1. Introduction
`
`2. Respiratory syncytial virus
`
`3. RNA interference and
`respiratory syncytial virus:
`a historic alliance
`marching forward
`
`4. Expert opinion
`
`For reprint orders,
`please contact:
`ben.fisher@informa.com
`
`Review
`
`Peptides, Proteins & Antisense
`
`Prospects of RNA interference
`therapy in respiratory viral
`diseases: update 2006
`
`Sailen Barik† & Vira Bitko
`†University of South Alabama College of Medicine, Department of Biochemistry and Molecular Biology,
`307 University Blvd, Mobile, Alabama 36688-0002, USA
`
`Respiratory viruses, such as influenza, parainfluenza and respiratory syncytial
`virus (RSV), claim millions of lives annually. At present, there is no completely
`effective vaccine or drug against these highly mutable RNA viruses. Passive
`antibody therapies for RSV, despite their limited application and staggering
`cost, enjoy a virtual monopoly in a multibillion-dollar global market.
`Recently, however, pioneering discoveries have launched RNA interference as
`a novel, nucleic acid-based therapy against viral pathogens. Specifically, small
`interfering RNAs (siRNAs) offered protection against respiratory syncytial
`virus, parainfluenza and influenza. siRNA against RSV has entered Phase I
`clinical trials
`in humans, and preliminary reports are promising.
`If
`appropriately
`formulated
`for
`improved
`specificity, delivery and
`pharmacokinetics, siRNAs may indeed become effective antivirals in the
`clinics of the future. This paper provides an overview of the prospects and
`hurdles facing the antiviral siRNA drugs, with special emphasis on RSV.
`
`Keywords: antiviral, influenza, intranasal delivery, respiratory syncytial virus, RNA interference,
`small interfering RNA
`
`Expert Opin. Biol. Ther. (2006) 6(11):1151-1160
`
`1. Introduction
`
`By virtue of its natural function, the lung is constantly exposed to airborne agents,
`such as toxic chemicals, particulate matter and pathogens, including viruses.
`Respiratory viruses are in fact the predominant causative agent of lower respiratory
`tract infections in infants and premature babies [1-4]. In the US, respiratory syncytial
`(RSV) alone causes ∼ 95,000 hospital admissions/year, costing
`virus
`∼ US$300 million at
`the average rate of US$1000/day/infant. Although
`traditionally considered paediatric viruses, RSV and parainfluenza virus (PIV) can
`cause life-threatening respiratory infection in individuals with heart conditions and
`chronic obstructive pulmonary disease, in immunocompromised patients, such as
`organ recipients and AIDS victims, and in older people. Together, they infect up to
`65% of babies in the first year of life and, essentially, 100% within the first 2 years.
`In a recent report from the WHO, the global annual infection and mortality figures
`for RSV were ∼ 64 million and 160,000, respectively [1]. Not surprisingly, the
`WHO continues to designate RSV as a major target of research and therapy
`throughout the globe. The recent emergence of the highly lethal severe acute
`respiratory syndrome coronavirus (SARS-CoV) [5] and H5N1 bird flu virus [6,7]
`further underscores the critical need for prevention and therapy of the respiratory
`viral diseases.
`immunology have
`features of respiratory viral genetics and
`Multiple
`contributed to the difficulties of prevention and treatment, and this review
`summarises some of them. First of all, RNA genomes, such as those of RSV, PIV,
`
`10.1517/14712598.6.11.1151 © 2006 Informa UK Ltd ISSN 1471-2598
`
`1151
`
`Downloaded by [National Library of Medicine] at 11:02 12 April 2016
`
`1
`
`MTX1031
`
`

`

`Prospects of RNA interference therapy in respiratory viral diseases: update 2006
`
`Family:
`Paramyxoviridae
`
` Subfamily:
`Paramyxovirinae
`
`Subfamily:
`Pneumovirinae
`
`Genus: Pneumovirus
`Respiratory syncytial
`Mouse pneumonia
`
`Genus: Metapneumovirus
`Human metapneumovirus
`Avian metapneumovirus
`
`Genus: Respirovirus
`Parainfluenza virus 1, 3
`Sendai
`Simian virus 10
`
`Genus: Morbillivirus
`Measles
`Canine distemper
`Rinderpest
`
`Genus: Rubulavirus
`Avian paramyxovirus 2-9
`Newcastle disease
`Parainfluenza virus 2, 4
`Mumps
`Simian virus 5, 41
`
`Figure 1. The two relevant subfamilies under the
`Paramyxoviridae family are shown with a few clinically
`significant viruses as examples (see section 2.1). Some
`viruses have alternative names in literature that are not shown
`here; for example, avian metapneumovirus was initially named
`turkey rhinotracheitis virus and is sometimes also called avian
`pneumovirus. Mouse pneumonia virus
`is also known as
`pneumonia virus of mice.
`
`influenza and SARS-CoV, mutate at a high rate due to the
`lack of a proof-reading mechanism, presenting a significant
`hurdle for vaccine or antiviral development [7-9]. In RSV, the
`exact roles of many of the 11 proteins produced by the viral
`genome remain ill-defined [10]. The viruses have a complex
`interaction with a myriad of cellular proteins, adding to the
`complexity of obtaining pure viral material [11-14]. Finally,
`the immunopathology of these viruses, particularly that of
`RSV, is exquisitely complex [2-4,15-17] and includes the
`unique
`phenomenon
`of
`immunopotentiation
`or
`vaccine-enhanced disease [16,17].
`The inadequacy of existing treatments of RSV has been
`reviewed previously [1,3,4], and this review briefly updates the
`present status in this area. Over the last 20 years, a variety of
`potential antiviral drugs have been tested against RSV that
`includes synthetic and natural small molecules, monoclonal
`antibodies, vaccines, and antisense oligodeoxynucleotides.
`However, essentially all are now abandoned and the few
`residual efforts focus on the development of humanised
`monoclonal
`antibodies
`for
`passive
`immunisation,
`recombinant cDNA-based vaccines and attenuated viruses for
`primary immunisation [1]. At present, the most prescribed
`treatment
`for RSV-infected
`infants
`is
`the humanised
`antibody, palivizumab, marketed by MedImmune, which
`remains an expensive and difficult treatment regimen [1,18-20].
`
`Faced with these challenges to traditional approaches, RNA
`interference (RNAi) has been introduced as an antiviral tool
`[21].
`using RSV as
`the
`first proof-of-concept
`target
`Subsequently, these results were duplicated and confirmed in
`a number of other viruses that include all the respiratory
`viruses mentioned earlier. Here, the authors focus mainly on
`continued recent work on RNAi as an antiviral tool against
`RSV and its pharmaceutical potential [22-25], and briefly
`summarise other respiratory viruses as well.
`
`2. Respiratory syncytial virus
`
`2.1 Respiratory syncytial virus taxonomy and
`gene expression
`The RNA viruses of the Mononegavirales order contain
`single-stranded RNA genomes that are antimessage or
`negative-sense [10]. Members of this order are important
`human and animal pathogens, examples of which include
`viruses
`causing measles, mumps,
`rabies,
`influenza,
`parainfluenza and RSV. The Ebola and Hantaan viruses are
`also members of this order, which have recently emerged and
`are among the most pathogenic viruses known to mankind.
`This order
`is comprised of seven families,
`including
`Orthomyxoviridae and Paramyxoviridae. Although the viral
`genomes in the former family are segmented, as exemplified
`by the influenza virus, the latter possesses non-segmented
`genomes (Figure 1). RSV is a member of the Pneumovirus
`genus in the Paramyxoviridae family and shares the genus with
`only one other major virus (Figure 1), underscoring its genetic
`uniqueness. Coronaviruses, such as SARS-CoV, belong to the
`Coronaviridae family and contain positive-stranded genome
`that can be directly translated [5].
`In this review, the authors present the salient features of viral
`gene expression with an emphasis on RSV [22,26]. As the
`genome
`is negative-sense (anti-mRNA),
`it can not be
`translated. Moreover, as human cells lack template RNA
`copying activity, all RNA viral genomes – positive- or
`negative-sense – must encode such an activity. Whereas
`retroviruses encode reverse transcriptase that copies the
`genomic RNA
`into DNA, other RNA viruses encode
`RNA-dependent RNA polymerase (RdRP) activity. In the
`Paramyxoviridae (such as PIV and RSV), the functional RdRP
`is composed of the large protein L and phosphoprotein P, and is
`packaged in the mature virion particle along with the genomic
`RNA wrapped in N protein (Figure 2). Immediately following
`infection, the RdRP transcribes the N-RNA template to
`produce gene-specific mRNAs that are then translated for
`de novo viral protein synthesis. The new L and P proteins form
`more RdRP and, thus, the viral transcription escalates further.
`When sufficient N protein has been synthesised, it starts
`wrapping the 5´ end of the newly transcribed nascent RNA and
`this somehow instructs the leading RdRP to switch from
`transcription to replication mode in which the RdRP ignores
`the stop signals between genes and produces full-length
`complementary positive-strand (mRNA-sense) RNA. The
`
`1152
`
`Expert Opin. Biol. Ther. (2006) 6(11)
`
`Downloaded by [National Library of Medicine] at 11:02 12 April 2016
`
`2
`
`

`

`Barik & Bitko
`
`RSV
`N-RNA
`complex
`
`L
`
`P
`
`G
`
`Infection
`
`F
`
`Cell
`
`L
`
`P
`
`N protein
`
`3'
`Genes: NS1 NS2 N P M1
`
`SH
`
`FG
`
`M2(1,2)
`
`L
`
`5'
`
`Transcription
`
` RdRP (L, P)
`
`mRNA
`(Molar ratio):
`
`100 95
`
`90 68 52
`
`32 21 18
`
`15
`
`3
`
`Translation
`
`De novo viral proteins
`
`More transcription
`and translation,
`followed by replication
`
`Figure 2. A simplified molecular life cycle of RSV. L and P (with help from M2-1) transcribe the genomic RNA wrapped with N protein
`(the N-RNA complex). The relative molar ratios of the mRNAs are shown, with the most abundant NS1 mRNA taken as 100. The most
`essential genes (N, P, F, M2-1, L) are considered highly prospective siRNA targets and are underlined (details in section 2.1).
`RSV: Respiratory syncytial virus; siRNA: Small interfering RNA.
`
`latter serves as a template for more negative-sense genomic
`RNA, adding to overall viral gene expression.
`An important feature of non-segmented negative-strand
`virus transcription is polarity, in which the viral genes close to
`the promoter (3´ end of the genome) are transcribed more
`abundantly than those that are distal [14,27,28]. Thus, the NS1
`and 2 mRNAs are the most abundant in RSV, and the
`L mRNA the least, in all viruses of this family (Figure 2). This
`is important to remember when choosing a particular gene
`mRNA as target. The other relevant point is the function of
`the various gene products that is described in section 3.1.
`
`2.2 Clinical features of the respiratory syncytial
`virus disease
`As a rule, respiratory viral infections are seasonal. RSV
`epidemics, for example, peak in winter and early spring in
`temperate regions and during the monsoon in tropical
`climates [1]. Humans are the only known reservoir for the
`human strain of RSV, although specific RSV strains also exist
`for non-human animals, such as cattle, goats and sheep.
`Premature babies,
`immunocompromised
`adults,
`and
`individuals with bronchopulmonary dysplasia and congenital
`heart disease are at elevated risk for severe RSV disease with
`pneumonia and bronchiolitis, and are often hospitalised.
`Essentially, all children experience one episode of RSV disease
`
`by 2 years of age [1-4]. Overall, 25 – 40% of RSV-infected
`young adults and healthy older children develop lower
`respiratory
`tract
`infections with many developing
`bronchiolitis or pneumonia.
`Immunity
`is generally
`incomplete and, as reviewed in detail earlier, there is a vast
`unmet need for a reliable vaccine or antiviral [1].
`
`3. RNA interference and respiratory syncytial
`virus: a historic alliance marching forward
`
`3.1 Specific respiratory syncytial virus genes as
`RNA interface target
`RNAi is a physiological pathway found in all metazoan cells
`examined to date, in which double-stranded (ds)RNA causes
`the silencing of the target mRNA [29]. The dsRNA is
`processed by Dicer to produce small interfering RNA
`(siRNA), which is ∼ 21 – 23 base pairs long with
`single-stranded 2-nt long 3´ extensions on both strands. The
`antisense strand of the siRNA engages the target mRNA as
`part of
`a
`large
`ribonucleoprotein
`complex
`called
`RNA-induced silencing complex (RISC). A specialised RNase
`component of RISC, named Argonaute, then cleaves the
`target mRNA, abrogating translation. RNAi, therefore, is a
`post-transcriptional silencing mechanism that
`leads to
`‘knock-down’ of gene expression. In recent years, RNAi has
`
`Expert Opin. Biol. Ther. (2006) 6(11)
`
`1153
`
`Downloaded by [National Library of Medicine] at 11:02 12 April 2016
`
`3
`
`

`

`Prospects of RNA interference therapy in respiratory viral diseases: update 2006
`
`Table 1. RSV proteins (gene products).
`
`Protein/gene symbol
`
`Name and known (or potential) function
`
`NS1
`
`NS2
`N
`
`P
`M1
`
`NS1 and NS2 (below) are small proteins made abundantly in the infected cell, but do not become part of
`the virion structure (hence their name); they probably function together and suppress innate immunity and
`IFN to promote virus growth.
`NS2; promotes RSV growth by suppressing IFN.
`Nucleoprotein; wraps full-length genomic and antigenomic RNA, and is essential for their
`template function.
`Phosphoprotein; an essential subunit for viral RdRP.
`Matrix protein; an important link between the packaged nucleocapsid and the viral envelope, and may
`inhibit viral transcription in the late stage of viral life cycle.
`Short hydrophobic glycoprotein of unknown function and dispensable for growth.
`Glycoprotein of the viral envelope, but dispensable for growth.
`Fusion protein; an envelope glycoprotein, essential for cell fusion.
`Matrix 2 (ORF1); previously called 22K (for its size), it is a transcription antiterminator, essential for optimal
`viral RdRP function and viral replication.
`Matrix 2 (ORF2); the second protein made from the M2 gene is of unknown function and dispensable for
`viral growth.
`Large protein; the largest viral protein (∼ 242 kDa) and the essential main subunit of viral RdRP.
`The RSV proteins and their best known functions are listed in [9,10].
`The proteins written in bold are absolutely essential for viral growth and, therefore, are preferred targets for antiviral siRNA. See text for details.
`NS: Non-structural protein; ORF: Open reading frame; RdRP: RNA-dependent RNA polymerase; RSV: Respiratory syncytial virus; siRNA: Small interfering RNA.
`
`SH
`G
`F
`M2-1
`
`M2-2
`
`L
`
`been called on to silence deviant or disease-causing genes,
`such as viral genes or cancer-causing cellular genes [23,24,30,31].
`RSV holds a unique place in the annals of RNAi as it was
`the first virus in which the anti-infective, therapeutic power of
`RNAi was established in 2001 [21]. In the original paper, viral
`F and P mRNAs were targeted [21]; however, as none of the
`viral genes have any significant sequence homology with the
`human genome, specific siRNAs can, in principle, be
`designed against any viral mRNA. Here, the authors offer a
`comprehensive and rational assessment of all RSV genes as
`potential RNAi targets based on their role(s) in virus growth
`and/or host–virus interactions (Table 1).
`By far the three most fundamentally important viral
`proteins, namely N, P and L, make up the core RNA
`synthetic machinery in both RSV and PIV [14,26]. As
`mentioned, L and P constitute the RdRP that copies the viral
`N-RNA template [14] and, hence, silencing of any of the three
`would have similarly strong antiviral effect. siRNAs against P
`have been indeed successfully used as antiviral against RSV
`and PIV in cell culture, as well as in animals [21,25], but there is
`no published report
`targeting
`the other
`two genes.
`Nonetheless, due to its large size (∼ 10 times the size of P), the
`L mRNA offers many sequence options for siRNA design,
`and its low abundance offers an advantage [32]. M2-1 is
`unique to RSV and important for viral RNA synthesis [33],
`although it is yet to be tested as an RNAi target. As shown in
`Table 1, of the three envelope glycoproteins, only F is truly
`essential and, thus, a good siRNA target [34-36]. Silencing of F
`by siRNA led to the expected loss of cell fusion (syncytium)
`ex vivo [21], which may inhibit cell-to-cell transfer of progeny
`
`virions, limiting the disease. In the past, chemical blockers of F
`exhibited strong anti-RSV activity and were considered
`clinically promising [37-40]. The other two glycoproteins, G and
`SH, as well as M2-2, are non-essential for virus growth and,
`therefore, are not recommended targets [41-44]. Knockdown of
`accessory proteins haemagglutinin-neuraminidase and F of PIV
`abrogates syncytia in cell culture, but its effect on viral growth
`remains untested [22].
`The non-structural proteins NS1 and 2 present an
`interesting scenario. Deletion analysis has shown that they are
`not absolutely essential
`for virus replication
`in cell
`culture [42,43]. However, RNAi targeting NS1 protected mice
`from RSV (see section 3.2) [45]. Recent results suggest that
`these ‘accessory’ proteins play important roles in suppressing
`the host innate immune response, such as the IFN response,
`and thereby improve virus growth [46-50]. NS2, in particular,
`promotes degradation of signal transducer and activator of
`transcription (STAT)2, inhibiting type I IFN-dependent gene
`expression, which underlies the ability of RSV to subvert
`type I IFN synthesis and signalling in the infected cell [48,50].
`Thus, anti-NS siRNAs may produce antiviral effects by
`reinstating the host’s antiviral IFN response. This mechanism
`remains to be tested. Combining a siRNA against NS (or
`another accessory gene, such as F) with one against an
`essential function, such as P, is also worth testing. It is to be
`noted that the NS genes are unique to RSV, and in other
`paramyxoviruses,
`including PIV,
`the
`IFN-antagonistic
`proteins are produced from an alternative translational frame
`within the P gene, accessed by insertion of a few extra
`G residues to the mRNA during transcription of a ‘slippery’
`
`1154
`
`Expert Opin. Biol. Ther. (2006) 6(11)
`
`Downloaded by [National Library of Medicine] at 11:02 12 April 2016
`
`4
`
`

`

`Barik & Bitko
`
`Table 2. Comparison between DNA-directed versus synthetic siRNA-directed RNAi.
`
`DNA-directed RNAi
`
`Synthetic siRNA-directed RNAi
`
`Continuously expressed over time; stability is not an issue.
`
`Suitable for long-term treatment.
`Expression can be regulated by inducible promoters.
`
`Stable expression rules out alteration of sh/siRNA sequence.
`
`Constitutive expression may cause lethality or toxicity.
`Long ds/shRNAs may trigger IFN response.
`Chemical modifications impossible.
`
`Short life span due to degradation – can be improved
`by modifications.
`Repeated application necessary in chronic infections such as AIDS.
`Unregulated when applied, but the application itself can be
`planned; synthesis can be scaled up.
`New siRNA sequence can be quickly made to encounter
`mutational resistance of the virus.
`Short-term use of unstable siRNA minimises toxic effects.
`21- to 25-mer synthetic siRNAs do not trigger IFN response.
`A variety of modifications can optimise the pharmacokinetics
`of siRNA.
`
`ds: Double-stranded; RNAi: RNA interference; shRNA: Short hairpin RNA; siRNA: Small interfering RNA.
`
`sequence by the viral RdRP [51,52]. Thus, it will be exceedingly
`difficult to design siRNAs against these accessory proteins
`without silencing the P mRNA. In a variation of this theme,
`one can target cellular genes that are relatively non-essential
`for cellular physiology, at least in the short term, but critical
`for viral replication. Knockdown of profilin, for example,
`strongly inhibited RSV morphogenesis [11,12] and, thus,
`antiprofilin siRNA can be tested in combination with an
`antiviral siRNA. Such treatments will provide simultaneous
`protection from other viruses that may also need profilin.
`Successful antiviral
`siRNAs were generated against
`influenza virus matrix protein [53], transcriptional gene PA,
`and the nucleocapsid gene N [54,55]. In SARS-CoV, effective
`targets included the non-structural protein 1 [56], the spike
`protein [57] and, interestingly, non-coding leader RNA [58]. In
`a more comprehensive approach, a total of 48 siRNAs
`covering the entire SARS-CoV genome were first screened for
`antiviral activity in cell culture [59] and the two most potent
`siRNAs were selected for further analysis [60]. The clinical
`testing of these and other siRNAs are detailed later.
`
`3.2 Therapeutic short interfering RNAs against
`respiratory viruses: design and application
`To apply as an antiviral drug, intracellular siRNA can be
`generated by a number of means (Figure 1) [61]. The siRNA
`precursors, in the form of dsRNA or short hairpin RNA
`(shRNA), can be transcribed from transfected recombinant
`DNA (a technique sometimes called DNA-directed or
`-derived RNAi. The ds/shRNA is eventually processed by
`cytoplasmic Dicer to generate siRNA as mentioned earlier. A
`more common approach, however, is to chemically synthesise
`siRNAs and apply directly. The authors offer a brief
`comparison of the pros and cons of the two approaches in
`Table 2. For respiratory RNA viruses that mutate rapidly and
`produce acute short-term infections, synthetic siRNAs seem
`to offer significant advantages.
`All anti-RSV studies published so far [21,22,25] have
`employed synthetic siRNA without chemical modifications.
`
`Although it is likely that chemical modifications will
`improve stability and pharmacokinetics [24,61,62], this is yet
`to be tested intranasally against respiratory viruses. Based on
`standard recommendations, the successful anti-RSV siRNAs
`to date [21,22,25] have included the following sequence
`features: 19-mer double-stranded core sequence with
`2-nucleotide 3´ extensions; AU-rich 5´ end and GC-rich
`3´ end of the guide (antisense) strand to facilitate its
`preferential
`incorporation
`into RISC;
`lack
`of
`homonucleotide runs; and lack of homology with human
`genome, ascertained by Basic Local Alignment Search Tool
`search. At least two recent studies have suggested that
`27- to 29-mer dsRNAs may be much more potent than the
`classic 19- to 21-mers [63,64]; however, such longer siRNAs
`remain untested against RSV. The high mutability of the
`RNA viral genome may
`lead
`to
`the
`selection of
`siRNA-resistant mutants during treatment. This can be
`minimised by designing siRNAs against sequences that have
`remained conserved in multiple isolates and strains of the
`virus. The other option is to use a cocktail of multiple
`siRNAs, akin to the multi-antibiotic treatment of bacterial
`infections or multi-drug therapy of AIDS.
`As a rule, siRNAs designed against conserved domains of
`RNA viral proteins are less likely to encounter mutational
`resistance. In the clinical arena, promising progress has been
`made in RNAi therapy against a number of respiratory
`viruses. With RSV, in particular, anti-P siRNA, which was
`originally screened in cell culture [21], was subsequently tested
`nasally, both as drops and as aerosol [23-25], in BALB/c mice.
`Essentially, all mice resisted subsequent challenge by RSV as
`judged by a variety of parameters, such as 103- to 104-fold
`lower pulmonary viral titre, retention of body weight, normal
`breathing, and reduced concentration of leukotriene, a
`common allergy marker, in bronchoalveolar lavage fluid [25].
`The siRNA also had a curative effect. When applied at
`different days after RSV inoculation, the siRNA always
`improved health and prognosis, compared with control mice
`not receiving siRNA.
`
`Expert Opin. Biol. Ther. (2006) 6(11)
`
`1155
`
`Downloaded by [National Library of Medicine] at 11:02 12 April 2016
`
`5
`
`

`

`Prospects of RNA interference therapy in respiratory viral diseases: update 2006
`
`In the same study using the mouse model [25], strong
`protection was also achieved against PIV with intranasal
`siRNA against the PIV P mRNA. Moreover, a mixture of the
`anti-RSV and -PIV siRNAs offered protection against joint
`challenge by the two viruses, paving the way for multi-siRNA
`therapy against a combination of respiratory viruses.
`Interestingly, although the NS proteins were not absolutely
`required for RSV growth in cell culture, DNA-derived RNAi
`targeting NS1 provided significant protection against RSV
`infection in mice [45]. It remains to be determined whether
`the protection was due to the known role of NS proteins in
`cell culture (i.e., suppression of the host IFN response and
`innate immunity).
`DNA-derived RNAi, as well as siRNA, also provided
`significant protection against influenza and SARS-CoV in
`appropriate animal models [54,55,60]. In two parallel studies in
`mice,
`intranasal and systemic administration of RNAi
`targeting viral PA and N genes produced an anti-influenza
`state, monitored by
`lung viral
`titre
`and
`clinical
`outcomes [54,55]. The two siRNAs that exhibited the highest
`efficacy against SARS-CoV in cell culture [59] were tested in
`the rhesus macaque (Macaca mulatta) model. Both showed
`significant prophylactic and therapeutic effects as indicated by
`relief from SARS-CoV-induced fever, lower viral levels and
`reduced alveoli damage [60].
`Due to their relatively large size, siRNAs do not freely
`diffuse into cells and, generally, need carrier(s) to cross the cell
`membrane. In cell culture, mainly cationic lipid formulations,
`such as Oligofectamine™ (Invitrogen) and TransIT-TKO®
`(Mirus), have been used with success. Interestingly, a number
`of animal studies have shown strong activity of uncomplexed,
`‘naked’ siRNA dissolved in physiological solutions, such as
`water, PBS, 5% glucose solution in water or infasurf (a natural
`lung surfactant) [25,60,65]. It has been speculated that the
`success of intranasal ‘naked siRNA’ against respiratory viruses
`is due, in part, to the compromised membrane integrity of the
`[25]. Regardless of
`its
`virus-infected pulmonary cells
`mechanism, the use of naked siRNA eliminates toxicity
`concerns of the carriers, such as untoward immune response,
`which
`is
`an
`important
`advantage
`in
`therapeutic
`applications [66]. Moreover, it lends itself easily to a simple
`inhaler-based application.
`In terms of drug development, intranasal siRNA technology
`against RSV [21,25] is being tested at present in Phase I clinical
`[101]. Nastech
`trials
`by Alnylam
`Pharmaceuticals
`Pharmaceuticals, a leader in nasal delivery, recently announced
`in vitro preclinical screening results that identified highly potent
`siRNAs with IC50 values between 20 and 500 pM against
`representative human and avian influenza strains, including the
`deadly H5N1 strain [102]. Alnylam, in collaboration with
`Novartis, also initiated RNAi studies against influenza [101].
`Optimisation of anti-SARS-CoV siRNA studies are being
`continued by Intradigm, in collaboration with Qiagen, NIAID,
`and investigators in Hong Kong and Guangzhou of China [103].
`Recently, Sirna Therapeutics and GlaxoSmithKline announced
`
`an exclusive multi-year strategic alliance on discovery,
`development and commercialisation of novel RNAi-based
`therapeutics for respiratory diseases, although the exact targets
`were not made public [104].
`
`3.3 New antiviral siRNA formulations
`It is now universally realised that the medical applications of
`RNAi are not going to be as straightforward as in vitro use.
`The major obstacles to developing siRNA as a drug are
`specificity, stability and delivery [24]. Fortunately, there are a
`variety of strategies that can potentially improve these
`parameters. Here the authors summarise some of them and
`the little available information regarding their application in
`respiratory viruses.
`Specificity is improved by screening multiple siRNAs
`against rational viral sequences, as was done for RSV and
`SARS-CoV (see above). The siRNA(s) showing the highest
`potency
`can
`then be used
`at
`low dosage
`(e.g.,
`> 80% knockdown
`at
`low nanomolar
`concentration),
`minimising toxicity. Specific inhibition of the same virus by
`multiple siRNAs designed against the same target sequence
`also assures that the antiviral effect is truly specific and
`RNAi-mediated [25]. A major source of nonspecificity is
`various degrees of mismatch tolerance by RNAi, leading to
`the silencing of unintended genes, sometimes referred as
`off-target effects [67]. Potentially, silencing of off-target
`mRNAs can also be caused by a RISC containing the sense
`strand of the siRNA. To prevent this, various modifications of
`the siRNA on either strand have been advocated that inhibit
`its RISC engagement without compromising the engagement
`or function of the antisense strand. In the most recent study,
`the 2´-OMe substitution of ribose at position 2 of the guide
`strand significantly reduced off-target effects [68]. Invitrogen
`claims that its Stealth™ technology only allows the antisense
`strand to efficiently enter the RNAi pathway to eliminate
`concerns about sense strand off-target effects; the nature of
`the modification(s) remains proprietary [105].
`in
`stability
`Various modifications
`improve
`siRNA
`fundamental
`[24,61,62]. A
`biological
`tissues and
`fluids
`requirement of siRNA function is that the antisense strand
`must either have a free hydroxyl or a phosphate at the
`5´-terminus and, therefore, this terminus cannot be modified.
`Of all the internal modifications, substitutions of the 2´-OH
`of ribose remain the best studied, and include -H, -OMe and
`-F. In general, these modifications improve stability and
`silencing activity, but need to be tested intranasally against
`respiratory viruses. Replacement of the phosphodiester
`(P = O) backbone with phosphorothioates (P = S) or
`boranophosphonates (P = B) also augments nuclease
`resistance, but excessive substitution may lead to reduced
`silencing activity and increased toxicity. siRNAs containing a
`synthetic RNA-like nucleotide analogue, known as ‘locked
`nucleic acid’, are difficult to synthesise, but do exhibit
`improved stability and enhanced silencing. The relative
`difficulty of their chemistry may pose a technical and
`
`1156
`
`Expert Opin. Biol. Ther. (2006) 6(11)
`
`Downloaded by [National Library of Medicine] at 11:02 12 April 2016
`
`6
`
`

`

`Barik & Bitko
`
`economic hurdle in drug development. As mentioned,
`unmodified siRNA has already progressed to clinical trials;
`however, it is worth testing whether any of the aforesaid
`modifications fare even better. Nasally applied unmodified
`siRNA stays within the lung (authors’ unpublished results)
`and does not escape into the general circulation, which is an
`ideal attribute for respiratory viral therapy. Detailed tissue
`status of modified siRNAs remains unknown.
`
`4. Expert opinion
`
`Significant progress has been made in applying RNAi to
`respiratory viral prevention and therapy, as evidenced by the
`fact that the clinical trial against RSV begun < 5 years after
`RNAi was first established as an antiviral tool in cell
`culture [21]. Similar progress is expected to follow in other
`respiratory
`viral
`ailments,
`including
`influenza
`and
`SARS-CoV, in the near future. Although there is a sense of
`déjà vu in reminiscing similar excitement in the DNA-based
`antisense technology only ∼ 10 years ago, the resulting
`frustration is also a fresh memory. To ensure that RNAi is not
`doomed to a similar fate, the following landmarks are
`recommended during the development process:
`
`(cid:127) Perform a methodical screening to obtain the most potent
`siRNAs sequence(s) with IC50 in the low nanomolar or
`even picomolar range in cell culture.
`(cid:127) Titre siRNAs by monitoring both target mRNA and
`protein levels.
`(cid:127) Target RNA sequences that code for essential amino acids,
`invariant in various strains and isolates of the virus.
`(cid:127) Chemically modify the candidate siRNAs to improve in vivo
`stability and potency without compromising specificity.
`
`(cid:127) Test various carriers, including nanoparticles, for optimal
`delivery, absorption and stability of the siRNAs.
`(cid:127) In the homology search, allow for 2 – 3 mismatches with
`off-target genes.
`(cid:127) Conduct specificity testing in animals and humans,
`including measurement of IFN response (e.g., STAT,
`protein kinase R and phospho-eIF2α) and inflammatory
`markers (e.g., chemokines and cytokines).
`(cid:127) Use a scrambled siRNA control.
`(cid:127) Take account of the applied siRNA in various tissues and
`subcellular locations to ensure appropriate delivery and
`effective transport from endosome to cytoplasm.
`(cid:127) Exercise vigilance against viral mutants resistant to the first
`siRNA and, anticipating that this may happen, keep
`alternative siRNAs ready in the arsenal. Simultaneous
`multi-siRNA therapy is also an option, especially in
`emergency situations, such as under a bioterrorism attack,
`but overdosing may saturate the cellular RISC machinery
`and should be avoided [22,25,69].
`(cid:127) As nasal congestion is common in respiratory infections,
`nasal delivery of therapeutic siRNA may be challenging. A
`potential solution is to combine the siRNA formulation
`with decongestants, antihistamines or corticosteroids.
`
`As with any new technolog