@ A
`
`NNUAL REVIEW OF
`BIOCHEMISTRY
`
`VOLUME 55 , 1986
`
`
`CHARLES C. RICHARDSON, Editor
`Harvard Medical School
`
`PAUL D. BOYER, Associate Editor
`
`University of California, Los Angeles
`
`IGOR B. DAWID, Associate Editor
`
`National Institutes of Health
`
`ALTON MEISTER, Associate Editor
`
`Cornell University Medical College
`
`
`
`ANNUAL REVIEWS, INC.
`
`4139 EL CAMINO WAY
`
`PAID ALTO. CALIFORNIA 94306 USA
`
`1
`
`MTX1 042
`
`1
`
`MTX1042
`
`

`

`@l
`
`ANNUAL REVIEWS INC.
`
`Palo Alto, California, USA
`
`COPYRIGHT © 1986 BY ANNUAL REVIEWS 1Nc., PALO ALTO, CALIFORNIA, USA. ALL
`
`RIGHTS RESERVED. The appearance of the code at the bottom of the first page of an
`article in this serial indicates the copyright owner’s consent that copies of the
`article may be made for personal or internal use, or for the personal or internal use
`of specific clients. This consent is given on the condition, however, that the copier
`pay the stated per-copy fee of $2.00 per article through the Copyright Clearance
`Center, Inc (21 Congress Street, Salem, MA 01970) for copying beyond that
`permitted by Sections 107 or 108 of the US Copyright Law. The per-copy fee of
`$2. 00 per article also applies to the copying, under the stated conditions, of articles
`published1n any Annual Review serial before January 1, 1978. Individual readers,
`and nonprofit libraries acting for them, are permitted to make a single copy of an
`article without charge for use in research or teaching. This consent does not extend
`to other kinds of copying, such as copying for general distribution, for advertising
`or promotional purposes, for creating new collective works, or for resale. For such
`uses, written permission is required. Write to Permissions Dept. Annual Reviews
`Inc., 4139 El Camino Way, Palo Alto, CA 94306 USA.
`
`International Standard Serial Number: 0066-4309
`International Standard Book Number: 0—8243—0855-7
`
`Library of Congress Catalog Card Number: 50—13143
`
`Annual Review and publication titles are registered trademarks of Annual Reviews
`Inc.
`
`Annual Reviews Inc. and the Editors of its publications assume no responsibility
`for the statements expressed by the contributors to this Review.
`
`Typesetting by Kachina Typesetting Inc., Tempe, Arizona; John Olson, President
`Typesetting coordinator, Janis Hoffman
`
`PRINTED AND BOUND IN THE UNITED STATES OF AlvaRICA
`
`2
`
`

`

`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`Ann. Rev. Biochem. 1986. 55:569—97
`Copyright © 1986 by Annual Reviews Inc. All rights reserved
`
`THE ROLE OF ANTISENSE RNA IN
`
`GENE REGULATION
`
`Pamela J. Green] , Ophry Pines, and Masayori Inouye
`
`Department of Biochemistry, State University of New York at Stony Brook, Stony
`Brook, New York 11794
`
`
`
`CONTENTS
`
`PERSPECTIVES AND SUMMARY ..............................................................
`PROKARYOTIC ANT[SENSE REGULATORY SYSTEMS .................................
`Plasmid Replication ......................................................................
`
`:..................................
`Tn10 Transposition..................................
`
`
`Bacterial and Phage Gene Expression ..............................................
`
`Artificial Antisense RNA (micRNA) ..................................................
`mic Immune System ..........................................................................
`
`ANTISENSE RNA IN EUKARYOTIC SYSTEMS ........................................
`
`Amisense Genes .......................................................
`Microinjecfion QfAmisense RNA ...................................
`
`Possible Existence ofAntisense RNA in Eukalyotic Cells .................
`PROSPECTS ...........................................................................................
`
`569
`570
`570
`574
`576
`579
`582
`584
`584
`588
`590
`592
`
`PERSPECTIVES AND SUMMARY
`
`Gene expression in both prokaryotes and eukaryotes is controlled by the
`products of regulatory genes. According to a great number ofstudies in the past
`20 years, the products of such genes were determined to be proteins termed
`activators or repressors. Recently, naturally occurring regulatory genes have
`been discovered that direct the synthesis of RNA which can directly control
`gene expression. These newly discovered RNA repressors are highly specific
`inhibitors of gene expression. The regulatory RNA contains a sequence that is
`complementary to the target RNA, and binding of the two RNAs occurs by base
`
`xPresent Address: Laboratory of Plant Molecular Biology, The Rockefeller University, 1230
`York Avenue, New York, New York 10021
`
`569
`
`0066—4154/86/0701-0569$02.00
`3
`
`3
`
`

`

`570
`
`' GREEN, PINES & INOUYE
`
`pairing. The term “antisense RNA” has been coined to designate this regulatory
`RNA. The genes that direct the synthesis of antisense RNA are designated
`antisense genes.
`Antisense genes were initially discovered in prokaryotes. Within a relatively
`short period of time, antisense RNA has been identified in the regulation of
`diverse and complex phenomena in bacteria such as plasmid replication and
`incompatibility, Tn 10 transposition, osmoregulation of porin expression, regu-
`lation of phage reproduction, and autoregulation of CAMP-receptor protein
`synthesis. At present, naturally occurring antisense genes in eukaryotes have
`not been identified.
`
`Three lines of research have emerged following the discoveries of antisense
`genes: (a) searches for additional systems in which antisense RNA regulation is
`naturally involved, (b) the development of systems in which artificially con~
`structed antisense genes regulate a cellular or viral gene of interest, and (c)
`studies probing the mechanisms by which antisense RNA affects gene expres-
`sion.
`
`The finding that antisense RNA can inhibit gene expression in natural
`systems led to the development of strategies to artificially regulate genes using
`antisense RNA. With relatively simple manipulations, antisense RNA com-
`plementary to a chosen mRNA can be synthesized in vivo and may be used to
`inhibit the expression of the respective target gene. The function of endogenous
`genes has now been suppressed both in prokaryotes and eukaryotes by artificial
`antisense genes. In eukaryotes, direct microinjection of antisense RNA (syn-
`thesized in vitro) into cells has also resulted in the specific inhibition of gene
`expression. The potential use of antisense RNAs, for not only basic research but
`also applied research, is demonstrated by the types of genes that have been the
`successful targets of antisense RNA—mediated inhibition. These include, for
`example, developmentally regulated genes and genes that code for products
`that may be harmful for the host such as oncogenes and genes that are required
`for virus production. Thus, it should be possible to use antisense RNAs not only
`for gene therapy but also for preventing viral infection.
`As mentioned earlier, some of the requirements for antisense RNA regula-
`tion are understood. However, many aspects of the mechanism are still unclear.
`This ,review will attempt to cover the accumulated data and our current un-
`derstanding of antisense RNA regulation.
`
`PROKARYOTIC ANTISENSE REGULATORY SYSTEMS
`
`Plasmid Replication
`
`The regulatory role of small complementary RNAs first became apparent
`through the study of plasmid replication in Escherichia coli. For members of
`the ColEl and F11 Plasmid groups, replication is negatively controlled by a
`
`4
`
`

`

`ROLE OF ANTISENSE RNA
`
`571
`
`distinct untranslated RNA species approximately 100 nucleotides long (1, 2).
`In addition to regulating plasmid copy number, the small RNAs prohibit the
`stable maintenance of two similar plasmids in the same cell, a phenomenon
`known as incompatibility. Recently two small RNAs were shown to regulate
`the replication of a Staphylococcus aureus plasmid, pT181, in a similar way
`(3). Yet despite sharing a common purpose,
`the ColEl, FH, and pTl81
`regulatory RNAs have different modes of action.
`
`ColEl The negative control of replication of ColEl-type plasmids is accom-
`plished by RNA I which is transcribed in the opposite direction from the same
`DNA that encodes the primer RNA for DNA replication at the origin (1 , 4—6).
`Replication of ColEl initiates when the primer precursor (RNA H) hybridizes to
`its DNA template and is cleaved by ribonuclease H at the origin (7). Using the
`cleaved RNA as a primer, DNA synthesis by DNA polymerase I can ensue.
`RNA I functions to block processing of the primer precursor by hybridizing
`with it (1, 4) to prevent the formation of a suitable ribonuclease H substrate (i.e.
`a DNA : RNA hybrid). RNA I is complementary to nucleotides -552 to —447
`of RNA II (8) which initiates 555 nucleotides upstream (at -555) of the
`replication origin (7).
`Many mutant studies, which have been reviewed previously (9, 10), and
`ribonuclease sensitivity experiments (8, 11, 12) indicate that RNA I exists in a
`tRNA-like cloverleaf structure containing «three stem-and-loops. The three
`stem-and-loops (l , 6, 13) function together with the 5' end of RNA I (l l , 14,
`15) to achieve hybrid formation with RNA H. Detailed in vitro kinetic studies
`support an initial transient interaction between the loop regions of RNA I and
`RNA II followed by pairing beginning at the 5' end of RNA I as depicted in
`Figure 1A (11). These initial contacts then lead to complete hybrid formation.
`An alternative model for RNA 1 secondary structure where RNA I forms a
`dimer (by pairing in an antiparallel fashion) has also been proposed (9). The
`RNA I dimer has a double-stranded rather than a single-stranded cloverleaf
`structure and contains three interior loops. Closure of the interior loops has been
`proposed as a mechanism by which mutations might inactivate RNA 1. While
`this model provides an attractive explanation for the effects of several mutations
`(9), it is important to realize that a dimer of RNA I has not been observed thus
`far.
`
`Maximal negative control of plasmid replication by RNA I requires a ColEl-
`encoded protein consisting of 63 amino acid residues (16—19), referred to as
`Rop (16) or Rom (17). A target site involved in conferring sensitivity to
`Rop-RNA I-mediated inhibition exists between 52 and 135 nucleotides down-
`stream of the RNA H transcription start site (18, 19). Recent work indicates that
`Rop protein aids in the binding of RNA I to RNA H by specifically enhancing
`the initial reversible interaction between the two species (17, 20). Rop protein
`
`5
`
`

`

`572
`
`GREEN, PINES & INOUYE
`
` Iv
`
`V {-‘I
`
`Z .
`
`-" E;
`Ev
`
`WNW
`
`MMWMW__
`
`III
`
`(399
`Eli?
`
`VI
`
`40(11— _. _
`
`MNMNM‘MWIMMM _ __
`
`a
`
`p-
`I
`\
`bar: > f
`r ”(c
`Au.“
`AA‘
`5:?
`M:
`RM”
`A-l;
`u-A
`u-
`U-A
`c-5
`Mac
`u-c
`223
`“I:
`€36
`kt?
`CUU'A
`AC
`/U\
`g'g
`m
`a
`J
`
`c Iljiuggulgm“IllirllllII
`IIIIII
`leiWIiHIlji:ijfliz'“ ‘
`I
`5'
`.
`$26
`NEW ctAF
`u-g
`2’;
`\ /
`‘
`4“"er
`i'-NIJ: ~«mmmf‘dfm
`AW;“*“"9"“ “mm“
`“cccmm—-= I'm»:
`
`‘~.-n
`
`r
`
`IIIWEC—
`
`'
`
`M
`
`‘
`
`A. m
`m
`
`in the presence of RNA I may also elicit transcription termination or pausing as
`transcripts consistent with termination 200 (and possibly 110) nucleotides
`downstream of the primer start site have been observed in vitro (21). Such an
`effect has not been observed with ColEl (17). Rop requires RNA I for activity,
`but RNA I can also function alone. The contribution of Rop varies depending
`upon the conditions or the parameter being assayed. For example, Rop has been
`observed to enhance binding of RNA I and RNA II by twofold (l7), and
`decrease copy number (17) or in vitro replication (21) by threefold. In the
`absence of Rop approximately four times more RNA I is required to block
`pMBl primer processing in vitro (21).
`
`HI I Plasmids that belong to the FII incompatibility group are also regulated by
`a small untranslated RNA, but in contrast to the ColEl situation, the regulatory
`
`6
`
`

`

`ROLE OF ANTISENSE RNA
`
`573
`
`C
`
`G
`
`A
`
`U
`
`G
`
`C
`
`+l|
`
`crp mRNA
`
`A Divergent RNA
`+2
`
`+2
`A
`
`U
`+l4
`
`U
`
`A
`
`G
`
`C
`
`C
`
`G
`
`U
`
`A
`
`U
`
`A
`
`U
`
`U
`
`U
`
`+2
`
`crp mRNA
`
`RNA
`
` Divergent
`
`DNA
`
`Figure I Proposed interactions of three prokaryotic antisense RNAs and their respective target
`RNAs.
`.
`(A) Model for the stepwise binding of ColEl RNA I to RNA 11 (l 1). In step I RNA I and RNA 11
`interact at the loops of their folded structures. This interaction facilitates pairing (step II) that starts
`at the 5‘ end of RNA I (step III). At this stage, the loop-to-loop contacts may be broken. Progressive
`pairing continues as the stem-and-loop structures unfold (steps IV and V). It should be noted that
`the three pairs of loops do not necessarily interact simultaneously, and RNA 11 may interact in an
`alternative folded structure (11). This figure was adapted from references 11 and 20.
`(B) Model for hybrid formation between micF RNA and ompF mRNA (51). The complementary
`region of the micF RNA is sandwiched by two stable stem-and-loop structures: (a) A G = - 12.5
`and (b) A G = —4.5. Hybrid formation blocks translation of the ompF mRNA because the
`Shine-Dalgamo sequence and the AUG initiation codon are within the base-paired region. The
`formation of the hybrid may also promote rapid degradation of the ampF mRNA. This figure is
`adapted from Ref. 51.
`(C) Model for the inhibition of crp transcription by antisense RNA (57). Top: structure of the
`hybrid between the 5' end of the crp mRNA and the first 14 nucleotides of the antisense RNA is
`shown. The RNA : RNA hybrid resembles the initial segment of rho-independent terrninators (58).
`Bottom: the complete structure of the rho-independent terminator is shown. The structure consists
`ofthe crp mRNA-antisense RNA hybrid followed by an A-U—rich crp mRNA-DNA hybrid. Like in
`a rho-independent terminator, the RNA-RNA hybrid causes the polymerase to pause and the
`instability of the A-U—rich RNA-DNA hybrid leads to release of the crp transcript and termination
`of transcription. This figure was adapted from Ref. 57.
`
`7
`
`

`

`574
`
`GREEN, PINES & INOUYE
`
`RNA [called RNA 1 (2), RNA E (22), or CopA RNA (23)] likely acts as a
`translational repressor (22, 24, 25). The target of RNA 1 is the mRNA for the
`essential plasmid replication protein, RepAl (24, 26, 27). The repression
`mechanism is somewhat complicated by the fact that two RNA transcripts,
`RNA—CX and RNA-A (RNA-II) (24, 28—30), each contain the RepAl coding
`sequence and a 91—nucleotide upstream region that is complementary to RNA 1.
`RNA-CX and RNA-A are transcribed in the same direction from the same DNA
`
`but RNA-CX initiates about 380 bp upstream of RNA-A. In vivo, however, at
`or above the normal copy number, only RNA-CX appears to be a primary target
`for RNA I—mediated repression (31). The other RepAl mRNA, RNA-A, is
`controlled by the presence of excess amounts of the CopB protein (30a)
`(RepA2) (31), a transcriptional repressor (32, 32a, 32b). The 5' region of
`RNA-CX (not included in RNA—A) contains the coding sequence for CopB;
`thus in the RNA-CX transcript, the RNA I complementary region is sand—
`wiched between the RepAl and CopB coding sequences (24, 33). It has been
`proposed that the hybridization between RNA I and RNA-CX induces a change
`in RNA-CX secondary structure in the region of the RepAl ribosome binding
`site that blocks translation (24). The DNA sequences of several mutants
`indicate that a single-stranded loop in the RNA I secondary structure may be an
`important determinant for hybrid formation (27, 32a, 33, 34).
`
`pT181 Like the ColEl and F11 plasmids discussed above, the S. aureus plas-
`mid, pT181, has evolved a replication control mechanism whereby small
`complementary RNAs play a pivotal role. Two regulatory RNAs (RNA I and
`RNA II) (3) appear to repress the replication of pT181 by inhibiting the
`translation of the RepC protein, an essential replication factor (35—37). The
`RepC coding sequence is contained in both RNA III and RNA IV, which have
`different 5' ends (3). RNA I and RNA 11 differ at their 3' ends and are therefore
`complementary to the 5' untranslated regions of the two RepC mRNAs to
`different extents. The regulatory contributions of the individual pT181 tran-
`scripts have not been determined. RNA I and RNA II share no significant
`sequence homology with their ColEl or FII group counterparts; however there
`may be some similarities at the level of RNA secondary structure. Again, the
`analysis of copy number mutants (38) supports the contention that the primary
`interaction between regulatory and target transcripts involves contacts between
`complementary single-stranded loops (3).
`
`Tn] 0 Transposition
`
`Detailed analysis of the transposon TnlO has revealed that the activity of this
`element can also come under antisense RNA control. Transposition of TnlO is
`mediated by functions specified primarily by IS lO-Right (39, 40), one of two
`inverted repeat sequences that comprise the ends of the transposon. A long open
`
`8
`
`

`

`ROLE OF ANTISENSE RNA
`
`575
`
`reading frame encoding the transposase spans nearly the entire length of
`IS lO-Right from the outside towards the inside of the element. Transcription of
`the transposase gene is directed by a weak promoter called pIN (41, 42). The
`much stronger pOUT promoter, located just inside pIN, specifies the synthesis
`of an antisense RNA (RNA-OUT) that is complementary to the first 36 nucleo—
`tides of the transposase mRNA (RNA-IN) transcribed from pIN. Recently a
`70-nucleotide-long RNA was identified as the major in vivo RNA made from
`pOUT of ISlO-Right (42a). The concentration of this RNA was estimated at
`5—10 molecules per cell containing a single copy of TnlO.
`Several observations indicate that RNA-OUT negatively regulates the syn-
`thesis of the 1810 transposase by forming a hybrid with the transposase transla-
`tion initiation site region of RNA-IN (43, 44). First, single-copy TnlO elements
`transpose at much lower frequencies in E. coli when IS lO-Right is present on a
`multicopy plasmid (43). This phenomenon has been designated multicopy
`inhibition (MCI). Deletion and mutational analyses have limited the position of
`the determinants responsible for MCI to the outermost 180 bp of IS lO-Right.
`This region contains the pOUT transcription unit (41 , 42). Second,
`it has
`become clear that the mechanism of MCI involves inhibition of transposase
`gene expression. In addition, the mode of MCI is primarily posttranslational.
`MCI plasmids have been shown to decrease the expression of a transposase-
`lacZ gene fusion (Le. a translational fusion) by more than 90% while operon
`fusions (i.e. transcriptional fusions) were inhibited by less than 50% (43).
`Finally, pairing between RNA-IN and RNA-OUT has been demonstrated in
`vitro (44). Many mutant derivatives of RNA-OUT have been analyzed in this
`manner to determine the mechanism by which hybrid formation occurs. The
`results are most easily interpreted in terms of a potential stem-and—loop struc-
`ture in RNA-OUT that begins near the 5' end of the molecule and includes the
`region capable of pairing with RNA-IN. This structure can be divided into two
`domains: a 21-nucleotide loop domain and a stem domain consisting of about
`45 nucleotides. The 5 ' end region of RNA—IN is complementary to part of the
`RNA-OUT loop and one strand of the stem. Of all the mutations tested in an in
`vitro pairing assay, only two were found to decrease the ability of RNA-OUT to
`pair with wild-type RNA-IN. Both of the pairing-defective mutants clearly
`increase the base-pairing potential within the loop domain of RNA-OUT.
`RNA—OUT mutations that do not affect in vitro pairing between RNA-IN and
`RNA-OUT include one mutation in the loop that does not alter the base-pairing
`potential and several others that are located in the strand of the RNA-OUT stem
`that is not complementary to RNA-IN. While the latter set of mutations do not
`alter the efficiency of in vitro hybrid formation, several do decrease the
`calculated stability of the stem domain. Their lack of effect on the in vitro
`pairing reaction indicates that hybrid formation does not initiate at the base of
`the RNA-OUT stem. Based on these and other observations, a model has been
`
`9
`
`

`

`576
`
`GREEN, PINES & INOUYE
`
`proposed (J. D. Kittle, N. Kleckner, unpublished results) which predicts that
`pairing begins with a rate-limiting interaction between the 5' end of RNA-[N
`and the loop domain of RNA-OUT. After this initial pairing, one strand of the
`RNA-OUT stem is displaced as hybrid formation proceeds.
`It has been demonstrated previously (45) that transposition frequencies in
`vivo are related to transposase levels. Thus, it is easy to understand how the
`pOUT transcript, when synthesized from a multicopy plasmid, might control
`TnlO transposition frequency by inactivating the transposase mRNA. TnlO
`also contains other overlapping transcripts such as tetD RNA and tetC RNA
`(46, 47). However, at present, the activities of these RNAs as well as antisense
`RNA species possibly encoded by other nansposons (see Refs. cited in 42, 48,
`and 49) are more obscure.
`
`Bacterial and Phage Gene Expression
`
`Antisense RNAs are naturally encoded by both bacterial and phage genomes.
`The structural and functional analyses of these species and those described in
`the preceding two sections have led to a better understanding of the differing
`roles of antisense RNAs in complex regulatory schemes.
`
`micF The first natural E. coli antisense RNA to be identified was the micF
`
`RNA (50, 51). It was discovered during the characterization of ompC, a gene
`for one of the two major E. coli outer membrane porins. Surprisingly, several
`ampC promoter clones were found to repress the synthesis of the other major
`outer membrane porin, OmpF. The region responsible for this activity was
`localized to a 300-bp fragment upstream of the ampC promoter. Sequence
`analysis revealed that a stretch of approximately 80 bp was 70% homologous to
`the DNA encoding the 5' end of the ompF mRNA including the ribosome
`binding site and the ompF initiation codon. From 81 mapping and lacZ fusion
`studies, it became clear that the homologous DNA is transcribed in the opposite
`direction from ompC giving rise to a l74-nucleotide RNA that is com-
`plementary to a region of the ompF mRNA encompassing the translation
`initiation site. It was proposed that the complementary RNA, designated
`micRNA (mRNA-interfering complimentary RNA), inhibits ompF expression
`by forming a hybrid with the ompF mRNA (50, 51) (see Figure 18). This model
`predicted that artificial micRNAs could be used to regulate selected genes as
`discussed in the next section.
`
`Figure 13 shows that 44 nucleotides of the 5' untranslated leader region
`including the Shine-Dalgarno sequence, and 28 nucleotides of the coding
`region of the ompF mRNA, are included in the hybrid. Several small bulges and
`internal loops are present in the hybrid due to the short regions of nonhomology
`between the micF and ompF genes. The base-paired region of the micRNA
`(micF RNA) is flanked by two stable stem-and-loops, a and b (see Figure 18).
`
`10
`
`10
`
`

`

`ROLE OF ANTISENSE RNA
`
`577
`
`Based on this structure, it seems most likely that micF RNA inhibits OmpF
`synthesis by repressing translation.
`Northern blot experiments with a micF probe have detected the micF RNA
`(50) and two other smaller RNA species in cells containing micF on a multicopy‘
`plasmid. The T1 digestion products of these RNAs indicate that smaller RNAs
`are derived from micF and that the major micF RNA species (4.58) is missing
`about 80 nucleotides from the micF RNA 5' end (J. Andersen, N. Delihas,
`unpublished results). The physiological significance of the smaller RNAs and
`’the events leading to their production are not yet known. In similar ex-
`periments, the production of the micF RNA (50) was associated with a greatly
`diminished level of ompF mRNA. This could be due to degradation or pre-
`mature termination of the ompF message upon hybrid formation.
`Transcription of micF appears to be coordinated with the complex induction
`pathway leading to ompC mRNA synthesis (50). Two regulatory proteins,
`OmpR and EnvZ, control the expression of ompC and ompF in response to the
`osmolarity of the culture medium (for a review see Ref. 52). While the total
`amount of OmpC and OmpF in the outer membrane remains constant, the
`proportions of the individual proteins vary depending on the culture conditions.
`For example, in high-osmolarity medium, the outer membrane contains mainly
`OmpC and very little OmpF. The simultaneous transcription of ompC and micF
`predicts that the role of micF RNA is to rapidly repress OmpF synthesis as
`OmpC induction proceeds. The phenotype of a chromosomal mutation deleting
`ompC and the region upstream including micF (53) is consistent with this
`assessment. The deletion mutant synthesizes OmpF constitutively but has little
`effect on B-galactosidase transcription from the ompF promoter. The trans-
`lational regulatory function responsible for the OmpF constitutive phenotype of
`the deletion mutant is apparently present in the region upstream of ompC due to
`the fact that ompF is osmoregulated in ompC mutants.
`Although the inhibitory activity of micF RNA can adequately explain these
`results, several new observations indicate that the picture is somewhat more
`complicated. As discussed above, the chromosomal mutation that deletes both
`these genes along with ompC gives rise to constitutive OmpF synthesis (53).
`Yet when micF was replaced with the gene for kanamycin resistance, no
`marked effect on the osmoregulation of ompF was detected (54). However, the
`ompC gene of this mutant was expressed even in low osmolarity (54). A
`detailed analysis of the transcriptional regulation of micF is clearly necessary to
`reconcile these results which were obtained using different strains grown under
`different conditions.
`
`Synthesis of the E. coli cAMP receptor protein (CRP) is autoregulated
`crp
`both in vivo (55) and in vitro (56). Recently, cAMP—CRP was shown to repress
`transcription of the crp gene indirectly, by inducing the synthesis of an anti-
`
`11
`
`11
`
`

`

`578
`
`GREEN, PINES & INOUYE
`
`sense RNA (57). The transcription start sites of the antisense RNA and the crp
`mRNA are 3 bp apart and on opposite DNA strands. Although the sequence
`coding for the antisense RNA and the crp mRNA do not overlap, nucleotides
`2—6 and 10—14 of the antisense RNA are complementary to nucleotides 2—11 of
`the crp mRNA (see Figure 1C). CRP activates transcription of the antisense
`RNA in vivo and in vitro in the presence of cAMP. In vitro synthesis of
`antisense RNA specifically inhibits transcription from the crp promoter. This
`was demonstrated in two ways. First, the antisense promoter was inactivated by
`a linker insertion that did not affect the binding of CRP to its nearby site. In the
`absence of antisense transcription, in vitro transcription of crp from the muta-
`genized template was no longer repressed by the addition of CAMP-CRP.
`Moreover, the addition of purified antisense RNA to the reaction specifically
`inhibited crp transcription from either the normal or the mutagenized template.
`As shown in Figure 1C , the proposed hybrid formed between the antisense
`RNA and the crp mRNA is followed by an A-U—rich stretch. The resemblance
`of this structure to a rho-independent terminator (58) has led to the proposal that
`the antisense RNA acts by inducing premature termination of the crp mRNA.
`For this reason, the regulatory RNA has been designated antisense attenuator
`RNA. It is likely that antisense attenuator RNA will provide an additional
`approach to artificially regulate other genes of interest.
`
`PHAGE A Antisense RNA also seems to be one of the regulatory factors
`involved in A development. A promoter, designated PaQ, has recently been
`shown to direct antisense transcription of part or all of the amino-terminal
`region of the XQ gene coding sequence (59). Q protein is the antitemiinator
`required for A late gene transcription. The antisense transcript is thought to
`repress the synthesis of the Q protein by interfering with Q transcription or
`translation. Originally implied by its DNA sequence (D. Court, as cited in 59),
`the existence of PaQ was recently confirmed by in vitro transcription and genetic
`analysis (59). Transcription initiation at PaQ requires CH protein and a putative
`CH binding site has been identified. CH protein is also required to activate
`transcription of the integrase and A repressor genes which are necessary for the
`establishment of lysogeny. A mutation in the CH binding site (paql), which
`does not effect Q function, resulted in the relief of inhibition of late gene expression
`(in a )\ c185 7 cr020 phage) which is dependent upon CH. From these data it was
`concluded that PaQ was responsible in part for this phenomenon which has been
`designated “ClI dependent inhibition” (59). Furthermore, PaQ transcription appears
`to favor the lysogenic over the lytic pathway because the pan mutation causes
`)th857 to form clear rather than turbid plaques at low temperature.
`An additional antisense RNA which may play a role in x development is the
`GOP RNA ()t 48 RNA) (60, 61). The role of GOP RNA is unknown at present
`but its structure has been well characterized (62, 63). The 3' end of the
`
`12
`
`12
`
`

`

`ROLE OF ANTISENSE RNA
`
`579
`
`77—nucleotide-long oop transcript is complementary to the last 17 amino acid
`codons of the C11 mRNA coding sequence. The region of GOP RNA that could
`hybridize to 10 out of the 17 complementary CII codons is contained within the
`putative secondary structure thought to be the COP RNA terminator. 00p
`transcription initiates in the intercistronic region between the 0 and (:11 genes,
`just upstream of the Shine-Dalgarno sequence for the 0 protein. The cap gene is
`located near the x ori region but it has been shown to be dispensable for the
`normal function of the origin (64) and therefore is unlikely to serve as the primer
`for the initiation of DNA replication. At present, there is no evidence that OOP
`RNA functions to repress the expression of CI], but the possibility could add an
`interesting twist to the emerging role of the antisense RNA transcribed from PaQ
`which is induced by CH (see above).
`
`PHAGE P22 Another intricate regulatory system involving a small antisense
`RNA has been identified in the Salmonella phage P22 (65). The small antisense
`RNA, called sar RNA, is complementary to the Shine-Dalgarno sequence of
`the mRNA for the P22 antirepressor protein, Ant, as well as the region between
`ant and the cotranscribed gene arc. The sar RNA contains no open reading
`frame and has been shown to inhibit antirepressor synthesis in trans. Mutations
`in the sar promoter that decrease promoter strength in vitro (66) give rise to
`higher levels of antirepressor synthesis in vivo. It has been proposed that sar
`RNA negatively regulates ant expression by forming a hybrid with the ant
`mRNA which represses translation (65).
`
`Artificial Antisense RNA (micRNA)
`
`The discovery of micF RNA prompted the development of a system designed to
`artificially regulate bacterial gene expression with antisense RNA (67). In
`addition to facilitating the controlled expression of various bacterial genes, the
`artificial “mic" system has provided insight as to the mechanism of antisense
`regulation and the characteristics of effective antisense RNAs. It was reasoned
`that an artificial mic gene2 could be created by positioning a DNA fragment
`coding for a portion of an mRNA between a suitable promoter and transcription
`terminator,
`in the antisense orientation. The mic transcripts from such a
`construct would be complementary to the target mRNA over the region covered
`by the cloned DNA fragment.
`The mic cloning vector pJDC402, shown in Figure 2A, has been used to
`efficiently construct and regulate a number of antisense genes in E. coli (67,
`
`2The term “mic gene" has been used to define a gene capable of encoding an RNA com-
`plementary to all or part of a specific mRNA (67). The target gene for the micRNA is shown in
`parentheses. For example, transcription of a mic (lpp) gene will result in the production of RNA
`complementary to the lpp mRNA.
`
`13
`
`13
`
`

`

`580
`
`GREEN, PINES & INOUYE
`
`PP gene
`
`
`
`X= _'/////////////.Ipl'
`
`.
`
`mm
`
`. mic 1
`
`ene
`
`( PP) 8
`Ipfl
`
`"canon mlo Xba! sin a! pJDC402
`I
`lu'o
`,(
`W \\\\\\\\\\\\\-m
`
`s’
`
`a:
`
`b:
`
`n
`n
`
`s'
`
`l
`
`'
`
`transcripluon
`0M
`i
`i
`
`on
`
`s'
`l'
`
`[23‘ Inn-l
`
`(I): Dual)
`
`.
`
`Figure 2 Structure of the mic vector pJDC402 and construction of a mic(lpp) gene
`(A) Structure of pJDC402. lpp" (cross—hatches) and lac?“ (solid dots) represent the lipoprotein
`promoter and the lactose promoter operator, respectively. The promoters are separated from the lpp
`transcription termination region (open dots) by a unique Xbal site (X). The PvuII (P) and the EcoRI
`(E) sites are also indicated. Ampr designates the ampicillin resistance gene.
`(B) Construction of a mic(lpp) gene using pIDC402. Open arrows represent promoters. The open
`bar represents the region coding for the 5' nontranslated region of the lpp mRNA. The slashed bar
`and the solid bar correspond to