Molecular & Biochemical Parasitology 126 (2003) 275–279

Short communication

Development of RNA interference revertants inTrypanosoma bruceicell lines generated with a double stranded RNA expression

construct driven by two opposing promotersYili Chen, Chien-Hui Hung, Thomas Burderer, Gwo-Shu Mary Lee∗Department of Pathology, New York University School of Medicine, 550 First Avenue, New York, NY 10016, USA

Received 14 June 2002; received in revised form 28 October 2002; accepted 28 October 2002

Keywords: RNA interference;Trypanosoma brucei; Revertants

1. Introduction

Double stranded RNAs (dsRNAs) direct gene-specificsilencing by selectively degrading the cognate mRNA, inmany organisms ranging from trypanosomes to mice[1,2].This type of post-transcriptional gene silencing is calledRNA interference (RNAi) and has been used effectivelyto down-regulate gene expression inTrypanosoma brucei[3–8]. Two types of trypanosome double stranded RNA ex-pression constructs have been reported. Type one constructscontain a single promoter driving head-to head arrangedgene fragments[3,5,6]. Type two constructs contain twoopposing promoters controlling the expression of a singlecopy of a gene fragment[7,8].

We have been using RNAi to evaluate the function of var-ious genes involved in transcription and protein trafficking.We used the Type two-construct pZJM described by Wanget al. [7] to express gene specific dsRNA inT. brucei. InpZJM, the dsRNA expression is controlled by two oppos-ing tetracycline-inducible T7 promoters and the plasmid isdesigned to integrate into the rDNA non-transcribed spacerregion [7]. The bloodstream form cell line 13-90 and theprocyclic form cell line 29-13 (obtained from Dr. Cross’slaboratory) that express T7 RNA polymerase and tetracy-cline repressor were used as host cell lines[9]. Based onthe RNAi impact on cell growth, we generally observedthree types of responses. Obviously, for non-essential genes,the growth of trypanosomes is not significantly affected bythe induction of RNAi. For some essential genes, induc-tion of RNAi rapidly stops cell growth and can lead to cell

∗ Corresponding author. Tel.:+1-212-263-8260; fax:+1-212-263-8179.E-mail address: leeg02@med.nyu.edu (G.-S.M. Lee).

death, as exemplified inFig. 1. Induction of RNAi againstthe Tb-292-an endoplasmic membrane network associatedprotein [10] in procyclic trypanosomes and induction ofthe RNA polymerase (pol) I largest subunit RNAi[11,12]in bloodstream form trypanosomes severely and rapidly af-fected cell growth. For many genes, induction of RNAi mod-erately affects the growth efficiency; and after a long pe-riod (14 days or longer) under the induction of RNAi, cellsbecome irresponsive to RNAi and resume a normal growthefficiency. We refer the cells that recovered from the RNAistress and exhibited improved growth efficiency to as RNAirevertants. Additionally, we found that the frequency of de-veloping RNAi revertants drastically increased in RNAi celllines that were cultured for a long period of time and the oc-currence of RNAi revertants in bloodstream form cell lineswas higher than that in procyclic cell lines. Thus, it seemedthat RNAi revertants spontaneously developed during thegrowth. Revertants with improved growth efficiency, com-pared to the RNAi transformants were selectively enrichedduring passages. We determined the frequency of RNAi re-vertant occurrence in both bloodstream form and procyclicform trypanosomes and identified the potential mechanismsleading to RNAi revertants.

We analyzed the procyclic form Tb-292 dsRNA express-ing cell line and the bloodstream form RNA pol I largestsubunit dsRNA expressing cell line which were transformedby pZJM-Tb-292 and pZJM-RNA pol I, respectively. Boththe Tb-292 and RNA pol I largest subunit genes are essentialfor trypanosomes. Induction of RNAi of Tb-292 and RNApol I largest subunit stops cell growth and leads to cell death(Fig. 1). The function of Tb-292 and RNA pol I largest sub-unit and the details of the phenotypes of transformed celllines upon the RNAi induction are beyond the scope of thispaper and will be described separately. Approximately 10

0166-6851/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.PII: S0166-6851(02)00276-1

276 Y. Chen et al. / Molecular & Biochemical Parasitology 126 (2003) 275–279

Fig. 1. Effect of RNAi on trypanosome cell growth. The procyclic form ofT. brucei was maintained in SDM-79 medium at 25◦C [14]. Bloodstreamform trypanosomes were maintained in HMI9 medium at 37◦C [15]. The procyclic trypanosome cell line 29-13 and the bloodstream form trypanosomecell line 13-90 (obtained from Dr. G. Cross’s laboratory;[9]) were used for transfomation using pZJM derived dsRNA expression plasmid. Ten or twentymiligrams linearized plasmid was electroporated into trypanosomes using a BTX electroporator as previously described[13]. Thirty-six and sixteen hourspost-electroporation of the procyclics and the bloodstream form trypanosomes, respectively, hygromycin B (40�g/ml for procyclics and 2 mg/ml for thebloodstream form) or phleomycin (2�g/ml for the procyclics and 1�g/ml for the bloodstream form) or G418 (20�g/ml for the procyclics and 1�g/ml forthe bloodstream form) was added into cell cultures to select stably transformed trypanosomes. The individually transformed drug resistant trypanosomeswere cloned by limiting dilution using microtiter dishes. For cloning procyclic cell lines, 5× 106 to 5× 106/ml wild type trypanosomes was added toeach well. For induction of dsRNA, trypanosomes were cultured in medium containing 1.0�g/ml tetracycline. The growth curves of trypanosomes celllines P-Tb-292 RNAi and BS-RNA pol I RNAi (two sets of experiment is shown) were determined in the absence (open circle or square) or in thepresence (closed circles and squares) of 1�g/ml of tetracycline (for the induction of dsRNA expression). Cells were continuously maintained at a logphase. If needed, medium was added to expand the culture. Total number of cells in each culture was calculated at different time points.

and 5 days after drug selection of stable transformants ofthe procyclic form and the bloodstream form, respectively,we performed limiting dilution cloning of individually trans-formed trypanosomes. In general, it takes approximately 15and 6 days to establish clonal populations of procyclic celllines and bloodstream form cell lines, respectively. We antic-ipated that RNAi revertants have already developed duringthe growth of each clonal population of transformants. Thus,immediately after the establishment of clonal cell lines, weanalyzed the abundance of RNAi revertants in each clonalpopulation. For all experiments including the selection ofRNAi revertants, transformants were maintained in the pres-ence of phleomycin, which was originally used to select thetransformed cell lines.

We performed a series of limiting dilutions of one initiallycloned procyclic Tb-292 dsRNA transformant in 96-wellplates. Ten aliquots of each incremental cell density, rang-ing from 106 trypanosomes per well (∼200�l) to 102 try-panosomes per well (∼200�l) were grown in each plate.Three sets of plates were prepared. Plate one was culturedin the presence of tetracycline for RNAi induction. Platetwo was grown under an induced condition and with theaddition of ∼106/ml of wild type procyclics in each wellto rescue revertants, because the growth efficiency of pro-cyclic trypanosomes is drastically reduced at a low cell den-sity. As a positive control, plate three was maintained underun-induced conditions with the addition of∼106/ml wildtype procyclic trypanosomes per well. After approximately

10 days of culture, we found that: (1) all of the cells in Platethree survived; (2) no survivals were developed from thePlate one; and (3) in Plate two, survival of cells was foundin 3 out of 10 wells that contained 104 trypanosomes and in7 out of 10 wells that contained 105 trypanosomes (Table 1).All wells containing 106 trypanosomes developed revertants.Thus, we estimated that approximately one revertant can befound in 2.5×104 cells. A similar frequency of revertant oc-currence was also found in cell lines exhibiting a moderategrowth effect upon RNAi induction. A similar experimentperformed using one cloned procyclic Tb-292 dsRNA cellline that was passage for approximately 3 weeks in the ab-sence of tetracycline, revealed that the frequency of revertantoccurrence was increased to∼10−2. We have found that, ingeneral under un-induced conditions, the expression level ofRNAi target genes in many RNAi transformed cell lines wasslightly lower than those observed in the host cell 29-13.This observation suggested that the dsRNA expression maynot be completely repressed under a un-induced conditionin transformed cell lines. Thus, the drastic increase of thefrequency of revertant occurrence during a routine passagemay result from selection of revertants with an increasedgrowth efficiency.

Similar limiting dilutions were performed using an ini-tially clonal bloodstream form RNA pol I largest subunitdsRNA transformed cell line. Since a low cell density doesnot limit the growth efficiency of bloodstream form try-panosomes, it is not needed to add wild type trypanosomes in

Y. Chen et al. / Molecular & Biochemical Parasitology 126 (2003) 275–279 277

Table 1Frequency of RNAi revertant occurrence in trypanosomes

P-dsTb-292 clone 1 BS-dsRNA pol I clone 1

No. of trypanosomesper well

Total no. ofwells

No. ofwells/revertants

No. of trypanosomesper well

Total no. ofwells

No. ofwells/revertants

1 × 106 10 10 1× 106 10 101 × 105 10 7 1× 105 10 101 × 104 10 3 1× 104 10 91 × 103 10 0 1× 103 10 11 × 102 10 0 1× 102 10 0

Series of limiting dilutions were performed under the induction of RNAi (in the presence of 1�g/ml of tetracycline). For P-dsTb-292 clone 1 cell line,wild type procyclics was added to maintain cell growth. If needed, medium will be added to maintain the growth.

the limiting dilutions of bloodstream form trypanosomes. Wefound that under an induced condition, 1 out of 10 wells from103 trypanosomes per well aliquots and 9 out of 10 wellsfrom 104 trypanosomes per well aliquots developed rever-tants (Table 1). All wells that contained >104 trypanosomesdeveloped revertants. Thus, we estimated that approximatelyone revertant can be found in 5× 103 bloodstream formRNAi cells. We performed the analysis twice to confirm thefrequency of revertants observed in each cell line.

Fig. 2. Southern blot analysis. Nuclear DNA was isolated from trypanosomes as described[16]. Following the digestion with Hind III restrictionendonuclease, the DNA was separated on a 0.8% agarose gel and transferred onto nitrocellulose filters. Panel A: C1, C2, P1 and 1–10 represent genomicDNAs derived from wild type trypanosomes, cell line 29-13, PdsTb-292 clone 1 and P-dsTb-292 revertant clones 1–10, respectively. Panel B: C1, P2and 1′ to 10′ represent genomic DNAs derived from wild type trypanosomes, BS-dsRNA pol I clone 1 and BS-dsRNA pol I revertant clones 1′ to 10′,respectively. The A and B blots were hybridized with a Tb-292 3′-coding region probe[10] and a RNA pol I largest subunit coding region probe[11,12],respectively. The final post-hybridizational wash was performed in 0.1× SSC, 0.1% SDS at 65◦C. Stars indicate endogenous Tb-292 and RNA pol Igene. Arrowheads indicated the gene fragment derived from the integrated plasmid DNA.

To further determine the potential mechanisms leadingtransformed trypanosomes to lose the ability to respondRNAi effect, we performed Southern genomic blot analysisof the 10 Tb-292 RNAi revertant cell lines developed from1×104 to 1×105 trypanosome aliquots and the 10 RNA polI largest subunit RNAi revertant cell lines developed from1× 103 to 1× 104 trypanosome aliquots (Fig. 2). In the ge-nomic DNA isolated from the original-clonal Tb-292 dsRNAexpressing cell line, we detected two sizes of Tb-292 frag-

278 Y. Chen et al. / Molecular & Biochemical Parasitology 126 (2003) 275–279

ments; the large fragment represents the endogenous Tb-292gene (Fig. 2, panel A, lane P1; indicated by star) and thesmall fragment of∼0.6 kb (Fig. 2, panel A, lane P1, indi-cated by arrow head) is derived from the integrated copy ofpZJM-Tb-292. The 0.6 kb Tb-292 fragment was not detectedin the genomic DNA derived from the wild type procyclictrypanosome nor the procyclic cell line 29-13 (Fig. 2, panelA, lanes C1 and C2). We found that 4 out of 10 procyclicTb-292 dsRNA revertants lost the small Tb-292 fragmentthat is originally cloned into the integrated pZJM-Tb-292construct (Fig. 2, panel A; lanes 2, 3, 9, and 10). Five celllines maintained the predicted structure for the correct in-tegration of pZJM-Tb-292 construct (Fig. 2, panel A; lanes4 to 8). One cell line exhibited an altered size for the in-tegrated pZJM-Tb-292 (Fig. 2, panel A; lane 1). We alsoexamined the revertants isolated from the bloodstream formRNA pol I largest subunit dsRNA transformants. Two RNApol I largest subunit hybridizing fragments were detectedin the genomic DNA from a parental cloned RNA pol Ilargest subunit dsRNA expressing cell line: the large frag-ment represents the endogenous RNA pol I largest subunitgene (Fig. 2, panel B, lane P2; indicated by star) and thesmall fragment is derived from the integrated pZJM-RNApol I construct (Fig. 2, panel B, lane P2; indicated by ar-row head). This small RNA pol I largest subunit fragmentwas not detected in the genomic DNA isolated from wildtype trypanosomes (Fig. 2, panel B, lane C1). We foundthat 8 out of 10 RNA pol I largest subunit dsRNA rever-tants lost the small RNA pol I largest subunit fragment thatwas originally cloned into the integrated pZJM-RNA polI construct (Fig. 2, panel B; lanes 1′, 2′, 5′, 6′, 7′, 8′, 9′and 10′); and two revertants maintained the correct struc-ture for the integrated pZJM-RNA pol I construct (Fig. 2,panel B; lanes 3′ and 4′). These results indicate that a sig-nificant number of revertants were generated by deleting thetarget gene fragment in the integrated pZJM derived con-struct, leading to a diminished response to the RNAi effect.Since all transformants were continuously maintained in thepresence of phleomycin, we expected that all cell lines in-cluding RNAi revertants would contain the pZJM plasmidderived phleomycin resistance gene. Further hybridizationof the Southern genomic blots with the phleomycin resis-tance gene probe demonstrated that all revertants containedthe phleomycin resistance gene. However, the phleomycinresistance gene-hybridizing fragment in each revertant cellline exhibited a size, which was different from that detectedin the corresponding normal cell line (data not shown). Ap-parently other unclear mechanisms without destroying theintegrated target gene fragment also involved in the devel-opment of RNAi revertants. We speculated that mutationsat the two T7/tetracyclin operator promoters or at the T7RNA polymerase or tetracycline repressors might also resultin the development of RNAi revertants. Further Southernblotting analysis of the P-Tb-292 revertant cell line #1 (in-dicated inFig. 2, panel A; lane 1) that exhibited an altered(larger) size for the integrated copy of Tb-292 fragment,

demonstrated that sequence mutations had occurred at theregion spanning one end of the Tb-292 derived insert andthe adjacent T7 promoter/tetracycline operator. These muta-tions resulted in the loss of one Hind III site, leading to theincreased size for the integrated copy of the Tb-292 frag-ment in the P-Tb-292 revertant cell line #1. We speculatedthat this mutation might have affected the function of theadjacent T7 promoter/tetracycline operator. It is also possi-ble that the size of the target fragment placed in between thetwo opposing promoters may affect the efficiency of deletingthe target fragment, though we have not yet performed thisanalysis.

In short, we report the stability of the RNAi transformantsgenerated using dsRNA expression constructs controlled bytwo opposing inducible promoters. The frequency of RNAirevertants occurred in the procyclic form is∼10−4 to 10−5

and in the bloodstream form is∼10−3 to 10−4. A long pe-riod of passage will drastically increase the abundance of re-vertants in the population, most likely due to the selection ofgrowth efficiency. Deleting the target gene insert seems to beone of the major mechanisms resulting in RNAi revertants.Our analysis also suggests that the initial-clonal cell linesshould be used to examine the phenotypic changes resultingfrom the RNAi effect. Analysis of prolonged cultivated celllines may mislead the results, due to genetic changes of thecell population.

Acknowledgements

We thank Dr. Englund for providing the pZJM vector andDr. Cross for providing the trypanosome procyclic 29-13 cellline and the bloodstream form 13-90 cell line. GSML whois Burroughs Wellcome fund New Investigator in Molec-ular Parasitology is supported by NIH grants AI31117 &AI28953 and a WHO grant A10328.

References

[1] Bass BL. Double-stranded RNA as a template for gene silencing.Cell 2000;101:235–8.

[2] Sharp PA. RNA interference—2001. Genes Dev 2001;15:485–90.[3] Ngo H, Tschudi C, Gull K, Ullu E. Double-stranded RNA induces

mRNA degradation inTrypanosoma brucei. Proc Natl Acad Sci USA1988;95:14687–92.

[4] Bastin P, Sherwin T, Gull K. Paraflagellar rod is vital for trypanosomemotility. Nature 1997;391:548.

[5] Bastin P, Ellis K, Kohl L, Gull K. Flagellum ontogeny in try-panosomes studies via an inherited and regulated RNA interferencesystem. J Cell Sci 2000;113:3321–8.

[6] Shi H, Djikeng A, Mark T, Wirtz E, Tschudi C, Ullu E. Ge-netic interference inTrypanosoma brucei by heritable and inducibledouble-stranded RNA. RNA 2000;6:1069–76.

[7] Wang Z, Morris JC, Drew ME, Englund PT. Inhibition ofTry-panosoma brucei gene expression by RNA interference using anintegratable vector with opposing T7 promoters. J Biol Chem2000;275:40174–9.

Y. Chen et al. / Molecular & Biochemical Parasitology 126 (2003) 275–279 279

[8] LaCount DJ, Bruse S, Hill KL, Donelson JE. Double-stranded RNAinterference inTrypanosoma brucei using head-to-head promoters.Mol Biochem Parasitol 2000;111:67–76.

[9] Wirtz E, Leal S, Ochatt C, Cross GA. A tightly regulated in-ducible expression system for conditional gene knock-outs anddominant-negative genetics inTrypanosoma brucei. Mol BiochemParasitol 1999;99:89–101.

[10] Le MG-S, Russell D, D’Alesandro PA, Van der Ploeg LHT. Iden-tification of membrane associated proteins inTrypanosoma brucei,encoding an internal EARLRAEE amino acid repeat. J Biol Chem1994;269:8408–15.

[11] Smith JL, Levin JR, Agabian N. Molecular characterization of theTrypanosoma brucei RNA polymerase I and III largest subunit genes.J Biol Chem 1989;264:18091–9.

[12] Jess W, Hammer A, Cornelissen AWCA. Complete sequence ofthe gene encoding the largest subunit of RNA polymerase I ofTrypanosoma brucei. FEBS Lett 1989;249:123–8.

[13] Rudenko G, Le Blancq S, Smith J, Lee G-SM, Rattray A, Vander Ploeg LHT. Procyclic acidic repetitive protein (PARP) geneslocated in an unusually small�-amanitin-resistant transcriptionunit: PARP promoter activity assayed by transient DNA trans-fection of Trypanosoma brucei. Mol Cell Biol 1990;10:3492–504.

[14] Brun R, Schonenberger M. Cultivation and in vitro cloning of pro-cyclic culture formTrypanosoma brucei in a semidefined medium.Acta Trop 1979;36:289–92.

[15] Huirumi H, Huirumi M. Continuous cultivation ofTrypanosoma bru-cei blood stream forms in a medium containing a low concentrationof serum protein without feeder cell layers. J Parasitol 1989;75:985–9.

[16] Van der Ploeg LHT, Bernards A, Rijswijk FAM, Borst P. Characteri-zation of the DNA duplication-transposition that controls the expres-sion of two genes for variant surface glycoproteins inTrypanosomabrucei. Nucleic Acids Res 1982;10:593–609.