Proc. Natl. Acad. Sci. USAVol. 90, pp. 9295-9299, October 1993Genetics

Transformation of Tetrahymena thermophila by microinjection of aforeign gene

(ciliate/neo gene/rRNA-encoding DNA/histone H4-I gene/homologous recombination)

ROBERT W. KAHN, Bo H. ANDERSEN, AND CLIFFORD F. BRUNKBiology Department and Molecular Biology Institute, University of California, Los Angeles, CA 90024-1606

Communicated by John R. Preer, Jr., June 29, 1993

ABSTRACT Tetrahymena thermophila has been trans-formed to paromomycin-resistant phenotypes by microinjec-tion of an aminoglycoside 3'-phosphotransferse (neo) geneunder the control of the T. thermophila histone H4-I promoter.This chimeric neo gene, by itself or on a vector containing arRNA-encoding DNA (rDNA) origin of replication, transformsT. thermophla. In cells transformed with the rDNA originvector, the neo gene is usually found integrated into theendogenous rDNA molecules and is present in high copynumber. In transformants obtained by micro'ijecting only thelinear chimeric gene, the neo gene is found to have replaced thehistone H4-I gene or is found integrated into the 5' flankingregion of the H4-I gene. The relative transcript levels of the neogene in T. thermophUa transformed by the linear chimeric geneare much higher than in cells transformed with the vector. Theneo gene provides an effective selectable marker for transfor-mation of T. thermophila.

The ciliated protozoan Tetrahymena thermophila has provento be a valuable biological system for studying the genetic andmolecular processes of eukaryotic cells. Research on T.thermophila, which elucidated the molecular mechanisms ofribozymes (1), DNA rearrangement (2, 3), and telomerestructure and function (4), has stimulated similar research inother organisms. Although the genetics of T. thermophila arewell characterized, little information about gene expressionin this organism is available (5-7). A transformation systemfor T. thermophila using a foreign gene would facilitatestudies of gene expression and enhance the utility of thisorganism.Like all other ciliated protozoa, Tetrahymena contains two

different types of nuclei in a common cytoplasm, a somaticmacronucleus and a germ-line micronucleus. The micronu-cleus is diploid and expressionally silent during vegetativegrowth, while the macronucleus is polyploid and responsiblefor somatic gene expression (8, 9). In T. thermophila, themacronucleus contains "'50 copies of each gene, except forthe rRNA genes (rDNA), which exist in ""10,000 copies (9).rDNAs are located in the nucleoli and reside on autono-mously replicating DNA molecules 21 kb long (10).

Transformation of T. thermophila has only been achievedby delivering the transforming gene to the macronucleus. T.thermophila has been transformed by the introduction ofnative rDNA that confers paromomycin resistance (11-13).Several different rDNA alleles have been characterizedbased on their origin of replication-e.g., rDNA moleculesisolated from T. thermophila strain C3 outreplicate rDNAmolecules from T. thermophila strain B, when they exist inthe same nucleus (14). Transformation of T. thermophila withrDNA genes requires that the transforming rDNAs outrep-licate the endogenous rDNA genes.

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

Transformation has also been achieved by introduction ofa modified ribosomal protein gene L29 that confers cyclo-heximide resistance (15, 16). This modified L29 gene trans-forms T. thermophila by intragenic recombination with theendogenous wild-type gene. The modified L29 gene productmust compete with the wild-type gene product in order toproduce the cycloheximide-resistant phenotype. Transfor-mation with the L29 gene and the rDNA gene depends on T.thermophila gene products that are resistant to cyclohexi-mide and paromomycin, respectively; thus, they have tocompete with wild-type gene products in order to confer drugresistance. The introduction of a foreign gene whose geneproduct inactivates paromomycin provides resistance with-out competition or interference from endogenous gene prod-ucts.We have developed a transformation system for T. ther-

mophila based on the neo gene from the bacterial transposonTnS (17). The neo gene product is an aminoglycoside 3'-phosphotransferase that phosphorylates antibiotics of theneomycin family, including paromomycin. T. thermophila ishighly sensitive to paromomycin, making the neo gene aneffective selectable marker. We have constructed a chimericgene containing the neo gene flanked by the upstream anddownstream sequences of the T. thermophila histone H4-Igene. The coding region of the TnS neo gene terminates inTGA, which is the only termination codon used by mostciliates (18, 19). When this chimeric neo gene is introducedinto T. thermophila cells, they are transformed to a paromo-mycin-resistant phenotype. This demonstrates the expres-sion of a foreign gene in T. thermophila.We describe construction of the chimeric neo gene, its

introduction into T. thermophila, and the configurations ofthe neo gene within the genome of transformed cells. Therelative transcript levels from the neo gene in differentgenomic configurations are also investigated. This studydemonstrates the utility of the neo gene as a selectablemarker in T. thermophila.

MATERIALS AND METHODSCells and Culturing Conditions. T. thermophila strain

SB2120 was provided by E. Orias (University of California,Santa Barbara). This strain is sensitive to paromomycin andcontains a C3-rmm 1 rDNA type, which is outreplicated bya C3 rDNA type (14). Cells were grown in PPY medium (5)with penicillin (250 ug/ml) and streptomycin (250 ,ug/ml) at30°C with gentle shaking.

Construction of the Chimeric neo Gene. A plasmid contain-ing the neo gene from the bacterial transposon Tn5 (pKm2l)was provided by H. Schaller (17). A clone containing the T.thermophila histone H4-I gene and flanking regions (p508.8)was provided by M. A. Gorovsky (20). The Hae III fragment('2 kb) from the histone gene 5' flank that terminates just

Abbreviation: rDNA, rRNA-encoding DNA.

9295

Dow

nloa

ded

by g

uest

on

Aug

ust 2

6, 2

020

Proc. Natl. Acad. Sci. USA 90 (1993)

within the coding region was isolated and HindIII linkers(New England Biolabs 1022) were ligated to it. This segmentwas then ligated into the Hindlll site ofpKm21 in the properorientation to place the neo gene under the control of thehistone promoter. An EcoRI/Dde I restriction fragment(-3.0 kb) was isolated from this construction. The 5' end ofthis fragment is the EcoRI site -150 bp upstream from the 5'proximal Hindlll site and the 3' end is the naturally occurringDde I site within the termination codon of the neo gene (17).The histone gene and flanks of p508.8 [from Sau3AI (5') to

Bgl II (3'), 2 kb] were inserted into pUC18, oriented withthe Bgl II site adjacent to the EcoRI site in pUC18. Thetermination codon of the histone gene in this subclone was

modified by in vitro mutagenesis to produce a Dde I site. TheDde I (termination codon)/EcoRI (in pUC18) restrictionfragment (-1.2 kb), including the 3' flank ofthe histone gene,was isolated from this modified subclone. This segment ofDNA and the EcoRI/Dde I segment (-3 kb) that contains theneo gene under the control of the histone promoter wereligated together. The resulting chimeric gene (-4 kb) con-taining the neo gene under the control ofthe histone promoterwas ligated into the EcoRI site of pUC19. The 1.9-kb BamHIfragment of prD4-1, containing a mutated T. thermophila C3rDNA origin of replication (courtesy of E. Blackburn; ref.21), was inserted into the vector to produce pH4neo (Fig. 1).

Microinjection and Transformation. A microinjection pro-cedure similar to that described by Tondravi and Yao (11)was used to introduce the vector into T. thermophila. AnEppendorf model 5170 micromanipulator attached to a Zeiss(Axiovert 35) inverted microscope and an Eppendorf model5242 microinjector (courtesy of J. Lusis, University of Cal-ifornia, Los Angeles) were used to microinject DNA (250,g/ml in 10 mM Tris.HCl/1 mM EDTA, pH 7.5) into mac-ronuclei of T. thermophila. Cells microinjected with pH4neowere cloned into the wells of a microtiter plate, allowed togrow to high cell density, and transferred daily to both freshmedium and medium with paromomycin (Parke-Davis; 100,ug/ml). Paromomycin-resistant (pmr) clones were identified6-14 days after microinjection. A vector containing only theH4 neo gene (without the rDNA origin inserted into theBamHI site) was digested with EcoRI and microinjected intothe macronuclei of T. thermophila (without removing thepUC19 sequences). These cells were cloned as describedabove, allowed to grow at least 4 generations, and challengedwith paromomycin at a final concentration of 100 ,g/ml. pmrcell lines were identified 2-3 days after microinjection. Thepmr level of each transformant was measured by determiningthe maximum concentration of paromomycin at which trans-

Eco RI

I Bam HIHind III

Hind III ' -Eco RI

FIG. 1. Schematic diagram of the vector pH4neo containing theneo gene under the control of the T. thermophila histone H4-Ipromoter and the T. thermophila rDNA origin of replication. Openarrow, neo gene coding region; thick solid lines, histone H4-Iflanking regions; open box with arrow, C3 origin of rDNA replica-tion; thin line, pUC19, orientation indicated by arrow. Restrictionendonuclease sites relevant to the construction are indicated.

formed cell lines can be maintained with daily 2-fold dilution(Table 1).

Nucleic Acid Isolation and Analysis. Total cellularDNA wasisolated from cultures of transformed T. thermophila cells(22). Southern blot analysis (23) was performed on DNAdigested with various restriction endonucleases using vac-uum blotting onto Zeta-Probe membrane (Bio-Rad) and hy-bridization to [a-32P]dATP oligonucleotide (Pharmacia LKB;catalogue no. 27-2166-01)-labeled probes (24). Total cellularRNA was isolated from cultures of transformed T. thermoph-ila cells (25) and Northern blot analysis was performed asdescribed by Williams and Mason (26) with the followingmodifications; after blotting to Zeta-Probe (Bio-Rad) themembrane was exposed for 30 sec to UV radiation. The neogene copy number and transcript levels were determined bydensitometric scan analysis of Southern and Northern auto-radiograms using a Javelin camera (model 602977) and theNational Institutes of Health IMAGE 1.43 software.

RESULTSTransformation of T. thermophila to Paromomycin-

Resistant Phenotypes by the neo Gene. The transformationvector, pH4neo, contains the Tn5 neo gene flanked by the 5'and 3' regions ofthe T. thermophila histone H4-I gene, the C3rDNA origin of replication, and pUC19 plasmid sequences(Fig. 1). When this vector or the EcoRI fragment containingonly the chimeric neo gene is microinjected into T. thermoph-ila, cells are transformed to pmr phenotypes. Table 1 lists themaximum concentrations of paromomycin in which eachtransformant can grow. All transformants will grow in 1 mgof paromomycin per ml, which is a concentration -30 timeshigher than wild-type cells will grow in. There is, however,substantial variation in paromomycin resistance (1-32 mg/ml) among the transformants.

Transformation with the Vector pll4neo. Fig. 2 showsSouthern analysis of 8 of 17 transformants (T10-T17) ob-tained by microinjection of the vector pH4neo. GenomicDNA digested with BamHI and hybridized with the neo gene

Table 1. neo gene configuration and paromomycin-resistancelevels of transformants

Transformant* (resistance)trDNA and Replacement

rDNA vector Replacement and flank FlankTi (32) T2 (8) LT2 (32) LT1 (8) LT4 (32)T5 (1) T3 (8) LT3 (2) LT5 (4) LT8 (2)T6 (1) T4 (16) LT7 (1) LT6 (32) LT9 (8)T9 (2) T7 (2) LT10 (1) LT12 (16) LT13 (1)T15 (2) T1l (4) LT11 (1) LT14 (4)

T12 (8) T8 (4)T13 (2) T10 (32)T14 (2)T16 (4)T17 (NA)

Genomic configuration of the neo gene in transformed cells:rDNA, integration into the rDNA genes; vector, high quantities ofpH4neo; replacement, replacement of the histone H4-I gene with theneo gene; flank, integration of the neo gene into the 5' flank of theH4-I gene.*T. thermophila, strain SB2120, was transformed with the vectorpH4neo (T) and the linear chimeric neo gene (LT). Wild-type cells(SB2120) are resistant to paromomycin (0.03 mg/ml) and cellstransformed with pmr C3rDNA are resistant to paromomycin (2mg/ml).

tResistance is defined as maximum concentration (mg/ml) of paro-momycin in which cells will double in a 24-hr period but will notgrow at twice these concentrations. Transformants were not chal-lenged with paromomycin concentrations >32 mg/ml. NA, data notavailable.

9296 Genetics: Kahn et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

6, 2

020

Proc. Natl. Acad. Sci. USA 90 (1993) 9297

AM V 10 11 12 13 14 15 16 17

6974-

2,3w2.0W.

BM

9.4.6.7.

4.4

2.3.2.0.

Cwildtype rDNA I

b 13.4 b

rDNA with 1 neo integration _

b 8.6 b 6.8 b 6.7 b

rONA with 2 neo integrations-_b 67 b 6.8 b 38 b 68 b 6.7 b

FIG. 2. Southern analysis of T. thermophila transformed with thevector pH4neo. Genomic DNA (-10 ,ug) isolated from transformedcells and 3 ng of vector DNA was digested with BamHI, separatedby electrophoresis on a 0.8% agarose gel, vacuum blotted onto anylon membrane, and hybridized to the appropriate probes. Lane V,vector DNA; lanes 10-17, DNA from transformants T10-T17, re-spectively. Marker lane (M) shows sizes in kb. (A) Southern analysesusing the neo gene coding region as a probe. (B) The same membranestripped and hybridized with the rDNA origin as a probe. Bandscorresponding to various DNA sources are indicated by letters on theright: E, endogenous rDNA molecules; S and D, rDNA moleculeswith single and double neo gene integrations, respectively; V,pH4neo vector. The faint band at -3.2 kb in lane 12 is of unknownorigin. (C) Schematic diagrams of the different alleles of the rDNAmolecules found in transformed cells. Symbols are the same as in Fig.1. Solid vertical bars, telomeric sequences; dashed line, center ofthepalindrome. BamHI sites are indicated by b and fragment lengths(kb) are shown between the sites.

as a probe resulted in a single band corresponding to a 6.8-kbfragment (Figs. 1 and 2A). The size of this fragment indicatesthat the neo gene and its flanking regions have not beenreorganized in these transformants (Fig. 2C). Although trans-formants T8 and T10 display substantial paromomycin resis-tance, the neo gene was not detected by this analysis (Fig.2A). However, when this filter was exposed to film for alonger time, a faint band was resolvable. Southern analysissimilar to that described below for the linear transformants

revealed that in transformants T8 and T10 the neo gene hasintegrated into the histone H4-I gene region (data not shown).Due to the high copy number of the rDNA molecules in T.

thermophila, the most probable site of integration of thevector is within the rDNA origin ofreplication. To investigatethe possibility that the vector integrated into the endogenousrDNA origin of replication, the same filter originally hybrid-ized to the neo gene was stripped and hybridized with theorigin as a probe. In wild-type cells the origin is on a 13.4-kbfragment (Fig. 2C). Each lane in Fig. 2B has a band (band E)corresponding to a fragment of this size; however, theirintensities vary among the transformants. Most of the laneshave additional bands (Fig. 2B, lanes 11-17). An integrationof the vector into one side of the palindromic rDNA wouldresult in origin sequences on 8.6- and 6.7-kb BamHI frag-ments (Fig. 2C). Bands S corresponding to fragments ofthesesizes are seen in Fig. 2B (lanes 11-17). A double integrationof the vector into both sides of the palindromic rDNA wouldresult in origin sequences on BamHI fragments of 6.7 and 3.8kb (Fig. 2C). Bands D corresponding to fragments of thesesizes are seen in Fig. 2B (lanes 11, 12, and 14-16). Thetransformation vector contains the rDNA origin of replica-tion on a 1.9-kb BamHI fragment (Fig. 1). A band (band V)corresponding to a fragment of this size is detected in Fig. 2B(lanes V, 12-14, 16, and 17).When T. thermophila is -transformed with the vector

pH4neo, the entire vector is usually found integrated into therDNA origin on one or both sides of the palindrome. Addi-tional Southern analyses with other endonucleases corrobo-rate these results (data not shown). In some, but not all,transformants the vector pH4neo is found in an unintegratedstate. However, none of the transformants contains only theunintegrated vector.

After transformant T2 had been grown for over a hundredgenerations, total DNA was prepared and used to transformEscherichia coli. A restriction enzyme map of the vectorpH4neo isolated from E. coli indicated that the vector had notbeen altered. Apparently, the vector can be replicated in T.thermophila without being reorganized.

Transformation with the Linear Chimeric neo Gene. The4.4-kb EcoRI fragment of the vector that contains only thechimeric neo gene (Fig. 1) was used to transform T. ther-mophila. Southern analysis of HindIII-digested genomicDNA from 9 of the 14 transformants (LT6-LT14) is shown inFig. 3. When the neo gene was used as a probe, a 6.2-kbfragment and a 9.5-kb fragment were detected (Fig. 3A).Lanes 7, 10, and 11 contain only the 6.2-kb fragment (bandR), while lanes 8, 9, and 13 contain only the 9.5-kb fragment

B7 8 9 10 11 12 13 14 W

cwildtype histone H4-1h h

[.M 6 7 8 9 10 11 12 13 14 W

F 4S

67.R.....

4 4ip:

78 h

H4-i

.F replacementh h h 6.2 h

I Il- t> ~ -L

neo

integration into 5' regionh h h

Cneo

9,5

H4-i

FIG. 3. Southern analysis of T. thermophila transformed with the linear chimeric neo gene. Genomic DNA isolated from transformed cells(w20 ,g) and wild-type cells (-30 ,ug) was digested with HindIII, separated by electrophoresis on a 0.8% agarose gel, vacuum blotted onto a

nylon membrane, and hybridized to the appropriate probes. Lanes 6-14, DNA from transformants LT6-LT14, respectively; lane W, DNA fromwild-type T. thermophila (SB2120). Marker lane (M) shows sizes in kb. Bands in lane 13 are less distinct due to degradation of the DNA. (A)Southern analysis using the neo gene coding region as a probe. Bands corresponding to various DNA sources are indicated by letters on theright: R, replacement of the histone H4-I gene by the neo gene; F, integration of the neo gene into the histone H4-I 5' flanking region. (B) Thesame membrane stripped and hybridized with histone H4 coding region as a probe. Bands corresponding to various DNA sources are indicatedby letters on the right: F, integration of the neo gene into the histone H4-I 5' flanking region; I, endogenous histone H4-I gene; II, endogenoushistone H4-II gene. (C) Schematic diagrams of different alleles of the histone H4-I genes found in transformed cells. Symbols are the same as

in Fig. 1 with the following additional symbols: boxes with arrows, histone H4-I genes; h*, HindIll site introduced by integration of the neo

gene. HindlIl sites are indicated by h and the fragment lengths (kb) are shown between the sites.

AM 6

6.7o..

4.4wh

[3'~~~~~~~~~~~~~~~~~~~~~~~~

-n

Genetics: Kahn et al.

-k

11

Dow

nloa

ded

by g

uest

on

Aug

ust 2

6, 2

020

Proc. Natl. Acad. Sci. USA 90 (1993)

(band F) and lanes 6, 12, and 14 contain both fragments. The6.2-kb band is consistent with the expected fragment sizeresulting from a replacement of the endogenous histone H4-Igene with the neo gene, while the 9.5-kb band suggests anintegration of the neo gene into the flanking region of thehistone H4-I gene (Fig. 3C).To further characterize the neo gene integrated into the

region of the histone H4-I gene, the filter was stripped andhybridized with the histone H4 coding region as a probe. TheH4 coding region hybridizes to fragments of 3.3, 7.8, and 9.5kb in transformants (Fig. 3B). All 9 transformants have the3.3-kb fragment (band II), which comigrates with the wild-typehistone H4-II gene fragment (Fig. 3B, lane W). All transform-ants contain various amounts of the 7.8-kb fragment (band I)(Fig. 3B, lane W), which comigrates with the wild-type histoneH4I gene fragment. The 9.5-kb fragment hybridizes with boththe neo gene (Fig. 3A, band F) and the histone H4 gene (Fig.3B, band F). The fact that the neo gene and the histone H4-Igene are linked on a single 9.5-kb Hindlll restriction fragmentindicates that the neo gene must be integrated into the 5'flanking region of the histone H4-I gene. If the neo gene wereintegrated into the 3' flanking region ofthe histone H4-I gene,the HindIII site at the beginning of the neo gene wouldseparate the genes onto different restriction fragments (Fig.3C, h*). Additional Southern blot analyses with other endo-nucleases corroborate these results (data not shown).Copy Number and Transcript Levels of the neo Gene in Both

Types of Transformants. The copy number of the neo genesand their relative transcript levels were determined for twocell lines (Ti and T2) transformed with the vector pH4neoand for two cell lines (LT2 and LT3) transformed with onlythe chimeric neo gene. Southern analysis was performed ongenomic DNA from transformants Ti and T2 digested withBamHI using the rDNA origin as a probe. The relativeamounts of native rDNA molecules, rDNA molecules withneo genes inserted (both single and double integrations), andvector molecules were determined from densitometric scananalysis. The copy number of each type of molecule wasestimated by assuming that the total number ofrDNA originsof replication is 20,000 (27). Table 2 shows the estimatednumber of neo gene copies present in Ti and T2 transform-ants.

In a similar experiment, Southern analysis was performedon genomic DNA from transformants LT2 and LT3 digestedwith HindIII and EcoRI using the histone H4-I 3' flankingregion as a probe. The relative amounts ofnative histone H4-Igenes and replacements of the histone H4-I by the neo genewere determined from densitometric scan analysis. In thesetransformants, virtually all of the native histone H4-I geneswere replaced by neo genes. Thus, the copy number of theneo genes in transformants LT2 and LT3 is assumed to beclose to 50 (Table 2; ref. 9).The level of the neo gene transcripts in different trans-

formants was estimated by normalizing the amount of neogene transcript to the amount of histone H4-II gene tran-script. Northern analysis was performed on RNA isolated

Table 2. neo gene copy number and transcript levelsin transformants

Copy Transcript Transcripts pmrTransformant no.* levelt per copy level

Ti 10,300 23,000 2.2 32T2 7,700 17,000 2.2 8LT2 50 2,700 54 32LT3 50 2,200 44 2

*Includes both single and double integrations into the rDNA in Ti

\- -oy- CJ F

FJi

FIG. 4. Northern analysis of transformants Ti, T2, LT2, andLT3. Total cellular RNA was isolated from these transformants asdescribed. Approximately 20 ug ofRNA from each transformant wasdenatured with glyoxal and separated by electrophoresis on a 1.5%agarose gel using 10 mM sodium phosphate buffer. RNA was thenvacuum blotted to a nylon membrane and hybridized to the neo geneas a probe.

from the transformants using the neo gene as a probe (Fig. 4).The same filter was then hybridized with the histone H4coding region as a probe. The relative amount of mRNAtranscripts from each gene was determined by densitometricscan analysis. For each transformant, the neo gene transcriptlevels were normalized to the transcript levels of the histoneH4-II gene. The relative amounts of neo gene transcriptsfrom Ti, T2, LT2, and LT3 are shown in Table 2.

DISCUSSION

When the neo gene under the control of the T. thermophilahistone H4-I promoter is introduced into T. thermophila, thecells are transformed to paromomycin-resistant phenotypes.All of the transformants display substantial levels of resis-tance compared to the wild-type cells; however, among thetransformants there is a remarkable variation in resistanceranging from 1 mg to >32 mg of paromomycin per ml. Thereare no obvious correlations between the paromomycin-resistance levels and the configuration of the neo gene withinthe genomes of the transformants. Transformants LT2 andLT3 have similar levels of neo gene mRNA; yet there is asignificant difference in the levels ofparomomycin resistance(Table 2). A similar situation exists for transformants Ti andT2. It is unclear precisely what produces the wide range ofparomomycin-resistance levels.The codon utilization in T. thermophila is highly skewed,

with some codons occurring very infrequently (28). Forexample, T. thermophila uses AGA to the virtual exclusionofthe other five codons specifying arginine. The neo gene has19 arginine residues, but none of them is specified by AGA.This dramatic mismatch in codon utilization may significantlydecrease the efficiency of translation, but it does not preventgene expression as evidenced by the high level of paromo-mycin resistance displayed by some of the transformants. Itis conceivable that the mismatch in codon utilization maycontribute to the variation in paromomycin resistances dis-played by different transformants.

In all of the transformants we have examined, the integra-tion of the neo gene into the genome appears to be mediatedby processes involving homologous recombination. Whentransformants are produced by injecting the vector pH4neo,recombination occurred between the rDNA origin on thevector and the origin on the native rDNA molecule in alltransformants except T8 and T10. These two transformantshave integrated into the genome via the homologous regionsof the 5' and 3' flanking sequences of the histone H4-I gene,which are similar in length to the rDNA origin sequence. Thefact that 15 of 17 transformants obtained with the vector havethe neo gene in the rDNA suggests that the vector is targetedto the nucleoli, where the rDNA molecules are located. Whenthe chimeric gene, as a linear DNA molecule, is used totransform cells, all integrations occur in the histone H4-I

and T2.tValues are normalized to the level of the endogenous histone H4-IIgene transcripts and expressed in arbitrary units.

9298 Genetics: Kahn et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

6, 2

020

Proc. Natl. Acad. Sci. USA 90 (1993) 9299

region. The neo gene replaces the histone H4-I gene orintegrates into the 5' flanking region. We have not observedintegrations into the 3' flanking region, but this may simplybe due to the limited number of transformants characterized.Our observation that the chimeric neo gene integrates into

the T. thermophila genome predominantly by homologousrecombination is consistent with previous reports of theintegration of modified rDNA molecules and ribosomal pro-tein genes into the T. thermophila genome (13, 15). Similarobservations have been reported in yeast and trypanosomes(29, 30). All these results are consistent with Capecchi'shypothesis (31) that homologous recombination is more prev-alent in lower eukaryotes because efficient mechanisms ofnonhomologous recombination are less common. Chlamy-domonas appears to be an exception to this trend (32).The number of neo gene transcripts per gene copy in

transformants with the neo gene integrated into the rDNA isonly 4-5% the number of neo gene transcripts per gene copyin transformants with the neo gene integrated into the histoneH4-I gene region (Table 2). This may be due to localizationof the neo genes on the rDNA molecules in the nucleoli.Transcription of the neo gene, which is under the control ofa RNA polymerase II promoter, may be inefficient in thenucleoli.We conclude that the neo gene is an excellent selectable

marker for the transformation of T. thermophila. Wild-typecells are sensitive to low levels of paromomycin and expres-sion of the neo gene confers high levels of resistance.Utilizing a foreign gene to transform T. thermophila avoidscompetition with endogenous wild-type gene products andthus simplifies the analysis of transformants. Coupling atransformation vector by using the neo gene as a selectablemarker with improved electrophoretic introduction of DNA(16) should allow development of a highly efficient transfor-mation system for T. thermophila and facilitate investigationsof gene expression.

This work was supported by grants from the National ScienceFoundation (BRS-8800805) and the Academic Senate of the Univer-sity of California.

1. Hershlag, D. & Cech, T. R. (1990) Nature (London) 344,405-409.

2. Yao, M. C., Zhu, S. G. & Yao, C.-H. (1985) Mol. Cell. Biol. 5,1260-1267.

3. Conover, R. K. & Brunk, C. F. (1986) Mol. Cell. Biol. 6,900-905.

4. Yu, G.-L., Bradley, J. D., Attardi, L. D. & Blackburn, E. H.(1990) Nature (London) 344, 126-132.

5. Orias, E. & Bruns, P. J. (1976) Methods Cell Biol. 13, 247-282.6. Brunk, C. F. & Sadler, L. A. (1990) Nucleic Acids Res. 18,

323-329.7. Stargeli, L. A., Karrer, K. M. & Gorovsky, M. A. (1990)

Nucleic Acids Res. 18, 6637-6639.8. Mayo, K. A. & Orias, E. (1981) Genetics 98, 747-762.9. Brunk, C. F. (1986) Int. Rev. Cytol. 99, 49-83.

10. Yao, M.-C. (1986) in The Molecular Biology of Ciliated Pro-tozoa, ed. Gall, J. G. (Academic, Orlando, FL), pp. 179-225.

11. Tondravi, M. M. & Yao, M.-C. (1986) Proc. Natl. Acad. Sci.USA 83, 4369-4373.

12. Brunk, C. F. & Navas, P. (1988) Exp. Cell Res. 174, 525-531.13. Yu, G.-L. & Blackburn, E. H. (1989) Proc. Natl. Acad. Sci.

USA 85, 5151-5155.14. Larson, D. D., Blackburn, E. H., Yeager, P. C. & Orias, E.

(1986) Cell 47, 229-240.15. Yao, M.-C. & Yao, C.-H. (1991) Proc. Natl. Acad. Sci. USA

88, 9493-9497.16. Gaertrig, J. & Gorovsky, M. (1992) Proc. Natl. Acad. Sci. USA

89, 9196-9200.17. Beck, E., Ludwig, G., Auersald, E. A., Reiss, B. & Schaller,

H. (1982) Gene 19, 327-336.18. Preer, J. R., Preer, L. B., Rudman, B. M. & Barnett, A. J.

(1985) Nature (London) 314, 188-190.19. Horowitz, S. & Gorovsky, M. A. (1985) Proc. Natl. Acad. Sci.

USA 82, 2452-2455.20. Bannon, G. A., Bowen, J. K., Yao, M.-C. & Gorovsky, M. A.

(1984) Nucleic Acids Res. 12, 1961-1975.21. Yu, G.-L. & Blackburn, E. H. (1989) Proc. Natl. Acad. Sci.

USA 86, 8487-8491.22. Brunk, C. F. & Hanawalt, P. C. (1967) Science 158, 663-664.23. Southern, E. (1975) J. Mol. Biol. 98, 503-517.24. Feinberg, A. P. & Vogelstein, B. (1984) Anal. Biochem. 137,

266-267.25. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D.,

Seidman, J. G., Smith, J. A. & Struhl, K. (1987) CurrentProtocols in Molecular Biology (Wiley/Greene, New York),Vol. 1, pp. 4.3.1-4.3.3.

26. Williams, J. G. & Mason, P. J. (1985) in Nucleic Acid Hybrid-isation: A Practical Approach, eds. Hames, B. D. & Higgins,S. J. (IRL, Washington, DC), pp. 139-159.

27. Yao, M.-C. (1982) in The Cell Nucleus, eds. Bush, H. &Rothblum, L. (Academic, New York), Vol. 12, pp. 127-153.

28. Martindale, D. W. (1989) J. Protozool. 36, 29-34.29. Hinnen, A., Hicks, J. B. & Fink, G. R. (1978) Proc. Natl.

Acad. Sci. USA 75, 1929-1933.30. Cruz, A. & Beverley, S. M. (1990) Nature (London) 348,

171-173.31. Cappechi, M. (1990) Nature (London) 348, 109.32. Mayfield, S. P. & Kindle, K. L. (1990) Proc. Natl. Acad. Sci.

USA 87, 2087-2091.

Genetics: Kahn et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

6, 2

020