AND METHODS

Bacterial strains and plasmid DNA. B. megaterium 7A17(metA4), 7A1, 7A2, 7A16, 7A24 (leuBI), and B. subtilis 1E6[thr-5 trpC2(pUB110)] were provided by the Bacillus Ge-netic Stock Center, Ohio State University, Columbus (Table1). B. megaterium pv. cerealis Hosford (ATCC 35075), thecausal agent of white blotch of wheat, was obtained fromAmerican Type Culture Collection, Rockville, Md. (11). Allcultures were maintained on Schaeffer's sporulation mediumat 4°C (22).

Plasmid pUB110, a 4.5-kb plasmid that confers kanamycinresistance, was isolated from B. subtilis 1E6. Strain 1E6 wasincubated with aeration (150 rpm) for 24 h in Luria-Bertani(LB) broth plus 50 ,ug of kanamycin sulfate per ml at 37°C,and plasmid DNA was isolated by the boiling lysis method(17). Plasmid DNA was purified by CsCl density gradientultracentrifugation. Purified DNA was resuspended in TEbuffer (1 mM Tris [pH 7.8], 0.1 mM EDTA), and the con-centration was determined spectrophotometrically.

Preparation of cells and microprojectiles for bombardment.B. megaterium strains were grown in LB broth for 15 h at35°C with aeration (250 rpm). A 50-ml sample of each culturewas centrifuged for 10 min at 8.93 x 102 x g at roomtemperature. The bacterial pellet was resuspended in 5 ml ofsupernatant, and then the cell density was determined spec-trophotometrically. Except where noted, 108 CFU werespread on the surface of each plate, containing LB mediumwith methionine (50 ,ug/ml), D-sorbitol (1.0 M), D-mannitol(0.75 M), and agar (15 g/liter). The surface of the plate wasslowly dried before bombardment to remove all visiblewater.M5 tungsten particles (Sylvania, GTE Products Corp.,

Towanda, Pa.) were used as microprojectiles. These parti-cles were characterized by a mean distribution size of 0.771pLm, with a median of 0.362 p.m and a mode in the range of

480

BIOLISTIC TRANSFORMATION OF BACILLUS MEGATERIUM 481

TABLE 1. Bacterial strains

Strain . Source or(BGSC' code) Original name reference

Bacillus megaterium7A1 899 10a7A2 ATCC 19213 197A16 QM B1551 24a7A17 PV2 24b7A24 JV75 9

Bacillus megaterium pv. WB28 (ATCC 35075) 11cerealis

Bacillus subtilis 1E6 BD366 10a Bacillus Genetic Stock Center, Ohio State University, Columbus.

0.1 to 0.2 Rm. pUB110 DNA was precipitated onto thetungsten particles as previously described (7) with minormodifications. Spermidine was used at 0.1 M (10 ,ul) insteadof 1.0 M (5 ,ul). Coated particles were washed after precip-itation with 70% ethanol (to remove the free CaC12 andspermidine) and were then resuspended in 100% ethanol tofacilitate rapid drying of the coated particles onto the launch-ing surface. After a brief water bath sonication to dispersethe particles, 3 ,ul of suspended particles was loaded in thecentral portion of the launch surface. Hence, approximately600 ,ug of tungsten coated with 0.8 jig of pUB110 was used tobombard each plate containing 108 CFU.

Particle accelerator. A new helium-driven biolistic device(20a) was used in this study to transform B. megaterium. Thenew acceleration device is driven by a shock wave ofcompressed helium. This shock wave is used to acceleratemicroprojectiles by three distinct mechanisms. In the gasentrainment mechanism, particles are loaded onto a nylonmesh which transects the path of the shock wave, and theparticles are launched by and entrained into the shock wavedirectly. In the ruptured membrane mechanism, particles areloaded onto an anchored aluminum foil disk which transectsthe path of the shock wave and are launched when the shockwave distends and then bursts the membrane. In the flyingdisk mechanism, particles are loaded onto an unanchored2-mil (ca. 51-,um) Kapton membrane which transects thepath of the shock wave and are accelerated as the flying diskis accelerated and are launched when the flying disk isstopped by a brass ring with a steel screen insert. In all casesa vacuum chamber surrounds the source of the helium shockwave, the particle launch surface, and the target cells.The helium-driven apparatus we used is merely a retrofit

of our previously described (14, 20) gunpowder-driven ap-paratus. The firing mechanism and acceleration barrel havebeen replaced by a mechanism which can generate a con-trolled helium shock. Helium is delivered from a standardgas tank through a regulator and high-pressure hose to abrass tube which replaces the acceleration barrel. At thelower end of the brass tube is a high-pressure cavity (ap-proximately 1 cm3), which is sealed at its lower end by oneor more layers of Kapton membrane. Each layer of Kaptonis 2 mils thick and will hold 300 lb/in2 of pressure. When thepressure overlying the membranes exceeds their combinedstrength (or when the membranes are mechanically rup-tured), the membranes burst, releasing a sharply definedshock wave. This shock wave propagates downward towardthe partition (stopping plate platform) of the old gunpowdersystem. Where the Lexan stopping plate was inserted in thisplatform, a new insert is placed which holds an internally

threaded aluminum sleeve, 2 cm deep and 2 cm in internaldiameter. Into this sleeve are screwed brass rings which areused to hold the launch surfaces (membranes or screens) inthe direct path of the gas shock. In the flying disk treatment,a stainless steel stopping screen is fixed within the sleeve 1cm below the launch site of the flying disk. All other physicalaspects of the system are the same or comparable to thecommercially available PDS-1000 gunpowder-driven sys-tem.

Particle accelerator conditions were optimized for strain7A17 and were then used without modification on otherstrains. To arrive at the optimum biolistic conditions for B.megaterium 7A17, we tested several physical variables ofparticle acceleration with respect to transformation effi-ciency. Those variables included helium pressure (600, 900,1,200, and 1,500 lb/in2), distance between the helium shocksource and launch site (1.6, 1.0, and 0.55 cm), distancebetween launch site and target cells (4.1, 6.1, and 8.1 cm),and launch mechanism (flying disk, ruptured membrane, andgas entrainment).Bombardment protocol and selection of transformants and

controls. A plate inoculated with B. megaterium was placedinside the bombardment chamber (with the lid removed),and a partial vacuum was drawn (29 in. [ca. 74 cm] ofmercury). The plate was bombarded with the DNA-coatedmicroprojectiles, the vacuum was released, and the platewas covered and then incubated for 2 h at 35°C. Afterincubation, each plate was overlaid with 15 ml of LBmedium containing methionine (50 jig/ml), kanamycin (50,ug/ml), and agar (15 g/liter) and was returned to the 35°Cincubator. Transformants were visible after 24 h, and finaltransformation counts (colonies per plate) were determinedafter 72 h of incubation.

Controls plates were included in all experiments. DNA-coated tungsten was resuspended in 0.01 M spermidineinstead of 100% ethanol and then mixed with cells. Themixture of cells and coated tungsten was spread over an agarsurface. The control plates were exposed to vacuum in thevacuum chamber but not to helium bombardment. In allexperiments, there were no spontaneous kanamycin trans-formants or mutants.

Optimization of biological variables affecting transforma-tion of strain 7A17. Four biological factors associated withthe target cells were tested with regard to transformationyield. These factors, growth phase of the target cells, con-centration of osmoticum in the bombardment medium, celldensity, and cell strain, had previously been shown to affecttransformation yield in Saccharomyces cerevisiae (1). Totest growth-phase effects, we spread cells from cultures inthe mid-logarithmic (6.5 h at 35°C), late logarithmic (15 h),stationary (24 h), and late stationary (40 h) phases ontoplates and bombarded them. Each treatment was replicatedfive times. Osmotic effects were tested by using variousconcentrations of sorbitol (0.5, 0.75, 1.0, and 1.25 M) andmannitol (0.75 M) and combined concentrations of these(0.75 M sorbitol plus 0.75 M mannitol, 1.0 M plus 0.75 M,and 1.25 M plus 0.75 M) as osmotic agents in the bombard-ment medium. The two best osmoticum concentrations werethen selected and were tested further at five different celldensities. The cell densities tested were S x 106, 1 X 107, 5X 107, 7.5 x 107, and 1 x 108 CFU/85-mm-diameter platefrom 15-h-old cultures. Five replicates per treatment were

bombarded, and the number of transformants per treatmentwas determined. Last, strain differences of B. megateriumwere tested. The optimum physical and biological conditionsof strain 7A17 were used to transform B. megaterium 7A1,

VOL. 57, 1991

APPL. ENVIRON. MICROBIOL.

1 2 : 4 5 6 7 8 9

Kb

23-

9.4-6.7-4.4-

2.3-2.0-

FIG. 1. Agarose gel electrophoresis of plasmid DNA isolatedfrom B. megaterium transformants. Plasmid DNA was isolated fromsix Kmr transformants of strain 7A17 (lanes 4 to 9), and pUB110 wasisolated from its host, B. subtilis 1E6 (lane 3), restricted with BamHIfor 1 h, subjected to electrophoresis, and stained with ethidiumbromide. A Hindlll markers (lane 1) and undigested pUB110 from aCsCl preparation (lane 2) were also included.

7A2, 7A16, 7A17, and 7A24 and B. megaterium pv. cerealiswith plasmid pUB110. Each treatment was replicated fivetimes, and the experiment was repeated five times.

Confirmation of transformation. Ten putative transfor-mants were tested for authenticity by Gram stain reactionand isolation and visualization of plasmid DNA on ethidiumbromide-stained agarose gels. Plasmid DNA was isolatedfrom overnight cultures of putative transformants by amini-boiling preparation method (17), a 1-h digestion withBamHI (Promega Biotec, Madison, Wis.) following themanufacturer's directions, and agarose gel electrophoresis(0.8% agarose, 90 V, 4 h).

RESULTS

Transformation of B. megaterium 7A17 was verified bygrowth in the presence of kanamycin, a positive Gram stainreaction, and plasmid isolation from putative transformantsfollowed by sizing of plasmid DNA by agarose gel electro-phoresis (Fig. 1). BamHI-restricted plasmid DNA fromputative transformants was identical in size to digestedpUB110 (4.5 kb), while nontransformed cells of the recipientstrain lacked any equivalent plasmid.The efficiency of transformation of B. megaterium by the

biolistic process is affected by numerous biological factors aswell as by physical factors associated with the accelerationprocess. The growth phase of the target cells, cell density,and concentration of osmotic agent in the bombardmentmedium were all found to be important biological factors anddrastically affected transformation efficiency. When equalnumbers of cells from different growth phases were bom-barded with an equal amount of DNA-coated particles, cellsfrom late-logarithmic cultures of strain 7A17 yielded signif-icantly (P = 0.05) more transformants (four to five times

TABLE 2. Transformation of B. megaterium 7A17 atdifferent growth phasesa

Growth phase Mean no. of SE No. of transformants/(h) transformants/plateb recipient cell

Expt 1Late log (15) 436.8 92.3 5.80 x 10-6Stationary (24) 87.7 31.1 1.16 x 10-6Late stationary (40) 71.4 31.9 9.52 x 10-7

Expt 2Mid-log (6.5) 832.4 454.9 8.32 x 10-6Late log (15) 90.2 35.3 9.02 x 10-7

" 7.5 x 107 CFU per plate and 1 x 108 CFU per plate and 10 and 5 replicatesper treatment were used in experiments 1 and 2, respectively. Cells werebombarded on medium containing 0.75 M sorbitol and 0.75 M mannitol.

b Each plate was bombarded with tungsten coated with 0.8 jLg of pUBllODNA.

more) than cells from stationary cultures (24 and 40 h) (Table2). Cells from mid-log-phase (6.5-h) cultures yielded signifi-cantly (P = 0.18) more transformants than cells from late-log-phase (15-h) cultures.

Increasing the molar concentration of sorbitol in thebombardment medium from 0 to 1.5 M increased the trans-formation efficiency of B. megaterium 7A17 sevenfold (Fig.2). Bombardment medium with a saturated concentration ofmannitol (0.75 M) yielded more transformants than the sameconcentration of sorbitol (Table 3). Increasing the overallconcentration of osmoticum by combining 0.75 M mannitolwith 1 M sorbitol gave a significantly (P = 0.25) highertransformation rate than medium containing the same sorbi-tol concentration alone. The greatest number of transfor-mants per plate was produced on medium with total osmoticconcentrations greater than a combined concentration of 1.5M (i.e., 1 M sorbitol plus 0.75 M mannitol).The number of target cells per plate also affected the

transformation rate. The optimum cell density tested for

100

I-a.

z

20LLC/3z

80

60

40

20

O '-

0.0 0.5 1.0 1.5

SORBITOL (M)FIG. 2. Effect of sorbitol concentration in the bombardment

medium on transformant yield in B. megaterium 7A17. A total of 108CFU per plate were bombarded with tungsten particles coated with0.8 ,ug of pUB110. Cells were bombarded by the flying disk launchmechanism at optimum pressure and distances. Transformants wereselected with a 10-ml LB agar overlay containing 50 ±g of kanamy-cin per ml.

482 SHARK ET AL.

BIOLISTIC TRANSFORMATION OF BACILLUS MEGATERIUM 483

TABLE 3. Effect of osmoticum concentration on transformation of B. megaterium 7A17'

Mean no. of No. of transformants/transformants/plateS recipient cell

Sorbitol 0.75 235C 154.1 3.13 x 10-7Sorbitol 1.25 328C 225.8 4.37 x 10-7Mannitol 0.75 584C 204 7.79 x 10-7Sorbitol + mannitol 0.75 + 0.75 1,085c.d 847 1.44 x 10-5Sorbitol + mannitol 1.00 + 0.75 5,693d 1,852.4 7.59 x 10-5Sorbitol + mannitol 1.25 + 0.75 3,584 .d 1,943.5 4.77 x 10-5

a 7.5 x 107 CFU per plate of a 15-h-old culture were spread on bombardment medium containing various concentrations of osmotic agents. Five replicates wereused per treatment.

b Each plate was bombarded with tungsten coated with 0.8 pLg of pUBllO DNA.c.d Means not followed by a common letter are significantly different from one another.

strain 7A17 was 108 CFU/85-mm-diameter plate (Table 4).However, there appeared to be some interaction betweenosmoticum concentration and cell density. At 5 x 107 CFUper plate, there were more transformants on medium con-taining 1.0 M sorbitol plus 0.75 M mannitol, but at the highercell densities tested (7.5 x 107 and 1.0 x 108 CFU per plate),there were more transformants on 1.25 M sorbitol plus 0.75M mannitol.

In addition to growth phase, cell density, and osmoticumconcentration, strain differences were observed to affecttransformation efficiency. The relative ranking among thestrains remained constant in five experiments, and the re-

sults of one experiment are presented in Table 5. Note thatthree of six strains had higher transformation rates thanstrain 7A17, the principal strain used in most studies. How-ever, only strain 7A1 yielded significantly (P = 0.075) moretransformants per plate than the other five strains. There wasno statistically significant difference in transformant yieldamong the remaining five strains.

Pressures and distances were empirically optimized forthe three launch mechanisms. The resulting optimized treat-ments (flying disk [900 lb/in2, 4.1 cm], gas entrainment [1,200lb/in2, 6.1 cm], and ruptured membrane [900 lb/in2, 6.1 cm])were then compared. The flying disk mechanism yielded 20and 40 times more transformants per plate than gas entrain-ment and ruptured membrane, respectively. Pressure (900,1,200, and 1,500 lb/in2), distance between helium source andmacroprojectile (1.6, 1.0, and 0.55 cm), and distance be-tween microprojectile launch site and target cells (4.1, 6.1,and 8.1 cm) were varied in different combinations to furtheroptimize the flying disk configuration. The flying disk treat-ment that gave the greatest yield Qf transformants was 900lb/in2, 1.6 cm between helium source and macroprojectile,

TABLE 4. Effect of cell density and osmoticum on

transformation of B. megaterium 7A17a

Mean no. of transformants/platebCell density/plate

Osmoticum A Osmoticum B

5x106 0 01 X 107 0.2 0.25 x 107 251.6 4.67.5 x 107 261.2 >20,000x 108 1,009.8 >20,000

a Cells from a 15-h-old culture were bombarded on medium containingeither 1.0 M sorbitol plus 0.75 M mannitol (A) or 1.25 M sorbitol plus 0.75 Mmannitol (B).

b Each plate was bombarded with tungsten coated with 0.8 Fig of pUB110DNA.

and 4.1 cm between microprojectile launch site and targetcells. Use of these optimum conditions often resulted in sucha large number of transformants (>20,000) that they some-times appeared as a lawn of confluent colonies too numerousto count except when counted with a binocular dissectingscope at 48 to 65 h of incubation (Fig. 3).

DISCUSSION

Since its inception in the early 1980s (14, 20, 21), biolistictechnology has been successfully used to genetically trans-form a wide variety of eucaryotic cells and their organellesand has gained recognition and credibility as an effectivemethodology. MonoFpts including corn (12), rice, and wheat(27) as well as dicots such as tobacco (13) and soybean (5, 27)have all been transformed by DNA introduction throughparticle bombardment. Cultured animal cells have beentransformed utilizing this technology (28). Also, nucleartransformation ; f eucaryotic microorganisms includingyeasts and filamentous fungi (1) has been shown. Biolistictechnology has been unique in its capability to directly andreproducibly transform both mitochondria (8) and chloro-plasts (2, 3).The previous success of biolistic technology in transform-

ing a diversity of cell types, together with the need for arapid, easily adaptable method for the transformation ofprocaryotic cells, led to our desire to determine whetherbacteria could be transformed via particle bombardment. Itwas not clear that the use of high-velocity microprojectileswould be effective in transformation of procaryotic cells.The procaryotic cells provided a markedly smaller targetthan eucaryotic cells; outer cell structures and the genetic

TABLE 5. Transformation of six strains of B. megateriuma

Strain Mean no. of SE No. of transformants/transformants/plate S recipient cell

7A1 1,568.2 526.1 1.56 x 10-57A2 4.2 1.6 4.20 x 10-87A16 0.8 0.8 8.00 x 10-97A17 36.8 9.3 3.68 x 10-87A24 217.2 114.7 2.17 x 10-6pv. cerealis 253.4 160.3 2.53 x 10-6

a Cells from 15-h-old cultures were bombarded on medium containing 1.0 Msorbitol plus 0.75 M mannitol. Rates were generally low in this experiment, as

reflected by only 37 colonies for 7A17, which often gave >104 colonies perplate. However, relative ranking of these strains was consistent over fiveexperiments.

b Each plate was bombarded with tungsten coated with 0.8 ,ug of pUBllODNA.

VOL. 57, 1991

APPL. ENVIRON. MICROBIOL.

associated with DNA uptake and it was previously recog-nized as a hard-to-transform species needing improvedtransformation methodology (25). Recently, an efficient pro-toplasting transformation method for B. megaterium hasbeen published (26), although this method is relatively te-dious and time-consuming compared with the biolisticmethod. In our experiments, transformation of B. megate-rium was initially accomplished by using a commerciallyavailable gunpowder-driven device (PDS-1000; Dupont).The rates were, however, too low to be workable (less thanone colony per plate). The improved helium-driven biolisticdevice described here and elsewhere (20a) was found to bedramatically more effective than the PDS-1000 unit, so allthe work shown in this report employed that device.

B. megaterium 7A17 was easily transformed with plasmidDNA via the biolistic process. Under optimum biologicaland physical conditions, greater than 1 x 104 transformantsper plate were often produced, equivalent to 1 x 10-4transformants per recipient cell or 8 x 103 transformants per,ug of DNA. For B. megaterium 7A17, an osmoticum con-centration of >1.5 M was optimum and cells in logarithmicgrowth were more efficiently transformed than cells ofstationary cultures. As in fungal systems, osmoticum in thebombardment medium affects transformation rates (1). Un-like yeasts, B. megaterium cells in the log phase are moreefficiently transformed than stationary cells. Although all B.megaterium strains tested were successfully transformed,transformation rates varied between some strains. For max-imum numbers of transformants, biological factors such asgrowth phase, osmoticum concentration, and cell densityshould be optimized for each species. After optimum condi-tions are determined, cell density or DNA load can bereduced to decrease the number of transformants per plateso that colonies are well separated and easily quantified.

High-efficiency biolistic transformation of a gram-positivebacterium was described here. Similar results have beenfound for a gram-negative bacterium (E. coli) (23a). Theseresults and the relative ease with which this technology canbe employed suggest a potential for use of the biolisticprocess in the transformation of procaryotic cells.

FIG. 3. Typical shotgun pattern of B. megaterium 7A17 (A) and7A1 (B) transformants 72 h after bombardment.

structure of the cells are fundamentally different. For manyyears, it was believed the biolistic process might only beeffective in very large cell types (14). The weight andvelocity (momentum) of impinging DNA-carrying particlesneeds to be sufficient to penetrate the cell wall and mem-brane, yet particles must be small enough that they do notirreparably damage the cell. We felt that the need forsufficiently high momentum compounded with the require-ment for a small frsiop size might create a minimal effectivetarget size. Therefore, this study was initiated strictly todetermine, the feasibility of biolistic transformation of pro-caryotic cells and to demonstrate that the biolistic processcould be applied tp even very small cellular targets. Initialexperiments using the PDS-1000 gunpowder-driven devicewere begun with E. coli. The initial rates of E. coli biolistictransformation were exceedingly low and were only compa-rable with the background rates of spontaneous DNA uptakein our negative controls. Therefore, we chose B. megateriumas a model procaryote for our proof-of-concept experiments,as there was virtually no background rate of transformation

ACKNOWLEDGMENTSWe thank R. Marrero, S. Zahler, and P. Van der Horn for

valuable information at the inception of this study.This work was supported by a grant from DuPont Co. F.S. was

supported by Public Health Service grant ROI-GM 41426-01 fromthe National Institutes of Health.

REFERENCES1. Armaleo, D., G. N. Ye, T. M. Klein, K. B. Shark, J. C. Sanford,

and S. A. Johnston. 1990. Biolistic nuclear transformation ofSaccharomyces cerevisiae and other fungi. Curr. Genet. 17:97-103.

2. Blowers, A. D., L. Bogorad, K. B. Shark, G. N. Ye, and J. C.Sanford. 1989. Studies on Chlamydomonas chloroplast transfor-mation: foreign DNA can be stably maintained in the chromo-some. Plant Cell 1:123-132.

3. Boynton, J. E., N. W. Gillham, E. H. Harris, J. P. Hosler, A. M.Johnson, A. R. Jones, B. L. Randolph-Anderson, D. Robertson,T. M. Klein, K. Shark, and J. C. Sanford. 1988. Chloroplasttransformation of Chlamydomonas using high velocity micro-projectiles. Science 240:1534-1538.

4. Brown, B. J., and B. C. Carlton. 1980. Plasmid-mediatedtransformation in Bacillus megaterium. J. Bacteriol. 142:508-512.

5. Christou, P., D. E. McCabe, and W. F. Swain. 1988. Stabletransformation of soybean callus by DNA-coated gold particles.Plant Physiol. 87:671-674.

484 SHARK ET AL.

BIOLISTIC TRANSFORMATION OF BACILLUS MEGATERIUM 485

6. Cohen, S. N., A. C. Y. Chang, and L. Hsu. 1973. Nonchromo-somal antibiotic resistance in bacteria: genetic transformation ofEscherichia coli by R-factor DNA. Proc. Natl. Acad. Sci. USA69:2110-2114.

7. Daniell, H., J. Vivelzananda, B. L. Nielsen, G. N. Ye, K. K.Tewari, and J. C. Sanford. 1990. Transient foreign gene expres-sion in chloroplasts of cultured tobacco cells after biolisticdelivery of chloroplast vectors. Proc. Natl. Acad. Sci. USA87:88-92.

8. Fox, T. D., J. C. Sanford, and T. W. McMullin. 1988. Plasmidscan stably transform yeast mitochondira totally lacking endog-enous mtDNA. Proc. Natl. Acad. Sci. USA 85:7288-7292.

9. Garbe, J. C., G. F. Hess, M. A. Franzen, and P. S. Vary. 1984.Genetics of leucine biosynthesis in Bacillus megaterium QMB1551. J. Bacteriol. 157:454-459.

10. Gryczan, T. J., S. Contente, and D. Dubnau. 1978. Characteri-zation of Staphylococcus aureus plasmids introduced by trans-formation into Bacillus subtilis. J. Bacteriol. 134:318-329.

10a.Halvorson, H. 0. Unpublished data.11. Hosford, R. M. 1982. White blotch incited in wheat by Bacillus

megaterium pv. cerealis. Phytopathology 72:1453-1459.12. Klein, T. M., M. E. Fromm, A. Weissinger, D. Tomes, S. Schaaf,

M. Sleeten, and J. C. Sanford. 1988. Transfer of foreign genesinto intact maize cells using high velocity microprojectiles.Proc. Natl. Acad. Sci. USA 85:4304-4309.

13. Klein, T. M., E. C. Harper, Z. Svab, J. C. Sanford, M. E.Fromm, and P. Maligna. 1988. Stable genetic transformation ofintact Nicotiana cells by the particle bombardment process.Proc. Natl. Acad. Sci. USA 85:671-674.

14. Klein, T. M., E. D. Wolf, R. Wu, and J. C. Sanford. 1987.High-velocity microprojectiles for delivering nucleic acids intoliving cells. Nature (London) 327:70-73.

15. Luchansky, J. B., P. M. Muriana, and T. R. Klaenhammer.1988. Application of electroporation for transfer of plasmidDNA to Lactobacillus, Lactococcus, Leuconostoc, Listeria,Pediococcus, Bacillus, Staphylococcus, Enterococcus and Pro-pionibacterium. Mol. Microbiol. 2:637-646.

16. Mallonee, D. H., and R. A. Speckman. 1989. Transformation ofBacillus polymyxa with plasmid DNA. Appl. Environ. Micro-biol. 55:2517-2521.

17. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

18. Priest, F. G. 1977. Extracellular enzyme synthesis in the genusBacillus. Bacteriol. Rev. 41:711-753.

19. Rode, L. J., and J. W. Foster. 1960. Mechanical germination ofbacterial spores. Proc. Natl. Acad. Sci. USA 46:118-128.

20. Sanford, J. 1988. The biolistic process-a new concept in genetransfer and biological delivery. Trends Biotechnol. 6:229-302.

20a.Sanford, J. C., et al. Unpublished data.21. Sanford, J. C., T. M. Klein, E. D. Wolf, and N. Allen. 1987.

Delivery of substances into cells and tissues using a particlebombardment process. J. Part. Sci. Technol. 5:27-37.

22. Schaeffer, P., J. Millet, and J.-P. Aubert. 1965. Catabolicrepression of bacterial sporulation. Proc. Natl. Acad. Sci. USA54:704-711.

23. Schafer, A., J. Kalinowski, R. Simon, A.-H. Seep-Feldhaus, andA. Puhler. 1990. High-frequency conjugal plasmid transfer fromgram-negative Escherichia coli to various gram-positive cory-neform bacteria. J. Bacteriol. 172:1663-1666.

23a.Smith, F. D., et al. Unpublished data.24. Trieu-Cuot, P., C. Carlier, P. Martin, and P. Courvalin. 1987.

Plasmid transfer by conjugation from Escherichia coli to gram-positive bacteria. FEMS Microbiol. Lett. 48:289-294.

24a.U.S. Army, Quarter master R&D Center. Unpublished data.24b.Vary, P. Unpublished data.25. Vary, P. S., and Y.-P. Tao. 1988. Development of genetic

methods in Bacillus megaterium. Genet. Biotechnol. Bacilli2:403-407.

26. Von Tersch, M. A., and H. L. Robbins. 1990. Efficient cloning inBacillus megaterium: comparison to Bacillus subtilis and Esch-erichia coli cloning hosts. FEMS Microbiol. Lett. 70:305-310.

27. Wang, Y. C., T. M. Klein, M. Fromm, J. Cao, J. C. Sanford, andR. Wu. 1988. Transformation of rice, wheat, and soybean by theparticle bombardment method. Plant Mol. Biol. 11:433-439.

28. Zelenin, A. V., A. V. Titomirov, and V. A. Kolesnikov. 1989.Genetic transformation of mouse cultured cells with the helpof high velocity mechanical DNA injection. FEBS Lett. 244:65-67.

VOL. 57, 1991