Derivation and function of small interfering RNAs and microRNAs

Virus Research 102 (2004) 3–9

Derivation and function of small interfering RNAs and microRNAs Bryan R. Cullen∗ Department of Molecular Genetics and Microbiology, Howard Hughes Medical Institute, Duke University Medical Center, Durham, NC 27710, USA

Abstract Small interfering RNA (siRNA) duplexes are generally produced by Dicer cleavage of double-stranded RNAs of frequently exogenous origin and can induce the cleavage and degradation of mRNAs bearing an identical sequence. In contrast, microRNAs (miRNAs) are encoded within the eukaryotic genome as short RNA hairpin structures. While these pre-miRNAs are also processed by Dicer, mature miRNAs appear to function primarily by inhibiting the translation of mRNAs bearing multiple, partially mismatched target sites. Nevertheless, recent data argue that the posttranscriptional regulatory machinery utilized by siRNAs and miRNAs is largely or entirely identical. In this review, I will discuss recent progress in unraveling the RNA processing pathway utilized for the biosynthesis of mature miRNAs and argue that this pathway offers at least three distinct entry points for the functional expression of artificial siRNAs in vertebrate cells. While each of these entry points offers distinct advantages and disadvantages, they all have the potential to induce the effective knock-down of specific genes either in cell culture or in experimental animals. © 2004 Elsevier B.V. All rights reserved. Keywords: miRNA; RNA interference; RNA processing; siRNA

1. Introduction RNA interference (RNAi) was first defined in C. elegans as a mechanism that induced the posttranscriptional silencing of genes in response to the introduction of long double-stranded RNAs (dsRNAs) of identical sequence (Fire et al., 1998). It was soon realized that RNAi was closely related to previously described, but poorly understood, gene inactivation pathways in plants and certain fungi (reviewed by Bernstein et al., 2001b). While long dsRNAs also proved highly effective at inducing RNAi in Drosophila (Hammond et al., 2000), initial efforts to demonstrate RNAi in vertebrate cells were largely unsuccessful. This failure resulted from the fact that long dsRNAs also activate the vertebrate interferon response, a complex defense system, lacking in non-vertebrate species, that leads to a global posttranscriptional inhibition of gene expression (Sen, 2001). Efforts to understand how RNAi works in plants and invertebrates demonstrated that long dsRNAs are initially cleaved into characteristic ∼21 nt dsRNAs, bearing 2 nt 3 overhangs, termed small interfering RNA (siRNA) duplexes (Fig. 1) (Hamilton and Baulcombe, 1999; Hammond et al.,

∗

Tel.: +1-919-684-3369; fax: +1-919-681-8979. E-mail address: [email protected] (B.R. Cullen).

0168-1702/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.virusres.2004.01.009

2000). This cleavage is mediated by a processive cytoplasmic RNase III family enzyme called Dicer (Bernstein et al., 2001a; Grishok et al., 2001; Hutvágner et al., 2001; Knight and Bass, 2001). One strand of this siRNA duplex is then selectively incorporated into a large protein complex termed the RNA induced silencing complex or RISC (Hammond et al., 2000; Nykänen et al., 2001). This RNA strand then acts as a guide RNA to target RISC to homologous mRNAs (Martinez et al., 2002; Schwarz et al., 2003). Once RISC is bound, a currently unidentified ribonuclease component cleaves the target mRNA opposite the center of the bound guide siRNA (Fig. 1). RISC is then released, to seek additional mRNA targets, while the two fragments of the mRNA are degraded by cellular exonucleases (Hutvágner and Zamore, 2002). A critical discovery, made initially in Drosophila and subsequently confirmed in vertebrate cells, is that synthetic siRNA duplexes can also program RISC and induce the inactivation of specific target genes (Elbashir et al., 2001a,b). This result was made possible in vertebrate cells by the fact that the interferon response is activated by dsRNAs of >30 bp (Manche et al., 1992), while siRNA duplexes bear a double-stranded region of only ∼19 bp. This observation demonstrated that the RNAi pathway was indeed fully active in mammalian cells and raised the possibility, now confirmed, that RNAi could be used to perform reverse

4

B.R. Cullen / Virus Research 102 (2004) 3–9

Fig. 1. MicroRNA and siRNA processing and function. A proposed processing pathway for endogenously encoded miRNAs is shown at the center of the figure. This proceeds from transcription through nuclear processing, nuclear export, cytoplasmic processing by Dicer and finally to incorporation into RISC. This results in the initial primary miRNA transcript (pri-miRNA) being processed to a pre-miRNA and finally to a miRNA duplex intermediate, one strand of which then serves as a substrate for RISC incorporation. The assembled RISC can then target mRNAs bearing a perfectly complementary target site for degradation or can inhibit the translation of an mRNA that contains multiple, partly mismatched target sites. Entry points for designed siRNAs are shown at the left of the figure. These include synthetic siRNA duplexes as well as short hairpin RNAs, transcribed from expression plasmids, that closely mimic pre-miRNAs. Finally, artificial miRNA genes can also be used to express novel siRNAs. Long dsRNAs, which can generate siRNAs in invertebrate and plant cells, activate the interferon response in vertebrate cells and are therefore not useful in the latter system.

genetic analyses in human cells (reviewed by Dykxhoorn et al., 2003).

2. The discovery of microRNAs As noted above, long dsRNAs are processed by Dicer to give ∼21 bp siRNA duplexes, one strand of which is then incorporated into RISC. Efforts in the Drosophila and C. elegans systems to clone these short non-coding RNAs based on their size led to the surprising discovery that not only these organisms, but also vertebrates, express over 200 genomically encoded ∼21 nucleotide RNAs (Lagos-Quintana

et al., 2001; Lau et al., 2001; Lee and Ambros, 2001). These RNAs, which are now termed microRNAs (miRNAs), are predicted to form part of one strand of an imperfect precursor RNA hairpin structure of between 60 and 90 nucleotides in length. This contrasts with siRNAs, which are invariably derived from long dsRNAs. In addition, while miRNAs are by definition of endogenous origin, siRNAs can be endogenous or exogenous. While the discovery of this large family of miRNAs was unexpected, two miRNAs had in fact been known for several years in C. elegans. These miRNAs, termed let-7 and lin-4, emerged from mutational screens designed to identify genes involved in regulating larval development in C. elegans

B.R. Cullen / Virus Research 102 (2004) 3–9

(Lee et al., 1993; Reinhart et al., 2000). Both let-7 and lin-4 are 21 nt in length and both are initially transcribed as one arm of an ∼70 nt stem-loop RNA precursor. Both let-7 and lin-4 are expressed in a developmentally regulated manner and both inhibit the expression of specific, developmentally important mRNAs by forming duplexes with multiple conserved sequence elements present in their 3 untranslated region (3 UTR). However, unlike siRNAs, this interaction specifically blocks the translation of these target mRNAs, rather than inducing their degradation (Fig. 1) (Olsen and Ambros, 1999). Interestingly, known target sequences for let-7 and lin-4 bear central mismatches so that the bound let-7 and lin-4 miRNAs would be predicted to form bulges (Lee et al., 1993; Reinhart et al., 2000). Recent evidence indicates that it is these mismatches that prevent target mRNA cleavage by these miRNAs, and not some intrinsic difference in miRNA versus siRNA function (see below). Although many different miRNAs have been reported in a range of species, very few have as yet been assigned a function. Exceptions include not only let-7 and lin-4 in C. elegans but also two miRNAs termed bantam and miR-14 in Drosophila, miR-171 in Arabidopsis and miR-23 in human cells (Llave et al., 2002; Brennecke et al., 2003; Kawasaki and Taira, 2003; Xu et al., 2003). As in the case of let-7 and lin-4, these miRNAs are expressed in a developmentally regulated manner and inhibit the expression of target mRNAs bearing complementary sequences. Consistent with the hypothesis that other miRNAs may also play a role in development, many vertebrate and non-vertebrate miRNAs have been found to be expressed in a tissue specific and/or developmentally regulated manner (Lagos-Quintana et al., 2001, 2002; Lau et al., 2001; Lee and Ambros, 2001). Moreover, mutational disruption of genes encoding proteins that play a global role in miRNA biogenesis, such as Dicer, has profound deleterious affects on the development of mutant organisms (Grishok et al., 2001; Park et al., 2002).

3. Expression, processing and function of microRNAs The emerging importance of miRNAs in regulating the appropriate development and differentiation of multicellular organisms, and the similarity of miRNAs to siRNAs, has prompted efforts to more fully understand miRNA derivation and function. Although these studies remain incomplete, the pathway delineated in Fig. 1 is likely to represent a fairly accurate overview of this process. miRNAs are found either as individual miRNA genes or in tandemly arrayed clusters that may suggest a functional relationship analogous to a bacterial operon (Lagos-Quintana et al., 2001; Lau et al., 2001). In any event, the miRNA genes that have been studied in detail thus far, such as human miR-30, appear to be transcribed as part of a several hundred nucleotide long primary transcript termed a pri-miRNA (Fig. 1) (Lee et al., 2002; Zeng and Cullen, 2003). In at least some cases, transcription is mediated by

5

RNA polymerase II, and it is therefore likely that these pri-miRNAs are capped and polyadenylated. It remains possible that other pri-miRNAs might be transcribed by other cellular polymerases. The first step in miRNA processing is the excision of the ∼70 nt miRNA hairpin intermediate, termed the pre-miRNA, from the much longer pri-miRNA (Fig. 1) (Lee et al., 2002; Zeng and Cullen, 2003). This cleavage, which occurs largely or exclusively in the nucleus, is mediated by an enzyme termed RNase III or Drosha, a member of the family of large RNases that also includes Dicer (Lee et al., 2003b). Analysis of the pre-miRNA cleavage product that results from processing of the miR-30 pri-miRNA shows that this cleavage both defines the 3 end of the miR-30 miRNA and leaves a 2 nt 3 overhang (Lee et al., 2002; Zeng and Cullen, 2003). Importantly, this cleavage does not occur at the base of the stem of the predicted 80 nt miR-30 precursor RNA hairpin but rather ∼7 nt above the base, resulting in a 63 nt pre-miRNA intermediate. The sequences that determine the sites of cleavage of the miR-30 pri-miRNA by RNase III remain to be defined. However, it is clear that the helical nature of the basal region of the miR-30 stem-loop precursor plays a critical role and that insertions or deletions in the precursor stem can result in a shift in the cleavage sites (Zeng and Cullen, 2003). The next step in the formation of the mature miR-30 miRNA is export of the pre-miRNA to the cytoplasm. Recent evidence (Yi et al., 2003) indicates that the factor responsible for nuclear export of pre-miRNAs is Exportin 5 (Exp5), a member of the karyopherin family of nucleocytoplasmic transport factors (Fig. 1) (reviewed by Cullen, 2003). Exp5 is known to mediate the nuclear export of a range of other small non-coding RNAs, including the adenovirus VA1 RNA and the human Y1 RNA (Gwizdek et al., 2003). Like other karyopherin family members involved in nuclear export, Exp5 is dependent on the GTP-bound form of the Ran co-factor for specific binding to its export substrate in the cell nucleus and is predicted to release its RNA cargo in the cytoplasm after hydrolysis of Ran·GTP to Ran·GDP by the cytoplasmic Ran GTPase activating protein. Although pre-miRNAs are recognized as specific substrates for Exp5 mediated nuclear export (Yi et al., 2003), the characteristics of the pre-miRNA that mediate this recognition remain to be determined. Dicer is a processive, ATP-dependent RNase that binds with high affinity to the ends of dsRNAs bearing 2 nt 3 overhangs, i.e. the product that it normally generates during processing of long dsRNAs (Zhang et al., 2002). Dicer then cleaves both RNA strands ∼21 nt from the bound end. The structure of the pre-miRNA, which consists of an RNA hairpin ending with a 2 nt 3 overhang (Zeng and Cullen, 2003), is closely analogous to intermediates generated during processing of long dsRNAs except that one end is closed by a loop. Importantly, the fact that Dicer cleaves at a predetermined distance from the end of the pre-miRNA hairpin means that the structure of the pre-miRNA predetermines

6

B.R. Cullen / Virus Research 102 (2004) 3–9

the sequence of the miRNA duplex intermediate that is generated by Dicer processing. This emphasizes the importance of the initial, nuclear RNA processing event in determining the final product of this pathway (Fig. 1). Although it is now clearly established that Dicer processing of pre-miRNAs gives rise to a duplex RNA intermediate, this is generally short-lived and therefore hard to detect. However, while the majority of miRNA precursors give rise to only a single mature miRNA, a small number, including miR-30, do give rise to mature miRNAs derived from both strands (Zeng et al., 2002). This allows the structure of the miRNA duplex intermediate to be clearly deduced. In fact, this miRNA duplex is identical in structure to an siRNA duplex, with the caveat that miRNA precursors frequently contain one or two mismatches in the stem, while siRNAs are generally perfectly double-stranded (Fig. 1). The short half-life of the miRNA duplex intermediate is apparently due largely to the fact that assembly of one strand of the duplex into RISC is rapid and efficient. It is believed that a helicase component of RISC binds to one end of the duplex and unravels the double-strand, a process that requires ATP (Nykänen et al., 2001). The strand whose 5 end is at the bound end of the duplex is then incorporated into RISC. This process is non-random, in that RISC binding strongly favors the end of the miRNA duplex intermediate that is less tightly helical (Schwarz et al., 2003). Therefore, it is commonly observed that the 5 end of the miRNA strand incorporated into RISC is poorly base paired in the predicted duplex intermediate, while the 5 end of the excluded strand forms a strong basepair, such as an G:C (Lagos-Quintana et al., 2001; Schwarz et al., 2003). While the composition of the RISC complex remains to be fully established, it is clear that RISC contains one or more members of the Argonaut family of proteins (Hammond et al., 2001; Carmell et al., 2002; Mourelatos et al., 2002). Certain Argonaut proteins have been found to be essential for miRNA function in C. elegans and for siRNA function in cultured human cells (Tabara et al., 1999; Doi et al., 2003). However, their actual role, and whether different human Argonaut proteins have different roles, remains to be determined. As noted above, the C. elegans let-7 and lin-4 miRNAs inhibit the expression of target mRNAs after interacting with multiple imperfect target sites in the 3 UTR. Importantly, this does not result in a marked drop in the cytoplasmic level of the target mRNA (Olsen and Ambros, 1999). Inhibition at the translational level has therefore been inferred, even though the polysome profile of the target mRNAs is largely unaffected. While the mechanism underlying this inhibition remains unclear, it appears to be cooperative in that the binding of multiple RISC complexes is required (Doench et al., 2003). Unlike mRNA cleavage by RISC, which allows turnover of RISC (Hutvágner and Zamore, 2002), this translational inhibition also appears to require stoichiometric levels of RISC and hence of the relevant miRNA (Zeng et al., 2003). Translational inhibition is therefore less effec-

tive, particularly at lower levels of miRNA/siRNA expression, than is inhibition via mRNA cleavage. However, the requirement for binding of multiple RISC complexes for effective mRNA translation inhibition does offer the potential for coordinate regulation of mRNA expression by multiple, distinct miRNAs. Although it initially seemed possible that miRNAs and siRNAs might be functionally distinct, with the former inducing translational inhibition and the latter mRNA cleavage, this is no longer believed to be true. Endogenously encoded or overexpressed human miRNAs have now been shown to induce the cleavage of artificial mRNA substrates bearing perfectly homologous target sites both in vitro and in vivo (Hutvágner and Zamore, 2002; Zeng et al., 2003). Similarly, artificial siRNAs have been shown to inhibit gene expression from mRNAs bearing imperfect targets without affecting mRNA expression levels (Doench et al., 2003; Zeng et al., 2003). Together, these data demonstrate that siRNAs and miRNAs are able to program RISC in a functionally indistinguishable manner (Fig. 1). On the other hand, it does remain possible that mRNA cleavage and mRNA translation inhibition are mediated by distinct forms of RISC, perhaps containing different members of the Argonaut protein family. However, in that case these alternate forms must be equivalently programmable by both siRNAs and miRNAs.

4. Strategies for expression of designed siRNAs The miRNA processing pathway delineated in Fig. 1 is conserved in most, possibly all, higher eukaryotic cells and can be used to promote the entry of exogenously encoded or synthetic siRNAs into this highly effective posttranscriptional pathway of gene inactivation. In many invertebrates, long dsRNAs are an effective means of inducing gene specific RNAi. However, as noted above, long dsRNAs activate the interferon response in vertebrate cells and are therefore not useful in these organisms. To overcome this problem, it is possible to program RISC by transfection of cells with synthetic siRNA duplexes that closely mimic the miRNA duplex intermediate (Elbashir et al., 2001a). This approach can result in the specific and effective destruction of target mRNAs (Fig. 1). However, synthetic siRNAs can be quite expensive and RNA transfection is not efficient in all cell types of interest, particularly primary cells. Moreover, because only a finite amount of the siRNA duplex is introduced into cells, inhibition is lost over a period of a few days due to dilution in the growing cell culture and/or degradation of the transfected siRNAs. To overcome this problem, several groups have developed plasmids that can express short hairpin RNAs that are structurally analogous to pre-miRNA hairpin intermediates, using promoters dependent on RNA polymerase III (Brummelkamp et al., 2002; Paddison et al., 2002; Sui et al., 2002). In a typical expression plasmid of this type, the U6 or H1 promoter drives the expression of an RNA

B.R. Cullen / Virus Research 102 (2004) 3–9

that is predicted to fold into a short RNA hairpin, with the sequence of one side of the stem being complementary to the mRNA target sequence of interest. Transcription termination is induced by a run of five thymidine residues, which causes termination of transcription by RNA polymerase III after the second thymidine residue. This allows the short hairpin to acquire a 2 nt 3 overhang, exactly like authentic pre-miRNAs, and recent evidence indicates that these short hairpin RNAs are exported from the nucleus by Exp5, just like an authentic pre-miRNA (Fig. 1) (Yi et al., 2003). Subsequent work has refined the parameters for design of short hairpin RNAs for maximal likelihood of successful RNAi. Thus, it is now clear that stems of from 25 to 29 bp in length give rise to better RNA processing and siRNA expression than do shorter stems (stems >29 bp in length can induce the interferon response). Because Dicer cleaves ∼21 nt from the base of the short hairpin RNA, after binding to the end of this RNA structure, it is the more basal sequence of the stem that gives rise to the siRNA duplex intermediate (Lee et al., 2003a). To promote incorporation of the antisense strand into RISC, it is important to ensure that the projected 5 end of the antisense strand is poorly basepaired (Schwarz et al., 2003). Even in the absence of these modifications, short hairpin RNA expression plasmids have proven to be quite effective in inducing RNAi. However, these plasmids again have practical problems in terms of the ability to effectively transfect cells of interest and in terms of the transience of the inhibitory effect observed. To overcome these problems, several groups have developed viral vectors, based on murine leukemia virus or human immunodeficiency virus that incorporate short hairpin RNA expression cassettes (Barton and Medzhitov, 2002; Lee et al., 2003a; Qin et al., 2003; Stewart et al., 2003). These vectors can be used to efficiently infect target cells in culture or in vivo and have been shown to give rise to long term expression of the encoded siRNA and, hence, to the long term suppression of the target mRNA of interest. Most recently, lentivirus-based short hairpin RNA expression vectors have been used to effectively and stably suppress the expression of genes in transgenic mice (Rubinson et al., 2003; Tiscornia et al., 2003), thus hopefully heralding a new era of simple reverse genetic experiments in experimental animals. A final approach to the expression of siRNAs uses as its entry point the first step in the expression pathway for endogenous miRNAs (Fig. 1). Because pri-miRNAs contain all the sequences required for miRNA processing in cis, it should in principle be possible to switch the part of the miRNA gene that encodes the mature miRNA, while leaving the processing signals intact, and hence use the entire miRNA processing machinery to produce an siRNA, or artificial miRNA, of interest. In fact, using the miR-30 gene as a starting point, this has indeed been achieved using several different siRNA sequences (Zeng et al., 2002; Zeng and Cullen, 2003). A major advantage of this approach to siRNA expression is that it relies on transcription via RNA polymerase II and thus has the potential to permit the use of

7

tissue-specific or developmentally-regulated promoters for expression of particular siRNAs in vivo. In contrast, the level of expression of siRNAs transcribed from vectors that depend on RNA polymerase III is not easily controlled, although this approach does permit the constitutive expression of high siRNA levels.

5. Conclusions Although RNAi plays a critical role in antiviral defense in plants, and is important in inhibiting transposon activation in invertebrates (Plasterk, 2002), it remains unclear whether RNAi serves similar functional roles in vertebrate species. What is clear is that miRNAs are expressed in all higher eukaryotes and, at minimum, play a critical role in the posttranscriptional regulation of genes during differentiation and development. The pathway that mediates miRNA processing and function is becoming increasingly well understood and it is now apparent that this pathway can support several different strategies for the introduction of functional, exogenously produced siRNAs (Fig. 1). While it remains uncertain whether RNAi plays any role in controlling virus infections in animals, there is no question that the artificial induction of RNAi can be used to target viral transcripts, or host genes that produce co-factors critical for virus replication, and thereby effectively block virus replication in culture. Whether RNAi will lead to effective antiviral treatments in the future remains to be seen. What is clear, however, is that RNAi is an extremely powerful tool to identify and characterize cellular factors that are required for effective virus replication. In doing so, RNAi has the potential to identify targets for chemotherapeutic intervention that would be very difficult to define using more conventional approaches.

References Barton, G.M., Medzhitov, R., 2002. Retroviral delivery of small interfering RNA into primary cells. Proc. Natl. Acad. Sci. U.S.A. 99, 14943– 14945. Bernstein, E., Caudy, A.A., Hammond, S.M., Hannon, G.J., 2001a. Role for bidentate ribonuclease in the initiation step of RNA interference. Nature 409, 363–366. Bernstein, E., Denli, A.M., Hannon, G.J., 2001b. The rest is silence. RNA 7, 1509–1521. Brennecke, J., Hipfner, D.R., Stark, A., Russell, R.B., Cohen, S.M., 2003. Bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila. Cell 113, 25–36. Brummelkamp, T.R., Bernards, R., Agami, R., 2002. A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550–553. Carmell, M.A., Xuan, Z., Zhang, M.Q., Hannon, G.J., 2002. The argonaute family: tentacles that reach into RNAi, developmental control, stem cell maintenance, and tumorigenesis. Genes Dev. 16, 2733–2742. Cullen, B.R., 2003. Nuclear RNA export. J. Cell Sci. 116, 587–597. Doench, J.G., Petersen, C.P., Sharp, P.A., 2003. siRNAs can function as miRNAs. Genes Dev. 17, 438–442.

8

B.R. Cullen / Virus Research 102 (2004) 3–9

Doi, N., Zenno, S., Ueda, R., Ohki-Hamazaki, H., Ui-Tei, K., Saigo, K., 2003. Short-interfering-RNA-mediated gene silencing in mammalian cells requires Dicer and eIF2C translation initiation factors. Curr. Biol. 13, 41–46. Dykxhoorn, D.M., Novina, C.D., Sharp, P.A., 2003. Killing the messenger: short RNAs that silence gene expression. Nat. Rev. 4, 457–467. Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., Tuschl, T., 2001a. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Elbashir, S.M., Lendeckel, W., Tuschl, T., 2001b. RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 15, 188–200. Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E., Mello, C.C., 1998. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811. Grishok, A., Pasquinelli, A.E., Conte, D., Li, N., Parrish, S., Ha, I., Baillie, D.L., Fire, A., Ruvkun, G., Mello, C.C., 2001. Genes and mechanisms related to RNA interference regulate expression of the small temporal RNAs that control C. elegans developmental timing. Cell 106, 23–34. Gwizdek, C., Ossarch-Nazari, B., Brownawell, A.M., Doglio, A., Bertrand, E., Macara, I.G., Dargemont, C., 2003. Exportin-5 mediates nuclear export of minihelix-containing RNAs. J. Biol. Chem. 278, 5505– 5508. Hamilton, A.J., Baulcombe, D.C., 1999. A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286, 950–952. Hammond, S.M., Bernstein, E., Beach, D., Hannon, G.J., 2000. An RNA-directed nuclease mediates post-transcriptional gene silencing in Drosophila cells. Nature 404, 293–295. Hammond, S.M., Boettcher, S., Caudy, A.A., Kobayashi, R., Hannon, G.J., 2001. Argonaute 2, a link between genetic and biochemical analyses of RNAi. Science 293, 1146–1150. Hutvágner, G., McLachlan, J., Pasquinelli, A.E., Bálint, É., Tuschl, T., Zamore, P.D., 2001. A cellular function for the RNA-interference enzyme Dicer in the maturation of the let-7 small temporal RNA. Science 293, 834–838. Hutvágner, G., Zamore, P.D., 2002. A microRNA in a multiple-turnover RNAi enzyme complex. Science 297, 2056–2060. Kawasaki, H., Taira, K., 2003. Hes1 is a target of microRNA-23 during retinoic-acid-induced neuronal differentiation of NT2 cells. Nature 423, 838–842. Knight, S.W., Bass, B.L., 2001. A role for the RNase III enzyme DCR-1 in RNA interference and germ line development in Caenorhabditis elegans. Science 293, 2269–2271. Lagos-Quintana, M., Rauhut, R., Lendeckel, W., Tuschl, T., 2001. Identification of novel genes coding for small expressed RNAs. Science 294, 853–858. Lagos-Quintana, M., Rauhut, R., Yalcin, A., Meyer, J., Lendeckel, W., Tuschl, T., 2002. Identification of tissue-specific microRNAs from mouse. Curr. Biol. 12, 735–739. Lau, N.C., Lim, L.P., Weinstein, E.G., Bartel, D.P., 2001. An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science 294, 858–862. Lee, M.-T., Coburn, G.A., McClure, M.O., Cullen, B.R., 2003a. Inhibition of HIV-1 replication in primary macrophages using Tat or CCR5-specific small interfering RNAs expressed from a lentiviral vector. J. Virol. 77, 11964–11972. Lee, R.C., Ambros, V., 2001. An extensive class of small RNAs in Caenorhabditis elegans. Science 294, 862–864. Lee, R.C., Feinbaum, R.L., Ambros, V., 1993. The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75, 843–854. Lee, Y., Ahn, C., Han, J., Choi, H., Provost, P., Kim, S., Kim, V.N., 2003b. The nuclear RNase III drosha initiates microRNA processing. Nature 425, 415–419. Lee, Y., Jeon, K., Lee, J.T., Kim, S., Kim, V.N., 2002. MicroRNA maturation: stepwise processing and subcellular localization. EMBO J. 21, 4663–4670.

Llave, C., Xie, Z., Kasschau, K.D., Carrington, J.C., 2002. Cleavage of scarecrow-like mRNA targets directed by a class of Arabidopsis miRNA. Science 297, 2053–2056. Manche, L., Green, S.R., Schmedt, C., Mathews, M.B., 1992. Interactions between double-stranded RNA regulators and the protein kinase DAI. Mol. Cell. Biol. 12, 5238–5248. Martinez, J., Patkaniowska, A., Urlaub, H., Lührmann, R., Tuschl, T., 2002. Single-stranded antisense siRNAs guide target RNA cleavage in RNAi. Cell 110, 563–574. Mourelatos, Z., Dostie, J., Paushkin, S., Sharma, A., Charroux, B., Abel, L., Rappsilber, J., Mann, M., Dreyfuss, G., 2002. miRNPs: a novel class of ribonucleoproteins containing numerous microRNAs. Genes Dev. 16, 720–728. Nykänen, A., Haley, B., Zamore, P.D., 2001. ATP requirements and small interfering RNA structure in the RNA interference pathway. Cell 107, 309–321. Olsen, P.H., Ambros, V., 1999. The lin-4 regulatory RNA controls developmental timing in Caenorhabditis elegans by blocking LIN-14 protein synthesis after the initiation of translation. Dev. Biol. 216, 671–680. Paddison, P.J., Caudy, A.A., Bernstein, E., Hannon, G.J., Conklin, D.S., 2002. Short hairpin RNAs (shRNAs) induce sequence-specific silencing in mammalian cells. Genes Dev. 16, 948–958. Park, W., Li, J., Song, R., Messing, J., Chen, X., 2002. Carpel factory, a Dicer homolog, and HEN1, a novel protein, act in microRNA metabolism in Arabidopsis thaliana. Curr. Biol. 12, 1484–1495. Plasterk, R.H.A., 2002. RNA silencing: the genome’s immune system. Science 296, 1263–1265. Qin, X.-F., An, D.S., Chen, I.S.Y., Baltimore, D., 2003. Inhibiting HIV-1 infection in human T cells by lentiviral-mediated delivery of small interfering RNA against CCR5. Proc. Natl. Acad. Sci. U.S.A. 100, 183–188. Reinhart, B.J., Slack, F.J., Basson, M., Pasquinelli, A.E., Bettinger, J.C., Rougvie, A.E., Horvitz, H.R., Ruvkun, G., 2000. The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 403, 901–906. Rubinson, D.A., Dillon, C.P., Kwiatkowski, A.V., Sievers, C., Yang, L., Kopinja, J., Zhang, M., McManus, M.T., Gertler, F.B., Scott, M.L., Van Parijs, L., 2003. A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat. Gen. 33, 401–406. Schwarz, D.S., Hutvágner, G., Du, T., Xu, Z., Aronin, N., Zamore, P.D., 2003. Asymmetry in the assembly of the RNAi enzyme complex. Cell 115, 199–208. Sen, G.C., 2001. Viruses and interferons. Annu. Rev. Microbiol. 55, 255– 281. Stewart, S.A., Dykxhoorn, D.M., Palliser, D., Mizuno, H., Yu, E.Y., An, D.S., Sabatini, D.M., Chen, I.S.Y., Hahn, W.C., Sharp, P.A., Weinberg, R.A., Novina, C.D., 2003. Lentivirus-delivered stable gene silencing by RNAi in primary cells. RNA 9, 493–501. Sui, G., Soohoo, C., Affar, E.B., Gay, F., Shi, Y., Forrester, W.C., Shi, Y., 2002. A DNA vector-based RNAi technology to suppress gene expression in mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 99, 5515–5520. Tabara, H., Sarkissian, M., Kelly, W.G., Fleenor, J., Grishok, A., Timmons, L., Fire, A., Mello, C.C., 1999. The rde-1 gene, RNA interference, and transposon silencing in C. elegans. Cell 99, 123–132. Tiscornia, G., Singer, O., Ikawa, M., Verma, I.M., 2003. A general method for gene knockdown in mice by using lentiviral vectors expressing small interfering RNA. Proc. Natl. Acad. Sci. U.S.A. 100, 1844–1848. Xu, P., Vernooy, S.Y., Guo, M., Hay, B.A., 2003. The Drosophila microRNA Mir-14 suppresses cell death and is required for normal fat metabolism. Curr. Biol. 13, 790–795. Yi, R., Qin, Y., Macara, I.G., Cullen, B.R., 2003. Exportin-5 Mediates the nuclear export of pre-microRNAs and short hairpin RNAs. Genes Dev. 17, 3011–3016.

B.R. Cullen / Virus Research 102 (2004) 3–9 Zeng, Y., Cullen, B.R., 2003. Sequence requirements for micro RNA processing and function in human cells. RNA 9, 112– 123. Zeng, Y., Wagner, E.J., Cullen, B.R., 2002. Both natural and designed micro RNAs can inhibit the expression of cognate mRNAs when expressed in human cells. Mol. Cell 9, 1327–1333.

9

Zeng, Y., Yi, R., Cullen, B.R., 2003. MicroRNAs and small interfering RNAs can inhibit mRNA expression by similar mechanisms. Proc. Natl. Acad. Sci. U.S.A. 100, 5779–5784. Zhang, H., Kolb, F.A., Brondani, V., Filipowicz, W., 2002. Human Dicer preferentially cleaves dsRNAs at their termini without a requirement for ATP. EMBO J. 21, 5875–5885.