Clinical Potential of Respirable Antisense Oligonucleotides (RASONs) in Asthma

Am J Pharmacogenomics 2003; 3 (2): 97-106MOLECULAR MEDICINE 1175-2203/03/0002-0097/$30.00/0

© Adis Data Information BV 2003. All rights reserved.

Clinical Potential of Respirable AntisenseOligonucleotides (RASONs) in AsthmaHoward A. Ball,1,2 Anthony Sandrasagra,1,2 Lei Tang,1,2 Mike Van Scott,1,2 James Wild1,2 andJonathan W. Nyce1,2

1 EpiGenesis Pharmaceuticals Inc., Cranbury, North Carolina, USA2 Department of Physiology, Brody School of Medicine, East Carolina University, Greenville, North Carolina, USA

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

1. Novel Disease Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

2. Target Validation: Identifying Targets Involved in the Disease Process Using Antisense Oligonucleotides (ASONs) . . . . . . . . . . . . . . 99

3. Factors Contributing to the Effectiveness of Respirable ASONs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

4. Optimal Models for ASON Testing and Human Target Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

5. What Are the Optimal Properties of Respirable ASONs for Clinical Therapeutics? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

5.1 Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

5.2 Chemistry Backbone and Metabolic Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

5.3 Immunological Adverse Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

5.4 Clinical Efficacy of EPI-2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

5.5 Development Hurdles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

The human genome project, as well as advances in our understanding of asthma susceptibility, are yieldingAbstractnovel candidate targets for disease intervention. The normalization of up-regulated gene expression may treat or

improve the disease outcome. However, only some of these gene product targets may be ‘tractable’, i.e.

amenable to blockade by small, orally active, organic molecules. The remainder have been termed ‘non-tract-

able’.

For over a decade, antisense oligonucleotides (ASONs) have been used as tools to evaluate the importance of

specific gene products in vitro. In recent years evidence has accumulated indicating their potential as a viable

new therapeutic approach in their own right, being able to block ‘non-tractable’ targets as well as ‘tractable’

targets.

Distribution, cell-specific uptake, and effectiveness of aerosolized phosphorothioate ASONs are currently

being evaluated in animal models. The results demonstrate broad distribution throughout the lung, and uptake by

all of the cell types examined to date. Functionality has been demonstrated against diverse targets, including

nuclear transcription factors, tyrosine kinases, G-protein coupled receptors, cytokine receptors, growth factors,

and chemokines.

EPI-2010, a respirable ASON (RASON) against the adenosine A1 receptor, is the first test case for this new

class of respiratory therapeutics. The rationale for EPI-2010 is that overactivity of the adenosine-signaling

pathway in asthmatic lungs contributes to airway inflammation and hyperresponsiveness. EPI-2010 binds to the

98 Ball et al.

initiation codon of the adenosine A1 receptor mRNA, and thereby blocks translation and targets the message for

degradation by RNase. EPI-2010 is apparently metabolized locally by endogenous nucleases confining its

activity to the airways. Phase I clinical trials have shown EPI-2010 to be well-tolerated, with indications of

efficacy.

In conclusion, one important application of RASONs is in addressing up-regulated disease targets, only some

of which are ‘tractable’ by small molecules. It is hoped that this will yield new therapeutic options to the benefit

of patients with asthma and allergic disorders.

Asthma is a common lung disease characterized by symptoms sion analysis approaches are being used extensively to identify

of coughing or wheezing, shortness of breath, and excess mucus genes and pathways whose expression patterns are altered in aproduction. Although the symptoms of asthma have long been variety of model systems that are meant to approximate humanrecognized, the discovery that asthma was characterized by an asthma. The model systems used in these target identificationinflammatory process in the bronchial air passages of the lungs programs as sources of mRNA for expression profiling are variedwas only realized in the 1970s. Between 100 and 150 million and include in vitro model systems of inflammation (e.g. humanpeople around the globe – roughly the equivalent of the population and rodent primary cultures and cell lines), in vivo models ofof the Russian Federation – suffer from asthma, and this number is human asthma (e.g. tissues from rodent and other challenge basedrising. Worldwide, deaths from this condition have reached over models of allergic asthma), and biopsy samples from human180 000 annually.[1] An estimated 17 million Americans suffer asthmatics. These types of studies have yielded hundreds of poten-from asthma, nearly 5 million are under age 18 years. It is the most tial respiratory disease candidate genes, by virtue of the fact thatcommon chronic childhood disease, affecting more than one child they are up- or down-regulated in relation to the appropriatein 20. Asthma is the only chronic disease, besides AIDS and

controls, for further evaluation and prioritization. For example, intuberculosis, with an increasing death rate. Each day 14 Ameri-

an ovalbumin-sensitized mouse model of allergy,[3] analysis ofcans die from asthma. From 1979–1992, asthma death rates in-

12 423 genes revealed a 2-fold or greater change in mRNAcreased 58% overall. The death rate for children 19 years and

abundance in 3.41% (424) of the total genes. Three hundred andyounger increased by 78% between 1980 and 1993.[1] According

forty or 2.74% of genes showed an increase in expression levels,to the Center for Disease Control in Atlanta, the incidence of

and 0.67% (83) of the genes showed a decrease in expressionasthma has nearly doubled in the US in the past decade, for reasons

levels. Most interestingly these data indicate that in this modelthat remain unclear. Similar increases in the incidence of asthma

there is a 4-fold difference between the up- and down-regulatedhave been reported for most developed and developing countries.

genes. Functional classification of differentially expressed genesThere were approximately 2 million asthma-related emergencyshowed immune-related genes occur predominantly in the groupvisits/admissions in 1998 in the US clearly suggesting that the

available classes of asthma drugs (table I) are not sufficient to

control the disease (Asthma Facts, Asthma and Allergy Founda-

tion of America[2]).

These drugs address less than a handful of disease mechanisms.

This might be considered surprising given the multifactorial nature

of asthma.

1. Novel Disease Targets

Advances in our understanding of the biochemistry of inflam-

mation and hyperresponsiveness, genetics of asthma susceptibil-

ity, as well as the human genome project, are yielding novel

molecular targets for disease intervention. Whole genome expres-

Table I. Principal drug classes used to treat asthma

Class Pharmacologic activity

β2-adrenergic receptor Bronchodilatoragonists (short and longacting)

Corticosteroids Anti-inflammatory

Xanthines Anti-inflammatory

Chromones Mast cell stabilizers

Antihistamines Histamine receptor blockade

Leukotriene modifiers Leukotriene receptor blockade orinhibition of synthesis

Antimuscarinic antagonists Reduction of cholinergicbronchoconstrictor airway tone

© Adis Data Information BV 2003. All rights reserved. Am J Pharmacogenomics 2003; 3 (2)

Clinical Potential of RASONs in Asthma 99

with increased expression levels (39%) compared with the group

exhibiting a decrease in expression (7%).

It should be noted, however, that not all studies have indicated a

preponderance of up-regulated genes. In a recent study using

Ascaris-sensitive Cynomolgus monkeys,[4] 40 000 cDNA ele-

ments were screened for expression changes following aerosol

allergen challenge. 149 genes exhibited a greater than 2.5-fold

change in expression following challenge. Thirty five to 50% of

these cDNAs were up-regulated depending on the timepoint used

for tissue harvest. The number of genes determined to be up-

regulated in this study was smaller than that found in the above

study using allergic mice, perhaps because of the more stringent

Table II. Difficulties in developing small molecule therapeutic agents for‘non-tractable’ targets

Properties of protein targets rendering them difficult to block bysmall molecules

Complex signaling between proteins involving multiple binding sites

Members of multi-subtype protein families (problem of selectivity)

Ubiquity of intracellular signaling factors leading to non-target tissueadverse effects

Common difficulties with small molecule approach for receptor/enzyme blockade

Insufficient receptor subtype selectivity

Prohibitively complex chemical synthesis

Poor pharmacokinetic properties (low dissolution/absorption rate, highclearance)

criteria used to define alteration of expression or because of some

unique characteristics of the model. Of note, however, 35% of theantisense oligonucleotides, because they act in a sequence-depen-

cDNAs exhibiting altered expression were novel.dent manner at the level of mRNA, are well positioned to take up

One hypothesis is that therapy directed to inhibit or normalizethis challenge. Moreover, these specific attributes also enable the

the elevated gene expression patterns (or at least of some of therapid discovery and development of antisense oligonucleotide

critical genes) seen in these disease models will be useful intherapeutics against the vast majority of candidate targets to be

treating or improving patients’ asthma or other allergic disorders.found in the post-genome era, i.e. those of unknown function.

However, it is likely that only a subset of the products of these up-Two problems must be addressed in the development of new

regulated gene targets will be both relevant and amenable ortherapeutic agents for asthma. First, a system for target validation,

‘tractable’ to therapy by small, orally active, organic moleculesto determine which novel targets are truly important in the asthma

(table II and table III).disease process, must be established. Secondly, once a target has

Here, it is perhaps worth expanding on this concept. In general,been shown to be important, a therapeutic agent must be identified.

drug targets can be classified as ‘tractable’ or ‘non-tractable’.In theory, use of antisense oligonucleotides (ASONs) should facil-

Generally large pharmaceutical companies will target tractableitate the process by addressing both of these problems. In contrast

molecules when developing novel therapeutics. Tractable targetsto small molecules, ASONs are designed from genomic informa-

are those that can be modulated or blocked by small orally activetion, which eliminates the need for screening large libraries to find

organic molecules possessing ‘drug-like properties’. Such proper-candidate molecules. Consequently, once a potential target is

ties have been described more formally by Lipinski’s rule of 5[5]

sequenced, an ASON can quickly be generated for use in targetwhich addresses the optimal molecular weight, lipophilicity, and

validation. Following validation, the ASON can be developed as athe number of hydrogen donors and acceptors. Examples of such

clinical antisense therapy candidate, thereby having the potentialtargets are G protein-coupled receptors, kinases, ion channels, etc.

to circumvent many of the limitations in development of conven-(table III).

tional small molecule therapeutics.The remaining targets have been termed ‘non-tractable’ or

‘non-drugable’ (table III) by pharmaceutical scientists, and clearly 2. Target Validation: Identifying Targets Involved inwarrant a new approach representing one of the biggest post- the Disease Process Using Antisensegenome challenges. Examples of such targets are the interaction Oligonucleotides (ASONs)sites of cytokines, growth factors or chemokines with their recep-

tors and the interaction sites of intracellular transcription factors. Gene expression requires transcription of genomic DNA

These targets are at present difficult to disrupt with conventional sequences into mRNA, which in turn are translated into proteins.

therapeutic agents because large proteins with multiple interaction Novel asthma targets, transcription factors, and other protein

sites are involved. While the variety of new approaches to this classes, will be the gene products of abnormally up-regulated

problem is beyond the scope of this review we believe that genes associated with an asthma phenotype. Targets that are often


100 Ball et al.

Table III. Examples of ‘tractable’ and ‘non-tractable’ or ‘difficult’ targets in developing small molecule therapeutic agents

Tractable Non-tractable or difficult

G protein-coupled receptors Interaction sites of cytokines, growth factors or chemokines with their receptors binding sites

Protein kinases Interaction site of intracellular transcription factors

Ion channels

Enzymes

Proteases

not tractable with small molecules, may be evaluated using sine A2b receptor nor bradykinin B2 receptor expression was

ASONs. affected.

Since the first report, a number of animal studies (table IV)In vitro studies conducted over the last 10 years have providedhave used antisense therapy (or a modification thereof) where theextensive evidence for the usefulness of ASONs in inhibitingrole of certain signal transduction molecules and other proteinspecific gene expression and protein production.[6] ASONs aretargets would have been difficult to evaluate using conventionaldesigned to hybridize via Watson-Crick base pairing to theirsmall molecule approaches. The ASONs were given via the air-‘sense’ counterparts. This hybridization blocks the templateways either by mechanical aerosolization or intranasal administra-properties of the mRNA, preventing its translation into the proteintion. The table does not include studies in which ASONs werefor which it codes. How it does this is not clear. It may involve theadministered by other routes, nor non–antisense-based oligonucle-simple blockade of the translation machinery, degradation of theotide modalities (e.g. CpG oligonucleotides) designed to modulatetarget mRNA by one or more of several intracellular RNasethe immune system.[8] As indicated in the footnotes to table IV,enzymes, or by other means. To achieve specificity and ability toeither a mismatch or nonsense (random) oligonucleotides are usedcross the cell membrane, the optimal length of ASONs rangesas a control for possible sequence-independent effects.from 14–25 base pairs long (unpublished data). Under intracellular

The main conclusions from these studies are that:conditions, any ASON that is not targeted to highly conserved or

• Down-regulation of gene expression following antisense ther-repetitive sequence structures should theoretically hybridizeapy has been shown in well-controlled studies against a diversi-uniquely to its target mRNA.ty of target types, including transcription factors, tyrosine ki-Disease target validation, by its very nature, often requires innases, G-protein coupled receptors, cytokine receptors, growthvivo administration of validating agents in appropriate diseasefactors, chemokines and ion channels.models. In vitro, ASONs have generally required the presence of

• Airway administered antisense therapeutics are able to pene-cationic proteins to facilitate their cellular uptake. The first reporttrate and act in a variety of lung cell types.showing that airway administered antisense therapy was effective

was in 1997 with the publication[7] of preclinical data on a 21-mer

phosphorothioate ASON called EPI-2010, designed to reduce 3. Factors Contributing to the Effectiveness ofadenosine A1 receptor expression. The rationale for EPI-2010 is Respirable ASONsthat overactivity of the adenosine-signaling pathway in asthmatic

lungs contributes to airway inflammation and hyperresponsive-The following anatomical and physiological properties of theness. EPI-2010 binds to the initiation codon of the adenosine A1

respiratory tract are likely to contribute to the effectiveness ofreceptor mRNA, and thereby blocks translation and targets therespirable ASONs (RASONs) in treatment of pulmonary disease:message for degradation by RNase. This article reported the

• aerosolization of ASONs provides direct access to the targetsurprising finding that short, single-stranded naked ASONs weretissuetaken up by lung tissues in an allergic rabbit model. A dose-

• a large surface area to surface liquid volume ratio in the lungsdependent inhibition of A1 receptor binding could be measured exfacilitates absorptionvivo for EPI-2010 but not for the corresponding mismatch control

sequence indicating a selective reduction in receptor expression. • cationic lipids, such as those found in pulmonary surfactant,

This effect was specific for the target mRNA since neither adeno- facilitate cellular uptake.



Table IV. Summary of effects of airway administered antisense oligonucleotides on antigen challenge response

Target Target type Species Route Size/control Effect Cell type References

GATA-3 Nuclear Mouse Intranasal 18-merab ↓ protein Th2 cells 9transcription factor ↓ eosinophil infiltration

↓ hyperresponsiveness

c-kit (SCF) Growth factor Mouse Intranasal 18-mera ↓ protein Cultured fibroblasts and 10keratinocytes

↓ IL-4 production Lung interstitium andepithelial cells

↓ eosinophil infiltration

Syk tyrosine Kinase Rat Aerosol 60-mer liposome ↓ mRNA and protein Alveolar macrophages 11,12kinase complexb

↓ markers of pulmonaryinflammation

β-chain of Receptors for BN Rat Intratracheal 18-mera ↓ mRNA and protein ↓ No cell type localization 13IL-3, IL-5, cytokine/ hyperresponsivenessGM-CSF chemokinesreceptor

Gob-5 Ca2+-dependent Mouse Intratracheal Adenovirus- ↓ mucus overproduction Epithelia 14chloride channel antisense gob-5c

↓ hyperresponsiveness

Adenosine GPCR Rabbit Aerosol 21-mera ↓ protein Smooth muscle 7A1R ↓ hyperresponsiveness

Bradykinin GPCR Rabbit Aerosol 21-mer ↓ protein Smooth muscle 7B2R

a Mismatch control.

b Nonsense control.

c No effect of adenoviral vector alone.

GM-CSF= granulocyte-macrophage colony stimulating factor; GPCR = G-protein coupled receptor; IL= interleukin.

Generation of aerosols with a specific particle size range pro- small volume of liquid coupled with the large surface area of the

vides a simple, safe, and effective method for delivering ASONs to respiratory tract facilitates delivery of ASONs to cells in and

respiratory targets. Aerosol particles between 0.01 and 100 mi- around the airways. Consider the local concentration that results

crons in diameter deposit at different locations within the respira- when 10mg of a 20-mer ASON with a phosphorothioate backbone

tory tract depending on their size. Particles larger than 5 microns in is dissolved in saline to a concentration of 5 mg/mL, aerosolized

diameter deposit in the nasopharynx due to impaction, whereas and inspired. Assuming 2% deposition,[15] a 10μm deep layer of

particles with diameters between 1 and 5 microns deposit in the liquid on the airway and alveolar surface, and surface area of 3m2

alveoli and small airways by sedimentation. Particles between 1 (rabbit), the concentration of ASON in the airway surface liquid

and 3 microns in diameter exhibit significant deposition in alveoli. could reach 0.3μM. Yet the total body load would be only 200 μg.

Thus, aerosolization of ASONs to a specific range of particle sizes Thus, a large surface area maximizes the chance for deposition,

provides a mechanism to maximize delivery to a target site within and a small amount of airway surface liquid minimizes dilution of

the respiratory tract. the deposited drug.

The surface of the respiratory tract is covered by a thin film of In addition to achieving high concentrations at the airway/

aqueous liquid, and upon contact with the surface, the aerosol alveolar surface, the alveolar surface liquid, and airway surface

particle dissolves into this liquid layer. Continuous absorption of liquid to a lesser extent, contain surfactant, which is composed of

salt and water by the alveolar and airway epithelium limit the cationic lipids similar to those used in vitro to facilitate uptake of

volume of this layer, ensuring that the airways remain patent. The naked oligonucleotides into cells.


102 Ball et al.

4. Optimal Models for ASON Testing and Human reach a cellular target, then failure to modulate the disease out-

Target Validation come is not immediately interpreted as a lack of a role for that

target. The use of the small animal models also provides a relative-

ly inexpensive and time conservative means to initially screen theAn optimized approach to ASON development and their use inin vivo efficacy of the test ASON, and evaluate dynamic andtarget validation for asthma includes development of an in vitrokinetic parameters of interest such as potency and half-life.human cell expression system for screening a library of ASONs

Subject to a positive outcome in the preliminary target valida-followed by in vivo testing in suitable animal models of asthma.tion studies, selected ASONs can be evaluated in primate modelsThe first step for ASON library screening is to utilize anof asthma leading to a firmer conclusion of whether a particularexisting human cell line or develop an human cell expressiontarget is validated, i.e. plays a role in human asthma or does not.vector which can be induced to express the target mRNA andWe believe that the Cynomolgus monkey allergen model,[17] beingprotein. Once this system has been established in a reproduciblea non-human primate, has a high likelihood of clinical predictivity.fashion, then a library of ASONs of optimal length (14–25 baseThis conclusion is based on the highly conserved genetic relation-pairs), designed to hybridize with specific regions of targetship between Cynomolgus monkeys and humans (95% con-mRNA, can be screened in vitro in the presence of cationic lipidsserved), and outcome measurements of the model indicating manyfor their ability to block the specific target message and protein.similarities with human asthma. Because of the genetic similarityThe ASONs determined in the screening process to be mostto humans, this model has the advantage that the ASON targetefficacious, can then be scheduled for further testing in the nextsequence will likely be conserved, and thus the test ASON, assum-step leading to target validation, in vivo screening.ing efficacy in Cynomolgus monkeys can then be developed forSeveral models of allergic asthma utilizing mice, rats, guineaclinical trials without further sequence modification.pigs, rabbits, sheep, and non-human primates, have been devel-

The ability to rapidly generate specific RASONs to validateoped.[16] In general, the models based on rodents and rabbits arenovel targets based on genomic information, coupled with theless expensive and allow more rapid screening than the largerability to rapidly progress those compounds directly into theanimal models. Also mechanistic studies are more easily per-clinics due to their specificity and low potential for toxicity, holdsformed in the smaller animal models as the animals are generallythe potential to markedly reduce the time required to generatesacrificed following a certain time course of modeling events anddisease-relevant target-specific drugs. To capitalize on this poten-multiple cell and tissue samples can be obtained. In mouse, rat,tial, preclinical testing must be done in animal models exhibitingguinea pig or rabbit models, depending on which species presentstarget mRNA sequences that are conserved in humans. This re-the most conserved mRNA and specific ASON sequence relativequirement may be best met through development of nonhumanto humans, the time-course of the expression of the relevantprimate models that accurately reproduce pulmonary diseases inmRNA/protein can be measured using bronchoalveolar lavagehumans.cells or in lung tissue upon sacrifice following antigen challenge.

After selection of the optimal time (plateau/peak) point a group of

5. What Are the Optimal Properties of Respirableanimals are pretreated with ASON and challenged with antigen,

ASONs for Clinical Therapeutics?and inhibition of the message/protein expression determined. In a

parallel group of animals, pulmonary function tests (PFT) orThe preceding section indicates that RASONs can be success-measurement of inflammatory cells and mediators (depending on

fully used in the experimental setting to validate novel targets inthe anticipated pharmacological effect) are performed at appropri-models of asthma. To develop an ASON for the clinic and reallyate timepoints. These preliminary target validation studies arecapitalize on this approach, many aspects of the research anduseful for several reasons. A key consideration in construction ofdevelopment must be considered (table V). The choice of se-the decision matrix is the desire to avoid false-negative or false-quence that results in the desired antisense effect while minimiz-positive interpretations of the data. Thus, it is important to measureing adverse effects is beyond the scope of this review.[18]a specific reduction of target message and protein produced by

ASON treatment. In the event of a finding that the ASON has not An extensive body of data has been accumulated in the last few

specifically inhibited target message translation or has failed to years on the safety of ASONs in humans, with supporting toxicol-



Table V. Key points in developing a therapeutic respirable antisense oligonucleotide (RASON) from an antisense tool

Property Goal Proposed approaches

High potency and Sufficient target inhibition without unacceptably large Choose targets with appropriate message, protein and cellselectivity deliverable doses and non-specific adverse effects turnover rates

Increase binding affinity of ASON target-mRNA complex usingnewer chemistriesScreen in silico against whole human genome sequence toavoid undesirable hybridization

Formulation and Acceptable purity and shelf life Optimize chemistries and formulationdelivery

Adequate delivery to cell target Confirm appropriate intracellular distributionMatch particle size to desired lung distribution and targetlocation

Drug metabolism and Appropriate physical and chemical properties conferring Use phosphorothioate backbone to increase stability and reducepharmacokinetics chemical and metabolic stability, and a duration of metabolic clearance

action consistent with clinically acceptable dose and (Metabolism by RNase is sequence-independent: thereforedose regimen similar pharmacokinetic properties expected for all similarly-sized

RASONs)

Acceptable safety and Acceptable safety margin between minimal effective Investigate toxicology studies in species with conservedtoxicology therapeutic dose and those causing adverse effects sequence to see target-dependent toxicities

Deliver by aerosol to minimize systemic levels and avoidadverse effects seen following intravenous administrationAvoid CpG motifs in new molecules to avoid immune systemstimulationExplore new chemistries to reduce back-bone related chemicaltoxicity

Clinical benefit Address those targets considered ‘non-tractable’ or Select patients with appropriate diagnosis where currentnon-amenable to treatment by conventional means therapeutic options are limited where there is an acceptable risk/

benefit ratio

ogy in a number of species.[19] The principle driver has been in cases of life-threatening disease where there is arguably an

acceptable risk to benefit ratio.clinical studies with ASONs in the treatment of cancers.[20,21]

The local delivery of ASONs represents one solution to achiev-These studies have generally relied on intravenous infusion, and as

ing effective disease tissue concentrations while minimizing sys-a consequence, some adverse effects related to backbone modifi-

temic dose effects. This approach has been used effectively in thecations and immune stimulation have been observed such as

complement activation (table VI). This has proved to be tolerable

Table VI. Summary of the sequence dependent and independent effects of phosphorothioate antisense oligonucleotides (ASONs)

Effect Sequence related? Mechanism Comment

Target inhibition in airways Yes Hybridization Desired effect

Target inhibition in non-airway tissue Yes Hybridization Avoided by airway administration of ASON

Inhibition of non-intended target containing Yes Hybridization Minimized by in silico screening of humanehomologous sequence genome sequence

CpG motif and G-quartet effects: Yes Non-hybridization Minimized by avoidance of such sequenceB cell mitogen motifsImmune stimulation

Complement activation and hypotension – No; non-specific Non-hybridization Minimized by selection of appropriate back-boneprolongation of clotting time interactions related to chemistries and avoidance of intravenous route

negatively chargedback-bone


104 Ball et al.

development of Fomivirsen™1 the first antisense therapeutic to submucosal compartments (unpublished observation). The

reach the market. Fomivirsen™ is an antisense treatment for RASONs are rapidly taken up by diverse cell types, including

AIDS-associated cytomegalovirus (CMV) retinitis, and is deliv- epithelial and smooth muscle cells, macrophages, lymphocytes,

ered locally by intravitreal injection.[22] Local delivery, i.e. the eosinophils, and neutrophils, and translocated to the nuclei. Con-

airway route, is also the one we advocate for the clinical adminis- sistent with decay of the radioactive signal in rabbits, the fluores-

tration of ASONs for the treatment of asthma. Our working cent signal is lost in most cell types over an 8-hour period in vivo.

hypothesis is that this will avoid adverse effects as can be seen Although inflammatory cells clearly take up the RASON, the

with EPI-2010. At efficacious doses EPI-2010 will not appear in practicality of targeting a migratory cell is questionable since

the circulation in sufficient amounts to cause adverse effects as ‘untreated’ cells continually enter the lungs from the circulation.

shown in table VI (unpublished observation). Phosphorothioate oligonucleotides do not appear in the circula-

Two key areas will be discussed in further detail; distribution tion when given orally. There is evidence that this is not related to

following airway administration (covering our experience with poor solubility or metabolic stability but to poor absorption char-

EPI-2010), and chemistry and metabolic stability. acteristics of the molecule across the gastrointestinal epitheli-

um.[23] It has been argued that this is to be expected based on the5.1 Distribution high molecular weight and hydrophilicity (log P approximately

–3.5).A study of the distribution of radiolabeled EPI-2010 in rabbit

lungs indicated a fairly circumscribed delivery to the lung. Autora-5.2 Chemistry Backbone and Metabolic Stability

diographic analysis indicated that radiolabeled RASON,

EPI-2010, delivered by nebulization resulted in uniform depo- Most antisense molecules in clinical trials are DNA (2′-deox-

sition of drug throughout the large and small airways of the rabbit ynucleotidic acid) oligonucleotides with a phosphorothioate back-

lung. In contrast to studies involving parenteral delivery of an- bone (where one of the non-covalently bonded oxygen atoms is

tisense therapeutic agents where the highest concentrations of drug replaced by a sulphur atom) resulting in increased in vivo resis-

are found in the liver and kidney, the majority of EPI-2010 derived tance to nucleases relative to standard phosphodiester oligonucleo-

radioactivity was found in the target lung tissue. There was very tides. However, the phosphorothioate backbone modification has

little extrapulmonary distribution of EPI-2010 derived radioactivi- been linked to toxic adverse effects when oligonucleotides con-

ty. In fact the EPI-2010-derived [35S]-radioactivity found in the taining them have been administered in high doses. Although local

heart, liver, and kidney combined was only 6.9% of the adminis- delivery of RASONs to the lung can minimize many of these

tered dose.[15] In addition, post-labeling studies indicated that adverse effects, it is desirable to evaluate new generation

intact EPI-2010 could only be detected in the lung. These data oligonucleotide chemistries that enhance hybridization to target

provide additional evidence that RASONs do not effectively es- mRNAs, reduce backbone related chemical toxicity, and inhibit

cape the lung as intact molecules. EPI-2010 was found to be the oligonucleotide susceptibility to nuclease degradation in

relatively short-lived in the lung; our data demonstrate that only vivo.[18,24,25]

approximately 3% of the delivered drug remained in the lung after It is worth noting that the metabolism of phosphorothioate72 hours and that it’s elimination half-life in the lung was approxi- oligonucleotides is by nucleases, and not by cytochrome P450mately 30 hours.[15] The studies described here measure (CYP) enzymes that metabolize small organic molecules. Clinical-EPI-2010-derived radioactivity, and as such, provide data for ly, drug-drug interactions are a relatively common finding withEPI-2010 and its metabolites. The EPI-2010 elimination half-life small molecules because of their common metabolic pathway.is at the lower end of the range (20–120 hours, depending on tissue Because of nuclease metabolism, this is not anticipated to be aor organ) reported for phosphorothioates delivered via parenteral problem with antisense drugs, and the lack of drug-drug interac-routes.[19] tions represents a key advantage. It should be noted that all

Subsequent studies using fluorescent-labeled RASON have potential antisense drug candidates are screened in silico against

demonstrated effective distribution throughout the mucosal and the mRNA sequence of all members of the CYP family of en-

1 The use of tradenames is for product identification purposes only and does not imply endorsement.



zymes to avoid possible effects on the synthesis of the cytochrome Long-term toxicity will need to be addressed in the context of

enzymes. target blockade, the chemical backbone of the oligonucleotide, and

delivery. The effect of long-term sequence-specific target block-5.3 Immunological Adverse Effects ade is no different than what is expected with long-term use of

small molecule drugs, and will require the same evaluation pro-Certain sequence motifs merit attention in the design of

cess. Toxicity, due to the oligonucleotide and its metabolites, hasRASONs since they have been postulated to mimic DNA motifs

been a concern because of the polyanionic charge of fullyserving biological function. One example is the dinucleotide

phosphorothioated molecules, which can interact with proteinsCpG[26,27] of which greater than 90% are methylated at C-5 in

nonspecifically. Advances in backbone chemistry are being madeanimal DNA, in contrast to bacterial DNA, where CpG dinucleo-

and the long-term safety of these molecules is still being evalu-tides remain essentially unmethylated. This difference in C-5

ated.methylation within CpG dinucleotides acts as a signal to animal

A further consideration concerns chemical stability and formu-immune systems, with the detection of nonmethylated CpGs

lation selection (particle size) and use of an appropriate inhalation(characteristic of bacterial DNA) inducing a potent TH1-based

device to deliver the ASONs. Oligonucleotides of this sizeimmune response. Oligonucleotides with unmethylated CpGs

(19–23-mers) range are chemically stable and can be synthesizedwithin certain sequence contexts resemble bacterial DNA and

on a large scale within acceptable cost limits. Since the physicalinduce an immune response. This characteristic can be exploited

properties of oligonucleotides are considered to be sequence inde-therapeutically,[8] but can also confound observations in the ab-

pendent the development of an efficient airway delivery device issence of appropriately designed control ASONs. Another example

anticipated to benefit all future RASON therapies.of confounding sequence motifs are G-rich oligonucleotides capa-

ble of forming ‘G-quartet’ secondary structures that confer non- 6. Conclusionantisense-based mechanisms of activity.[28]

We now have a better understanding of which particular fea-5.4 Clinical Efficacy of EPI-2010 tures of the development process are critical in developing a

therapeutic RASON from a laboratory antisense tool.In phase I studies, aerosolized doses of EPI-2010 did not affect

Clearly this holds potential in addressing those up-regulatedany coagulation or cardiovascular parameter, supporting the hypo-

novel disease targets that are amenable to antisense therapy but arethesis that the local delivery ensures minimal systemic bioavaila-

‘non-tractable’ by conventional small molecules. It is hoped thatbility. Preliminary evidence suggests that the long duration of

this will yield new therapeutic options translating into improvedbiological effect (one week or longer) that was observed in ani-

asthma control for patients.mals may hold true for humans. While a scientific report on

clinical findings has not yet been published, a significant decrease Acknowledgementsin the need of patients to use bronchodilator therapy (36%, p =

The authors at Epigenesis Pharmaceuticals, Inc. have a number of0.026), concomitant with reduction in a symptom scores (p =RASONs in research and development.0.032) was observed (data to be published elsewhere). EPI-2010 is

currently in phase II clinical trials, and thus more conclusiveReferencesinformation on safety and clinical efficacy is expected in 2003.

1. WHO/OMS. WHO information: bronchial asthma fact sheets [online]. WHO FactSheet No 206. Rev 2000 Jan. Available from URL: http://www.who.int/inf-fs/

5.5 Development Hurdles en/fact206.html [Accessed 2003 Mar 3]

2. Asthma and Allergy Foundation of America [online]. Available from URL: http://www.aafa.org [Accessed 2003 Mar 3]The development of ASONs is still in its infancy with only

3. Zimmermann N, Moulton EA, Aronow BJ, et al. Gene expression profile analysisFomivirsen™ available on the market. Some lessons are clear. Thein experimental asthma. J Allergy Clin Immunol 2002 Jan; 109 (1 Suppl.):

choice of a target having a conserved sequence homology between S296,905

4. Zou J, Young S, Zhu F, et al. Microarray profile of differentially expressed genes inin vitro cell culture, animal and human target sequences allowsa monkey model of allergic asthma. Genome Biol 2002; 3 (5): 20.1-20.13

rapid evaluation of pharmacology. Similarly, toxicology is ideally5. Walters WP, Ajay, Murcko MA. Recognizing molecules with drug-like properties.

performed in species with appropriate target sequence homology. Curr Opin Chem Biol 1999 Aug; 3 (4): 384-7


106 Ball et al.

6. Dove A. Antisense and sensibility. Nat Biotechnol 2002; 20: 121-4 18. Agrawal S. Importance of nucleotide sequence and chemical modifications of

antisense oligonucleotides. Biochim Biophys Acta 1999 Dec 10; 1489 (1):7. Nyce JW, Metzger WJ. DNA antisense therapy for asthma in an animal model

53-68[published erratum appears in Nature 1997 Nov 27; 390 (6658): 424]. Nature1997 Feb 20; 385 (6618): 721-5 19. Levin AA. A review of the issues in the pharmacokinetics and toxicology of

phosphorothioate antisense oligonucleotides. Biochim Biophys Acta 1999 Dec8. Kandimalla ER, Yu D, Agrawal S. Towards optimal design of second-generation10; 1489 (1): 69-84immunomodulatory oligonucleotides. Curr Opin Mol Ther 2002 Apr; 4 (2):

122-9 20. Cho-Chung YS. Antisense DNAs as targeted therapeutics for cancer: no longer a

dream. Curr Opin Investig Drugs 2002 Jun; 3 (6): 934-99. Finotto S, De Sanctis GT, Lehr HA, et al. Treatment of allergic airway inflamma-tion and hyperresponsiveness by antisense-induced local blockade of GATA-3 21. Tamm I, Dorken B, Hartmann G. Antisense therapy in oncology: new hope for anexpression. J Exp Med 2001 Jun 4; 193 (11): 1247-60 old idea? Lancet 2001 Aug 11; 358 (9280): 489-97

10. Finotto S, Buerke M, Lingnau K, et al. Local administration of antisense 22. de Smet MD, Meenken CJ, van den Horn GJ. Fomivirsen: a phosphorothioatephosphorothioate oligonucleotides to the c-kit ligand, stem cell factor, sup- oligonucleotide for the treatment of CMV retinitis. Ocul Immunol Inflammpresses airway inflammation and IL-4 production in a murine model of asthma. 1999; 7: 189-98J Allergy Clin Immunol 2001 Feb; 107 (2): 279-86

23. Nicklin PL, Bayley D, Giddings J, et al. Pulmonary bioavailability of a11. Stenton GR, Kim MK, Nohara O, et al. Aerosolized Syk antisense suppresses Syk phosphorothioate oligonucleotide (CGP 64128A): comparison with other deliv-

expression, mediator release from macrophages, and pulmonary inflammation. ery routes. Pharm Res 1998 Apr; 15 (4): 583-91J Immunol 2000 Apr 1; 164 (7): 3790-7

24. Wang H, Cai Q, Zeng X, et al. Antitumor activity and pharmacokinetics of a12. Stenton GR, Ulanova M, Dery RE, et al. Inhibition of allergic inflammation in the

mixed-backbone antisense oligonucleotide targeted to the RIα subunit of pro-airways using aerosolized antisense to syk kinase. J Immunol 2002 Jul 15; 169

tein kinase A after oral administration. Proc Natl Acad Sci U S A 1999; 96:(2): 1028-36

13989-9413. Allakhverdi Z, Allam M, Renzi, et al. Inhibition of antigen-induced eosinophilia

25. Crooke ST. An overview of progress in antisense therapeutics. Antisense Nucleicand airway hyperresponsiveness by antisense oligonucleotides directed against

Acid Drug Dev 1999; 8: 115-22the common β chain of Il-3, Il-5, GM-CSF receptors in a rat model of allergicasthma. Am J Respir Crit Care Med 2002; 165: 1015-21 26. Branch AD. A hitchhiker’s guide to antisense and nonantisense biochemical

pathways. Hepatology 1996; 24: 1517-2914. Nakanishi A, Morita S, Iwashita H, et al. Role of gob-5 in mucus overproductionand airway hyperresponsiveness in asthma. Proc Natl Acad Sci U S A 2001 Apr 27. Coulson JM, Poyner DR, Chabtry A, et al. A nonantisense sequence-selective24; 98 (9): 5175-80 effect of a phosphorothioate oligodeoxynucleotide directed against the epider-

mal growth factor receptor in A431 cells. Mol Pharmacol 1996; 50: 314-2515. Ali S, Leonard SA, Kukoly CA, et al. Absorption, distribution, metabolism, andexcretion of a respirable antisense oligonucleotide for asthma. Am J Respir Crit 28. Dapic V, Bates PJ, Trent JO, et al. Antiproliferative activity of G-quartet-formingCare Med 2001; 163: 989-93 oligonucleotides with backbone and sugar modifications. Biochemistry 2002

Mar 19; 41 (11): 3676-8516. Karol MH. Animal models of occupational asthma. Eur Respir J 1994 Mar; 7 (3):555-68

17. Van Scott MR, Hooker JL, Kukoly C, et al. Characterization of TruPrimateTM aCorrespondence and offprints: Dr Howard A. Ball, EpiGenesis Pharmaceuti-

nonhuman primate model for identification and validation of gene targets incals, Inc., 7 Clarke Drive, Cranbury, NC 08512, USA.allergic asthma. Keystone Conference Rethinking the Pathogenesis of Asthma;

2002 Feb 14, Taos (NM) E-mail: [email protected]


Documents

Clinical Potential of Respirable Antisense Oligonucleotides (RASONs) in Asthma