Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://rnp-group.genebee.msu.su/pdfs/MolBio6_00KopylovLO.pdf Äàòà èçìåíåíèÿ: Mon Oct 21 17:39:11 2002 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 19:43:44 2012 Êîäèðîâêà:

Molecular Biology, Vol. 34, No. 6, 2000, pp. 940­954. Translated from Molekulyarnaya Biologiya, Vol. 34, No. 6, 2000, pp. 1097­1113. Original Russian Text Copyright © 2000 by Kopylov, Spiridonova.

UDC 577.271.34

Combinatorial Chemistry of Nucleic Acids: SELEX
A. M. Kopylov1 and V. A. Spiridonova
1 2

2

Department of Chemistry, Moscow State University, Moscow, 119899 Russia Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow, 119899 Russia
Received May 16, 2000

Abstract--A method of irrational oligonucleotide design, SELEX, is considered. Individual SELEX products, aptamers, are small molecules (40­100 nt) that have a unique three-dimensional structure, which provides for their specific and high-affinity binding to targets varying from low-molecular-weight ligands to proteins. Thus, the sophisticated biosynthesis of recognizing protein elements, antibodies, can be emulated in vitro via selection and synthesis of principally new recognizing elements based on nucleic acids. Key words: nucleic acids, combinatorial chemistry, aptamers

INTRODUCTION Utilization of the information written in the oligonucleotide primary structure has allowed the elaboration and rapid progress of the atisense technology. However, this one-dimensional information also determines a unique, complex tertiary structure of single-stranded oligonucleotides. In 1990, this property has been put into use in a method elaborated independently by two research groups to isolate nucleic acid (NA) molecules of interest from a combinatorial library of more than 1015 individual molecules [1, 2]. This method of irrational oligonucleotide design has been termed SELEX (Systematic Evolution of Ligands by EXponential enrichment); individual selected products have been termed aptamers (from the Latin aptus, fitting). To select RNA aptamers, a DNA template is transcribed; the RNA binds to an immobilized target; the unbound fraction is removed; the bound fraction, which is enriched in a sequence of interest, is eluted and used as a template for cDNA synthesis; and the cDNA is amplified by PCR to an amount sufficient for the next round (Fig. 1). To select DNA aptamers, asymmetrical PCR is carried out instead of transcription. Six to ten such rounds yield aptamers which specifically and efficiently bind with a given ligand. Aptamers have a size of 40­100 nt depending on the size of the random region in the combinatorial DNA library. Since SELEX is considered in several reviews [3-16], here we analyze only certain aspects rather than all modern applications and the great progress made with this method in studying the structure and functions of NAs, proteins, and their complexes.

SELEX In combinatorial chemistry, any experiment, including SELEX, involves three steps: synthesis of a library, selection or screening, and structural analysis of the resulting aptamer­target complex. Synthesis of Nucleic Acid Libraries To construct a combinatorial library, the initial template is obtained via chemical synthesis of DNA fragments, each consisting of a random sequence of 30­60 nt flanked with defined sequences. Several strategies have been proposed for constructing combinatorial libraries [17]. With DNA, the most common method is automatic stepwise synthesis with four, rather than one, syntons introduced at each step. With a degenerate sequence of n nt, a library of Nmax = 4n individual DNA molecules is obtained. The actual degeneracy of a library depends on the amount of DNA used for selection rather than on Nmax. Relatively small libraries are currently preferred to giant ones. This is explained by "the weak lead" phenomenon; i.e., aptamers displaying moderate binding are major in pools obtained at the intermediate steps of selection. These aptamers determine the apparent total positive signal, and mask signals from better ones, which occur in far lower concentrations. Selection To select RNA aptamers, a combinatorial library is obtained via transcription of a DNA combinatorial library. Hence DNA fragments must each contain a promoter at the 5' end, the unique T7 promoter being most commonly used. For instance, we have constructed such a library to select aptamers to amylase BLMA (Fig. 2) [18]. Hybridization of primer 2, which

0 0026-8933/00/3406-0940$25.00 © 2000 MAIK "Nauka / Interperiodica"


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX
DNA COMBINATORIAL LIBRARY

941

RNA COMBINATORIAL LIBRARY Binding

Transcription

ENRICHED DNA FRACTION

Replication and PCR

COLUMN WITH IMMOBILIZED LIGAND

BOUND RNA MOLECULES

Elution UNBOUND RNA MOLECULES Fig. 1. Selection of RNA aptamers to a ligand immobilized on a sorbent.

contains the T7 promoter, with the initial template 1, which contains a 30-nt random sequence, yielded partial DNA duplexes (step I). The duplexes were made perfect with DNA polymerase II (step II). To obtain an RNA library, the duplexes were transcribed with T7 RNA polymerase (step III). Selection of aptamers binding with a target is the key step in SELEX, as aptamers account for only a small fraction of the initial library (one aptamer per 109 to 1013 molecules) [6]. Aptamer selection takes advantage of specific physical (or functional) properties of the aptamer­target complex. In affinity chromatography, a ligand is immobilized on a polymeric carrier. Aptamers are bound with the ligand on a column and are thereby separated from unbound molecules. In the above example, BLMA was immobilized on thiopropyl-activated Sepharose. An RNA fraction enriched in aptamers was eluted and used to synthesize cDNA with reverse transcriptase (Fig. 2, step IV). The cDNA was amplified in PCR for the next round of selection. The amount of RNA binding with the sorbent increases with every next round, demonstrating the enrichment in aptamers (Fig. 3). Another selection method involves sorption of proteins and protein­NA complexes on nitrocellulose filters. Gel electrophoresis utilizes the difference in mobility between free and target-bound molecules.
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

Separation of complexes and unbound molecules via equilibrium dialysis is less common. Several rounds of selection yield a fraction enriched in aptamers, with relative binding increasing several orders of magnitude compared with the initial library. Isotherms of thrombin binding have been compared between initial libraries and aptamer pools obtained in 11 rounds of selection (Fig. 4). The apparent dissociation constant (aKd) has been estimated at 8.3 and 5.0 µM for the libraries with a random region of 30 and 60 nt, respectively, and at 2.3 and 3.0 nM for the corresponding aptamer pools. Thus, the overall affinity increases by three orders of magnitude upon selection, regardless of the size of the degenerate region. Finally, an aptamer pool is cloned by the standard methods; individual aptamers are sequenced, and their affinity for a ligand estimated. The aggregate results provide for choosing the best aptamer. Not only do the resulting aptamers efficiently bind with a target protein, but they also recognize its individual structural elements (epitopes), as observed for DNA aptamers to mutant -thrombin. Substitution of Glu for Arg70 in the mutant prevented its binding with aptamers to the wild-type protein. Aptamers to the mutant protein bound with the wild-type thrombin, apparently interacting with an epitope unaffected by the mutation [19, 20].


942 Template 1 65 nt Primer 2 Primer 3 43 nt 17 nt
5' 5' 5'

KOPYLOV AND SPIRIDONOVA
3' 3' 3'

Hybridization of template with primer 2 (step I) 3'
5' 3' 5'

DNA polymerase (step II)
3' 5' 5' 3'

T7 RNA polymerase (step III)
3' 5'

(N15 molecules)

Binding with ligand Reverse transcriptase + primer 3 (step IV)
3' 5' 5' 3'

PCR + primer 2 + primer 3 (step V)
3' 5' 5' 3'

Fig. 2. Major steps and enzymic reactions in SELEX illustrated by the example of a combinatorial library used to select aptamers to amylase BLMA. The T7 promoter for RNA polymerase in primer 2 is underlined.

Thus, aptamers can be obtained even to proteins that does not normally bind with NAs, such as thrombin. This demonstrates that SELEX is universal, as are immunochemical techniques. Strong specificity and high affinity for a target make aptamers similar to antibodies. To automate, the SELEX procedure has been adapted to robot-compatible microtitration plates [21]. Three targets (alkaline phosphatase, -thrombin, and platelet growth factor) were immobilized through

hydrophobic adsorption. Still carried out manually, SELEX yielded aptamers with Kd less than 0.3 nM. Structure of Aptamer­Target Complexes The structure of aptamers in a family obtained by SELEX can be optimized to enhance a certain property. This requires structural data on an aptamer and on its complex with a target. Their structures can be directly determined by the X-ray and nuclear magnetic resonance (NMR) analyses [22­24]. However,
MOLECULAR BIOLOGY Vol. 34 No. 6 2000


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX

943

complexes of aptamers with target proteins are commonly modeled with indirect methods. First, a putative secondary structure of an aptamer is predicted. One way involves consensus comparisons of the possible secondary structures of all aptamers in a family, as it is done in the comparative phylogenetic analysis elaborated by Woese and colleagues [25]. Then the tertiary structure is modeled [26]. The second step is to identify the aptamer sites that are in contact with a target via standard footprinting and damage-selection [27]. In addition, the deletion analysis of the aptamer is carried out. This often makes it possible to minimize the size of aptamers, because their ends of known primary structure are often not involved in selection. For instance, one of the best minimal aptamers to thrombin, GGTAGGGCAGGTTGG, binds with the target to produce a complex with aKd = 0.5 nM. The aptamer forms two G quartets and has additional fragments forming a standard DNA duplex (Fig. 5). Interestingly, the quartets are similar to those of telomeric DNA [28]. Analysis of chemical crosslinks with target proteins is also informative. When known for an aptamer and a ligand, the tertiary structure can be assumed for their complex. Introduction of reactive groups in certain nucleotides allows addressed modification of the target [29, 30]. On evidence of photoinduced crosslinking, a 5iododeoxyuridine derivative of aptamer T12 is in contact with Phe245 of thrombin (Fig. 5), suggesting a certain structure for their complex (Fig. 6). With two reactive groups introduced, the arrangement of an aptamer and a protein can be deduced with higher confidence, as demonstrated for selectin (Fig. 7) [10]. SELECTION WITH LIBRARIES OF NAs CONTAINING MODIFIED NUCLEOSIDES Aptamers obtained in classical SELEX display high affinity and specificity for a target, and their postselectional modification can further enhance these parameters. The ribose ring and the phosphodiester and glycosidic bonds have torsional limitations in nucleotides [31]. Complementary pairing further limits conformational changes of an aptamer. These are the disadvantages of NAs compared with proteins, which have greater torsion freedom and more easily change their conformation. Modification with additional chemical groups increases the conformational freedom and enhances the ability of aptamers to interact with their targets. Moreover, chemical groups can be varied to achieve a better aptamer­target contact in complexes of known structure.
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

30 25 20 15 10 5 0 1' 2' 3' 4' 5' 6' 7' 8' 9' 10' 11' Round

Fig. 3. Enrichment in aptamers binding with immobilized BMLA according to selection rounds [16].

Retention, % 100 80 60 40 20 0 10

­10

Binding, %

10

­9

10­8 10­7 Thrombin, M

10

­6

10

­5

Fig. 4. Isotherms of thrombin binding with ( ) an initial combinatorial library with a 30-nt random region and ( ) an aptamer fraction resulting from 11 rounds of selection; similar data ( and , respectively) have been obtained for a 60-nt random region [87].

Expansive libraries, i.e., those of NAs containing modified nucleosides, differ from libraries based on the four natural nucleosides. First, even a small amount of modified nucleosides greatly increases the complexity of a library. Second, modified nucleosides provide for a far greater conformational diversity of aptamers and thereby increase the potential of a library. Third, modified aptamers can have a more stable conformation, which makes it possible to reduce the aptamer size. Fourth, modification can increase the aptamer affinity for a target. Fifth, modified aptamers can be more resistant to hydrolysis, which expands the field of their application. Libraries of modified NAs can be obtained in two ways. When modified NTPs are used as substrates in the polymerase reactions, SELEX yields modified aptamers. The major drawback is that only certain modifications allow NTPs to be efficiently introduced in aptamers because of the substrate specificity of


944
60­18 (38) 3' A AG G

KOPYLOV AND SPIRIDONOVA

T G C T T C A G T G G G15 G
14

5' C A G T C C

60­18 (29) 60­18 (27)

G TA 9 G1 G G2 T3 T13 T12

C8 G7 G6
10

G5 A4

G11

233 241 ... RLKKWIQKVIDQFGE Fig. 5. Spatial structure of DNA aptamers to thrombin [87]. Aptamer sizes are indicated in parentheses. Nucleotides of the major motif of the two G quartets are numbered. Arrows indicate T12 of the 29-nt aptamer and Phe245 of thrombin which are linked together in the complex.

polymerases. Another approach is postselectional modification; i.e., modified syntons are used in combinatorial chemical synthesis to obtain derivatives of aptamers selected. This method does not have the above limitations and makes it possible to simultaneously analyze various derivatives. However, this causes the necessity of reselecting the modified aptamers. Then, to establish the structure of the resulting modified aptamer is not a trivial problem; this procedure has been aptly termed deconvolution. Only the progress in mass spectrometry (MS) makes deconvolution possible [32]. Aptamers can be at random modified with various hydrophobic, hydrophilic, and charged groups in the hope that some of their combinations would improve the aptamer properties. In contrast, rational design of an aptamer derivative is based on the structure assumed for an epitope recognized by the initial aptamer. Nucleosides for Aptamer Modification Various nucleosides can be used in both strategies. Since modification must not change folding nor inhibit interactions of an aptamer with its target, modifying groups are most commonly introduced at positions 5 in pyrimidines, 8 in purines, and 2' in nucleo-

sides. In addition, 2'-deoxypurines with a modifying group at position 7 can also be introduced in DNA by T7 DNA polymerase [33]. Modifying groups that do not require additional blocking/deblocking are considered optimal [34]. At present, about 100 various modified purine and pyrimidine phosphamidites are commercially available or can be easily synthesized [10, 35, 36] (Fig. 8). To modify 2'-deoxyuridine (Fig. 8a), a small hydrophobic group, such as vinyl (1) and 1-pentinyl (5), or a larger group, such as pyrenyl (13, 14), are introduced in a single step [37, 38] with the help of palladium catalyst (cited from [10]). Uridine is modified with keto (17, 18) [34] or amido (24, 27) [39] groups, the latter being capable of additional H-bonding. Most of these groups can also be used in postselectional modification. Similar derivatives can be obtained for cytidine. Relatively simple methods have been elaborated for modification of purine nucleosides [40] (Fig. 8b). Deoxyribo- and ribonucleotide derivatives have various modifying groups at position 8, including hydrophobic groups (28­30), the steric shielding of which can be regulated with an amide linker (31, 33, 34, 37, 38), hydrophilic (35), charged (36, 39), and affinity groups such as biotinyl (40). Versions of stereospecific methods have been elaborated to synthesize 2'-modified pyrimidine nucleosides [41­43]. This makes it possible to use virtually all hydroxylamines and many other amines. The above hydrophobic (41­43, 46, 49) and hydrophilic (45, 47) groups can also be introduced at position 2'; fucosyl group (45) can be used to obtain an affinity derivative (Fig. 8c). Chemical modification at position 2' is also promising. Uridine derivative 48 can be used as an intermediate substrate for substituting imidazole with various hydrophobic, hydrophilic, charged, and reactive groups at the pre-activation step in solid-phase synthesis of phosphoamidites. Since 2'-modified aptamer derivatives are highly structured, conformationally rigid molecules, specific methods have been elaborated for their purification [44]. Note that each 2'-modified uridine derivative can be transformed in a cytidine derivative. Complexity of Modified Aptamer Libraries Selection or postselectional modification with only the above 20 nucleoside derivatives would already result in giant libraries. For instance, for an aptamer of 20 modified nucleotides, a library would contain 2020 (about 1026) sequences; i.e., their number would be greater than Avogadro's number. Hence it is virtually impossible to use a complete library. As synthesis is
MOLECULAR BIOLOGY Vol. 34 No. 6 2000


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX

945

Fig. 6. Spatial structure assumed for the complex of thrombin with the 29-nt aptamer [87].

(b) (c) (a)
DMS Pt-POP
T Photocrosslinking to Phe O Crosslinking to Lys 87 C
44

via a Schiff base

H

CT G A T­A G­A G­C A A C­G A­T T­A G­A A­T C­G C­G O C H

Fig. 7. Modeling of the spatial structure of an aptamer­protein complex based on the mapping of intermolecular contacts [10]: (a) the secondary structure of a 26-nt DNA aptamer to L-selectin based on the footprinting and crosslinking data; (b) three-dimensional models of the aptamer (top) and L-selectin (bottom), with groups involved in crosslinking shown black; (c) three-dimensional model of the aptamer­selectin-L complex, with the components arranged according to the data on protein crosslinking with the loop and base of the aptamer (Phe44­T15, Lys87 are neighbors of the 3'-aldehyde group).

commonly run to 10­100 µmol of NA, the number of modifications per nucleotide position is restricted to less than ten, in order to generate a representative library. In addition, giant libraries require more efficient and rigorous selection.
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

Deconvolution of Aptamers Modified after Selection The progress in deconvolution is due to electrospray ionization MS (ESI-MS) [32], which has


946

KOPYLOV AND SPIRIDONOVA

revealed the structure of minor nucleotides even in cell RNAs [45, 46]. Pomeranets has first demonstrated that deconvolution of a mixture of postselectionally modified aptamers is principally possible with ESIMS (cited from [10]). A library of DNA dodecamers has been obtained and three positions randomized with three thymine analogs, yielding 27 individual molecules. A specific MS version, tandem MS, has been successfully used to identify each oligonucleotide. Although the library has a low degeneracy and consists of short oligonucleotides, these results suggest that ESI-MS can be used for deconvolution with more complex libraries if their 25- to 30-nt items are cleaved into 10- to 15-nt fragments. A library of aptamers cleaved into M fragments, each with degeneracy N, has complexity MN. Parameter N is the number of unique sequences that can be resolved by ESIMS without preliminary HPLC fractionation. For instance, a library of 390,625 aptamers can be split into four sets, each containing about 25 individual oligonucleotides. If selection would reduce the library size ten times, then a library of postselectionally modified aptamers can be extended to 106 molecules. McCloskey et al. [47] have considered modern achievements in establishing the oligonucleotide structure in complex mixtures via HPLC combined with MS. Examples Affinity for the basic fibroblast growth factor (bFGF) has been increased by substituting only one thymidine in DNA aptamers (cited from [10]). This thymidine is a low-conserved nucleotide located in a conserved "tetraloop" of the initial aptamer (Fig. 9). Its substitution with either 5-[N-(aminoethyl)-3-acrylamido]deoxyuridine or 5-[N-(aminohexyl)-3-acrylamido]deoxyuridine increase the affinity of aptamers compared with the initial one, the factor being higher than 5 with the latter. In an attempt to obtain nuclease-insensitive aptamers to the vesicular endothelial growth factor, an aptamer has been selected from a library modified with 2'-aminopurines [48]. The aptamer was used to construct 2'-O-methylpurine derivatives, each containing three or four substitutions. After selection, aptamer regions that allow substitutions were identified via electrophoretic analysis of alkaline hydrolysates. Based on the data obtained, an aptamer with substitutions in 10 out of 14 positions was synthesized (Fig. 10). The aptamer was nuclease-insensitive and had a 17 times higher affinity as compared with the initial one. The NA conformation is determined by relatively simple secondary-structure motifs, such as hairpins, which allows original design and organic synthesis of stable aptamers termed oligonucleotide mimetics. For

instance, Osborne et al. [49] have shown that disulfide bridges formed by N-3-thioethylthymidine at the base of a hairpin may lead to thermal stabilization of the secondary structure of a deoxyribonucleotide. Moreover, a conformation varying in an initial oligonucleotide is fixed in a modified one. Mimetics can be obtained not only via disulfide bridging [50­54], but also with other structural elements formed with terephthalamide [55], 9-(2-deoxy-5-O-triphospho-D-ribofuranosyl)-N6,N6-ethano-2,6-diaminopurine [56], psoralen-modified deoxyadenosine [57], and oligoethylene glycol [58­62] (Fig. 11). An interesting mimetic has been obtained with stilbene dicarboxamide which is capable of a photoinduced conformational change [63]. Being conformationally rigid, the derivative is as effective as a UUCG tetraloop in stabilizing RNA structures and far more effective than a TTTT tetraloop or a triethylene glycol linker in stabilizing DNA structures [64]. Its effect is so high that a 22-nt modified oligonucleotide binds with the Rev protein of HIV-1 almost as well as a 94-nt native fragment [64]. The mimetic is only 7.8 kDa, 2.1 kDa smaller than the smallest known Rev-binding element and 18.3 kDa smaller than the best aptamer, which is essential for elaboration of mimetic-based therapeutic agents. Covalent SELEX: Selection of Aptamers Capable of Covalent Linking to a Target Protein An interesting class includes modified aptamers that contain reactive groups forming crosslinks with a target protein and thereby preventing dissociation of the complex. Reversible crosslinks are easily adaptable to selection, for instance, with equilibrium binding. Such crosslinks can be obtained via imino, acetal, ester, and disulfide bridging and via bonding with ,-unsaturated carbonyl groups. Fig. 7 shows several synthetic nucleosides which can be used for this purpose, including 2'-deoxyuridine (3, 6­11, 14) and uridine (15­18) derivatives. Photoinduced crosslinking is also possible with modified aptamers [65]. Crosslinking is accidental and depends on the spatial arrangement and distance between reactive groups of an aptamer and a protein in the complex. Hence a special procedure has been elaborated to obtain aptamers that can be covalently crosslinked to a protein [10]. First, an aptamer library with one or several reactive group is selected. The condition for selection is linking in a complex; e.g., denaturing gel electrophoresis is used. Then complexes are isolated from a gel and aptamers amplified. Since SELEX is a thermodynamic process based on competitive binding, tight binding must be achieved before a chemical reaction. Covalent SELEX can be employed not only
MOLECULAR BIOLOGY Vol. 34 No. 6 2000


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX

947

f

OEt (a)
* * *

O
*

TMS 4

*

1 O
* *

2 O

3 O OEt
*

5 O
*

6 X = H; R =
*

7

8

TMS

9 O

*

10

O

11 O

O HN O O HO OH 15 O
*

*

*

R N X

*

12 O
* *

13 O OEt 16 O
*

14 O
*

O

17

TMS OH

18 O
*

N H 19 O

*

N H 20 O

N H 21 O

NH2

X = OH; R =
*

N H 22 O

OEt O

*

N H 23 O

*

N 24 O N O 26 27
*

HN

*

N H 25

N N H

*

HN

Fig. 8. Examples of synthetic modified nucleosides: (a) 2'-deoxyuridine and uridine derivatives with a modifying group at position 5; (b) 2'-deoxyadenosine, adenosine, and guanosine derivatives with a modifying group at position 8; (c) derivatives of 2'-modified uridine [10].

in studying NA­protein interactions, but also in constructing effective therapeutic agents. For instance, an amplification protocol with crosslinking through 5'-iodouridine (5-IU) [29, 66] has been used to obtain aptamers to HIV-1 Rev [65].
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

With all U substituted, aptamer 6a binds with Rev to yield a complex with Kd = 100 nM [67­69]; i.e., the binding is 100 times weaker than with the initial aptamer [70, 71]. A mixture of Rev and an RNA library with 5-IU randomly substituted for U in 35 positions has been exposed to monochromatic light of


948

KOPYLOV AND SPIRIDONOVA

(b) ,

OEt
* * * *

O
*

O N O O 32

HN 28 O NH2 N N N R N O HO O HN NH2 N N N O OH HO R OH
* * *

29 O

30 31 O
* *

HN R' R' = H or OH, R = OH
*

HN 34

HN OH 35 HN

HN N 36 OO
* (+)

33 HN

*

O 37 O N H H N O 40

O 38 S NH

O HN 39 HN NH H2N

HN

O

(c) d

*

H N O 41

*

H N O 42

O

CF
3

3

F3C CF

O HN O O HO OH N R R=
* *

H N O 43 H N

NO2
*

HO OH Me H N O O 45 O 47 OH OH O

H N O 44

Me
*

H
*

N O

H
*

46 O 48 H N

N

N

N O O N H

*

49

Fig. 8. (Contd.) MOLECULAR BIOLOGY Vol. 34 No. 6 2000


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX AC CG T G GC G GC A TC G G G G G C G 5' O HN O O HO OH HO OH N O N H NH2 HN O O N

949

G T T T* A C AA A C C C C G C 3' T O O N H NH2

Fig. 9. Secondary structure assumed for a DNA aptamer to the basic fibroblast growth factor. In postselectional modification, a thymidine base (indicated with an asterisk) has been substituted with either of the 2'-deoxyuridine derivatives shown below (cited from [10]).

a XeCl (308 nm) or HeCd (325 nm) laser. The complex was isolated via electrophoresis in 7 M urea and digested with proteinase K. Double SELEX was run for 13 cycles, with selection for binding in cycles 1­3 and 8­10 and for linking in cycles 4­7 and 11­13. This yielded two conserved aptamer families: class 1 with Kd 1­10 nM and 30­40% linking and class 2 with Kd 30­50 nM and 60­70% linking. The best minimal aptamer showed binding with Kd = 0.8 nM and 40% linking. Thus, covalent SELEX yields highaffinity aptamers with similarly arranged reactive groups. APPLICATION OF APTAMERS Compare again the advantages and drawbacks of aptamers and antibodies. Both antibodies and aptamers can be obtained to various targets, but only the latter ones allow low-immunogenic and toxic targets. Like antibodies, aptamers show a high specificity and a high affinity for their targets [72]. Aptamers are smaller and structurally simpler than antibodies. While antibodies are produced only in living systems, aptamers are in vitro synthesized with chemical and enzymic methods, and their production can be automated. Aptamers are more convenient and promising. Hence it is not surprising that application of aptamers, for instance, in diagnostics and therapy is investigated in recent years. Applied studies on SELEX are few because of the common, though wrong, opinion that
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

NAs are too complex and their synthesis is too expensive to be economically justified. Below we consider problems which must be solved before aptamers can be widely used in diagnostics and therapy. Aptamers as Diagnostic Reagents The immediate prospect is to employ aptamers in diagnostic tests and probing. Standard immunochemical methods have been adapted for aptamers, including dot hybridization [73], Western blotting [74], ELISA [75], affinity binding, fluorescence polarization (cited from [10]), and flow cytometry [10, 75, 76]. Standard binding with a radiolabeled aptamer allows quantitation of proteins, as demonstrated for protein kinase C [77] and human thyroid-stimulating hormone [73]. A 123I-labeled aptamer to thrombin has been used for in vivo detection of clots [78]. However, to be widely employed in clinical practice, aptamers must be detected via a nonradioisotope method with a comparable sensitivity; e.g., aptamers can be covalently linked to an enzyme [12, 21]. Fluorescent labels can also be used to detect aptamers. Drolet et al. [74] have synthesized a 5'-labeled fluorescent derivative of an aptamer to the vesicular endothelial growth factor, with a combined antibody­aptamer sandwich used in detection. This detection method seems most promising.


950 G AA u c 10G G c Gc G G G G u G
3'
14

KOPYLOV AND SPIRIDONOVA

7A

u u c c c A

6G

5'

Fig. 10. Secondary structure assumed for an aptamer to the vesicular endothelial growth factor which has been modified with 2'-O-methylpurines (boldface), purine riboderivatives (numbered), and 2'-aminopyrimidines (shown with small letters) (cited from [10]).

Fluorescent derivatives of aptamers can be detected in a solution, for instance, via flow cytofluorimetry. Davis et al. [75] have synthesized fluorescent aptamers to human neutrophilic elastase which detect the enzyme bound on polystyrene beads. A similar method has been used to detect specific marker antigens on the cell surface [79]. Fluorescent aptamers that change their conformation upon ligand binding can be used for direct detection of the ligand in a solution [12]. SELEX itself suggests an original approach to construction of principally new fluorescent probes [12]. With a label introduced in selection, SELEX may yield aptamers with exclusively ligand-dependent fluorescence. Ellington and colleagues [80] have designed biosensors which contain a fluorescent aptamer as a recognizing element [80]. In typical immunochemical analysis on microchips, an aptamer capable of photoinduced crosslinking can be used in place of both binding and detecting antibodies, which allows direct staining of a target protein [81]. Possibilities of Using Aptamers in Therapy As regards their therapeutic potential, aptamers can serve as drugs themselves or as a vehicle for specific delivery of other conjugated drugs. To be used as drugs, aptamers must meet several requirements. First, an aptamer must effectively and specifically bind to its target. Several best aptamers indeed form complexes with aKd = 0.1 nM [48, 82­84]; with most aptamers, the constant is comparable with that of complexes formed by the antigen-binding fragment (Fab) of antibodies [6]. Owing to a large size, aptamers can discriminate between epitopes of very similar proteins. For instance, aptamers differentially bind to protein kinase C isozymes having a 96% homology [85].

Second, an aptamer must specifically block the function of its target. For instance, DNA aptamers inhibit HIV reverse transcriptase with Ki = 1 nM [86]. RNA aptamers and their modified derivatives to bFGF block its binding to the membrane receptors at an extremely low (1 nM) concentration [82, 83]. An RNA aptamer to the vesicular endothelial growth factor inhibits its binding to the receptor at 20­40 nM [82]. Not only aptamers must suppress the function of their target, they also must change or inhibit the corresponding biological effect. For instance, aptamers to thrombin suppress coagulation [87­90]. An aptamer to human neutrophilic elastase inhibits interleukin 1-induced neutrophil-mediated degradation of the lung tissue in rat [91]. Aptamers to IgE block IgEmediated secretion of serotonin in a cell culture [92]. DNA aptamers to L-selectin inhibit migration of human and mouse lymphocytes [76]. At a low (0.092 nM) concentration, modified RNA aptamers to keratinocyte growth factor already suppress its mitogenic activity in a cell culture [93]. Aptamers selected for binding to virions of the Rous sarcoma virus are able to suppress virus infection [94]. Briefly mentioned, these examples clearly demonstrate the therapeutic potentialities of aptamers. However, several problems must be solved before aptamers are used in therapy. Solution of the most apparent problem of aptamer stability is in progress. Natural NAs, which consist of the four standard nucleotides bonded with the phosphodiester bonds, are relatively unstable both in cells and, for example, in serum. SELEX with modified nucleotides yields aptamers which not only possess new properties but also are surprisingly stable in vivo [35, 37, 95]. The most apparent chemical way to stabilize NAs is thiophosphate, rather than phosphodiester, bonding [48, 96]. Modification of the sugar moiety enhances nuclease resistance of aptamers [84, 95, 97]. Compared with initial aptamers, 2'-amino derivatives of aptamers to human neutrophilic elastase are more stable in the serum (20 h vs. 4 min) and urine (9 h vs. 8 min) [95] and those of aptamers to bFGF are at least 1000 times more stable in 90% human serum [84]. Modified aptamers to the vesicular endothelial growth factor remain stable for 17 h [48]. Even a greater nuclease resistance is achieved via conjugation with a liposome bilayer [98]. A drawback is that the 2'-amino modification destabilizes the structure of initial aptamers. This is not so with 2'-fluoropyrimidine derivatives, which are also nuclease-insensitive and have a higher thermostability [93, 99]. Such aptamers NX1838 to the vesicular endothelial growth factor (NeXstar) have shown adequate properties with cell cultures [100], suggesting further analysis of their therapeutic potential. Bartel and colleagues [101] have elaborated an original approach of constructing mirror-image
MOLECULAR BIOLOGY Vol. 34 No. 6 2000


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX

951

NH2 O O O O G C C U C A G C U C T S G A G T S N HO ON O O O O C GU GU A G C A A RNA-O HYR ON
HYR RNA-O

N ON O

RNA-O NH2 HYR O NH O

HYR RNA-O NH2

O SS O N

O

NO

RNA O-HYR

O-RNA HYR

OH

Fig. 11. Structure of a chimeric aptamer to HIV-1 Rev [64]. The terminal loop is substituted with polyethylene glycol, the base of the hairpin is stabilized via disulfide bridging, and uridines are substituted with 2'-aminoderivatives in the side loop.

aptamers [101]. They selected an aptamer to the enantiomer of a target (vasopressin) and then synthesized the enantiomer of the aptamer as a nuclease-insensitive ligand to the normal target. The method is further developed by the Noxxon company. The pharmacokinetic properties of aptamers have come to be studied [102]. The above NX1838 aptamer is the first candidate for clinical trial. Conjugated with polyethylene glycol, this aptamer inhibits the vesicular endothelial growth factor. Injected in monkeys in a single dose of 1 mg/kg, the aptamer occurs in the blood serum (25.5 µg/ml) and has a half-life 9.3 h [103]. Another problem to be solved is to elaborate a way of aptamer delivery. Aptamers are hardly suitable for oral administration and, most likely, are to be injected alone, in a conjugate with other biopolymers or polyethylene glycol, or packed in liposomes [104]. Lebruska and Maher [105] have proposed an interesting strategy of in vivo trapping by aptamers. The gist is to express an RNA aptamer in the nucleus of transformed cells, as done with an aptamer to transcription factor NF-B. Indeed, Rev-binding RNA
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

fragments corresponding to the Rev-responsive element have demonstrated the antiviral effect on HIV-1 when expressed in a cell culture [106, 107]. The problem of economical synthesis and production must be solved before aptamers are widely used. As many aptamers are 30­40 nt, new achievements and the modern level of automation in chemical synthesis of NAs will make their production economically feasible in the nearest future. Finally, potential resistance to the therapeutic effect is the major problem to be solved in constructing new drugs, including aptamers. The extremely high specificity is an unquestionable advantage and may prevent systemic side effects of therapeutic aptamers compared with other drugs. However, an aptamer may not block a protein altered by a mutation, which results in the resistance to its effect. To solve this problem, it is possible to select aptamers to a conserved protein region unaffected by mutations and thereby to block all potential mutants of the target protein. This property can be utilized in combined therapy, especially when a target becomes resistant to


952

KOPYLOV AND SPIRIDONOVA

a common drug as a result of monochemotherapy. Gold and colleagues [108] have shown that aptamers to reverse transcriptase of the feline immunodeficiency virus (HIV-1 analog) inhibit the polymerase activity of the enzyme not only from the wild-type virus, but also from its azidothymidine-resistant mutant. In addition, aptamers may serve as a vehicle for high-selective drug delivery when selected to bind a specific surface marker of a cell or a tissue. Aptamers to complex targets, such as cell lines or organs, may be used for tissue-specific delivery. As to the pharmacological application of aptamers in the more distant future, two notes should be made. Studies of the aptamer­target system will make it possible to identify not only the active center of a drug affecting a known target, but also targets that are still unknown. One way is to employ the potent method of photoinduced covalent crosslinking. Aptamers can be used as competitors in order to identify a metabolite that interacts with a drug possessing similar binding properties. Finally, aptamers may serve as functional mimetics of drugs. To conclude, we would like to cite the prognosis made by A.D. Ellington and colleagues, Aptamer Research Center, University of Indiana, United States [12]. "Just as monoclonal antibody facilities exist in many major corporate and academic research setting, it is possible that aptamer facilities will now begin to be set up. These facilities may reasonably be expected to include a robotic workstation that starts with purified targets or even target mixtures, and returns populations of binding species that can be quickly characterized by automated sequence acquisition and analysis. The selected aptamers can then be used to quantitate, localize, or inhibit proteins, even proteins whose function is unknown. As research reagents, aptamers may eventually contribute to nascent efforts in functional genomics, providing ready-made inhibitors of the multitude of new genes identified by genome projects." ACKNOWLEDGMENTS We are grateful to K.-H. Park and A.A. Bogdanov for help and support and to A.V. Golovin, T.I. Rassokhin, and Yu.Yu. Pakhomova for help in experiments and manuscript preparation. Our works on SELEX were supported by the Russian Foundation for Basic Research (projects nos. 9804-49 005, 99-04-39 072), the Universities of Russia program (project no. 015-05-0208), and the Russian Ministry of Science (grant no. 415/99).

REFERENCES
1. Tuerk, C. and Gold, L., Science, 1990, vol. 249, pp. 505­510. 2. Ellington, A.D. and Szostak, J.W., Nature, 1990, vol. 346, pp. 818­822. 3. Burke, J.M. and Bezzal-Herranz, A., FASEB J., 1993, vol. 7, pp. 106­112. 4. Klug, S.J. and Famulok, M., Mol. Biol. Rep., 1994, vol. 20, pp. 97­107. 5. Breaker, R.R. and Joyce, G.F., Trends Biotechnol., 1994, vol. 12, pp. 268­275. 6. Gold, L., Polisky, B., Uhlenbeck, O., and Yarus, M., Annu. Rev. Biochem., 1995, vol. 64, pp. 763­797. 7. Gold, L., J. Biol. Chem., 1995, vol. 270, pp. 13 581­ 13 584. 8. Iverson, B.L., Chem. Biol., 1995, vol. 2, pp. 67­70. 9. Bartel, D.P. and Szostak, J.W., RNA­Protein Interactions, Nagai, K. and Mattaj, I.W., Eds., Oxford: IRL, 1996, pp. 248­268. 10. Eaton, B.E., Gold, L., Hicke, B.J., et al., Bioorg. Med. Chem., 1997, vol. 5, pp. 1087­1096. 11. Gold, L., Brown, D., He, Y.-Y., et al., Proc. Natl. Acad. Sci. USA, 1997, vol. 94, pp. 59­64. 12. Osborne, S.E., Matsumura, I., and Ellington, A.D., Curr. Opin. Chem. Biol., 1997, vol. 1, pp. 5­9. 13. Famulok, M. and Jenne, A., Curr. Opin. Biol., 1998, vol. 2, pp. 320­327. 14. Famulok, M. and Mayer, G., Curr. Top. Microbiol. Immunol., 1999, vol. 243, pp. 123­136. 15. Famulok, M., Curr. Opin. Struct. Biol., 1999, vol. 9, pp. 324­329. 16. Kopylov, A.M., Spiridonova, V.A., and Park, K.-H., Ros. Khem. Zh., 1998, vol. 42, pp. 89­96. 17. Gordon, E.M., Barrett, R.W., Dower, W.J., et al., J. Med. Chem., 1994, vol. 37, pp. 1385­1401. 18. Kim, I.C., Cha, J.I., Kim, J.R., et al., J. Biol. Chem., 1992, vol. 267, pp. 22 108­22 114. 19. Padmanabhan, K., Padmanabhan, K.P., Ferrara, J.D., et al., J. Biol. Chem., 1993, vol. 268, pp. 17 651­17 654. 20. Tsiang, M., Gibbs, C.S., Griffin, L.C., et al., J. Biol. Chem., 1995, vol. 270, pp. 19 370­19 376. 21. Drolet, D.W., Jenison, R.D., Smith, D.E., et al., Comb. Chem. High Throughput Screen., 1999, vol. 2, pp. 271­ 278. 22. Patel, D.J., Curr. Opin. Chem. Biol., 1997, vol. 1, pp. 32­46. 23. Hermann, T. and Patel, D.J., Science, 2000, vol. 287, pp. 820­825. 24. Hermann, T. and Patel, D.J., Struct. Fold Des., 2000, vol. 15, pp. 47­54. 25. Pace, N., Thomas, B.C., and Woese, C.R., The RNA World, Gesteland, R.F., Cech, T.R., and Atkins, J.F., Eds., New York: CHSL, 1999, pp. 113­141. 26. Moore, P., The RNA World, Gesteland, R.F., Cech, T.R., and Atkins, J.F., Eds., New York: CHSL, 1999, pp. 381­ 401. 27. McGregor, A., Murray, J.B., Adams, C.J., et al., J. Biol. Chem., 1999, vol. 274, pp. 2256­2262.
MOLECULAR BIOLOGY Vol. 34 No. 6 2000


COMBINATORIAL CHEMISTRY OF NUCLEIC ACIDS: SELEX 28. Wang, K.Y., Swaminathan, S., and Bolton, P.H., Biochemistry, 1994, vol. 33, pp. 7517­7527. 29. Willis, M.C., Hicke, B.J., Uhlenbeck, O.C., et al., Science, 1993, vol. 262, pp. 1255­1257. 30. Hicke, B.J., Willis, M.C., Koch, T.H., and Cech, T.R., Biochemistry, 1994, vol. 33, pp. 3364­3373. 31. Cantor, C. and Shimmel, P., Biophysical Chemistry. Translated under the titile Biofizicheskaya khimiya, Moscow: Mir, 1984, vol. 1, pp. 287­294. 32. Ni, J., Pomerantz, S.C., Rozenski, J., et al., Anal. Chem., 1996, vol. 68, pp. 1989­1999. 33. Porter, K.W., Tomasz, J., Huang, F., et al., Biochemistry, 1995, vol. 34, pp. 11 963­11 969. 34. Dewey, T.M., Mundt, A., Crouch, G.J., et al., J. Am. Chem. Soc., 1995, vol. 117, pp. 8474­8475. 35. Eaton, B.E., Curr. Opin. Chem. Biol., 1997, vol. 1, pp. 10­16. 36. Eaton, B.E. and Pieken, W.A., Annu. Rev. Biochem., 1995, vol. 64, pp. 837­863. 37. Latham, J.A., Johnson, R., and Toole, J.J., Nucleic Acids Res., 1994, vol. 22, pp. 2817­2822. 38. Crouch, G.J. and Eaton, B.E., Nucleosides Nucleotides, 1994, vol. 13, pp. 939­944. 39. Dewey, T.M., Zyzniewski, M.C., and Eaton, B.E., Nucleosides Nucleotides, 1996, vol. 15, pp. 1611­1617. 40. Tu, C., Keane, C., and Eaton, B.E., Nucleosides Nucleotides, 1995, vol. 14, pp. 1631­1638. 41. Sebesta, D.P., O'Rourke, S.S., Martinez, R.L., et al., Tetrahedron, 1996, vol. 52, pp. 14 385­14 402. 42. McGee, D.P.C., Sebesta, D.P., O'Rourke, S.S., and Martinez, R.L., Tetrahedron Lett., 1996, vol. 37, pp. 1995­1998. 43. McGee, D.P.C., Vaughn-Settle, A., Vargeese, C., and Zhai, Y., J. Org. Chem., 1996, vol. 61, pp. 781­785. 44. Bridonneau, P., Bunch, S., Tengler, R., et al., J. Chromatogr. B. Biomed. Sci. Appl., 1999, vol. 726, pp. 237­ 247. 45. Limbach, P.A., Crain, P.F., and McCloskey, J.A., Nucleic Acids Res., 1994, vol. 22, pp. 2183­2196. 46. Felden, B., Hanawa, K., Atkins, J.F., et al., EMBO J., 1998, vol. 17, pp. 3188­3196. 47. McCloskey, J.A., Whitehill, A.B., Rozenski, J., et al., Nucleosides Nucleotides, 1999, vol. 18, pp. 1549­1553. 48. Green, L.S., Jellinek, D., Bell, C., et al., Chem. Biol., 1995, vol. 2, pp. 683­695. 49. Osborne, S.E., Volker, J., Stevens, S.Y., et al., J. Am. Chem. Soc., 1996, vol. 118, pp. 11 993­12 003. 50. Ferentz, A.E. and Verdine, G.L., J. Am. Chem. Soc., 1991, vol. 113, pp. 4000­4002. 51. Gao, H., Yang, M., and Cook, A., Nucleic Acids Res., 1995, vol. 23, pp. 285­292. 52. Cain, R.J., Zuiderweg, E.R.P., and Glik, G.D., Nucleic Acids Res., 1995, vol. 23, pp. 2153­2160. 53. Goodwin, J.T. and Glik, G.D., Tetrahedron Lett., 1994, vol. 35, pp. 1647­1650. 54. Goodwin, J.T., Osborn, S.E., Swanson, P.C., and Glik, G.D., Tetrahedron Lett., 1994, vol. 35, pp. 4527­ 4530. 55. Salunke, M., Wu, T., and Letsinger, R.L., J. Am. Chem. Soc., 1992, vol. 114, pp. 8768­8772.
MOLECULAR BIOLOGY Vol. 34 No. 6 2000

953

56. Cowart, M. and Benkovic, S.J., Biochemistry, 1991, vol. 30, pp. 788­796. 57. Pieles, U., Sproat, B.S., Neuner, P., and Cramer, F., Nucleic Acids Res., 1989, vol. 17, pp. 8967­8978. 58. Durand, M., Peloille, S., Thuong, N.T., and Maurizot, J.C., Biochemistry, 1992, vol. 31, pp. 9197­9204. 59. Rumney, S. and Kool, E.T., Angew Chem. Int. Ed. Engl., 1992, pp. 1617­1619. 60. Rumney, S. and Kool E.T., J. Am. Chem. Soc., 1992, vol. 117, pp. 5635­5646. 61. Ma, M.Y.-X., Reid, L.S., Clime, S.C., et al., Biochemistry, 1993, vol. 32, pp. 1751­1758. 62. Ma, M.Y.-X., McCallum, K., Clime, S.C., et al., Nucleic Acids Res., 1993, vol. 21, pp. 2585­2589. 63. Letsinger, R.L. and Wu, T., J. Am. Chem. Soc., 1995, vol. 117, pp. 7323­7328. 64. Nelson, J. S., Giver, L., Ellington, A.D., and Letsinger, R.L., Biochemistry, 1996, vol. 35, pp. 5339­5344. 65. Jensen, K.B., Atkinson, B.L., Wills, M.C., et al., Proc. Natl. Acad. Sci. USA, 1995, vol. 92, pp. 12 220­12 224. 66. Stump, W.T. and Hall, K.B., RNA, 1995, vol. 1, pp. 55­ 63. 67. Tuerk, C., MacDougal, S., Hertz, G., and Gold, L., The Polymerase Chain Reaction, Ferre, R., Mullis, K., Gibbs, R., and Ross, A., Eds., New York: Birkhauser, 1993, pp. 233­243. 68. Tuerk, C. and MacDougal-Waugh, S., Gene, 1993, vol. 137, pp. 33­39. 69. Jensen, K.B., Green, L., MacDougal-Waugh, S., and Tuerk, C., J. Mol. Biol., 1994, vol. 235, pp. 237­247. 70. Bartel, D., Zapp, M., Green, M., Szostak, J.W., Cell, 1991, vol. 67, pp. 529­536. 71. Giver, L., Bartel, D., Zapp, M., et al., Nucleic Acids Res., 1993, vol. 21, pp. 5509­5516. 72. Foote, J. and Eisen, H.N., Proc. Natl. Acad. Sci. USA, 1995, vol. 92, pp. 1254­1256. 73. Lin, Y., Nieuwlandt, D., Magallanez, A., et al., Nucleic Acids Res., 1996, vol. 24, pp. 3407­3414. 74. Drolet, D.W., Moon-Dermott, L., and Romig, T.S., Nat. Biotechnol., 1996, vol. 14, pp. 1021­1025. 75. Davis, K.A., Abrams, B., Lin, Y., and Jayasena, S.D., Nucleic Acids Res., 1996, vol. 24, pp. 702­706. 76. Hicke, B.J., Watson, S.R., Koenig, A., et al., J. Clin. Invest., 1996, vol. 98, pp. 2688­2692. 77. Conrad, R. and Ellington, A.D., Anal. Biochem., 1996, vol. 242, pp. 261­265. 78. Dougan, H., Hobbs, J.B., Weitz, J.I., and Lyster, D.M., Nucleic Acids Res., 1997, vol. 25, pp. 2897­2901. 79. Davis, K.A., Lin, Y., Abrams, B., and Jayasena, S.D., Nucleic Acids Res., 1998, vol. 26, pp. 3915­3924. 80. Potyrailo, R.A., Conrad, R.C., Ellington, A.D., and Hieftje, G.M., Anal. Chem., 1998, vol. 70, pp. 3419­ 3425. 81. Brody, E.N., Willis, M.C., Smith, J.D., et al., Mol. Diagn., 1999, vol. 4, pp. 381­388. 82. Jellinek, D., Green, L.S., Bell, C., and Janjic, N., Biochemistry, 1994, vol. 33, pp. 10 450­10 456. 83. Jellinek, D., Linott, C.K., Rifkin, D.B., and Janjic, N., Proc. Natl. Acad. Sci. USA, 1993, vol. 90, pp. 11 227­ 11 231.


954

KOPYLOV AND SPIRIDONOVA 97. Pieken, W.A., Olsen, D.B., Benseler, F., et al., Science, 1991, vol. 253, pp. 314­317. 98. Willis, M.C., Collins, B.D., Zhang, T., et al., Bioconjug. Chem., 1998, vol. 9, pp. 573­582. 99. Biesecker, G., Dihel, L., Enney, K., and Bendele, R.A., Immunopharmacology, 1999, vol. 42, pp. 219­230. 100. Bell, C., Lynam, E., Landfair, D.J., et al., In Vitro Cell Dev. Biol. Anim., 1999, vol. 35, pp. 533­542. 101. Williams, K.P., Liu, X.H., Schumacher, T.H., et al., Proc. Natl. Acad. Sci. USA, 1997, vol. 94, pp. 11 285­ 11 290. 102. Reyderman, L. and Stavchansky, S., Pharm. Res., 1998, vol. 15, pp. 904­910. 103. Tucker, C.E., Chen, L.S., Judkins, M.B., et al., J. Chromatogr. B Biomed. Sci. Appl., 1999, vol. 732, pp. 203­ 212. 104. Konopka, K., Duzgunes, N., Rossi, J., and Lee, N.S., J. Drug Target, 1998, vol. 5, pp. 247­259. 105. Lebruska, L.L. and Maher, J. III, Biochemistry, 1999, vol. 38, pp. 3168­3174. 106. Lee, T.C., Sullenger, B.A., Gallardo, H.F., et al., New Biol., 1993, vol. 4, pp. 66­74. 107. Lee, S.W., Gallardo, H.F., Gilboa, E., and Smith, C., J. Virol., 1995, vol. 68, pp. 8254­8264. 108. Chen, H., McBroom, D.G., Zhu, Y.Q., et al., Biochemistry, 1996, vol. 35, pp. 6923­6930.

84. Jellinek, D., Green, L.S., Bell, C., et al., Biochemistry, 1995, vol. 34, pp. 11 363­11 372. 85. Conrad, R., Keranen, L.M., Ellington, A.D., and Newton, A.C., J. Biol. Chem., 1994, vol. 269, pp. 32 051­ 32 054. 86. Schneider, D.J., Feigon, J., Hostomsky, Z., and Gold, L., Biochemistry, 1995, vol. 34, pp. 9599­9610. 87. Tasset, D.M., Kubik, M.F., and Steiner, W., J. Mol. Biol., 1997, vol. 272, pp. 688­698. 88. Kubik, M.F., Stephens, A.W., Schneider, D., et al., Nucleic Acids Res., 1994, vol. 22, pp. 2619­2626. 89. Block, L.C., Griffin, L.C., Latham, J.A., et al., Nature, 1992, vol. 355, pp. 564­566. 90. Griffin, L.C., Tidmarsh, G.F., Bock, L.C., et al., Blood, 1993, vol. 81, pp. 3271­3276. 91. Smith, D., Kischenheuter, G.P., Charlton, J., et al., Chem. Biol., 1995, vol. 2, pp. 741­750. 92. Wiegand, T.W., Williams, P.B., Dreskin, S.C., et al., Immunology, 1996, vol. 157, pp. 221­230. 93. Pagratis, N.C., Bell, C., Chang, Y.-C., et al., Nat. Biotechnol., 1997, vol. 15, pp. 68­73. 94. Pan, W., Craven, R.C., Qiu, Q., et al., Proc. Natl. Acad. Sci. USA, 1995, vol. 92, pp. 11 509­11 513. 95. Lin, Y., Qiu, Q., Gill, S.C., and Jayasena, S.D., Nucleic Acids Res., 1994, vol. 22, pp. 5229­5234. 96. Tam, R.C., Wu-Pong, S., Pai, B., et al., Antisense Nucleic Acid Drug Dev., 1999, vol. 9, pp. 289­300.

MOLECULAR BIOLOGY

Vol. 34

No. 6

2000