sign in sign up
<<
Home , Plasmid, T7 RNA polymerase

Proc. Nail. Acad. Sci. USA Vol. 88, pp. 7303-7307, August 1991 Prevention of human immunodeficiency type 1 expression in by a (intracediular/RNase HI) MOULDY SIOUD*t AND KARL DRLICA** *Public Health Research Institute, 455 First Avenue, New York, NY 10016; and tInstitute of Immunology and Rheumatology, Rikshospitalet, Fr. Qvamsgt. 1, 0172 Oslo, Norway Communicated by Thomas Cech, May 2, 1991

ABSTRACT are potentially very powerful pug/ml) and Casamino acids (0.5%), diluted 1:50, grown to an agents for perturbing intracellular expression. However, optical density of 0.4 at 600 nm, and induced with isopropyl pilot experiments in have met with mixed success. f3-D-thiogalactoside (IPTG) at 1 mM. Cells were harvested by We now report that a ribozyme designed to cleave the integrase at 40C. In some experiments, integrase was gene of the human immunodeficiency virus (HIV), when tran- expressed as a fusion from p22KS6 obtained from the scribed from a in Escherichia coli, led to destruction of National Institutes of Health AIDS Research and Reference integrase RNA and complete blockage of integrase protein Program (contributed by S. Goff, Columbia University). This synthesis. These results indicate that ribozymes can be used to derivative ofpBR322 contains the 5' region ofthe E. coli trpE study intracellular in and that the gene fused to a short portion of the 3' end of the HIV-1 HIV-1 integrase gene may be a useful target for therapeutic gene plus the entire gene for HIV-1 ribozymes. integrase. For induction of this fusion, cells were grown overnight at 37°C in M9 medium supplemented with tryp- Ribozymes are RNA molecules that catalyze RNA cleavage tophan (20,g/ml), Casamino acids (0.5%), and ampicillin (50 (1-3). Those that act in trans first hybridize to a specific ,ug/ml), diluted 1:25 in M9 medium lacking tryptophan, sequence in a target RNA and then cleave the target within grown to an optical density of0.4 at 600 nm, and induced with that sequence. The hammerhead type refined by Haseloffand indoleacrylic acid (IAA) at 5 ,ug/ml. Gerlach (4) possesses a 22-nucleotide-long catalytic domain Plasmid pMPD48, expressing ribozymefB, was constructed that contains the consensus sequences responsible for cleav- by inserting a 118-base-pair synthetic DNA encoding a T7 age activity. This catalytic domain is embedded within flank- and ribozyme ,B into the unique EcoRI site of ing sequences that are complementary to those surrounding pMPD51 (TetR), a derivative ofptrpl84 (10) lacking an EcoRI the cleavage site; they enable the ribozyme to interact with fragment. Plasmid pMPD52, which encoded ribozyme A3, a specific target sequence. The only sequence requirement was constructed in a similar manner. for the cleavage site is that it contain the trinucleotide GUC Nudeic Acid Analysis. Ribozyme activity was detected in (or a limited number of other triplets). Since appropriate vitro by incubating 2 ,ug of ribozyme with 20 ,ug of total target trinucleotides are common in , there are many cellular RNA extracted from strain MPD45 after induction of sites where ribozymes can be directed to cut. the trpE-integrase fusion gene (5 ,ug of IAA per ml at 30°C). To explore the possibility that bacterial cells could be used Mixtures oftarget RNA and ribozyme in 50 mM Tris-HCl (pH to refine ribozymes directed against viral RNA targets, we 8) were heated to 90°C for 1 min, rapidly cooled, and then examined the ability ofa ribozyme to cleave RNA containing incubated in 20 mM MgCl2 at 50°C or 37°C for 60 min. After the integrase gene ofthe human immunodeficiency virus type incubation, samples were treated with pancreatic DNase I 1 (HIV-1). The integrase gene, which is located within the 3' (ribonuclease free) for 15 min at 37°C. They were then domain ofthe HIV-1 pol gene, encodes a protein required for separated by electrophoresis through a 1% agarose gel con- the integration of viral DNA into the host (5, 6). taining 6.1% formaldehyde, transferred to a Hybond-N mem- Since integration is essential for productive , a brane (Amersham), and hybridized with a [32P]DNA probe ribozyme that interferes with the expression of integrase derived from the HIV-1 integrase gene [a 1631-base-pair Kpn should also block production of infectious HIV-1. The ribo- I/Sal I fragment of pNL4-3 (11)]. In some cases, the probe zyme that we designed completely blocked expression of was removed after hybridization by incubation in 10 mM integrase in Escherichia coli. Tris-HCI, pH 7.5/0.1% SDS/10 mM EDTA at 85°C for 2 hr. A second probe specific to trpE was then hybridized to the MATERIALS AND METHODS RNA on the membrane. Ribozyme activity occurring in vivo was detected after Bacterial Strains, , and Culture Conditions. The induction of expression of both ribozyme and target RNA. plasmids and bacterial strains used in this study are listed in For analysis of RNA, total RNA was prepared by an acid Table 1. Integrase was expressed from plasmid pLJS10 guanidinium thiocyanate procedure (12). In experiments in obtained from J. Groarke, J. Hughes, and L. Shaw (Sterling which both RNA and DNA were recovered, samples were Research Group). This derivative of pBR322 contains the collected by centrifugation at 4°C and suspended in 2 mM HIV-1 (BH10) int gene to which codons for methionine and EDTA/1% SDS/1% (vol/vol) glycerol/0.01% bromophenol glycine have been added at the 5' end; the gene is under blue. Each sample was mixed with 1/9th vol of 37% form- control of a T7 promoter. To express inte- aldehyde, placed in a boiling water bath for 2 min, and then grase, cells containing pLJS10 were grown overnight at 37°C analyzed by electrophoresis through agarose gels containing in M9 minimal medium (9) supplemented with ampicillin (50 6.1% formaldehyde. Integrase RNA and trpE-integrase fu-

The publication costs of this article were defrayed in part by page charge Abbreviations: HIV-1, human immunodeficiency virus type 1; IPTG, payment. This article must therefore be hereby marked "advertisement" isopropyl ,-D-thiogalactoside; IAA, indoleacrylic acid. in accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 7303 Downloaded by guest on September 29, 2021 7304 Biochemistry: Sioud and Drlica Proc. NatL Acad. Sci. USA 88 (1991) Table 1. Plasmids and bacterial strains promoter (14). Ribozyme f3, which was identical to ribozyme Plasmid gene a except that it possessed a bacteriophage 17 Strain Plasmid expressed terminator (15) at its 3' end, was synthesized in vitro (16) by transcription from a double-stranded template cut from plas- MPD45 p22KS6 trpE-int mid pMPD48. Ribozyme (3 was also synthesized in vivo from MPD61 pATH1 trp pMPD48 introduced into E. coli strain BL21(DE3), which MPD77 pLS10 int contains a chromosomal copy of the T7 RNA MPD92 pMPD48 Ribozyme (3 gene under control of the lacUVS promoter (8). Thus, the MPD97 pMPD51 resulting strain could be induced to synthesize ribozyme 13 by MPD98 pMPD52 Ribozyme A3 adding IPTG to the growth medium. As a control for possible MPD99 pLJS10; pMPD48 int; ribozyme ( antisense effects, we constructed ribozyme AP, in which the MPD100 pLJS10; pMPD52 int; ribozyme A,& entire catalytic domain of ribozyme ,B was replaced by a MPD101 pLJS10; pMPD51 int single guanosine. Both ribozyme P and ribozyme A&P were MPD102 p22KS6 trpE-int detected in extracts from E. coli cells by Northern hybrid- MPD103 p22KS6; pMPD48 trpE-int; ribozyme (3 ization analyses after induction with IPTG (Fig. 1B). In some MPD104 p22KS6; pMPD52 trpE-4nt; ribozyme AP experiments, read-through transcripts were also present MPD105 p22KS6; pMPD51 trpE-int (data not shown). MPD146 pUS10; pMPD48 int; ribozyme (3 Cleavage of Integrase RNA by Ribozymes in Vifro. The MPD148 pWS10 int activities of ribozyme a and ribozyme 1 were examined in The host for MPD45 and MPD61 is HB101 (7); for all others it is vitro using as a target total RNA extracted from cells BL21(DE3) (8). Strains MPD146 and MPD148 are rnclOS glyA::TnS induced to transcribe a trpE-integrase fusion gene. Samples derivatives in which the rmc and gly alleles were transduced from were incubated with each ribozyme, RNA species were strain NC124, obtained from Donald Court (Frederick Cancer Re- separated by electrophoresis, and target and product RNAs search Facility, Frederick, MD), by selection for kanamycin resis- were detected by hybridization with a radioactive probe for tance and screening for accumulation of precursor rRNA. the HIV-1 integrase gene. Ribozyme a and ribozyme 1 (both prepared in vitro) cleaved the 2000-nucleotide-long target sion RNA were detected by Northern blot analysis (13), using RNA into fragments of 1500 and 500 nucleotides; cleavage the Kpn I/Sal I fragment encompassing the integrase gene as was much more extensive when incubation was at 50TC (Fig. a hybridization probe. 2A, lanes 2 and 3) than when at 37c (lanes 6 and 7). Two RESULTS observations confirmed that cleavage occurred at the ex- pected site within the integrase portion ofthe fusion gene, 500 Ribozymes Designed to Cleave HIV-1 Integrase. We pre- nucleotides from the 3' end of the RNA. First, the two pared several ribozymes directed at the GUC located at the cleavage products had the expected sizes. Second, a probe position corresponding to nucleotides 4027-4029 of HIV-1 specific for the E. coli trpE gene detected only the 1500- (ribozyme sequences are described in Fig. LA). Ribozyme a nucleotide fragment (Fig. 2B). Thus, both ribozymes cleaved was synthesized in vitro by transcription from a synthetic, the trpE-integrase target RNA in vitro in the presence of single-stranded DNA template containing a bacteriophage 17 unrelated RNAs. A GR-U-G. 20 G U U, G.U ,G C C GUCR G R 30 10 G R UA oMlcdon U., G' / 'U R, A

R U A 6 A U 40 RIBOZYME C6 Ga B TARGET C6 G6 1 2 3 4 5 6 7 C G

5' /C A

50 s60 c * u 469 A-A-C-C-C-C-U-U-G-G-G-G-C-C-U R 462 U-U-G-G-G-G-R -G U -U-C -U-G-G -G U 80 70 C-R

U T7 trasrption temnator

3'

FIG. 1. Anti-integrase ribozymes. (A) Ribozyme bound to its target in HIV-1 integrase mRNA. Ribozyme forms two distinctive structures: a catalytic domain (nucleotides 14-35) and a bacteriophage 17 transcription terminator (nucleotides 45-86). Cleavage of the target RNA occurs on the 3' side of the cytidine indicated by the arrow [corresponding to nucleotide 4029 in GenBank data base (entry HIVBH102)]. Ribozyme a is identical to ribozyme (, except that it has a single adenosine in place of the transcription terminator. Ribozyme AP is identical to ribozyme (3, except that it has a single guanosine in place of the catalytic domain. (B) Expression of ribozymes. E. coli strains containing a gene for no ribozyme (MPD97; lanes 1 and 4), ribozyme (MPD92; lanes 2, 5, and 6), or ribozyme AP (MPD98; lanes 3 and 7) were grown without (lanes 1-3) or with (lanes 4-7) IPTG for 30 min. RNA was extracted, separated by electrophoresis, and detected by Northern blotting. Lane 5 is identical to lane 6 except the sample was treated with RNase A (50 ,ug/ml for 30 min). Arrowheads indicate molecular size standards (RNA length in nucleotides); from these it is estimated that ribozymes and A(3, when extracted from cells, are 85 and 65 nucleotides long, respectively. Downloaded by guest on September 29, 2021 Biochemistry: Sioud and Drlica Proc. Natl. Acad. Sci. USA 88 (1991) 7305 ribozymie: I) 13 A15 Ui 0 I Hi (f A lane no.: 1 2 3 4 5 6 7 8 9 10 o0 15 30' 60'

.,Admak. 'A B C l;)' IA B C DA E C D"ACD 4 ,*XWI .mow -Imp-. A 4 .9 __Wqm 4 1.6 1.6 o S.. _At~ -2-_I._ 4 1.0 1.0 P -I 0.6k 4 0.6

B C o0 15' 60' 120' 1 2 3 4 5

- 4 DNA Itb. j -4 DNA 2.8 * 1.9 P 1.6 o ^ RNA 4 RNA FIG. 2. Cleavage of trpE-integrase RNA by ribozymes in vitro. 1.0 (A) Incubation products detected by a probe for the integrase gene. The following ribozymes were added to target RNA: no ribozyme (lane 1), ribozyme a (2 ,ug) synthesized in vitro (lanes 2, 6, and 10), 0.6 P ribozyme (3(2 jtg) synthesized in vitro (lanes 3 and 7), cellular RNA (10 Ag) extracted from strain MPD92 treated with IPTG for 30 min to induce synthesis of ribozyme ,( (lanes 4 and 8), cellular RNA (10 Aig) extracted from strain MPD98 treated with IPTG for 30 min to induce synthesis of ribozyme A,8 (lanes 5 and 9). The sample in lane FIG. 3. Elimination of HIV-1 integrase RNA by a ribozyme in 10 was not preheated at 900C before incubation at 370C. Lanes 1-5 vivo. (A) Detection of target RNA. Four strains were treated with were incubated at 500C and lanes 6-10 were incubated at 370C. IPTG to induce transcription of HIV-1 integrase and in some cases Arrowheads indicate molecular size standards (RNA length in kilo- to simultaneously transcribe a ribozyme. At different times after bases). (B) Incubation products detected by a probe for the E. coli induction with IPTG, the RNA in the cells was extracted (12), trpE gene. After hybridization with the integrase gene probe, the separated by electrophoresis, and detected by Northern blotting. The RNA-containing membranes shown in A were washed free of the strain used and the time of induction are identified above each lane DNA probe and rehybridized with the Pvu I/BamHI fragment of the of the gel. Lanes: A, strain MPD77 (expressing integrase RNA); B, trpE gene obtained from pATH1 (17). strain MPD101 (expressing integrase RNA in the presence of a plasmid that encodes no ribozyme); C, strain MPD100 (expressing Elimination of RNA in Vivo: Coordinate Expression of integrase RNA and ribozyme Afl); D, strain MPD99 (expressing Target and Ribozyme. As shown in Fig. 1B, both ribozyme ( integrase RNA and ribozyme(). Arrowheads indicate molecular size and its deleted form, ribozyme AJB, could be expressed in standards (in kilobases). (B) Detection of target RNA and plasmid encoding target RNA. The E. coli strains described inA were induced vivo. RNA extracted from cells induced for expression of to coordinately express both the target and the ribozyme for the times ribozyme f3 cleaved RNA in vitro (Fig. 2A, lanes 4 and 8). indicated. Samples were treated as described in Materials and Under the same conditions of induction and analysis, cells Methods. (C) Activity ofribozyme (3 in rnc- mutant. Strain MPD146, induced for expression of ribozyme AI3 exhibited no activity an rnclO5 mutant capable ofexpressing both integrase and ribozyme (lanes 5 and 9). Since ribozyme A,3 was expressed at the same ,8 (lanes 1-3) and strain MPD148, an rncJO5 mutant able to express level as ribozyme f (Fig. 1B), this mutant ribozyme could be integrase but not ribozyme (3 (lanes 4 and 5), were treated with IPTG used as a control for possible antisense activities associated for 30 min (lanes 1, 3, and 5) to induce expression of integrase and, with ribozyme f3. when present, ribozyme (. Samples were prepared and analyzed as To examine the action of ribozyme f3 on HIV-1 integrase described in Materials and Methods. The sample in lane 3 was also RNA in vivo, we introduced two compatible plasmids into incubated with RNase A prior to electrophoresis. strain BL21(DE3): pMPD48, which expresses ribozyme f, of ribozyme /8 is due to its ability to cleave RNA rather than and pLJS10, which contains the HIV-1 integrase gene. Both to antisense properties. The molar ratio ofribozyme to target the ribozyme and the integrase gene were under the control was -100 as estimated of a bacteriophage 17 promoter so both could be induced from Northern blots (data not shown). simultaneously by the addition of IPTG to the growth me- To rule out loss ofthe plasmid encoding the target RNA as dium. The results (Fig. 3) demonstrate that ribozyme (3 is an explanation for the absence of target RNA in cells ex- active inside E. coli cells. Transcripts containing the HIV-1 pressing ribozyme /B, we extracted nucleic acids under con- target sequence were present after induction in cells that did ditions in which plasmid DNA would be recovered. The data not express an active ribozyme (Fig. 3A, lanes A-C). How- show that the plasmid responsible for integrase expression ever, when ribozyme (3 was expressed (lanes D), target RNA was present at roughly the same copy number in all strains was not recovered. In these experiments the products of (Fig. 3B). ribozyme activity were not observed; cleaved integrase RNA T7 RNA polymerase is not completely repressed in strain is apparently degraded by cellular nucleases. When ribozyme BL21(DE3) (8); consequently, some expression of ribozyme /3 was replaced by the mutant ribozyme A/3, there was only and target is expected in the absence ofIPTG. None was seen a small reduction in the level of integrase RNA (Fig. 3, lanes in the experiment shown in Fig. 3A (lane A, 0 min); however, C). Since Northern blot experiments similar to those shown the use of a more efficient RNA recovery method revealed a in Fig. 1B showed that ribozymes / and Ad8 were expressed low level of target RNA expression in uninduced cells (Fig. at similar levels (data not shown), we suggest that the activity 3B, lane A, 0 min). Target RNA was absent in uninduced cells Downloaded by guest on September 29, 2021 7306 Biochemistry: Sioud and Drlica Proc. Natl. Acad. Sci. USA 88 (1991) that contained plasmids encoding integrase and ribozyme (3 eliminated these cleavage products and much of the target (lane D, 0 min). Thus, ribozyme (3is effective even when both RNA (lane 7). These observations indicate that ribozyme (3 ribozyme and target RNA are present at low concentrations. cleaves its target and that the products ofcleavage are rapidly Since RNase III activity of E. coli cleaves RNA in the degraded. vicinity of double-stranded regions (18), the possibility ex- Prevention of HIV-1 Integrase Synthesis. To determine isted that ribozyme P forms a double-stranded structure with whether ribozyme (3blocks synthesis ofthe integrase protein, its target that leads to cleavage by RNase III. To test this, we cell lysates were analyzed for the presence of integrase constructed an rnc- strain. Ribozyme activity was present in protein by Western blotting. When integrase RNA and ri- the mutant (Fig. 3C); thus, RNase III does not account for the bozyme (3 were induced simultaneously, the ribozyme pre- elimination ofintegrase RNA associated with the presence of vented expression of the 31-kDa integrase protein (Fig. 5B, ribozyme (3. lane 3). This is the result expected for an activity that rapidly Elimination of RNA in Vivo: Independent Expression of destroys integrase RNA. The antiserum used to detect inte- Target and Ribozyme. We also studied the function of ribo- grase also reacted with the 66-kDa trpE-integrase fusion zyme 13 in cells in which the expression ofthe ribozyme could protein (Fig. SA), allowing us to use the fusion protein to be controlled separately from the expression ofa target RNA examine the effect ofinducing ribozyme expression before or containing a trpE-integrase fusion gene. Expression of the after induction of target RNA. The fusion protein was de- fusion RNA was induced by the addition of IAA, and tected when induction of ribozyme followed the induction of expression of the ribozyme was induced by the addition of the fusion gene (Fig. 5D, lane 5), indicating that the fusion IPTG. Simultaneous induction of ribozyme B and fusion mRNA was present and translated into protein. Simultaneous RNA resulted in partial loss ofthe target RNA (Fig. 4A, lanes expression ofboth the ribozyme and the target RNA led to a 3 and 4); preinduction of ribozyme (3 led to the complete reduction in the amount of fusion protein (Fig. SC, compare elimination of the target RNA (lane 6). lanes 2 and 5). Induction of ribozyme before and during When trpE-integrase fusion RNA was used as the target, induction ofthe target RNA was more effective at preventing it was possible that the absence of target RNA in cells expression of the fusion protein than was induction of the expressing ribozyme could have resulted from competition ribozyme after extensive expression of the fusion gene (Fig. between 17 RNA polymerase and cellular RNA polymerase SD, compare lanes 3 and 5). These results indicate that when for nucleotides (8). However, this is unlikely to have oc- ribozymes are synthesized in cells before induction of target curred since the synthesis of ribozyme l( had little effect on RNA, protein expression is inhibited. the amount of target RNA present in the cells (Fig. 4B, lanes 4 and 5). Moreover, by building up the concentration oftarget DISCUSSION RNA in the cells and then briefly inducing the synthesis of ribozyme (, we detected one of the cleavage products, the Site-specific cleavage of RNA by ribozymes offers an addi- 500-nucleotide fragment (lane 6). A population of RNA tional way to interfere with the expression of in living molecules of heterogeneous size was also detected. Increas- cells. Our results demonstrate that a ribozyme designed to ing the time during which the ribozyme was expressed cleave HIV-1 integrase RNA blocks expression of integrase in vivo: target RNA was eliminated in cells in which the A B ribozyme was synthesized (Figs. 3 and 4), and integrase protein expression was inhibited (Fig. 5). When the target 1 2 3 4 56 1 2 3 4 5 6 7 gene encoded integrase, simultaneous expression of ri-

DNA 1 bozyme and target was sufficient to observe the f1ll effect of the ribozyme (Figs. 3 and SB). The more complex target

4 created by the fusion of integrase to trpE required preinduc- 1. tion of the ribozyme for full effect (Figs. 4 and 5 C and D). Several considerations support the general conclusion that RNAP, ,4 0.t. ribozyme ( led to the elimination of integrase mRNA and fI. thereby blocked integrase synthesis. First, it is unlikely that I ribozyme action occurred during cell lysis rather than in vivo. Cells were chilled and harvested rapidly with chelating agents (in some cases with 100 mM EDTA), RNA was extracted quickly (within 10 min), and ribozyme induction blocked integrase protein synthesis and also destroyed integrase FIG. 4. Elimination of trpE-integrase RNA by a ribozyme in RNA. Second, the absence ofa catalytic domain in ribozyme vivo. (A) Induction of ribozyme expression before target expression. A(3 greatly lowered its effectiveness in vitro and in vivo (Figs. Strain MPD103 (lanes 2 and 4-6) was induced with IPTG to synthe- 2-5). The small effect seen with ribozyme AP, which was size ribozyme 8, either along with or before induction with IAA to present in cells at the same level as ribozyme (3 (Fig. 1B), synthesize trpE-integrase target RNA. Strain MPD102, which could probably arose from its antisense activity (20). Third, a not express ribozyme, was included as a control (lanes 1 and 3). product of target RNA cleavage was detected after a brief Lanes: 1 and 2, no induction ofRNA synthesis; 3 and4, simultaneous induction addition of IAA and IPTG for 30 min; 5, pretreatment with IPTG for induction of ribozyme synthesis (Fig. 4B). Longer 5 min before addition of IAA for 30 min; 6, pretreatment with IPTG periods led to the complete loss of the target RNA and its for 30 min before addition of IAA for 30 min. (B) Induction of target cleavage products, indicating that RNA degradation follows RNA expression before ribozyme expression. Strain MPD103 (lanes ribozyme action. Fourth, ribozyme (3 was detected in cell 3, 6, and 7) was induced with IPTG to synthesize ribozyme (, either extracts, both as a specific RNA cleaving activity (Fig. 2A), along with or after the induction of synthesis of target RNA by IAA. and as an RNA molecule ofthe expected size (Fig. 1B). Fifth, Strain MPD102, which does not encode a ribozyme (lanes 1 and 4), the RNA cleaving activity ofRNase III (18) does not account and strain MPD104, which encodes ribozyme A,8 (lanes 2 and 5), for the action ofribozyme ( since cells deficient in RNase III were included as controls. Lanes: 1-3, no induction of target RNA (3 synthesis; 4, simultaneous addition ofIAA and IPTG for 60 min; 5-7, exhibited cleavage ofintegrase mRNA when ribozyme was induction of target RNA synthesis with IAA for 40 min followed by induced (Fig. 3C). the induction of ribozyme (lane 5) with IPTG for 20 min or The ribozyme we designed blocked expression from a ribozyme ,B with IPTG for 10 min (lane 6) or 20 min (lane 7). specific gene without affecting the rate ofbacterial growth, as Arrowheads indicate molecular size standards (in kilobases). measured by turbidity and direct cell count (data not shown). Downloaded by guest on September 29, 2021 Biochemistry: Sioud and Drlica Proc. Natl. Acad. Sci. USA 88 (1991) 7307 A B shifts in patterns of gene expression. For each of these applications, multiple ribozymes could be used to interfere 1 2 3 4 5 6 1 2 3 4 5 with the expression of several genes sequentially or simul- -4 97 f 97 o taneously. * 66 Other laboratories have observed ribozyme activity in 66 . _ 4 42 eukaryotic cells. Ribozymes reduced expression ofthe chlor- amphenicol acetyltransferase (21) and the HIV-1 gag (22) 42 b O 31 genes, and they led to the cleavage (23) and disappearance (24) of target RNA when coinjected into Xenopus oocytes. Our results with bacteria solidify the interpretation of these C D in vivo ribozyme studies by showing that ribozymes are present in cells that exhibit loss of target RNA, that RNA 1 2 3 4 5 1 2 3 4 5 6 4 97 cleavage products can be detected even though they are 97 P rapidly degraded, and that elimination of a specific RNA by 66 P. 466 a ribozyme can prevent the expression ofits encoded protein. Bacteria offer safe and rapid systems in which ribozyme activity can be evaluated, and they provide access to pow- 42 P l42 erful genetic selection procedures for developing still more effective ribozymes. FIG. 5. Prevention of protein expression by ribozymes in vivo. (A) Characterization of antisera recognizing integrase. Strains capa- We thank Sanjay Tyagi for insightful advice on the design of the ble of expressing trpE (MPD61; lanes 1, 3, and 5) or a trpE-integrase experiments; Don Court, Christine Debouck, Stephen Goff, James fusion protein (MPD45; lanes 2, 4, and 6) were treated with IAA for Groarke, Joseph Hughes, Abe Pinter, and Lucinda Shaw for pro- 2 hr. The fusion protein was identified by Western blotting (19) using viding strains, clones, and antisera; and Fred Russell Kramer, the following antisera: antiserum from an AIDS patient (lanes 1 and Samuel Kayman, Ellen Murphy, and Thomas Cech for many helpful 2), polyclonal rabbit antiserum (obtained from Christine Debouck, comments on the manuscript. This work was supported by grants Smith Kline & French) raised against an E. coli GaIK-HIV-1- from the Aaron Diamond Foundation, the National Science Foun- polymerase fusion protein (lanes 3 and 4), and antiserum raised dation (PMB 8718115), and the U.S. Army Medical Research and against a synthetic peptide derived from the 12 C-terminal amino Development Command under Contract DAMD17-88-C-8126. acids of integrase (lanes 5 and 6). (B) Effect of ribozyme expression on the synthesis ofintegrase. Strains containing plasmids from which 1. Cech, T. (1987) Science 236, 1532-1539. integrase or ribozymes could be coordinately expressed were treated 2. Uhlenbeck, 0. C. (1987) Nature (London) 328, 596-600. with IPTG for 60 min. Integrase was identified by Western blotting 3. Forster, A. & Symons, R. (1987) Cell 49, 211-220. with an antiserum raised against a GaIK-HIV-1-polymerase fusion 4. Haseloff, J. & Gerlach, W. L. (1988) Nature (London) 334, protein. Lanes: 1, strain MPD77 (expressing only integrase); 2, strain 585-591. MPD101 (expressing integrase in the presence of a plasmid that 5. Schwartzberg, P., Colicelli, J. & Goff, S. P. (1984) Cell 37, encodes no ribozyme); 3, strain MPD99 (expressing both integrase 1043-1052. and ribozyme (8); 4, strain MPD100 (expressing both integrase and 6. Donehower, L. A. & Varmus, H. E. (1984) Proc. Natl. Acad. ribozyme A,8); 5, strain MPD77 (expressing only integrase; induction Sci. USA 81, 6461-6465. was for 30 min). (C) Effect of ribozyme expression on the synthesis 7. Maniatis, T., Fritsch, E. & Sambrook, J. (1982) Molecular ofthe trpE-integrase fusion protein. Strains containing plasmids that : A Laboratory Manual (Cold Spring Harbor Lab., Cold could express the trpE-integrase fusion protein and a ribozyme were Spring Harbor, NY). grown in the presence of IAA and IPTG for 90 min. The trpE- 8. Studier, F. & Moffatt, B. (1986) J. Mol. Biol. 189, 113-130. integrase fusion protein was identified by Western blotting with an 9. Miller, J. H. (1972) Experiments in Molecular (Cold antiserum raised against a GaIK-HIV-1-polymerase fusion protein. Spring Harbor Lab., Cold Spring Harbor, NY). Lanes: 1, strain MPD97 (expressing neither fusion protein nor 10. Bogosian, G. & Somerville, R. L. (1984) Mol. Gen. Genet. 193, ribozyme); 2, strain MPD105 (expressing only the fusion protein); 3, 110-118. strain MPD102 (expressing only the fusion protein); 4, strain 11. Adachi, A., Gendelman, H., Koenig, S., Folks, T., Willey, R., MPD104 (expressing the fusion protein and ribozyme AB); 5, strain Rabson, A. & Martin, M. (1986) J. Virol. 59, 284-291. MPD103 (expressing the fusion protein and ribozyme ,B). (D) Effect 12. Chomczynski, P. & Sacchi, N. (1987) Anal. Biochem. 162, ofinducing ribozyme synthesis before or after the induction oftarget 156-159. RNA synthesis. Strain MPD103 (which expresses both ribozyme 18 13. Thomas, P. (1980) Proc. Natl. Acad. Sci. USA 77, 5201-5205. and the trpE-integrase fusion protein) was grown under various 14. Milligan, J. F., Groebe, D. R., Witherall, G. W. & Uhlenbeck, conditions (lanes 2-6). As a control, strain MPD102 (which expresses 0. C. (1987) Nucleic Acids Res. 15, 8783-8798. only the fusion protein) was grown in the presence of both IPTG and 15. Rosenberg, A., Lade, B., Chui, D., Lin, S., Dunn, J. & Studier, IAA for 90 min (lane 1). Lanes: 2, preinduction with IPTG for 20 min W. (1987) Gene 56, 125-135. and then induction with both IPTG and IAA for 90 min; 3, prein- 16. Melton, D. A., Krieg, P. A., Rebagliati, M. R., Maniatis, T., duction with IPTG for 5 min and then induction with both IPTG and Zinn, K. & Green, M. R. (1984) Nucleic Acids Res. 12, 7035- IAA for 90 min; 4, induction with IAA for 90 min; 5, preinduction 7056. with IAA for 90 min and then induction with IAA and IPITG for 15 17. Tanese, N., Roth, M. & Goff, S. P. (1985) Proc. Natl. Acad. min; 6, preinduction with IAA for 90 min and then induction with Sci. USA 82, 4944-4948. IAA and IPTG for 30 min. The fusion protein was identified by 18. Krinke, L. & Wulff, D. (1990) Nucleic Acids Res. 18, 4809- Western blotting with an antiserum raised against GaIK-HIV-1- 4815. polymerase fusion protein. 19. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. Thus ribozymes are likely to be good agents for selectively 20. Mizuno, T., Chou, M.-Y. & Inouye, M. (1984) Proc. Natl. blocking gene expression in bacteria. In cases in which a Acad. Sci. USA 81, 1966-1970. protein has a high turnover rate, ribozymes could substitute 21. Cameron, F. & Jennings, D. (1989) Proc. Natl. Acad. Sci. USA or for 86, 9139-9143. for specific chemical inhibitors of protein activity 22. Sarver, N., Cantin, E. M., Chang, P. S., Zaia, J. A., Ladne, temperature-sensitive mutations. And since ribozymes act at P. A., Stephens, D. A. & Rossi, J. J. (1990) Science 247, the level of RNA, they could be used to selectively block the 1222-1225. synthesis of new protein without affecting preexisting pro- 23. Saxena, S. & Ackerman, E. (1990) J. Biol. Chem. 265, 17106- tein, a feature that may be useful for studying essential 17109. that increase their rate of synthesis during major 24. Cotten, M. & Birnstiel, M. (1989) EMBO J. 8, 3861-3866. Downloaded by guest on September 29, 2021