Stable Transfection of the Human Parasite Leishmania Major Delineates a 30-Kilobase Region Sufficient for Extrachromosomal Replication and Expression GEOFFREY M

Home , Macronucleus

MOLECULAR AND CELLULAR BIOLOGY, Mar. 1990, p. 1084-1094 Vol. 10, No. 3 0270-7306/90/031084-11$02.00/0 Copyright © 1990, American Society for Microbiology Stable Transfection of the Human Parasite Leishmania major Delineates a 30-Kilobase Region Sufficient for Extrachromosomal Replication and Expression GEOFFREY M. KAPLER, CARA M. COBURN, AND STEPHEN M. BEVERLEY* Department ofBiological Chemistry and Molecular Pharmacology, Harvard Medical School, 250 Longwood Avenue, Boston, Massachusetts 02115 Received 29 August 1989/Accepted 16 November 1989

To delineate segments of the genome of the human protozoan parasite Leishmania major necessary for replication and expression, we developed a vector (pR-NEO) which can be reproducibly introduced into L. major. This DNA was derived from a 30-kilobase extrachromosomal amplified DNA bearing the dihydrofolate reductase-thymidylate synthase gene, with the coding region for neomycin phosphotransferase substituted for that of dihydrofolate reductase-thymidylate synthase and a bacterial origin of replication and selectable marker added. G418-resistant lines were obtained at high efficiency by electroporation of pR-NEO (approaching 10-4 per cell), while constructs bearing an inverted neo gene or lacking Leishmania sequences did not confer resistance. pR-NEO replicated in L. major and gave rise to correctly processed transcripts bearing the trans-spliced miniexon. Molecular karyotype analysis showed that in some lines pR-NEO DNA exists exclusively as an extrachromosomal circle, a finding supported by the rescue of intact pR-NEO after transformation of Escherichia coli. These data genetically localize all elements required in cis for DNA replication, transcription, and trans splicing to the Leishmania DNA contained within pR-NEO DNA and signal the advent of stable transfection methodology for addressing molecular phenomena in trypanosomatid parasites.

Trypanosomatid parasites are an important group of hu- that the circular DNAs contain all elements required in cis man pathogens that employ diverse mechanisms for survival for DNA replication and expression (6, 8, 16, 17, 27). and evasion of host defenses, thereby causing several severe However, it is possible that the amplified DNAs are tran- diseases. These protozoans exhibit numerous unusual mo- scriptionally silent but in some manner potentiate the lecular and biochemical properties, including obligatory expression of the chromosomal locus, perhaps by titration of trans splicing of a 39-base 5'-leader sequence onto protein- trans-acting factors. A similar effect in trans has been coding RNAs (11), antigenic variation mediated by gene observed in studies of the HSP70 locus in cultured mamma- rearrangement (11, 18), RNA editing of mitochondrial tran- lian cells (25). Moreover, owing to the wide variety ofunique scripts (38), polycistronic transcription and pre-mRNAs (24, mechanisms often employed by protozoans in storing and 33), and contain unique organelles such as the kinetoplast expressing genetic information, nonconventional possibili- (37) and the glycosome (34). These novel features have ties exist for the generation and maintenance of amplified attracted the interests of scientists from many diverse fields DNAs in Leishmania major. For example, the amplified and have frequently served as paradigms for the study of DNAs could be continually generated from a modified higher eucaryotes. Unfortunately, attempts to dissect both chromosomal locus. Examples of genetic systems involving the biological and molecular properties of trypanosomatid programmed gene amplification include the formation of the protozoa have been hindered because neither traditional macronucleus in ciliated protozoa (10, 12), ribosomal gene genetic analysis nor DNA transfection-based methods have amplification in amphibians (30), or chorion gene amplifica- been routinely available (11, 15, 18). tion in Drosophila melanogaster (26, 39). Thus, the function- One genetic approach that has been successfully utilized is ality of the amplified Leishmania DNAs was an open ques- the selection and analysis of mutants, especially in the tion. human parasitic genus Leishmania. These organisms can To address these issues and other questions concerning employ many diverse mechanisms of resistance when ex- trypanosomatid genetics, we have developed methods for posed to drugs such as the antifolate methotrexate (MTX), stable DNA transfection. We were encouraged by recent including specific gene amplifications (8, 16, 19; unpublished reports indicating that the related species Leptomonas sp. data). In most instances, the amplified DNAs are extrachro- and Leishmania enrietti were capable of transiently taking mosomal and circular and are present in up to 100 copies per up and expressing exogenous DNA constructs containing cell (6, 8, 17, 20, 35a). These DNAs can be maintained genomic sequences fused to the coding region of the chlor- indefinitely in the presence of drug pressure and for varying amphenicol acetyltransferase gene (3, 29). We used the amounts of time in the absence of selection. amplified dihydrofolate reductase-thymidylate synthase Because the circular amplified DNAs are maintained at (DHFR-TS) gene contained within the extrachromosomal high copy number and are associated with overexpression of circular amplified R region (6) as the starting point for the encoded RNAs and proteins, we and others have proposed construction of a DNA transfection vector, since a great deal is known about the protein-coding and transcriptional properties of this amplified DNA (7, 27; T. E. Ellenberger, Ph.D. * Corresponding author. thesis, Harvard University, Cambridge, Mass. 1989). How- 1084 VOL. 10, 1990 STABLE TRANSFECTION OF PARASITES 1085 ever, our previous studies have shown that transcriptional lated after digestion with SpeI and ligated with pR DNA organization of the amplified R region is complex, exhibiting which had previously been digested with SpeI and treated at least 10 RNAs including overlapping and antisense RNAs with phosphatase. The orientation of presumptive recombi- (27). Moreover, the locations of cis-acting elements required nants was confirmed by restriction mapping and DNA se- for transcription, processing, trans splicing, or replication quencing. pR-NEO carries the NPT-coding region in the have not been genetically defined for this or any other same orientation as DHFR-TS, while pR-OEN carries it in trypanosomatid gene. Accordingly, we sought to modify the the opposite orientation (Fig. 1B). amplified DNA minimally, assuming that these elements Cells. Leishmania is a digenetic trypanosomatid genus that were indeed contained within the amplified R region. In this resides as the spherical amastigote form within the phagoly- communication, we report the successful introduction of a sosome of vertebrate macrophages and is transmitted as the modified amplified DNA construct into Leishmania major flagellated extracellular promastigote form by phlebotomine and demonstrate its ability to function autonomously within sand fly vectors (44). All cells utilized were promastigotes the parasite. derived from the LT252 line of L. major (9). Cells were grown in M199 medium supplemented with 10% fetal bovine MATERIALS AND METHODS serum, 100 p.M adenine, 10 ,ug of heme per ml, 40 mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic Construction of pR-NEO. (i) Insertion of a SpeI site up- acid) (pH 7.4), 50 units of penicillin per ml, and 50 ,ug of stream of DHFR-TS. The 2-kilobase (kb) EcoRI-EcoRV streptomycin per ml. Clonal derivatives of this line and fragment from pLTS-S45 (7) containing the 1,560-base-pair transfected lines were obtained by plating on M199 medium- (bp) protein-coding region of the DHFR-TS gene of L. major 1% agar plates containing 0.6 ,ug of biopterin (Calbiochem- was excised and inserted into the EcoRI-EcoRV-digested Behring, La Jolla, Calif.) per ml and 16 ,ug of G418 per ml as vector pIC19H (31) and designated pRRv2O. The 2-kb indicated. Filter lifts of Leishmania colonies were made by EcoRI-HindIII fragment from this construct was then in- using a protocol developed by B. Sina in this laboratory serted into the EcoRI-HindIII sites of M13mpl9. Using the (unpublished data). Disks of nitrocellulose or Gene-Screen 32-mer (5') CGGGTCCGAGCACTAGTCGAAGATGTCC Plus membranes were pressed onto a plate containing colo- AGGG, an SpeI site was introduced 5 bp upstream of the nies; the filters were carefully peeled away and laid on filter initiation ATG codon of the DHFR-TS gene by the method paper soaked in 0.5 M NaOH-1.5 M NaCl for 10 min and of Taylor et al. (40), yielding the construct B1-Spe-9. Se- then on filter paper soaked in 0.5 M Tris-1.5 M NaCl (pH quencing confirmed the expected sequence from the EcoRI 7.4) for 10 min, and then on filter paper soaked in 2x SSPE to SpeI sites (see Fig. 7C). (lx SSPE is 180 mM NaCl, 10 mM NaPO4, and 1 mM (ii) Construction of the pR vector backbone. Plasmid pK300 EDTA; pH 7.0). The DNA was then fixed to the filter by containing the entire amplified R region (4) was digested with baking in vacuo or by UV cross-linking and subjected to EcoRV, ligated to the oligonucleotide CACTAGTG, cleaved hybridization analysis by standard methods. with SpeI, and recircularized, thus introducing an SpeI site Electroporation. Cells were grown to the late log phase at position 1737. This plasmid (pK300-Spe) was digested (0.5 x 107 to 1.0 x 107/ml), collected by centrifugation, and with EcoRI in the presence of 100 ,ug of ethidium bromide suspended at a density of 1 x 108/ml in HEPES-buffered per ml to obtain a population of singly-cut molecules, the saline (21 mM HEPES, 137 mM NaCl, 5 mM KCl, 0.7 mM ethidium bromide was removed, and the DNA was digested NaH2PO4, 6 mM glucose, pH 7.4 [36]) in an electroporation to completion with SpeI. Contour-clamped homogeneous cuvette (Bio-Rad Laboratories, Richmond, Calif.) and incu- electric field (CHEF) gel electrophoresis (2-s pulse time, 6 bated on ice for 10 min; 0.4 to 0.8 ml of cells was used. V/cm) was used preparatively to isolate the 30.6-kb SpeI- Chilled DNA (1 mg/ml in 10 mM Tris-1 mM EDTA, pH 7.4) EcoRI fragment, which was then ligated to the 0.2-kb was added, and then electroporation was immediately per- EcoRI-SpeI fragment from B1-Spe-9 and transformed into formed with a Bio-Rad Gene Pulser apparatus, using capac- Escherichia coli. The desired recombinant was designated itances and voltages as indicated below. After electropora- pK300-Spe-ADHFR. The pUC fragment of this plasmid was tion, cells were incubated for 10 min on ice and transferred moved from the KpnI site to a BglII site located at position to 10 ml of drug-free medium. Selection in G418 (geneticin; -18.7 kb by single-site digestion with BglII in the presence GIBCO Laboratories, Grand Island, N.Y.) was performed of ethidium bromide as described above, followed by diges- as described in the text. tion to completion with KpnI. The appropriate fragments Molecular karyotype analysis. Chromosomal DNAs were were preparatively isolated after CHEF gel electrophoresis prepared in low-melting-temperature agarose plugs, and and ligated together with BglII-cut and phosphatased pUCT chromosomes were separated by CHEF gel electrophoresis (pUC'r is a derivative of pUC8 in which the polylinker has (14) as described previously (4, 5). Gamma-irradiation anal- been replaced by EcoRI and HindIII digestion with the ysis was performed with a 'Co source as described previ- polylinker from the vector piAN7 [M. Fromm, personal ously (5, 35). communication]) to obtain the vector backbone designated RNA isolation and S1 nuclease and primer extension anal- pR. The lac promoter of the pUC vector is oriented in the ysis. Total cellular RNA was prepared by lysing parasites in same direction as DHFR-TS. 5 M guanidinium isothiocyanate followed by precipitation of (iii) Construction of pR-NEO and pR-OEN. SpeI sites were RNA with 4 M lithium chloride (13). Poly(A)+ RNA was added to the 0.9-kb MluI-BamHI fragment from pMClneo- prepared by chromatography on oligo(dT)-cellulose (1). POLYA containing the neo gene (Stratagene, La Jolla, Northern (RNA) blotting was performed as previously de- Calif.), and this fragment was inserted into the Bluescript scribed (27). Double-stranded DNA probes were radioac- vector SK- (Stratagene), yielding the plasmids pSpe-NEOA tively labeled with [a-32P]dCTP by random priming (20). and pSpe-NEOB (the neomycin phosphotransferase [NPT]- S1 nuclease mapping was performed as previously de- coding region is inserted in the same orientation from the lac scribed (28) with 2,ug of total cellular poly(A)+ RNA and the promoter in pSpe-NEOA and in the opposite orientation in 597-bp NarI-XhoI DNA restriction fragment shown in Fig. pSpe-NEOB). The 0.9-kb neo resistance cassette was iso- 7B and labeled at the Narl 5' end with T4 polynucleotide 1086 KAPLER ET AL. MOL. CELL. BIOL.

0 -8 -16

A 2.3 2:8 DHFR-TS 13 1.7s 6.2s 232u8 _V*.~~~okr~ -T..*~*-~ R 0 region B, 82 \BI K B3 04 'I \ -r---l I RiOOO I I D7B-RiOOO \ I\ SpeI B V pR vector B2 A K 83 B4a B4b 4- NEO -OEN FIG. 1. DHFR-TS gene and R region of L. major and design of pR-NEO. (A) Map and features of the chromosomal R region. The restriction map for the chromosomal R region is adapted from reference 6. The open box represents the coding region of the DHFR-TS enzyme (7), and the hatched bars correspond to the locations of repeated sequences that recombine to form the R-region amplification in the R1000-3 and other MTX-resistant lines (6; unpublished data). The locations of BgIII (B) and KpnI (K) restriction sites are shown, and the map positions (in kilobases) relative to the start codon of the DHFR-TS gene are shown at the top of the figure. Above the restriction map, the locations of poly(A)+ RNAs transcribed from this region are shown by wavy lines, with arrows denoting the direction of transcription (27). Below the restriction map are shown the segments amplified in the R1000-3 and other amplified lines (R1000 [6, 8, 21; unpublished data]) and the D7B-R1000 line (D7B-R1000 [8, 23]). (B) Map of the pR vector backbone and pR-NEO and pR-OEN derivatives. The map of the circular pR vector backbone is shown, linearized at the hatched boxes. pR contains all the segments between the hatched boxes in the map of panel A, except that (i) the DHFR-TS-coding region has been replaced by an SpeI restriction site, juxtaposing bases -5 and + 1771, and (ii) a pUC sequence has been inserted into the BgII site B4, giving rise to an additional BglII site (the orientation of the lac promoter of the pUC vector is the same as DHFR-TS). In the plasmid pR-NEO, the neo gene cassette from pSpe-NEO was inserted into the SpeI site, with the NPT-coding region in the same orientation as DHFR-TS; in pR-OEN, the NPT-coding region is in the opposite orientation. The neo gene cassette derives from pMClneo-POLYA (Stratagene) (41) and contains 34 bp of the HSV tk 5' region (lacking the TATA element [32]), an 18-bp synthetic translation initiation sequence fused to the NPT-coding region (2). and 55 bp containing a functional HSV tk polyadenylation sequence (45). See the Materials and Methods for details of these manipulations. kinase and [-y-32P]ATP. Protected fragments were separated tion consisting of 39 cycles of 2 min at 55°C, 2 min at 72°C, on 4% acrylamide-8 M urea gels. Primer extension was and 1 min at 94°C, followed by 1 cycle of 2 min at 55°C and performed with 2 Kg of poly(A)+ RNA and annealed with 5 7 min at 72°C in an Ericomp programmable cycle reactor. pmol of oligonucleotide primer (labeled at the 5' end with T4 The smaller Neo-DHFR-TS PCR product was recovered polynucleotide kinase) for 3 h at 55°C in 20 .l of 80 mM from a 3% Nusieve agarose-1% ME agarose gel (FMC KCl-80 mM Tris (pH 8.3). The volume was adjusted to 50 ,Il, Corp., Philadelphia, Pa.), incubated with T4 DNA polymer- final concentrations of 50 mM Tris (pH 8.3), 50 mM KCl, 6 ase in the presence of all four deoxynucleotide triphos- mM MgCl2, 10 mM dithiothreitol, 500 p.M each dATP, phates, digested with BamHI, and cloned into the BamHI dGTP, dCTP, and dTTP, 80 jig of actinomycin D per ml, and site of M13mpl9, and the sequence was determined by the 16 U of avian myelobastosis virus reverse transcriptase dideoxy method. The 5'-end sequence of the normal DHFR- (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) TS gene was similarly determined, using the oligonucleotide were added, and the mixture was incubated for 1 h at 42°C. GCCGTGCTGCATATCGAG (nucleotide positions 114 to The reaction mixture was extracted once with a 1:1 phenol- 97) in conjunction with the miniexon primer. chloroform mixture and precipitated with ethanol. The pre- cipitate was suspended in 6 [l of 10 mM Tris (pH 7.4)-10 RESULTS mM EDTA, incubated with 50 jig of RNase A per ml for 10 Strategy for design of transfection vector pR-NEO. Most min at 67°C, and subjected to electrophoresis as for Si DHFR-TS amplifications encompass a specific 30-kb seg- analysis. The oligonucleotide primer GGGAATTCGGATC/ ment of chromosomal DNA (the R region, map positions CATCAGAGCAGCCGATTGTCTG contained a 13-nucleo- +8.5 to -21.5 kb; Fig. 1A) (6, 8, 21; unpublished data) which tide 5' extension encoding EcoRI and BamHI restriction is rearranged and joined to form an extrachromosomal sites and 22 nucleotides of NPT-coding sequence (nucleotide circular DNA. In some lines, different rearrangement sites positions 108 to 87; the 5' end of this oligonucleotide are utilized, which can be interpreted as delimiting a core corresponds to position +121 of pR-NEO). region required for DHFR-TS expression and maintenance PCR amplification. Primer extension was performed as as a circular DNA. Specifically, the D7B-R1000 line (8, 23) described above, except that the oligonucleotide used was lacked the segment of DNA between positions -16 and not radiolabeled. cDNA from 0.2 jig of poly(A)+ RNA was -21.5 kb of the prototypic R region (Fig. 1A). Accordingly, subjected to polymerase chain reaction (PCR) amplification we inserted a pUC vector in this region (Fig. 1B). Assuming with the Neo oligonucleotide described above and the oligo- that the DHFR-TS coding region only provided protein- nucleotide GGAATTCGGATCC/AACGCTATATAAGTTA coding information, we replaced it with a unique SpeI TCAG, which contains a 14-nucleotide extension with restriction site, yielding the vector backbone pR (Fig. 1B). EcoRI and BamHI sites and nucleotides 5 to 23 of the L. The neo gene encoding NPT, which confers resistance to major miniexon (J. Miller, submitted for publication). Am- aminoglycosides such as G418 (2), was inserted into the SpeI plification was performed with 2 U of Taq polymerase site of pR in either the DHFR-TS orientation (pR-NEO) or (Perkin-Elmer-Cetus, Norwalk, Conn.) in a 40-cycle reac- the opposite orientation (pR-OEN; Fig. 1B). VOL. 10, 1990 STABLE TRANSFECTION OF PARASITES 1087

Transfection of pR-NEO: G418 resistance. Electroporation electroporations of pR-NEO DNA into L. major, all of of DNAs into L. major was performed as described in the which successfully yielded G418-resistant lines. Materials and Methods, using a Bio-Rad Gene Pulser appa- pR-NEO DNA in transfected L. major. We discuss here the ratus, 4 x 107 to 8 x 107 cells suspended in HEPES-buffered results obtained from four independent experiments trans- saline, voltage gradients of 2,250 to 3,500 V/cm, and capac- fecting pR-NEO: experiment 1, with the LT252 line, 25 ,uF, itances of 25 to 500 ,uF. These conditions were selected 3,000 V/cm, and selection protocol A; experiments 2 and 3, because they resulted in significant cell lethality, 50% or under standard electroporation conditions and with selection greater, which from studies in other organisms signals con- protocol B; and experiment 2-25, with 25 ,uF, 3,500 V/cm, ditions appropriate for the introduction of foreign DNAs and selection protocol B; the last three experiments em- (36). ployed the clonal CC-1 line as the recipient. The lines arising Following electroporation, 1 day was allowed for expres- from these transfections are referred to by the experiment, sion of the resistance gene before imposing drug selection. whether they are a population of cells (pool) or clonal As the DHFR-TS mRNA is not abundant (27, 28, 42) and derivatives obtained by plating, and in parentheses is found transfected genes frequently are expressed less well than the final G418 concentration (micrograms per milliliter) in endogenous genes, we anticipated low levels of expression which they were propagated, e.g., E1-pool(8) or E2-25- of the Neo-DHFR-TS hybrid mRNA. We employed a grad- B1(8). ual stepwise selection, beginning at 2 ,ug of G418 per ml and Southern blot hybridization of BglII-digested DNAs from progressing through three 1:10 dilutions of cells into 4 ,ug/ml transfectant lines with a neo-specific probe identified the and then 8 ,ug/ml (protocol A). The gradual nature of this expected 9.4-kb fragment only in cells transfected with selection is shown by the facts that the 50% effective pR-NEO and growing in G418 (Fig. 3A; the results of five concentration for G418 action was 2 ,ug/ml and that mock- representative clonal lines are shown). Hybridization of the transfected cells continued to grow in this selection protocol same blot with a pR-NEO probe (Fig. 3B) revealed frag- until the second passage in 8 ,ug of G418 per ml. In contrast, ments unique to the endogenous R region (7 kb and two pR-NEO recipients could be maintained indefinitely in 8 ,ug poorly resolved larger fragments) and fragments shared by of G418 per ml. In later experiments, we found that electro- both pR-NEO and the endogenous R region (3.4 and 12.5 porated cells could be passaged directly at a 1:100 dilution kb), with all these fragments being present in all lines. into 8 ,ug of G418 per ml (protocol B), a protocol that mock Additionally, the pR-NEO-transfected lines exhibited novel or control plasmid-transfected cells could not survive for one fragments derived exclusively from pR-NEO (9.4, 3.8, and passage. Although all electroporations with the ranges of 2.7 kb). Scanning densitometry of lines grown in 8 ,ug of voltage and capacitance listed above yielded G418-resistant G418 per ml revealed that the pR-NEO copy number was L. major, 2,250 V/cm and 500 ,uF yielded initial cultures that about 2 relative to the endogenous DHFR-TS gene in lines grew and attained stationary phase more rapidly, and we derived by selection protocol A and about 10 to 12 in lines refer to these as our standard electroporation conditions. derived from three different experiments by selection proto- Electroporation under the standard conditions followed by col B; the copy number was the same in pools and clonal selection with protocol B with plasmids bearing only the neo derivatives from the same experiment. gene cassette inserted in the E. coli plasmid vector in either It is surprising that the copy number of pR-NEO obtained orientation (pSpeNEOA, pSpeNEOB), pR, or pR-OEN in lines derived by protocol A in experiment 1 was about failed to yield G418-resistant lines. Mixed transfections of fivefold less than those derived by protocol B in experiments these DNA preparations in combination with pR-NEO DNA 2 and 3, since all lines were grown in the same final drug successfully yielded G418-resistant lines, showing that the concentration (8 ,ug/ml). The pooled samples were analyzed nonfunctional DNA preparations were not themselves inhib- as soon as possible after growth in G418 was established, itory. These data indicate that G418 resistance after electro- suggesting that the difference is a primary event and must poration requires Leishmania sequences and a correctly therefore arise from differences in the electroporation con- oriented neo gene. ditions or between gradual versus one-step selection proto- Quantitative plating studies of a clonal derivative of the cols. This difference in copy number was also maintained LT252 line (CC-1) electroporated in the presence of various through two rounds of cloning on agar plates (approximately amounts of pR-NEO DNA and grown for 1 day in drug-free 50 cell doublings). If the results with protocol A are inter- medium revealed that many independent G418-resistant col- preted to mean that only two copies of pR-NEO are required onies were obtained (Fig. 2A and C); no colonies were to confer drug resistance, these observations suggest that the observed in platings of mock- or control-transfected cells transfectants derived by protocol B have maintained excess (Fig. 2B). The colonies obtained were shown to be L. major copies of pR-NEO during this period. It remains to be by subculture and microscopic examination, and filter lift determined whether the copy number will decline over a hybridizations showed the presence of neo-specific se- prolonged period, however. quences in each colony (data not shown). Transfectants In mixed transfections, e.g., pR-NEO plus pR-OEN, only were obtained at efficiencies of up to 62 colonies per ,ug of pR-NEO-specific sequences were revealed in the G418- pR-NEO DNA, with up to 1,150 colonies recovered per plate resistant lines by Southern blotting with appropriate en- (Fig. 2D); this latter value corresponds to a total cellular zymes and probes (data not shown). This provided no transfection frequency of 2.8 x 10' (7 x 10' for cells evidence for stable cotransfection. surviving electroporation). The efficiency of transfection To determine whether the copy number of pR-NEO could was reduced fourfold at higher DNA levels but was not be increased, we selected the E1-pool(8) line for resistance saturated at the highest DNA level tested (88 ,ug). The shape to 128 and then 500 ,ug of G418 per ml. Analysis of these lines of this curve suggests that there are two different routes for by Southern blotting and densitometry revealed that the transfection, one of higher efficiency and saturable; whether copy number of pR-NEO increased from 2 in the El-pool(8) this reflects different routes of DNA entry or subpopulations line to 20 to 25 in the El-pool(128) line and to 40 to 50 in of cells (perhaps due to the cell cycle) is currently unknown. E1-pool(500) line (Fig. 3C), showing that cells bearing ele- To date, we have performed more than 50 independent vated copy numbers can be readily obtained. We did not 1088 KAPLER ET AL. MOL. CELL. BIOL.

A B

C D

/ /: ~~~~/I o-

A A, -1 - 11 14 t. "-.. FIG. 2. Plating assay of transfection of L. major. Cells (4 x 107) of the CC-1 line of L. major LT252 were electroporated with various amounts of pR or pR-NEO DNA as indicated (500 ,uF, 2,250 V/cm), grown for 1 day in drug-free medium, and plated on M199 agar plates containing 16 pLg of G418 per ml. (A) pR-NEO DNA (88 ,ug). Heterogeneity in colony size is commonly observed in platings of wild-type cells on drug-free plates (data not shown). (B) pR DNA (50 ,ug). (C) pR-NEO DNA (11 ,ug). (D) Quantitative plating efficiency. Shown are numbers of colonies and the calculated transfection efficiency as a function of the amount of pR-NEO DNA electroporated. The means of three replicate electroporations for each DNA amount are shown, with standard deviations of colony numbers being 5% or less of the mean. determine whether the increase in copy number results from which are also sensitive to the dam methylation (data not selection of cells that initially received high numbers of shown). These findings are presumptive evidence for repli- pR-NEO during electroporation or from subsequent ampli- cation of pR-NEO within L. major. fication of pR-NEO within a cell. Extrachromosomal localization of pR-NEO in transfectants. Replication of transfected DNA. Initial data indicated that Separation of transfectant chromosomes by pulsed-field the pR-NEO sequences could be maintained within trans- electrophoresis followed by Southern blot hybridization with fe¢ted L. major for many cell doublings in the presence of a neo-specific probe revealed the presence of DNAs exhib- G418, suggesting that pR-NEO was replicating. To test this, iting pulse-time-dependent mobility and migration outside we examined the methylation state of the pR-NEO DNA the apparent migration path of the linear chromosomes, two within L. major, since DNA in E. coli is methylated at the A hallmarks of circular DNA (4, 22, 23). These circular DNAs of GATC sequences by the dam methylase, whereas Leish- such as the mania DNA is unmethylated (unpublished data). Replication were seen only in the pR-NEO-transfected lines of the transfected DNA was assayed with restriction enzyme clonal derivatives E2-A4(8), E2-A6(8), E3-C4(8), and E3- isoschizomers that are differentially sensitive to this modifi- D1(8) (Fig. SA; the circular DNA is marked by an arrow; cation: DpnI, requiring methylated DNA, and MboI, requir- data for other pulse times are not shown). The patterns for ing unmethylated DNA. As expected, the input transfecting lines derived from experiment 1 were more complex and will pR-NEO DNA was sensitive to DpnI and resistant to MboI, be discussed elsewhere. Weak hybridization was also ob- while the total genomic DNA of the E1-pool(128) line was served to the sample well, probably reflecting the occurrence resistant to DpnI and sensitive to MboI (Fig. 4A). Southern of nicked or relaxed circular DNAs which are trapped in the blot hybridization with a neo-specific probe revealed that the well in pulsed-field electrophoresis (4), but no other hybrid- pR-NEO DNA within the E1-pool(128) line uniformly ization was evident (Fig. SA). This suggested that integration showed the same sensitivity as total genomic DNA and thus of pR-NEO into a chromosome had not occurred. Hybrid- was unmethylated (Fig. 4B). Similar data were obtained with ization with a pR-NEO probe similarly identified the circular other transfected lines and the enzymes BclI and XbaI, DNAs in addition to the 500-kb wild-type DHFR-TS chro- VOL. 10, 1990 STABLE TRANSFECTION OF PARASITES 1089

A ¢ z ¢ ¢ o C El-Pool Tv ¢wo3B~,26~~~~~~~~ ~ ~ c~~ (G418,ug/mr

9.3-23-Lt ~ ~~~~~~~~~~9qUCLUN2 N.< 0W L l,.L;V *LeN<- 3z I 3.

93__ _* ~~~~9.4-- _94 .@-1~~~~~~~2.54~~* ~~-6.811 4.3- _ _-328- 4 40 0 @-3.4- _ _ v -_ 2 3- 20- Probe: NEO pR-NEO 4 QR-NC-O FIG. 3. Southern blot analysis of transfected L. major. Genomic DNAs from the indicated_2~~7.-lines were digested with BglII and separated on 0.8% agarose gels; pR-NEO refers to the input plasmid DNA. Molecular weight markers (in kilobases) are shown on the left. (A) neo-specific hybridization probe (0.9-kb NPT-containing fragment purified from pSpe-NEOB; see Materials and Methods). Faint hybridization is seen to the 2.7-kb Bluescript vector sequence owing to contamination of the neo fragment. (B) pR-NEO hybridization probe (entire pR-NEO plasmid, depicted in Fig. 1B; same blot as in panel A). (C) Amplification of pR-NEO sequences. E1-pool(8) was selected for growth in 128 and then 500 ,ig of G418 per ml, and the DNA was analyzed as in panels A and B. The pR-NEO hybridization probe was used.

mosome, whose size and levels were the same in the sis followed by Southern blot hybridization with a neo- transfectant and parental lines (Fig. SB). specific probe revealed discrete linear DNAs whose appear- To confirm the presence and measure the size of the ance depended on irradiation (Fig. 5C; compare with Fig. circular pR-NEO DNA within the transfected L. major, we SA). In most lines, the circular DNAs detected by this used gamma-irradiation analysis to introduce double-strand approach were of the same length as the input pR-NEO breaks and thereby form discrete linear DNAs (5, 35, 43). DNA, 32 kb; however, in several lines, such as the E3- Separation of irradiated DNAs by pulsed-field electrophore- pool(8), larger DNAs were observed, corresponding to oli- gomers of the 32-kb unit (Fig. SC). Irradiated DNA from clonal derivatives of this pool revealed unit-length DNAs, A B showing that this pool was a heterogeneous population with Ei-P oo respect to oligomer length (Fig. SC). CI28) X PR- NJE0 W1 3 " The presence of circular DNAs was also indicated by M: D 'H3 MA D ' _ . -s reintroduction of transfected pR-NEO DNA back into E. sz;.-. [-24- coli. Chromosomal DNA preparations from several transfectants were transformed into E. coli, yielding ampicillin- -4,3- resistant colonies which contained intact pR-NEO DNA (data not shown). This additionally demonstrated the ability to shuttle pR-NEO between species. -2.0- Transfected pR-NEO is transcribed and gives rise to a correctly processed hybrid Neo-DHFR-TS mRNA. Northern blot analysis with a neo-specific probe identified a 2.4-kb RNA in poly(A)+ RNA prepared from the E1-pool(128) line (Fig. 6A, lane 2); no hybridization was observed to RNAs prepared from the wild-type (data not shown) or the MTX- resistant R1000 (Fig. 6A, lane 1) lines. Significantly, 2.4 kb is the size predicted for a hybrid Neo-DHFR-TS mRNA, assuming use of the processing sites of the normal 3.2-kb DHFR-TS mRNA (Fig. 6B). Correspondingly, a hybridization probe specific for the 3' nontranslated region of the wild-type DHFR-TS mRNA (probe 3'NT in Fig. 6B) identified the new 2.4-kb mRNA in the E1-pool(128) line and the endogenous 3.2-kb wild-type DHFR-TS mRNA in both lines Pr5-t- NE. (Fig. 6A, lanes 3 and 4). FIG. 4. Replication of pR-NEO in L. major. (A) Genomic DNA Since the 3'NT probe has comparable extents of homology from the E1-pool(128) line or pR-NEO plasmid DNA was digested with MboI (M) or Dpnl (D) and separated on a 2% agarose gel. to both the hybrid Neo-DHFR-TS and endogenous DHFR- Molecular weight markers (in kilobases) are shown on the side. (B) TS mRNAs, densitometric scanning was used to quantitate Southern blot analysis of the E1-pool(128) lanes from panel A with the relative abundance of these RNAs. The 2.4-kb Neo- the neo-specific probe (see the legend to Fig. 3A). DHFR-TS mRNA was only 50% as abundant as the 3.2-kb 1090 KAPLER ET AL. MOL. CELL. BIOL.

E5 ;., .& a5 a. - 30 .,Z, 77, 1. ,, 7 'z2l < < u 01- f 8 CN- -, -, U.- " LLi ,1-1 _j W Llj U Ad 6- '- C 40 40 do 0 --.:-i,: . in A& 4m 4. 4b Or:igin .-

--m-

I 4 qi

v . __

Probe: -w- Nw~F FIG. 5. Molecular karyotype analysis. (A) Chromosomes of the indicated lines were separated by pulsed-field electrophoresis with a CHEF apparatus (14), using a pulse time of 160 s for 16 h followed by a pulse time of 80 s for 16 h. The gel was blotted and hybridized with the neo-specific probe (see the legend to Fig. 3A). Molecular weight markers (kilobases) derive from polymers of lambda DNA. The arrows mark the positions of the circular pR-NEO DNAs. (B) Southern hybridization of the blot shown in panel A with the pR-NEO probe. (C) Gamma-irradiation analysis. Chromosome preparations from the indicated lines as well as pR-NEO DNA were irradiated with 100 krads and separated with a CHEF apparatus, employing a pulse time of40 s for 20 h followed by 20 s for 16 h. The positions of the lambda ladder markers (48.5-kb intervals) are shown on the left; unit corresponds to the monomeric 32-kb pR-NEO DNA.

DHFR-TS mRNA in the clonal E1-B1(8) and E1-C6(8) lines with E1-pool(128) but not R1000 RNA (Fig. 7A, panel PE), (Fig. 6A, lanes 5 and 6). In contrast, the Neo-DHFR-TS with the smaller product being more abundant. These prod- mRNA was fivefold more abundant than the DHFR-TS ucts map to positions approximately 40 nucleotides beyond mRNA in the E1-pool(128) line (Fig. 6A, lane 3). As dis- those identified by Si nuclease analysis (Fig. 7B), suggesting cussed earlier, the pR-NEO DNA was 2-fold amplified in the that these RNAs contain an additional 5' extension corre- E1-B1(8) and E1-C6(8) lines and 20-fold amplified in the sponding to the miniexon sequence of L. major. Accord- E1-pool(128) line. These data show that the relative abun- ingly, PCR amplification of cDNA derived from RNA of the dances of the hybrid Neo-DHFR-TS mRNA per pR-NEO E1-pool(128) line with the neo-specific oligonucleotide used gene copy (0.25) were comparable to that observed for in the primer extension analysis (Fig. 7A) and an oligonucle- DHFR-TS mRNA (0.19) in the R1000-3 line (15-fold mRNA otide from the L. major miniexon sequence yielded products increase, 80-fold R-region amplification [16, 27, 28]), al- whose sizes agreed with that obtained by primer extension though the R1000 line exhibits less mRNA per DHFR-TS (data not shown). DNA sequencing of the smaller PCR gene copy than the wild-type line. product showed that it contained the miniexon sequence and We next determined whether the hybrid Neo-DHFR-TS the 5' nontranslated region of the DHFR-TS gene fused to mRNA was correctly processed at the 5' end. SI nuclease the NPT-coding sequence as expected and that the site of analysis with a probe specifically labeled within neo-derived addition of the miniexon was identical for the Neo-DHFR- sequences revealed protected fragments of 405 and 530 TS hybrid and wild-type DHFR-TS mRNAs (Fig. 7C). In nucleotides from E1-pool(128) poly(A)+ RNA (Fig. 7A, total, these data demonstrated transcription of the trans- panel SI; see Fig. 7B for a map of this region and the location fected pR-NEO DNA and correct processing of a hybrid of Si probes), while no protected fragments were observed Neo-DHFR-TS mRNA, employing the sites present in pR- with R1000 poly(A)+ RNA (data not shown). These products NEO that are normally used by the endogenous DHFR-TS map to positions -267 and -392 in pR-NEO, which corre- gene. spond to positions -231 and -356 in the endogenous DHFR- Northern blot analysis with a pUC plasmid DNA probe TS gene and agree with those previously mapped for the identified in transfected lines three RNAs of 5.0, 5.5, and 9.0 normal DHFR-TS mRNA, positions -245 and -375 (Fig. kb (Fig. 6A, lane 8) that were absent in wild-type (data not 7B) (28). In both the DHFR-TS and hybrid Neo-DHFR-TS shown) and R1000 (Fig. 6A, lane 7) lines; the region into mRNAs, the smaller Si products were quantitatively more which the pUC DNA was inserted is normally transcribed abundant (Fig. 7A and reference 28). Primer extension into two 6.2-kb opposite-strand poly(A)+ RNAs (Fig. 1A) analysis with an oligonucleotide located within the neo (27). This suggested that transcription was occurring across sequence yielded two products of 430 and 560 nucleotides the inserted pUC sequences, since the size of the largest of VOL. 10, 1990 STABLE TRANSFECTION OF PARASITES 1091

A RNAs transcribed from regions shared by both the R region x and pR-NEO were observed (13 kb, 1.7 kb, and the down- 7- - ccc stream RNAs [27]), other than an elevation in level in the C),1 a- 0Lo O E1-pool(128) line consistent with that observed for the 1 7xr W_ ! Neo-DHFR-TS hybrid mRNA described above.

DISCUSSION -9- We showed that a molecular construct (pR-NEO) derived .~~~~- from the amplified circular DHFR-TS gene of MTX-resistant L. major can be successfully introduced into wild-type cells by electroporation, thereby conferring G418 resistance. Quantitative plating studies demonstrated that the efficiency 32.21- .~ ~ ~ of transfection of this 32-kb DNA approaches 10-4 per cell, 2. which compares well with efficiencies obtained in electroporations of cultured mammalian cells (36). It seems likely that -2.4- 6 the efficiency can be increased since we have not systemat- ically optimized these protocols. Given the high transfection efficiency of pR-NEO, it is surprising that our previous ut-~~~~~7 attempts to introduce the native circular R-region amplified Frof/tM N1 EO 3 NT pUC 6A 6B DNA into Leishmania species by electroporation followed by MTX selection did not succeed, since similar protocols succeed in cultured mammalian cells (unpublished data). This may be due to peculiarities of the effects of MTX on Leishmania species undergoing drug selection or to the B possible existence of targets other than DHFR-TS (Ellen- 90kb berger, Ph.D. thesis). To date, we have maintained pR-NEO in cells under G418

1-- -- ..4 pressure for more than 200 cell doublings; in these lines, the 81 B2 Neo B4, pUCir Be pR-NEO gene copy number remained at similar levels (unpublished data), suggesting that dilution of nonreplicating 6A pUCir 6B pR-NEO DNA was not taking place. Correspondingly, ,NEO methylation analysis confirmed that pR-NEO replicates FIG. 6. Northern blot analysis of RNAs from transfected L. within L. major. In several lines, pR-NEO is maintained major. (A) Poly(A)+ RNA (100 ng) from the indicated cell lines was analyzed by Northern blot analysis with the probes shown in panel exclusively as an extrachromosomal circular DNA, and in B. Molecular weight markers (kilobases) consisted of mouse rRNA. these there is no evidence for integration into any chromo- (B) Description of probes and hybrid RNAs. An abbreviated restric- some, including DHFR-TS gene which contains more than tion map of the plasmid pR-NEO is shown, with the locations of the 29 kb of sequence homology with pR-NEO (Fig. 1). It is probes used in panel A indicated below the map. NEO, 0.9-kb therefore not surprising that we did not observe cotransfec- neo-containing SpeI fragment from pSpe-NEOB; 3'NT, 1.3-kb SphI tion of nonfunctional molecular constructs when mixed with fragment within the 3'-noncoding region of the DHFR-TS mRNA pR-NEO, as stable cotransfection in cultured mammalian (located at positions 1.6 to 2.9 kb), 1.2 kb of which is retained in the cells requires covalent joining of the cotransfected DNAs plasmid pR-NEO as shown; pUC, pUC plasmid used in construction followed by integration into a chromosome. of pR-NEO; 6A, 1.1-kb PstI fragment (positions -16.2 to -17.3 kb [27]), 6B, 4.8-kb BglII-SalI fragment (positions -18.7 to -23.5 kb); Current data indicate that transcription from pR-NEO 2.8 kb of this probe is present in the amplified R region of the R1000 occurs in a manner that accurately reflects normal transcrip- line and pR-NEO, with the remainder corresponding to nonampli- tional processes. First, successful transfection depended on fied flanking sequences. B, BgIII (see Fig. 1). Hatched boxes Leishmania DNA sequences and required that the neo gene correspond to segments joined to form the circular amplified R within the pR backbone be in the same orientation as region in the R1000 line (Fig. 1). Arrows above the map denote the DHFR-TS. Second, pR-NEO-transfected L. major ex- predicted hybrid RNAs (Neo-DHFR-TS mRNA, 2.4 kb; pUC-6.2, pressed a hybrid Neo-DHFR-TS mRNA whose size corre- 9.0 kb). sponded to that predicted for an RNA that utilizes the normal DHFR-TS-processing sites present in pR-NEO. Ac- these RNAs corresponded to the predicted size for a hybrid cordingly, Si nuclease protection, primer extension, and RNA containing both pUC and the 6.2-kb RNA sequences sequence analysis of cDNAs indicated that the Leishmania (Fig. 6B). Hybridization with a probe located to the left of miniexon sequence was added to the 5' end of the hybrid the pUC insertion site (probe 6A, Fig. 6B) identified three Neo-DHFR-TS RNA, at precisely the sites utilized in the RNAs of 2.8, 6.2, and 9 kb (Fig. 6A, lane 9), while a probe normal DHFR-TS transcript (28). The accuracy of this located to the right of the pUC sequences (probe 6B, Fig. process is emphasized by the fact that both miniexon addi- 6B) identified four RNAs of 5, 5.5, 6.2, and 9 kb (Fig. 6A, tion sites were utilized in relative proportions similar to that lane 10). The location and sizes of the 5-, 5.5-, and 2.8-kb seen for the normal DHFR-TS gene (28). Third, preliminary RNAs relative to the normal pattern of transcripts from this studies of a construct bearing the chloramphenicol acetyl- region suggests the possibility that these RNAs arise from transferase gene correctly inserted into the SpeI site of the the introduction of a signal in the vicinity of the left end of plasmid pR showed low levels of enzyme activity in transient the pUC DNA, although whether the smaller RNAs arise assays, and it seems likely that G418 resistance conferred by from processing of the larger 9-kb hybrid RNA or by another pR-NEO is mediated by synthesis of NPT. Fourth, on a per mechanism is unknown. No alterations in the size of other gene copy basis the level of expression of the hybrid 1092 KAPLER ET AL. MOL. CELL. BIOL. A B

0l -r w - o.

iL I.. W.; .

At .1 ec ---530

.Wz C S |~~~~~~~~~~~~~~~~~~~IA-CY-~

- A J . 1 A .G

,A A A T G A A '~ A,A A'' ' A Cv Si FIG. 7. 5'-End analysis of hybrid Neo-DHFR-TS RNAs. (A) Si nuclease and primer extension analysis. Si nuclease-resistant products (SI, lanes 1 to 3) were obtained by using the Narl-XhoI fragment shown in panel B, labeled at the 5' end of the Narl site, and El-pool(128) or no RNA as indicated; a sample of the undigested probe is also shown. Primer extension.r~~~~~~~~~~~~~~~~~~~~~~~~~-products (PE, lanes 4 and 5) were obtained by using the neo-specific oligonucleotide depicted in panel B and R1000 (lane 4) or El-pool(128) (lane 5) RNA. Size markers (in bases) are shown on the left, and the measured sizes of the products obtained are on the right. (B) Restriction map of the 5' end of the neo region of pR-NEO and summary of the results of Si nuclease protection and primer extension analyses. The restriction map is shown at the top of the figure; the vertical arrows correspond to the sites identified by Si nuclease analysis of the DHFR-TS gene in reference 28. Sequences derived from the neo gene cassette are indicated by the box, and L. major-derived DHFR-TS sequences are indicated by the thin line; position 1 in this map corresponds to the start codon of the NPT-coding region. The location of the Si nuclease probe (end labeled at the Narl site at position + 139) is shown by thin double lines, and the two protected products are shown by thick and thin single lines, respectively. The location of the oligonucleotide used in primer extension analysis is shown as a hatched box, with the two extension products indicated by thick and thin lines. (C) Sequence of cloned PCR products generated by amplification of cDNAs with a Leishmania miniexon-specific primer and either a neo- or DHFR-TS-specific primer. The full miniexon sequence is shown in the 5' box, with the miniexon oligonucleotide primer underlined. The 214 bp separating the 5'- and 3'-most sequences shown are identical in the two cDNAs and genomic DNA. neo-specific sequences are shown in the 3' box, and the ATG start codons of DHFR-TS and NPT are boldfaced.

Neo-DHFR-TS mRNA appeared to be similar to that ob- sequence from the HSV gene and a synthetic translation served for the DHFR-TS gene in amplified lines obtained by initiation sequence (32, 41). However, the 3'-flanking end MTX selection. This suggested that expression directed by contains a functional HSV polyadenylation signal (45). This pR-NEO is occurring in a quantitatively normal manner. sequence is specifically not recognized in L. major since we Fifth, pR-NEO contains regions that encode other R-region did not observe the predicted 1.3-kb hybrid Neo-DHFR-TS transcripts, and correspondingly, lines bearing high levels of mRNA (Fig. 6). These studies demonstrate a fundamental pR-NEO DNA such as E1-pool(128) showed elevated levels difference in the cis-acting elements required for polyadeny- of these transcripts. Moreover, a 9-kb RNA corresponding lation in trypanosomatids and higher eucaryotes and are in size to that predicted for a hybrid pUC-6.2-kb RNA was consistent with the failure to detect consensus eucaryotic found. Thus, the pattern and abundance of stable RNAs in polyadenylation elements in many trypanosomatid genes. pR-NEO-transfected lines were consistent with utilization of The fact that pR-NEO is capable of autonomous extra- normal transcriptional and/or processing signals located chromosomal replication unambiguously demonstrates the within the pR-NEO DNA. functionality of the Leishmania sequences present within The neo gene cassette within pR-NEO derives from the this DNA. This argues that the original extrachromosomal plasmid pMClneo-POLYA and contains portions of the circular amplified R-region DNA also replicates autono- herpes simplex virus (HSV) thymidine kinase (tk) gene in mously and is not continually generated from a modified addition to the NPT-coding region. The 5' end of the neo chromosomal locus. Similarly, the ability of pR-NEO to cassette employed contains 30 bp of functionless 5'-flanking efficiently direct synthesis of a hybrid Neo-DHFR-TS VOL. 10, 1990 STABLE TRANSFECTION OF PARASITES 1093 mRNA suggests that increased DHFR-TS expression in 10. Blackburn, E. H., and K. M. Karrer. 1986. Genomic reorgani- amplified MTX-resistant lines is directly mediated by a gene zation in ciliate protozoans. Annu. Rev. Genet. 20:501-521. dosage effect. Thus, this work established the functional 11. Borst, P. 1986. Discontinuous transcription and antigenic vari- described DHFR-TS ation in trypanosomes. Annu. Rev. Biochem. 55:701-732. characteristics of the previously (R- 12. Brunk, C. F. 1986. Genome reorganization in Tetrahvmena. Int. region) amplifications. Rev. Cytol. 99:49-83. The data presented in this paper signal the advent of 13. Cathala, G., J. F. Savouret, B. Mendez, B. West, M. Karin, J. A. methodologies and vectors suitable for the stable introduc- Martial, and J. D. Baxter. 1983. A method for isolation of intact, tion of DNAs into human parasitic protozoa. Molecular translationally active ribonucleic acid. DNA 2:329-335. dissection of expressing constructs such as pR-NEO will 14. Chu, G., D. Vollrath, and R. W. Davis. 1986. Separation of large allow a more precise localization of the cis-acting elements DNA molecules by contour-clamped homogeneous electric required for transcription, processing (including miniexon fields. Science 234:1582-1585. addition), and replication, in conjunction with the applica- 15. Clayton, C. E. 1988. The molecular biology of the Kinetoplas- tion of methods transient assays of and tidae. Genet. Eng. 7:1-56. for expression 16. Coderre, J. A., S. M. Beverley, R. T. Schimke, and D. V. Santi. replication. Moreover, our results showing the formation of 1983. Overproduction of a bifunctional thymidylate synthase- hybrid transcripts containing pUC-derived sequences sug- dihydrofolate reductase and DNA amplification in methotrex- gest that it will be possible to introduce other genes of ate-resistant Leishmania. Proc. Natl. Acad. Sci. USA 80: interest into appropriate locations within the pR-NEO vector 2132-2136. and obtain transcription and, potentially, translation. Fi- 17. Detke, S., G. Chaudhuri, J. A. Kink, and K. P. Chang. 1988. nally, since the efficiency of transfection is high, it is now DNA amplification in tunicamycin-resistant Leishlmania mexi- possible to contemplate transfection-based approaches for cana. Multiple copies of a single 63 kilobase supercoiled mole- the identification and analysis of genes responsible for many cule and their expression. J. Biol. Chem. 263:3418-3424. features the infectious 18. Donelson, J. E. 1989. DNA rearrangements and antigenic vari- interesting and unique of parasite ation in trypanosomes, p. 763-781. In D. E. Berg and M. M. cycle. Howe (ed.), Mobile DNA. American Society for Microbiology, Washington, D.C. ACKNOWLEDGMENTS 19. Ellenberger, T. E., and S. M. Beverley. 1987. Reductions in We thank D. D. Rogers for technical assistance, Hunt Potter for methotrexate and folate influx in methotrexate-resistant lines of access to the electroporator and advice, R. Johnson for pMClneo- Leis/lnania major are independent of R or H region amplifica- POLYA DNA, J. Miller for providing unpublished data. and D. E. tion. J. Biol. Chem. 262:13501-13506. Dobson and A. James for reading this manuscript. 20. Feinberg, A. P., and B. Vogelstein. 1983. A technique for This work was supported by Public Health Service grant AI-2903 radiolabelling DNA restriction endonuclease fragments to high to S.M.B. and BRSG S)7 RR 05381-27 to Harvard Medical School specific activity. Anal. Biochem. 132:6-13. from the National Institutes of Health. G.M.K. was supported by 21. Garvey, E. P., J. A. Coderre, and D. V. Santi. 1985. Selection training grants 5T-32-GM07306 and 5T-32-GM07196 from the Na- and properties of Leishmania resistant to 10-propargyl-5,8- tional Institutes of Health through much of this work. S.M.B. is a dideazafolate, an inhibitor of thymidylate synthase. Mol. Bio- Burroughs-Wellcome scholar in molecular parasitology. chem. Parasitol. 17:79-91. 22. Garvey, E. P., and D. V. Santi. 1986. Stable amplified DNA in LITERATURE CITED drug-resistant Leishmania exists as extrachromosomal circles. 1. Aviv, H., and P. Leder. 1972. Purification of biologically active Science 233:535-540. globin messenger RNA by chromatography on oligodeox- 23. Hightower, R. C., M. L. Wong, L. Ruiz-Perez, and D. V. Santi. ythymidylic acid cellulose. Proc. Natl. Acad. Sci. USA 69: 1987. Electron microscopy of amplified DNA forms in antifo- 1408-1412. late-resistant Leishmania. J. Biol. Chem. 262:14618-14624. 2. Beck, E., G. L. E. A. Auserwald, B. Reiss, and H. Schaller. 1982. 24. Johnson, P. J., J. M. Kooter, and P. Borst. 1987. Inactivation of Nucleotide sequence and exact localization of the neomycin transcription by UV irradiation of T. brucei provides evidence phosphotransferase gene from transposon Tn5. Gene 19:327- for a multicistronic transcription unit including a VSG gene. Cell 336. 51:273-281. 3. Beliofatto, V., and G. A. M. Cross. 1989. Expression of a 25. Johnston, R. N., and B. L. Kucey. 1988. Competitive inhibition bacterial gene in a trypanosomatid protozoan. Science 244: of hsp7O gene expression causes thermosensitivity. Science 1167-1169. 242:1551-1554. 4. Beverley, S. M. 1988. Characterization of the "unusual" mobil- 26. Kafatos, F. C., S. A. Mitsialis, N. Spoerel, B. Mariani, J. R. ity of large circular DNAs in pulsed field-gradient electrophore- Lingappa, and C. Delidakis. 1985. Studies on the developmen- sis. Nucleic Acids Res. 16:925-938. tally regulated expression and amplification of insect chorion 5. Beverley, S. M. 1989. Estimation of circular DNA size using genes. Cold Spring Harbor Symp. Quant. Biol. 50:537-547. gamma irradiation and pulsed-field gel electrophoresis. Anal. 27. Kapler, G. M., and S. M. Beverley. 1989. Transcriptional Biochem. 177:110-114. mapping of the amplified region encoding the dihydrofolate 6. Beverley, S. M., J. A. Coderre, D. V. Santi, and R. T. Schimke. reductase-thymidylate synthase of Leishmania major reveals a 1984. Unstable DNA amplifications in methotrexate-resistant high density of transcripts, including overlapping and antisense Leishmania consist of extra-chromosomal circles which relocal- RNAs. Mol. Cell. Biol. 9:3959-3972. ize during stabilization. Cell 38:431-439. 28. Kapler, G. M., K. Zhang, and S. M. Beverley. 1987. Sequence 7. Beverley, S. M., T. E. Ellenberger, and J. S. Cordingley. 1986. and S1 nuclease mapping of the 5' region of the dihydrofolate Primary structure of the gene encoding the bifunctional dihy- reductase-thymidylate synthase gene of Leishmania major. Nu- drofolate reductase-thymidylate synthase of Leishmania major. cleic Acids Res. 15:3369-3383. Proc. Natl. Acad. Sci. USA 83:2584-2588. 29. Laban, A., and D. F. Wirth. 1989. Introduction and expression 8. Beverley, S. M., T. E. Ellenberger, D. M. lovannisci, G. M. of transfection vector pALTI-1 in Leishmania enrietti. J. Cell. Kapler, M. Petrillo-Peixoto, and B. J. Sina. 1988. Gene amplifi- Biochem. 13E:84. cation in Leishmania, p. 431-448. In P. T. Englund and A. Sher 30. Long, E. O., and I. B. Dawid. 1980. Repeated genes in eukary- (ed.), Biology of parasitism. Alan R. Liss, Inc., New York. otes. Annu. Rev. Biochem. 49:727-764. 9. Beverley, S. M., R. B. Ismach, and D. McMahon-Pratt. 1987. 31. Marsh, J. L., M. Erfle, and E. J. Wykes. 1984. The pIC plasmid Evolution of the genus Leishmania as revealed by comparisons and phage vectors with versatile cloning sites for recombinant of nuclear DNA restriction fragment patterns. Proc. Natl. Acad. selection by insertional inactivation. Gene 32:481-485. Sci. USA 84:484-488. 32. McKnight, S. 1980. The nucleotide sequence and transcript map 1094 KAPLER ET AL. MOL. CELL. BIOL.

of the herpes simplex virus thymidine kinase gene. Nucleic 39. Spradling, A. C., and A. P. Mahowald. 1980. Amplification of Acids Res. 8:5949-5964. genes for chorion proteins during oogenesis in Drosophila 33. Muhich, M. J., and J. C. Boothroyd. 1988. Polycistronic tran- melanogaster. Proc. Natl. Acad. Sci. USA 77:1096-1100. scripts in trypanosomes and their accumulation during heat 40. Taylor, J. W., J. Ott, and F. Eckstein. 1985. The rapid genera- shock: evidence for a precursor role in mRNA synthesis. Mol. tion of oligonucleotide-directed mutations at high frequency Cell. Biol. 8:3837-3846. using phosphorothioate-modified DNA. Nucleic Acids Res. 34. Opperdoes, F. R. 1987. Compartmentation of carbohydrate 13:8765-8785. metabolism in trypanosomes. Annu. Rev. Microbiol. 41:127- 41. Thomas, K. R., and M. R. Capecchi. 1987. Site-directed muta- 152. genesis by gene targeting in mouse embryo-derived stem cells. 35. Petrillo-Peixoto, M. L., and S. M. Beverley. 1988. Amplified Cell 51:503-512. DNAs in laboratory stocks ofLeishmania tarentolae: extrachro- 42. Washtien, W. L., R. Grumont, and D. V. Santi. 1985. DNA mosomal circles structurally and functionally similar to the amplification in antifolate-resistant Leishmania. J. Biol. Chem. inverted H region amplification of methotrexate-resistant Leish- 260:7809-7812. mania major. Mol. Cell. Biol. 8:5188-5199. 43. White, T. C., F. Fase-Fowler, H. van Luenen, J. Calafat, and P. 35a.Petrillo-Peixoto, M. L., and S. M. Beverley. 1989. Amplification Borst. 1988. The H circles of Leishmania tarentolae are a of a new region of DNA in an unselected laboratory stock of unique amplifiable system of oligomeric DNAs associated with Leishmania tarentolae: the T region. J. Protozool. 36:257-261. drug resistance. J. Biol. Chem. 263:16977-16983. 36. Potter, H. 1988. Electroporation in biology: methods, applica- 44. WHO Expert Committee. 1984. The leishmaniasis, p. 1-146. tions and instrumentation. Anal. Biochem. 174:361-373. World Health Organization, Geneva. 37. Simpson, L. 1987. The mitochondrial genome of kinetoplastid 45. Zhang, F., R. M. Denome, and C. N. Cole. 1986. Fine-structure protozoa: genomic organization, transcription, replication and analysis of the processing and polyadenylation region of the evolution. Annu. Rev. Microbiol. 41:363-382. herpex simplex virus type 1 thymidine kinase gene by using 38. Simpson, L., and J. Shaw. 1989. RNA editing and the mitochon- linker scanning, internal deletion, and insertion mutations. Mol. drial cryptogenes of kinetoplastid protozoa. Cell 57:355-366. Cell. Biol. 6:4611-4623.