- Email: [email protected]
Selection and analytical applications of aptamers
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Trends
Selection and analytical applications of aptamers Camille L.A. Hamula, Jeffrey W. Guthrie, Hongquan Zhang, Xing-Fang Li, X. Chris Le Advances in systematic evolution of ligands by exponential enrichment (SELEX), a selection protocol for aptamers, have resulted in increased applications of DNA and RNA aptamers in developing analytical techniques. We review recent developments in SELEX techniques as well as new aptamerbased bioanalytical applications. ª 2006 Elsevier Ltd. All rights reserved. Keywords: Affinity binding; Aptamer; Bioanalytical assay; Biosensor; Chiral separation; In vitro selection; Protein; SELEX
Camille L.A. Hamula1 , Jeffrey W. Guthrie1 , Hongquan Zhang, Xing-Fang Li, X. Chris Le* Environmental Health Sciences, Department of Public Health Sciences and Department of Laboratory Medicine and Pathology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada, T6G 2G3
1
These two authors contributed equally. *Corresponding author. Tel.: +1 780 492 6416; Fax: +1 780 492 7800 E-mail: [email protected]
1. Introduction Nucleic acids possess desirable properties for use in analytical applications. Conventionally, they are used as probes that hybridize with and detect complementary target RNA and DNA sequences. In the last decade, short synthetic DNA and RNA sequences known as aptamers have been used as ligands to bind to non-nucleic acid targets with high specificity and affinity. Dissociation constants for typical aptamer binding are in the micromolar to low picomolar range [1,2]. This binding is not through sequence hybridization but through binding of stable, complex stemloop and internal loop structures formed by the nucleic acids [1]. Aptamers are commonly evolved in vitro via a combinatorial chemistry technique known as systematic evolution of ligands by exponential enrichment (SELEX) [3]. They can be selected using a variety of targets from small molecules to whole organisms [4]. The ability to synthesize aptamers for a variety of targets has contributed to a wide range of analyses using aptamers. Another factor allowing the widespread application of aptamers to bioanalysis is that aptamers offer benefits for analytical applications when used in place of antibodies [5]. These benefits derive mainly
0165-9936/$ - see front matter ª 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2006.05.007
from temperature stability and ease of aptamer production. Many of these analytical applications of aptamers, such as those based on chromatography, mass spectrometry, molecular beacons, and some label-free biosensors, have been reviewed previously [6–9], and will not be discussed further. This review focuses on recent developments in SELEX analytical applications, as well as variations in SELEX techniques and the ramifications of such variations for aptamer characteristics.
2. SELEX techniques Fig. 1 shows the basic principle of most SELEX procedures [10]. Attempts to improve SELEX have led to the development of different techniques, delineated by variations in starting library, target molecule, and experimental protocol. These variations result in differences in aptamer affinity and specificity as well as selection efficiency. 2.1. Libraries A typical SELEX library comprises singlestranded RNA or DNA, in which a central region of randomized sequence is flanked on either side by fixed primer sequences for PCR and/or RT-PCR amplification. The first SELEX procedures used randomized single-stranded RNA libraries, in which the random sequence region was 100 nucleotides (nt) in length [1] and 8 nt [3], respectively. Since then, many other SELEX procedures have been carried out using randomized ssRNA or ssDNA libraries, typically containing 1013–1015 different randomized sequences [11]. Although any length of randomized region 681
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Figure 1. Schematic of Systematic Evolution of Ligands by Exponential Enrichment (SELEX) [10]. Briefly, typical SELEX procedures involve binding a random nucleic-acid library to a target, separating the bound and unbound nucleic acids, and amplifying the bound nucleic acids by the polymerase chain reaction (PCR) for use in the next round of selection [1,3]. After each round of selection, a smaller pool of nucleic acid sequences binding to the target is retained and the unbound nucleic acids are discarded. Typically, 8–15 rounds of SELEX are carried out in order to generate a pool of aptamers with sequences enabling the highest binding affinity for the target. These aptamers can then be cloned and sequenced. (Reprinted, with permission, from the Annual Review of Medicine, Volume 56 ª2005 by Annual Reviews www.annualreviews.org)
may be used to provide 4n theoretical random sequences, in reality, synthetic limitations prevent most libraries from containing more than 1016 random sequences (e.g., for a library to contain one molecule each of 1016 random sequences, a minimum of 16.7 nmole nucleic acids (or 0.4 mg of nucleic acids containing a 40 nt random region and two 20 nt primer regions, 25 kDa) is required). In practice, therefore, randomized regions of 30–60 nt are most common [11]. DNA libraries are particularly useful for applications requiring increased aptamer stability, such as biosensing, environmental monitoring, and therapy. However, in many cases, RNA libraries yield aptamers with higher binding affinities than DNA libraries due to the ability of RNA to take on a wider variety of conformations than DNA. Several SELEX techniques have been developed using modified randomized ssRNA or ssDNA libraries. Photocrosslinking procedures, described in more detail later, use ssDNA libraries substituted with fluorophore682
http://www.elsevier.com/locate/trac
modified nucleotides, such as 5 0 -BromodeoxyuridineUTP or 5-iodouracil [12,13]. Upon irradiation with a laser, these nucleotides facilitate crosslinking of an aptamer to the target. Genomic SELEX procedures make use of ssDNA or ssRNA libraries derived from the genome of an organism of interest to probe for genome sequences that interact with a specified target [14–16] Currently, there are Escherichia coli, Saccharomyces cerevisiae and human genomic DNA libraries that are able to be transcribed into RNA or amplified by PCR [14]. In all other respects, genomic SELEX is similar to conventional SELEX. Unfortunately, genomic SELEX procedures tend to be biased towards strong or highly abundant interactions and yield many interactions that may not be important in vivo. So far, genomic SELEX has yielded aptamers with relatively low binding affinities and high dissociationconstant (Kd) values of 1.9–12.8 lM in 5 rounds of selection [15].
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
2.2. SELEX targets SELEX has been applied to a variety of molecular targets. The first SELEX procedures involved generation of aptamers against organic dyes [1] and T4 DNA polymerase [3]. Since then, a myriad of other biomolecules have been targeted. Most of these are summarized by Ellington et al. in The Aptamer Database [17], which includes organic and inorganic small-molecule, nucleic acid, protein, peptide, aminoglycoside and carbohydrate targets, among others. SELEX is best suited to generating aptamers against large molecules, which offer the necessary surface area for interaction with aptamers. Target molecules must also be stable, and it is useful if they can be modified to enable partitioning. Hence, proteins and nucleic acids make good targets, a fact reflected by the abundance of literature reports involving such targets. The application of SELEX to small target molecules has not been as successful, although there are aptamers that bind Zn2+, AMP, ATP, tyrosinimide, arginine, theophylline, and carcinogenic aromatic amines [18–24]. These aptamers generally have Kd on the order of mM or lM, in contrast to the high binding affinities and the Kd of nM or pM for aptamers selected with larger targets. However, Plummer et al. have engineered a technique for selecting aptamers against small target molecules by binding the target to a presenter protein, thus enabling the selection of aptamers with Kd below 50 nM [25]. SELEX can also be applied towards complex cellular targets, an exciting, developing area of research. This application of SELEX involves selecting aptamers against cell-surface molecules, often proteins. Several types of recently studied cell SELEX targets are summarized in Table 1 [5,26–35]. The primary advantage of applying SELEX to complex targets, such as cells, is the ability to target a disease state, such as cancer, without prior knowledge of the structures or molecular changes associated with that state [29]. Aptamers are thought to be able to penetrate tumors rapidly and be cleared quickly from the blood due to their small size (8–15 kDa). Hence, they have
Trends
potential for use as anti-cancer therapeutics. Many aptamers also have an inhibitory effect on their targets, including inhibition of whole-cell differentiation. However, SELEX against complex cellular targets has drawbacks. Due to the abundance of targets on cell membranes, it is often impossible to know the precise target of an aptamer prior to its selection. Ligandmediated target purification using an aptamer is not always possible. In addition, selection favors abundant targets. Rare targets, even those with high aptamer affinity, are often under-represented in final aptamer pools. Selection of aptamers against purified protein targets prior to selection against more complex targets can help solve such problems, when the target is known and can be purified prior to selection. Cerchia et al. [32] and Ohuchi et al. [35] independently developed SELEX techniques in which recombinant protein is overexpressed on a mammalian cell surface and the whole cell is then used as a target to select aptamers against the recombinant protein. These aptamers had Kd values for their targets of 1–70 nM. Further refinement of such techniques will increase the future utlility of SELEX against complex cellular targets. Applications will probably be limited to cells for which the surface target molecules are known and can be over-expressed prior to selection. 2.3. Variations to SELEX protocols SELEX protocols may be classified in many ways. Here, we describe them in terms of both how the target is presented during the selection, and how the aptamertarget complexes are partitioned and separated. 2.3.1. Presentation of target. During the incubation step of SELEX, target molecules are either incubated with a library in free solution, or are bound to some sort of solid support. This dichotomy is reflected in the first two SELEX experiments, in which target was bound to an affinity column [1] or both aptamer and target were incubated in free solution [3]. Dissolving target in free solution is usually preferable as binding target molecules to solid supports has a
Table 1. Cellular SELEX targets and their applications Cellular SELEX Target
Application
References
Human red blood-cell ghosts Heat-killed anthrax spores Trypsanoma cruzi Transformed YPEN-1 endothelial cells Glioblastoma-derived U251 cells Differentiated P12 cells Receptor tyrosine kinase-expressing P12 cells Human osteoblasts Francisella tularensis antigens Recombinant growth factor-b Type III receptor-expressing CHO cells
Model system to determine if SELEX against complex targets is feasible Biosensors for detection of microbial agents Inhibition of T. cruzi invasion Detection of blood-vessel formation in tumors Aptamers against tumor-protein Tenascin-C Detection of cancer cells, identification of cancer-specific ligands Inhibition and reversal of cell transformation Enhancement of cell adhesion on surgical implants Detection of potential biological warfare agents Technique for selecting aptamers against cell-surface targets that cannot be purified
[26] [27] [5] [28] [29,30] [31] [32] [33] [34] [35]
http://www.elsevier.com/locate/trac
683
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
number of drawbacks. The first of these is that fixing a target molecule to a solid support prevents the conjugation side of the molecule from interacting with the library, hence limiting the evolution of high-affinity aptamers. The presence of linker molecules used to bind target to columns may also contribute to the cross reaction of non-specific binders. Further, a large amount of non-specific binding of the library to the solid surface can take place. Columns are difficult to prepare and maintain, and, when using them, it is difficult to elute the strongest binders among the nucleic acid sequences. The end result is that use of solid supports usually decreases the efficiency of SELEX and the binding affinities of the resultant aptamers. The SELEX techniques that result in the highest affinity aptamers, with sub-nanomolar and picomolar Kd values, all have incubation steps wherein target and library are both in free solution [13,36–38]. Typically during SELEX, the same target is used in each round. Exception is made for procedures where the goal is to affect aptamer specificity. Counter-selection is performed to select certain aptamer sequences and to exclude those that do not bind to the target of interest. During counter-selection, the target is switched with an undesirable molecule for one selection round, and the nucleic-acid sequences binding the undesirable molecule are removed from the pool prior to the next round of selection with the desired target. This increases the specificity of the aptamers for the target of interest. Counter-selection can also be carried out against the solid surface to which a target is fixed during selection, to decrease non-specific binders. Hence, counterselection can increase resultant aptamer-binding affinity for the desired target in cases where the target is bound to a solid surface. However, this adds complexity and decreases the efficiency of the procedure. Another instance in which the target molecule is switched during selection is the ToggleSELEX technique [38]. ToggleSELEX is used to select aptamers that are species cross-reactive. White et al. carried out alternating rounds of selection with human or porcine thrombin as a target; the final aptamers had high affinity for thrombin of both species, with Kd values of 1–4 nM for human and <1 nM for porcine thrombin. These aptamers were also able to inhibit thrombin-mediated plasma clot formation and platelet activation in both human and porcine samples. 2.3.2. Partitioning of aptamer-target complexes. Most vital in determining SELEX efficiency and key in resultant aptamer characteristics is the method via which aptamer-target complexes are separated from free target and unbound nucleic acids. The more traditional methods are columns or nitrocellulose membranes to separate aptamer-target complexes [1,3,15,16,26,38]. However, these techniques have low separation efficiencies and 684
http://www.elsevier.com/locate/trac
often require large amounts of target and library in the initial incubation. Aptamers resulting from using nitrocellulose membranes and columns as separation techniques have Kd values from the nanomolar range to as high as 600 lM. Improvements in separation efficiency are demonstrated when target is attached to magnetic beads as a separation matrix [39]. Use of magnetic beads requires less target, and yields aptamers with high binding affinity (Kd of 57–85 nM). Furthermore, these aptamers can potentially be used to check the surface occupancy of the beads and as linkers to join other molecules to the beads. Photocrosslinking is another method of partitioning aptamer-target protein complexes. Jensen et al. [13] first described this technique in developing ssRNA aptamers against HIV-1 Rev protein. It has also been used by Golden et al. [12] to create ssDNA aptamers against recombinant basic human fibroblastic growth factor (bFGF). Briefly, a nucleic-acid library substituted with a fluorophore, such as 5-iodouracil or 5-bromodeoxyuridine, is incubated with target. The libraries are then irradiated by 308-nm excimer laser light in the presence of target, which facilitates covalent cross-linking of the aptamer to the target. Both studies also used PAGE to partition the aptamer-target complexes. Aptamers obtained via cross-linking have high binding affinities and pM Kd values. They are also highly specific for their target. The aptamers for HIV-1 Rev show no binding to closely related proteins or serum components, and formed stable complexes with Rev in the absence of UV irradiation [13]. Golden et al. also found aptamers that formed non-dissociable complexes with target in the absence of UV irradiation [12]. The limit of detection (LOD) of these aptamers for bFGF was as low as 0.058 ppt, comparable to that obtained using a monoclonal antibody [12]. Photocrosslinking has been recently applied to create microarray chips of ssDNA aptamers capable of detecting 17 different proteins simultaneously in a complex sample matrix with LODs below 10 fM for interleukin-6, vascular endothelial growth factor, and endostatin [40]. A third of the aptamers exhibited high affinity for their target analytes, enabling a pM LOD, and proteins could be measured in serum. A limitation of the photocrosslinking technique is that it may not be successful against small-molecule targets because small targets may not offer functional groups for cross-linking. The recent application of capillary electrophoresis (CE) to SELEX has resulted in vast improvement of selection efficiency and aptamer affinity for protein targets [41]. CE separates target-aptamer complexes from the unbound library in free solution. Because of the high separation efficiency and low non-specific binding of CE, CE-SELEX takes only 2–4 rounds of selection to generate ssDNA aptamers with extremely high target affinities [36,41–43]. CE-SELEX is less laborious than other
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
methods and takes only 3 days to complete whereas most SELEX takes weeks to months [41]. Typical Kd values of aptamers obtained via CE-SELEX are in the low nM range but have been as low as 180 pM [36]. These aptamers also show high selectivity for their target proteins, binding unrelated or structurally-related but different proteins with Kd values several orders of magnitude higher than the desired target [41–43]. Unlike photocrosslinking, CE-SELEX has been successfully applied to small targets; Mendonsa and Bowser applied it to create aptamers to Neuropeptide Y, which is smaller than the DNA library used for selection [43]. Efficient selection of aptamers and the determination of kinetic and thermodynamic constants can be achieved using equilibrium capillary electrophoresis of equilibrium mixtures (ECEEM) and non-equilibrium capillary electrophoresis of equilibrium mixtures (NECEEM) techniques. Drabovich et al. pioneered the use of ECEEM for ssDNA aptamer selection [44], yielding aptamers binding MutS protein with Kd of 15 nM in 3 rounds of selection. This method enabled selection of aptamers with predefined kinetic/thermodynamic parameters (such as Kd), because the migration time of the aptamertarget complexes in the capillary depends on the Kd. NECEEM enables selection of ssDNA aptamers with nM Kd values in a single round of incubation, CE separation, and PCR amplification [37], and has the same advantages as ECEEM in selection of aptamers with predefined kinetic parameters. NECEEM has been used to develop an aptamerselection technique that does not rely on SELEX cycles [45] and produces ssDNA aptamers with Kd values of 0.3 lM for the hRas protein. This technique first involves the incubation of high concentrations of target protein and aptamer library, followed by CE separation and collection of the protein-aptamer complexes fraction. This fraction is incubated with additional protein and then subjected to another CE separation. The partitioning process is repeated until a small number of proteinaptamer complexes is obtained, after which the aptamers are dissociated from the complexes, PCR amplified, and cloned. All steps of this selection take 1 hour and can easily be automated. CE-SELEX techniques have a major drawback in that CE has an injection volume of nL, meaning that a limited amount of DNA library can be injected at a time without overloading the column, so that a lower number of possible DNA sequences can be used for selection, typically 1013 rather than the 101415 typical of conventional SELEX [36].
3. Bioanalytical applications of aptamers There are a myriad of applications for aptamers that have been reviewed before [6–9]. However, there have
Trends
since been new applications. The following sections deal with some of these new applications, including the detection of proteins and small molecules, and use as chiral selectors in separation techniques. Table 2 presents a comparison of the protein-detection methods reviewed below with respect to LODs and linear dynamic range, and Table 3 summarizes analytical figures for chiral separation using aptamers as the chiral stationary phase (CSP). 3.1. Protein detection 3.1.1. Affinity-PCR. Two methods based on affinityPCR have been developed for protein detection. One of these relies on the proximate ligation of adjacent targetbound aptamers, and the other does not. The ligation assays require that the target protein has two binding sites (epitopes), which are able to bind to two of the same, or two different aptamers. Each aptamer has an oligonucleotide extension. Upon binding of the aptamers to the target, the aptamer extensions come into close proximity, and are ligated by a connector ligand. The ligated sequences are then PCR-amplified and detected. Dual recognition of proteins results in excellent target selectivity, but only a few targets have the required pair of binding sites. The ligation assays using aptamers were first demonstrated for analysis of thrombin and a platelet derived growth factor (PDGF-BB) [46]. Subsequently, Di Gusto et al. used a proximity extension method, which employed a circular aptamer to act as a template for ligation of the bound aptamers [47]. Detection was by real-time PCR. Both thrombin and PDGF-BB were known to have two distinct binding sites, and appropriate affinity aptamers. Using their ligation methods, the authors were able to detect thrombin down to pM concentrations, and achieved linear dynamic ranges from pM to nM concentrations. Interfering proteins, such as bovine serum albumin (BSA), trypsin serine protease [47], PDGF-AA, PDGF-AB, PDGF-CC and PDGF-DD, and complex media, such as fetal calf serum, cerebral spinal fluid, and EagleÕs minimal essential medium [46], had no effect on target protein detection. Samples of human serum and human anaplastic thyroid carcinoma cells were also tested for PDGF-BB, and results were similar or better than conventional methods. Wang et al. [48] reported an exonuclease protection assay, in which one aptamer was bound to thrombin, thereby protecting it from hydrolysis by exonuclease I. The unbound aptamers were hydrolyzed by exonuclease I. The thrombin-bound aptamers were then used to ligate the two connector ligands, which formed a longer oligonucleotide suitable for detection using real-time PCR. Interfering proteins, such as BSA and Factor IX plasma serine protease, had no effect on detection down to 1 fM. http://www.elsevier.com/locate/trac
685
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Table 2. Comparison of analytical techniques with respect to target, limit of detection (LOD) and linear dynamic range (LDR) Method
Technique
Target
Kd (nM)
LOD
LDR
References
Affinity-PCR
Proximity ligation Proximity extension Exonuclease protection assay CE-PCR
0.5, 25 0.1 0.5, 25 25 1
N/A 3 fM 30 pM 1 fM 30 fM
>1000 >1000 10 pM–5 nM 10 lM–1 nM N/A
[46]
Affinity-PCR Affinity-PCR Affinity-PCR
Thrombin PDGF-BB Thrombin Thrombin HIV-RT
Electrochemical Electrochemical Electrochemical Electrochemical Electrochemical Electrochemical Electrochemical
Electrochemical sandwich Differential pulse voltammetry Cyclic and differential pulse voltammetry Cyclic and differential pulse voltammetry AC-voltammetry AC-voltammetry Square-wave voltammetry using QD-bound aptamers Aptamers bound to carbon-nanotube electrodes Impedance spectroscopy Chronopotentiometric stripping Impedance spectroscopy
25, 0.5 25 25 25 25 25 25 31 25 10 31 31
10 nM 0.5 nM 11 nM 10 nM 3 nM 6.4 nM 0.5 pM N/A 10 nM 0.1 nM 7 nM N/A
40–100 nM 5–35 nM 0–50.8 nM 0–50 nM 0–80 nM 50–768 nM 0.5–12.5 pM* N/A 0–100 nM 2.5–100 nM N/A N/A
[50] [51] [52] [53] [54] [55] [56]
Electrochemical Electrochemical Electrochemical Electrochemical
Thrombin Thrombin Thrombin Thrombin Thrombin Thrombin Thrombin, lysozyme Thrombin Human IgE lysozyme lysozyme
Nanoparticles Nanoparticles
Catalytic NP enlargement Aggregation of aptamer-GNPs
25 0.1 0.1 >10
QD-aptamer beacon
20 nM 35 nM 15 nM 2.5 nM 32 nM N/A
40–140 nM N/A N/A N/A N/A N/A
[61] [62]
Nanoparticles
Thrombin PDGF-BB PDGF-AB PDGF-AA PDGFR-b Thrombin
25
[47] [48] [49]
[57] [58] [59] [60]
[63]
N/A is not available. * The units were changed by the authors to allow easier comparisons.
Table 3. Comparison of the analytical figures of the chiral stationary phases (CSPs) References
Target
Kd
Enantioselectivity (a)
Retention Factor (kd)**
[68] [69]*
D-arginine-vasopressin Tyrosine a-Methyl-tyrosine DOPA Tryptophan 1-Methyl-tryptophan 5-Hydroxyl-trypophan N-Acetyl-tryptophan 3-Benzothienyl-alanine 2-Quinoyl-alanine 2-Naphtyl-alanine 1-Naphtyl-alanine D /L -arganine
1.1 lM 25 lM
N/A 0.57 0.59 0.83 1.42 1.50 1.09 0.49 2.13 1.43 5.29 3.03 7.4
N/A 28.06 21.91 7.20 4.09 6.98 3.62 9.48 1.73 1.88 3.06 1.32 1.43
[70]
412/60 lM
*
Other analytes were also tested, but results for enantioselectivity and retention factors were not obtained. The retention factor is based on the D -enantiomer, since the L -enantiomer eluted in the void volume.
**
Our group has recently developed an ultrasensitive aptamer-based affinity-PCR technique and demonstrated its application for the determination of trace amounts of HIV-1 Reverse transcriptase (RT) [49]. Fig. 2 shows the principle of operation of the assay. The typical CE injection of 10 nL of 30 fM HIV-1 RT represents approximately 180 molecules of HIV-1 RT, which is order of magnitude better LOD than the traditional 686
http://www.elsevier.com/locate/trac
methods. This method was shown to be specific to HIV-1 RT in the presence of human IgG and RNase H. Both the specific recognition by aptamers and the CE separation helped increase the specificity of the assay. Affinity-PCR assays are fast and extremely sensitive due to the PCR-amplification step. However, PCR is also a liability, since contamination can be a problem. One limitation of ligation assays is that the target must have
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Trends
Figure 2. Affinity aptamer amplification assay for proteins [49]. The protein was first incubated with the aptamer to form an aptamer-protein complex. Separation and fractionation of the protein-aptamer complex from excess aptamers was conducted by capillary electrophoresis (CE). After the fraction containing only the aptamer-protein complex was collected, the aptamer was dissociated from the protein and amplified by PCR. The amplified aptamers were indirect measures of the protein concentration.
two distinct binding sites, and appropriate aptamers to bind to the two sites. An advantage of this requirement is the increased specificity of the method. 3.1.2. Electrochemical sensors. Several electrochemical methods for protein detection have made use of aptamers [50–60]. The aptamers are usually coated on nanoparticles (NPs) or a gold-electrode surface with streptavidin before exposing the functionalized surface to a sample containing the target. Other surfaces, such as carbon nanotubes [57], streptavidin-coated magnetic beads [59], and streptavidin/polymer-coated indium-tin oxide electrodes [60], have also been used to immobilize aptamers. Amplification of the target is by preconcentration onto the electrode surface, and the target is detected by a conformational change in the aptamer that causes an electric current. Thrombin is the most commonly studied protein in using electrochemical methods [50–57], although lysozyme [56,59,60] and human IgE [58] have also been studied. Most of these electrochemical methods have achieved LODs below 10nM and a linear dynamic range of two orders of magnitude (Table 2). Xiao [54,55] developed ‘‘signal-on’’ and ‘‘signal-off’’ formats for detecting thrombin using immobilized aptamers on an electrode for binding and using methylene blue for electron transfer (Fig. 3). It was found that the ‘‘signalon’’ version was approximately 10-fold more sensitive than the ‘‘signal-off’’ version; however, the ‘‘signal-off’’ electrode was advantageous, since it could be regenerated, whereas the ‘‘signal-on’’ electrode was stable for only 24 hours, and had to be replaced every day. Other researchers noted that their electrodes could be regenerated a limited number of times [51,52], or were unusable after the first analysis [50]. The robustness of the methods was generally good, showing little to no response from interfering proteins, such as BSA [51,54,56], IgG [56], blood serum [55], IgG, streptavidin [51], elastase [57], beta-casein and ovalbumin [52], a mixture of albumin, cytochrome C and thrombin [60], and a similar mixture that also contained myoglobin,
L -tyrosine
and L -tryptophan [59]. Others showed significant changes in signal with IgG and HSA [53]. Electrochemical assays are inexpensive and, once the electrode has been prepared, they are also fast, usually being completed within a few minutes. One of the main challenges is to maintain the stability and the reusability of the electrode in order to minimize errors resulting from differently coated electrodes. 3.1.3. Aptamer-functionalized nanoparticles with optical detection. Aptamer-functionalized NPs and quantum dots (QDs) have been developed to overcome some of the difficulties of using organic fluorophore labels, which include narrow excitation spectra, short fluorescence lifetime and low quantum yield. Pavlov et al. [61] described a method that used aptamers bound to gold NPs for the detection of thrombin. In this assay, approximately 80 aptamers were bound to each NP. Binding of the thrombin to the aptamers on adjacent NPs resulted in an aggregated precipitate of NPs, which was subsequently separated from the supernatant by centrifugation, and amplified by catalytic enlargement. This resulted in an LOD of 2 nM and a linear dynamic range of 50–120 nM. The assay was shown to be selective for thrombin in the presence of BSA and human IgG. Similarly, Huang et al. described a colorimetric assay for the detection of a platelet derived growth factor (PDGF-BB) and its isoforms (PDGF-AB and PDGF-AA), which also used aggregation of NPs for detection (Fig. 4) [62]. Another approach reported by Levy et al. described a QD-based aptamer biosensor for thrombin, which was similar to a more traditional molecular beacon approach [63]. Approximately 40 quencher/aptamer pairs were bound to the QDs, which were initially in the ‘‘OFF’’ state. Upon binding to thrombin, the conformational change of the aptamer destabilized the hybridized quencher, released it, and turned the fluorescence of the QD ‘‘ON’’. One of the main advantages of coupling NPs with aptamers is the opportunity for significant signal
http://www.elsevier.com/locate/trac
687
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Figure 3. Schematic representation of signal-on (A) and signal-off (B) electrochemical sensors [54,55]. For the signal-on format, a rigid duplex DNA structure prevents the methylene blue (MB)-tagged oligomer from reaching the surface of the electrode. Thrombin binding destabilizes the duplex DNA structure by forcing the aptamer into the characteristic G-quartet structure, resulting in electron transfer from MB on the liberated oligomer to the electrode. For the signal-off architecture, the MB-tagged aptamer is initially able to reach the electrode surface and generate electrical signal. Upon thrombin binding, a conformational change in the aptamer results in the removal of the MB from the electrode surface and inhibition of the electron transfer.
amplification that is not available using conventional fluorophores. This has the potential to lower LODs greatly, which are currently in the low-nM range. A disadvantage of the assays dependent on aggregation is that the degree of aggregation and signal intensity depend on the concentration of both the NPs and the protein (Fig. 4). As a result, optimization of the concentration of NPs may have to be done for each sample, thus complicating the analysis of real samples. 3.2. Small-molecule sensors The use of aptamers as sensors for small molecules has not been studied nearly as extensively as protein targets. Small-molecule sensors are more difficult to develop than protein sensors since there are far fewer moieties for aptamer binding. There are only a few papers that aim to develop sensors for small molecules, and that may have implications for monitoring blood samples for illegal and prescription drugs, as well as enzyme cofactors and other molecules as biomarkers for disease states. 688
http://www.elsevier.com/locate/trac
A recently developed electrochemical sensor used a gold electrode coated with aptamers to detect cocaine in diluted fetal calf serum and human saliva, with LODs of about 10 lM [64]. This represents approximately a 200-fold lower LOD than the Scott Test for cocaine currently used in law enforcement. Other colorimetric sensors for small molecules include modular fluorescence sensors for ATP, flavin mononucleotide, theophylline [65], modified aptameric sensors with a 2 0 -Ribose-linked fluorophore for AMP, tyrosinimide, and arginimide [66], and molecular beacons for ATP [67] based on fluorescence-signal transduction. These sensors have been tested in a buffer mimicking the intercellular milieu [65] and in serum and urine [66] with good results. The LODs for small molecules are severely hampered by available aptamers, which have lM Kd values that result in LODs that are at best in the low lM range. This is orders of magnitude higher than is required to detect analytes in real samples. Despite the high Kd values, the aptamers can still be very selective. For example, the
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Trends
Figure 4. Schematic representation of the aggregation of aptamer gold nanoparticles (GNPs) in the presence of PDGFs at (A) low, (B) medium and (C) high concentrations [62]. (A) Low concentrations of PDGF result in very little aggregation of the GNPs and a very small color change. (B) A medium concentration of PDGF results in considerable aggregation of the GNPs, producing a large color change. (C) A high concentration of PDGF saturates the aptamer-binding sites on the GNPs, resulting in little to no aggregation, and only a very small color change.
theophylline aptamer is able to distinguish theophylline from caffeine, which differ by only a single methyl group [2]. 3.3. Chiral selectors in affinity chromatography Aptamers have been immobilized onto surfaces, typically as stationary phases in affinity chromatography, to provide separation of target and non-target compounds. In particular, enantiometric separations using CSPs are of increasing interest in both pharmaceutical and biological fields, because often one enantiomer of a chiral drug is pharmaceutically active, whereas the other is inactive or toxic. PeyrinÕs group used DNA aptamers for target-specific chiral separations [68]. They were able to separate the D and L -peptides of vasopressin using an HPLC column with a D -vasopressin-specific DNA aptamer immobilized on a highly porous polystyrene-divinylbenzene support via a biotin-streptavidin bridge. The D -enantiomer was retained while the L -enantiomer was eluted in the void volume. Although no quantitative data were presented, the authors reported that the enantioselectivity was better than conventional chiral selectors and comparable to using monoclonal antibodies. RNA aptamers have also been developed for CSP applications; however, the stability of the ‘‘natural’’
D -aptamers
is very low due to degradation by RNases. To address this problem, the L -RNA aptamer that was intrinsically resistant to enzyme degradation was used instead of the ‘‘natural’’ D -RNA aptamer [69,70]. This mirror-image strategy greatly increases the stability of RNA CSPs, which can otherwise be degraded over time by RNases in the sample. The L -oligonucleotides are not susceptible to degradation by the naturally-occurring enzymes, and have the same selectivity and specificity for the mirror-image of the target. Peyrin and coworkers used this strategy to develop target-specific RNA CSPs to separate enantiomers of arginine [70] and tyrosine and related compounds [69] with enantioselectivities of 7.4 and 10, respectively. The CSPs were stable for over 1500 column volumes of mobile phase (3 months of use). Table 3 presents some analytical data for these separation methods. For the RNA CSPs, the L -enantiomers were commonly eluted in the void volume, making it difficult to quantify the retention factors accurately, so the retention-factor data were obtained from the D -enantiomers. In most cases, the enantioselectivity was acceptable, except when it was very low, indicating incomplete separation. Despite the promise of aptamers for chiral separation, aptamer CSPs have lower binding capacities, lower efficiencies, and higher costs compared
http://www.elsevier.com/locate/trac
689
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
to conventional CSPs. An improvement may lie in the covalent binding of aptamers on chromatographic particles of lower diameter, which may significantly improve the chromatographic performance of the aptamer CSPs.
4. Concluding remarks Aptamers have great potential for use in bioanalytical applications. The main impetus for aptamer development is their ability to recognize virtually any target with high affinity and specificity. Antibodies are currently the reagent of choice in many assays because of their wide availability and proven success; however, new developments in SELEX techniques have shown promise for the generation of aptamers in shorter times with higher affinities for a wider array of targets. Aptamers can complement antibodies as affinity reagents in bioanalytical assays. Most studies have focused on protein targets, resulting in a lack of aptamer development for smaller inorganic and organic molecules, as well as complex targets. This is mainly because of the difficulty in obtaining aptamers with high binding affinity and specificity for these targets, which are both important factors to consider when selecting aptamers for a given application. Aptamers with high target-binding affinity do not necessarily have high target specificity; however, both of these factors will determine whether or not the aptamer could be used in a real sample with hundreds or thousands of potentially interfering components. Future studies should always assess the specificity of aptamers for other targets, as this element has been missing in many previous studies. Ideally, all assays should be tested in a medium representative of the matrix for which they were developed (e.g., biological fluids and environmental matrices). Active collaborations among different disciplines will help advance the field of aptamer technology, which will lead to new, innovative applications encompassing a number of different targets. Acknowledgement The authors acknowledge the financial support from the Natural Sciences and Engineering Research Council of Canada and Alberta Health and Wellness (to XCL and XFL), the Canada Research Chairs Program (to XCL), an Alberta Heritage Foundation for Medical Research Studentship and a Natural Sciences and Engineering Research Council of Canada Studentship (to CLAH), an Alberta Ingenuity Postdoctoral Fellowship (to JWG), and a Graduate Assistantship from the University of Alberta (to HZ).
690
http://www.elsevier.com/locate/trac
References [1] A.D. Ellington, J.W. Szostak, Nature (London) 346 (1990) 818. [2] R.D. Jenison, S.C. Gill, A. Pardi, B. Polisky, Science (Washington, DC) 263 (1994) 1425. [3] C. Tuerk, L. Gold, Science (Washington, DC) 249 (1990) 505. [4] C. Pestourie, B. Tavitian, F. Duconge, Biochimie 87 (2005) 921. [5] H. Ulrich, A.H.B. Martins, J.B. Perquero, Cytometry 59A (2004) 220. [6] S. Tombelli, M. Minunni, M. Mascini, Biosens. Bioelectron. 20 (2005) 2424. [7] V. Pavski, X.C. Le, Curr. Opin. Biotechnol. 14 (2003) 65. [8] W. Tan, K. Wang, T.J. Drake, Curr. Opin. Chem. Biol. 8 (2004) 547. [9] J.W. Guthrie, C.L.A. Hamula, H. Zhang, X.C. Le, Methods 38 (2006) 324. [10] S.M. Nimjee, C.P. Rusconi, B.A. Sullenger, Annu. Rev. Med. 56 (2005) 555. [11] S.E. Osborne, A.D. Ellington, Chem. Rev. 97 (1997) 349. [12] M.C. Golden, B.D. Collins, M.C. Willis, T.H. Koch, J. Biotechnol. 81 (2001) 167. [13] K.B. Jensen, B.L. Atkinson, M.C. Willis, T.D. Koch, L. Gold, Proc. Natl. Acad. Sci. USA 92 (1995) 12220. [14] B.S. Singer, T. Shtatland, D. Brown, L. Gold, Nucleic Acids Res. 25 (1997) 781. [15] T. Shtatland, S.C. Gill, B.E. Javornik, H.E. Johansson, B.S. Singer, O.C. Uhlenbeck, D.A. Zichi, L. Gold, Nucleic Acids Res. 28 (2000) E93. [16] T. Shimada, N. Fujita, M. Maeda, A. Ishihama, Genes Cells 10 (2005) 907. [17] J.F. Lee, J.R. Hesselberth, L.A. Meyers, A.D. Ellington, Nucleic Acids Res. 32 (2004) D95. [18] J. Kawakami, H. Imanaka, Y. Yokota, N. Sugimoto, J. Inorg. Biochem. 82 (2000) 197. [19] M. Koizumi, R.R. Breaker, Biochemistry 39 (2000) 8983. [20] P.E. Huizenga, J.W. Szostak, Biochemistry 34 (1995) 656. [21] E. Vianini, M. Palumbo, B. Gatto, Biochemistry 39 (2000) 946. [22] A. Geiger, P. Burgstaller, H. von der Eltz, A. Roeder, M. Famulok, Nucleic Acids Res. 24 (1996) 1029. [23] G.R. Zimmerman, T.R. Shields, R.D. Jenison, C.L. Wick, A. Pardi, Biochemistry 37 (1998) 918. [24] U. Brockstedt, A. Uzarowska, A. Montpetit, W. Pfau, D. Labuda, Biochem. Biophys. Res. Comm. 313 (2004) 1004. [25] K.A. Plummer, J.M. Carothers, M. Yoshimura, J.W. Szostak, G.L. Verdine, Nucleic Acids Res. 33 (2005) 5602. [26] K.N. Morris, K.B. Jensen, C.M. Julin, L. Gold, Proc. Natl. Acad. Sci. USA 95 (1998) 2902. [27] J.G. Bruno, J.L. Kiel, Biosens. Bioelectron. 14 (1999) 457. [28] M. Blank, T. Weinschenk, M. Priemer, H. Schluesener, J. Biol. Chem. 276 (2001) 16464. [29] D.A. Daniels, H. Chen, B.J. Hicke, K.M. Swiderek, L. Gold, Proc. Natl. Acad. Sci. USA 100 (2003) 15416. [30] B.J. Hicke, C. Marion, Y.-F. Chang, T. Gould, C.K. Lynott, D. Parma, P.G. Schmidt, S. Warren, J. Biol. Chem. 276 (2001) 48644. [31] C. Wang, M. Zhang, G. Yang, D. Zhang, H. Ding, H. Wang, M. Fan, B. Shen, N. Shao, J. Biotechnol. 102 (2003) 15. [32] L. Cerchia, F. Duconge, C. Pestourie, J. Boulay, Y. Aissouni, K. Gombert, B. Tavitian, V. de Franciscis, D. Libri, PLoS Biol. 3 (2005) e123. [33] K. Guo, H.P. Wendel, L. Scheideler, G. Ziemer, A.M. Scheule, J. Cell. Mol. Med. 9 (2005) 731. [34] J. Vivekananda, J.L. Kiel, Lab Invest. (2006) URL: http://www. nature.com/labinvest/journal/vaop/ncurrent/abs/3700417a.html;
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
[35] [36] [37]
[38] [39] [40]
[41] [42] [43] [44] [45] [46]
[47] [48] [49] [50]
jsessionid=B29434A4BA4D0E5D835424333F66AD77 [accessed 3 April 2006]. S.P. Ohuchi, T. Ohtsu, Y. Nakamura, Biochimie (2006) (doi:10.1016/j.biochi.2006.02.004).. K. Mosing, S.D. Mendonsa, M.T. Bowser, Anal. Chem. 77 (2005) 6017. M. Berezovski, A. Drabovich, S.M. Krylova, M. Musheev, V. Okhonin, A. Petrov, S.N. Krylov, J. Am. Chem. Soc. 127 (2005) 3165. R. White, C. Rusconi, E. Scardino, A. Wolberg, J. Lawson, M. Hoffman, B. Sulenger, Mol. Ther. 4 (2001) 567. R. Stoltenberg, C. Reinemann, B. Strehlitz, Anal. Bioanal. Chem. 383 (2005) 83. C. Bock, M. Coleman, B. Collins, J. Davis, G. Foulds, L. Gold, C. Greef, J. Heil, J.S. Heilig, B. Hicke, M.N. Hurst, G.M. Husar, D. Miller, R. Ostroff, H. Petach, D. Schneider, B. Vant-Hull, S. Waugh, A. Weiss, S.K. Wilcox, D. Zichi, Proteomics 4 (2004) 609. S.D. Mendonsa, M.T. Bowser, J. Am. Chem. Soc. 126 (2004) 20. S.D. Mendonsa, M.T. Bowser, Anal. Chem. 76 (2004) 5837. S.D. Mendonsa, M.T. Bowser, J. Am. Chem. Soc. 127 (2005) 9382. A. Drabovich, M. Berezovski, S.N. Krylov, J. Am. Chem. Soc. 127 (2005) 11224. M. Berezovski, M. Musheev, A. Drabovich, S.N. Krylov, J. Am. Chem. Soc. 128 (2006) 1410. S. Fredriksson, M. Gullberg, J. Jarvius, C. Olsson, K. Pietras, S.M. Gu´stafsdo´ttir, A. O¨stman, U. Landegren, Nat. Biotechnol. 20 (2002) 473. D.A. Di Giusto, W.A. Wlassoff, J.J. Gooding, B.A. Messerle, G.C. King, Nucleic Acids Res. 33 (2005) e64. X-L. Wang, F. Li, Y-H. Su, X. Sun, X.-B. Li, H.J. Schluesener, F. Tang, S.-Q. Xu, Anal. Chem. 76 (2004) 5605. H. Zhang, Z. Wang, X.-F. Li, X.C. Le, Angew. Chem. Int. Ed. 45 (2006) 1576. K. Ikebukuro, C. Kiyohara, K. Sode, Biosens. Bioelectron. 20 (2005) 2168.
Trends [51] A.-E. Radi, J.L.A. Sa´nchez, E. Baldrich, C.K. OÕSullivan, J. Am. Chem. Soc. 128 (2005) 117. [52] G.S. Bang, S. Cho, B.-G. Kim, Biosens. Bioelectron. 21 (2005) 863. [53] T. Hianik, V. Ostatna´, Z. Zajacova´, E. Stoikova, G. Evtugyn, Biorg. Med. Chem. Lett. 15 (2005) 291. [54] Y. Xiao, B.D. Piorek, K.W. Plaxco, A.J. Heeger, J. Am. Chem. Soc. 127 (2005) 17990. [55] Y. Xiao, A.A. Lubin, A.J. Heeger, K.W. Plaxco, Angew. Chem. Int. Ed. 44 (2005) 5456. [56] J.A. Hansen, J. Wang, A.-N. Kawde, Y. Xiang, K.V. Gothelf, G. Collins, J. Am. Chem. Soc. 128 (2006) 2228. [57] H.-M. So, K. Won, Y.H. Kim, B.-K. Kim, B.H. Ryu, P.S. Na, H. Kim, J.-O. Lee, J. Am. Chem. Soc. 127 (2005) 11906. [58] D. Xu, D. Xu, X. Yu, Z. Liu, W. He, Z. Ma, Anal. Chem. 77 (2005) 5107. [59] A.-N. Kawde, M.C. Rodriquez, T.M.H. Lee, J. Wang, Electrochem. Commun. 7 (2005) 537. [60] M.C. Rodriguez, A-N. Kawde, J. Wang, Chem. Commun. 34 (2005) 4267. [61] V. Pavlov, Y. Xiao, B. Shlyahovsky, I. Willner, J. Am. Chem. Soc. 126 (2004) 11768. [62] C.-C. Huang, Y.-F. Huang, Z. Cao, W. Tan, H.-T. Chang, Anal. Chem. 77 (2005) 5735. [63] M. Levy, S.F. Cater, A.D. Ellington, ChemBioChem 6 (2005) 2163. [64] B.R. Baker, R.Y. Lai, M.S. Wood, E.H. Doctor, A.J. Heeger, K.W. Plaxco, J. Am. Chem. Soc. 128 (2006) 3138. [65] M.N. Stojanovic, D.M. Kolpashchikov, J. Am. Chem. Soc. 126 (2004) 9266. [66] E.J. Merino, K.M. Weeks, J. Am. Chem. Soc. 127 (2005) 12766. [67] R. Nutiu, Y. Li, J. Am. Chem. Soc. 125 (2003) 4771. [68] M. Michaud, E. Jourdan, A. Villet, A. Ravel, C. Grosset, E. Peyrin, J. Am. Chem. Soc. 125 (2003) 8672. [69] C. Ravelet, R. Boulkedid, A. Ravel, C. Grosset, A. Villet, J. Fize, E. Peyrin, J. Chromatogr. A 1076 (2005) 62. [70] A. Brumbt, C. Ravelet, C. Grosset, A. Ravel, A. Vellet, E. Peyrin, Anal. Chem. 77 (2005) 1993.
http://www.elsevier.com/locate/trac
691
Trends
Selection and analytical applications of aptamers Camille L.A. Hamula, Jeffrey W. Guthrie, Hongquan Zhang, Xing-Fang Li, X. Chris Le Advances in systematic evolution of ligands by exponential enrichment (SELEX), a selection protocol for aptamers, have resulted in increased applications of DNA and RNA aptamers in developing analytical techniques. We review recent developments in SELEX techniques as well as new aptamerbased bioanalytical applications. ª 2006 Elsevier Ltd. All rights reserved. Keywords: Affinity binding; Aptamer; Bioanalytical assay; Biosensor; Chiral separation; In vitro selection; Protein; SELEX
Camille L.A. Hamula1 , Jeffrey W. Guthrie1 , Hongquan Zhang, Xing-Fang Li, X. Chris Le* Environmental Health Sciences, Department of Public Health Sciences and Department of Laboratory Medicine and Pathology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada, T6G 2G3
1
These two authors contributed equally. *Corresponding author. Tel.: +1 780 492 6416; Fax: +1 780 492 7800 E-mail: [email protected]
1. Introduction Nucleic acids possess desirable properties for use in analytical applications. Conventionally, they are used as probes that hybridize with and detect complementary target RNA and DNA sequences. In the last decade, short synthetic DNA and RNA sequences known as aptamers have been used as ligands to bind to non-nucleic acid targets with high specificity and affinity. Dissociation constants for typical aptamer binding are in the micromolar to low picomolar range [1,2]. This binding is not through sequence hybridization but through binding of stable, complex stemloop and internal loop structures formed by the nucleic acids [1]. Aptamers are commonly evolved in vitro via a combinatorial chemistry technique known as systematic evolution of ligands by exponential enrichment (SELEX) [3]. They can be selected using a variety of targets from small molecules to whole organisms [4]. The ability to synthesize aptamers for a variety of targets has contributed to a wide range of analyses using aptamers. Another factor allowing the widespread application of aptamers to bioanalysis is that aptamers offer benefits for analytical applications when used in place of antibodies [5]. These benefits derive mainly
0165-9936/$ - see front matter ª 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2006.05.007
from temperature stability and ease of aptamer production. Many of these analytical applications of aptamers, such as those based on chromatography, mass spectrometry, molecular beacons, and some label-free biosensors, have been reviewed previously [6–9], and will not be discussed further. This review focuses on recent developments in SELEX analytical applications, as well as variations in SELEX techniques and the ramifications of such variations for aptamer characteristics.
2. SELEX techniques Fig. 1 shows the basic principle of most SELEX procedures [10]. Attempts to improve SELEX have led to the development of different techniques, delineated by variations in starting library, target molecule, and experimental protocol. These variations result in differences in aptamer affinity and specificity as well as selection efficiency. 2.1. Libraries A typical SELEX library comprises singlestranded RNA or DNA, in which a central region of randomized sequence is flanked on either side by fixed primer sequences for PCR and/or RT-PCR amplification. The first SELEX procedures used randomized single-stranded RNA libraries, in which the random sequence region was 100 nucleotides (nt) in length [1] and 8 nt [3], respectively. Since then, many other SELEX procedures have been carried out using randomized ssRNA or ssDNA libraries, typically containing 1013–1015 different randomized sequences [11]. Although any length of randomized region 681
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Figure 1. Schematic of Systematic Evolution of Ligands by Exponential Enrichment (SELEX) [10]. Briefly, typical SELEX procedures involve binding a random nucleic-acid library to a target, separating the bound and unbound nucleic acids, and amplifying the bound nucleic acids by the polymerase chain reaction (PCR) for use in the next round of selection [1,3]. After each round of selection, a smaller pool of nucleic acid sequences binding to the target is retained and the unbound nucleic acids are discarded. Typically, 8–15 rounds of SELEX are carried out in order to generate a pool of aptamers with sequences enabling the highest binding affinity for the target. These aptamers can then be cloned and sequenced. (Reprinted, with permission, from the Annual Review of Medicine, Volume 56 ª2005 by Annual Reviews www.annualreviews.org)
may be used to provide 4n theoretical random sequences, in reality, synthetic limitations prevent most libraries from containing more than 1016 random sequences (e.g., for a library to contain one molecule each of 1016 random sequences, a minimum of 16.7 nmole nucleic acids (or 0.4 mg of nucleic acids containing a 40 nt random region and two 20 nt primer regions, 25 kDa) is required). In practice, therefore, randomized regions of 30–60 nt are most common [11]. DNA libraries are particularly useful for applications requiring increased aptamer stability, such as biosensing, environmental monitoring, and therapy. However, in many cases, RNA libraries yield aptamers with higher binding affinities than DNA libraries due to the ability of RNA to take on a wider variety of conformations than DNA. Several SELEX techniques have been developed using modified randomized ssRNA or ssDNA libraries. Photocrosslinking procedures, described in more detail later, use ssDNA libraries substituted with fluorophore682
http://www.elsevier.com/locate/trac
modified nucleotides, such as 5 0 -BromodeoxyuridineUTP or 5-iodouracil [12,13]. Upon irradiation with a laser, these nucleotides facilitate crosslinking of an aptamer to the target. Genomic SELEX procedures make use of ssDNA or ssRNA libraries derived from the genome of an organism of interest to probe for genome sequences that interact with a specified target [14–16] Currently, there are Escherichia coli, Saccharomyces cerevisiae and human genomic DNA libraries that are able to be transcribed into RNA or amplified by PCR [14]. In all other respects, genomic SELEX is similar to conventional SELEX. Unfortunately, genomic SELEX procedures tend to be biased towards strong or highly abundant interactions and yield many interactions that may not be important in vivo. So far, genomic SELEX has yielded aptamers with relatively low binding affinities and high dissociationconstant (Kd) values of 1.9–12.8 lM in 5 rounds of selection [15].
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
2.2. SELEX targets SELEX has been applied to a variety of molecular targets. The first SELEX procedures involved generation of aptamers against organic dyes [1] and T4 DNA polymerase [3]. Since then, a myriad of other biomolecules have been targeted. Most of these are summarized by Ellington et al. in The Aptamer Database [17], which includes organic and inorganic small-molecule, nucleic acid, protein, peptide, aminoglycoside and carbohydrate targets, among others. SELEX is best suited to generating aptamers against large molecules, which offer the necessary surface area for interaction with aptamers. Target molecules must also be stable, and it is useful if they can be modified to enable partitioning. Hence, proteins and nucleic acids make good targets, a fact reflected by the abundance of literature reports involving such targets. The application of SELEX to small target molecules has not been as successful, although there are aptamers that bind Zn2+, AMP, ATP, tyrosinimide, arginine, theophylline, and carcinogenic aromatic amines [18–24]. These aptamers generally have Kd on the order of mM or lM, in contrast to the high binding affinities and the Kd of nM or pM for aptamers selected with larger targets. However, Plummer et al. have engineered a technique for selecting aptamers against small target molecules by binding the target to a presenter protein, thus enabling the selection of aptamers with Kd below 50 nM [25]. SELEX can also be applied towards complex cellular targets, an exciting, developing area of research. This application of SELEX involves selecting aptamers against cell-surface molecules, often proteins. Several types of recently studied cell SELEX targets are summarized in Table 1 [5,26–35]. The primary advantage of applying SELEX to complex targets, such as cells, is the ability to target a disease state, such as cancer, without prior knowledge of the structures or molecular changes associated with that state [29]. Aptamers are thought to be able to penetrate tumors rapidly and be cleared quickly from the blood due to their small size (8–15 kDa). Hence, they have
Trends
potential for use as anti-cancer therapeutics. Many aptamers also have an inhibitory effect on their targets, including inhibition of whole-cell differentiation. However, SELEX against complex cellular targets has drawbacks. Due to the abundance of targets on cell membranes, it is often impossible to know the precise target of an aptamer prior to its selection. Ligandmediated target purification using an aptamer is not always possible. In addition, selection favors abundant targets. Rare targets, even those with high aptamer affinity, are often under-represented in final aptamer pools. Selection of aptamers against purified protein targets prior to selection against more complex targets can help solve such problems, when the target is known and can be purified prior to selection. Cerchia et al. [32] and Ohuchi et al. [35] independently developed SELEX techniques in which recombinant protein is overexpressed on a mammalian cell surface and the whole cell is then used as a target to select aptamers against the recombinant protein. These aptamers had Kd values for their targets of 1–70 nM. Further refinement of such techniques will increase the future utlility of SELEX against complex cellular targets. Applications will probably be limited to cells for which the surface target molecules are known and can be over-expressed prior to selection. 2.3. Variations to SELEX protocols SELEX protocols may be classified in many ways. Here, we describe them in terms of both how the target is presented during the selection, and how the aptamertarget complexes are partitioned and separated. 2.3.1. Presentation of target. During the incubation step of SELEX, target molecules are either incubated with a library in free solution, or are bound to some sort of solid support. This dichotomy is reflected in the first two SELEX experiments, in which target was bound to an affinity column [1] or both aptamer and target were incubated in free solution [3]. Dissolving target in free solution is usually preferable as binding target molecules to solid supports has a
Table 1. Cellular SELEX targets and their applications Cellular SELEX Target
Application
References
Human red blood-cell ghosts Heat-killed anthrax spores Trypsanoma cruzi Transformed YPEN-1 endothelial cells Glioblastoma-derived U251 cells Differentiated P12 cells Receptor tyrosine kinase-expressing P12 cells Human osteoblasts Francisella tularensis antigens Recombinant growth factor-b Type III receptor-expressing CHO cells
Model system to determine if SELEX against complex targets is feasible Biosensors for detection of microbial agents Inhibition of T. cruzi invasion Detection of blood-vessel formation in tumors Aptamers against tumor-protein Tenascin-C Detection of cancer cells, identification of cancer-specific ligands Inhibition and reversal of cell transformation Enhancement of cell adhesion on surgical implants Detection of potential biological warfare agents Technique for selecting aptamers against cell-surface targets that cannot be purified
[26] [27] [5] [28] [29,30] [31] [32] [33] [34] [35]
http://www.elsevier.com/locate/trac
683
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
number of drawbacks. The first of these is that fixing a target molecule to a solid support prevents the conjugation side of the molecule from interacting with the library, hence limiting the evolution of high-affinity aptamers. The presence of linker molecules used to bind target to columns may also contribute to the cross reaction of non-specific binders. Further, a large amount of non-specific binding of the library to the solid surface can take place. Columns are difficult to prepare and maintain, and, when using them, it is difficult to elute the strongest binders among the nucleic acid sequences. The end result is that use of solid supports usually decreases the efficiency of SELEX and the binding affinities of the resultant aptamers. The SELEX techniques that result in the highest affinity aptamers, with sub-nanomolar and picomolar Kd values, all have incubation steps wherein target and library are both in free solution [13,36–38]. Typically during SELEX, the same target is used in each round. Exception is made for procedures where the goal is to affect aptamer specificity. Counter-selection is performed to select certain aptamer sequences and to exclude those that do not bind to the target of interest. During counter-selection, the target is switched with an undesirable molecule for one selection round, and the nucleic-acid sequences binding the undesirable molecule are removed from the pool prior to the next round of selection with the desired target. This increases the specificity of the aptamers for the target of interest. Counter-selection can also be carried out against the solid surface to which a target is fixed during selection, to decrease non-specific binders. Hence, counterselection can increase resultant aptamer-binding affinity for the desired target in cases where the target is bound to a solid surface. However, this adds complexity and decreases the efficiency of the procedure. Another instance in which the target molecule is switched during selection is the ToggleSELEX technique [38]. ToggleSELEX is used to select aptamers that are species cross-reactive. White et al. carried out alternating rounds of selection with human or porcine thrombin as a target; the final aptamers had high affinity for thrombin of both species, with Kd values of 1–4 nM for human and <1 nM for porcine thrombin. These aptamers were also able to inhibit thrombin-mediated plasma clot formation and platelet activation in both human and porcine samples. 2.3.2. Partitioning of aptamer-target complexes. Most vital in determining SELEX efficiency and key in resultant aptamer characteristics is the method via which aptamer-target complexes are separated from free target and unbound nucleic acids. The more traditional methods are columns or nitrocellulose membranes to separate aptamer-target complexes [1,3,15,16,26,38]. However, these techniques have low separation efficiencies and 684
http://www.elsevier.com/locate/trac
often require large amounts of target and library in the initial incubation. Aptamers resulting from using nitrocellulose membranes and columns as separation techniques have Kd values from the nanomolar range to as high as 600 lM. Improvements in separation efficiency are demonstrated when target is attached to magnetic beads as a separation matrix [39]. Use of magnetic beads requires less target, and yields aptamers with high binding affinity (Kd of 57–85 nM). Furthermore, these aptamers can potentially be used to check the surface occupancy of the beads and as linkers to join other molecules to the beads. Photocrosslinking is another method of partitioning aptamer-target protein complexes. Jensen et al. [13] first described this technique in developing ssRNA aptamers against HIV-1 Rev protein. It has also been used by Golden et al. [12] to create ssDNA aptamers against recombinant basic human fibroblastic growth factor (bFGF). Briefly, a nucleic-acid library substituted with a fluorophore, such as 5-iodouracil or 5-bromodeoxyuridine, is incubated with target. The libraries are then irradiated by 308-nm excimer laser light in the presence of target, which facilitates covalent cross-linking of the aptamer to the target. Both studies also used PAGE to partition the aptamer-target complexes. Aptamers obtained via cross-linking have high binding affinities and pM Kd values. They are also highly specific for their target. The aptamers for HIV-1 Rev show no binding to closely related proteins or serum components, and formed stable complexes with Rev in the absence of UV irradiation [13]. Golden et al. also found aptamers that formed non-dissociable complexes with target in the absence of UV irradiation [12]. The limit of detection (LOD) of these aptamers for bFGF was as low as 0.058 ppt, comparable to that obtained using a monoclonal antibody [12]. Photocrosslinking has been recently applied to create microarray chips of ssDNA aptamers capable of detecting 17 different proteins simultaneously in a complex sample matrix with LODs below 10 fM for interleukin-6, vascular endothelial growth factor, and endostatin [40]. A third of the aptamers exhibited high affinity for their target analytes, enabling a pM LOD, and proteins could be measured in serum. A limitation of the photocrosslinking technique is that it may not be successful against small-molecule targets because small targets may not offer functional groups for cross-linking. The recent application of capillary electrophoresis (CE) to SELEX has resulted in vast improvement of selection efficiency and aptamer affinity for protein targets [41]. CE separates target-aptamer complexes from the unbound library in free solution. Because of the high separation efficiency and low non-specific binding of CE, CE-SELEX takes only 2–4 rounds of selection to generate ssDNA aptamers with extremely high target affinities [36,41–43]. CE-SELEX is less laborious than other
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
methods and takes only 3 days to complete whereas most SELEX takes weeks to months [41]. Typical Kd values of aptamers obtained via CE-SELEX are in the low nM range but have been as low as 180 pM [36]. These aptamers also show high selectivity for their target proteins, binding unrelated or structurally-related but different proteins with Kd values several orders of magnitude higher than the desired target [41–43]. Unlike photocrosslinking, CE-SELEX has been successfully applied to small targets; Mendonsa and Bowser applied it to create aptamers to Neuropeptide Y, which is smaller than the DNA library used for selection [43]. Efficient selection of aptamers and the determination of kinetic and thermodynamic constants can be achieved using equilibrium capillary electrophoresis of equilibrium mixtures (ECEEM) and non-equilibrium capillary electrophoresis of equilibrium mixtures (NECEEM) techniques. Drabovich et al. pioneered the use of ECEEM for ssDNA aptamer selection [44], yielding aptamers binding MutS protein with Kd of 15 nM in 3 rounds of selection. This method enabled selection of aptamers with predefined kinetic/thermodynamic parameters (such as Kd), because the migration time of the aptamertarget complexes in the capillary depends on the Kd. NECEEM enables selection of ssDNA aptamers with nM Kd values in a single round of incubation, CE separation, and PCR amplification [37], and has the same advantages as ECEEM in selection of aptamers with predefined kinetic parameters. NECEEM has been used to develop an aptamerselection technique that does not rely on SELEX cycles [45] and produces ssDNA aptamers with Kd values of 0.3 lM for the hRas protein. This technique first involves the incubation of high concentrations of target protein and aptamer library, followed by CE separation and collection of the protein-aptamer complexes fraction. This fraction is incubated with additional protein and then subjected to another CE separation. The partitioning process is repeated until a small number of proteinaptamer complexes is obtained, after which the aptamers are dissociated from the complexes, PCR amplified, and cloned. All steps of this selection take 1 hour and can easily be automated. CE-SELEX techniques have a major drawback in that CE has an injection volume of nL, meaning that a limited amount of DNA library can be injected at a time without overloading the column, so that a lower number of possible DNA sequences can be used for selection, typically 1013 rather than the 101415 typical of conventional SELEX [36].
3. Bioanalytical applications of aptamers There are a myriad of applications for aptamers that have been reviewed before [6–9]. However, there have
Trends
since been new applications. The following sections deal with some of these new applications, including the detection of proteins and small molecules, and use as chiral selectors in separation techniques. Table 2 presents a comparison of the protein-detection methods reviewed below with respect to LODs and linear dynamic range, and Table 3 summarizes analytical figures for chiral separation using aptamers as the chiral stationary phase (CSP). 3.1. Protein detection 3.1.1. Affinity-PCR. Two methods based on affinityPCR have been developed for protein detection. One of these relies on the proximate ligation of adjacent targetbound aptamers, and the other does not. The ligation assays require that the target protein has two binding sites (epitopes), which are able to bind to two of the same, or two different aptamers. Each aptamer has an oligonucleotide extension. Upon binding of the aptamers to the target, the aptamer extensions come into close proximity, and are ligated by a connector ligand. The ligated sequences are then PCR-amplified and detected. Dual recognition of proteins results in excellent target selectivity, but only a few targets have the required pair of binding sites. The ligation assays using aptamers were first demonstrated for analysis of thrombin and a platelet derived growth factor (PDGF-BB) [46]. Subsequently, Di Gusto et al. used a proximity extension method, which employed a circular aptamer to act as a template for ligation of the bound aptamers [47]. Detection was by real-time PCR. Both thrombin and PDGF-BB were known to have two distinct binding sites, and appropriate affinity aptamers. Using their ligation methods, the authors were able to detect thrombin down to pM concentrations, and achieved linear dynamic ranges from pM to nM concentrations. Interfering proteins, such as bovine serum albumin (BSA), trypsin serine protease [47], PDGF-AA, PDGF-AB, PDGF-CC and PDGF-DD, and complex media, such as fetal calf serum, cerebral spinal fluid, and EagleÕs minimal essential medium [46], had no effect on target protein detection. Samples of human serum and human anaplastic thyroid carcinoma cells were also tested for PDGF-BB, and results were similar or better than conventional methods. Wang et al. [48] reported an exonuclease protection assay, in which one aptamer was bound to thrombin, thereby protecting it from hydrolysis by exonuclease I. The unbound aptamers were hydrolyzed by exonuclease I. The thrombin-bound aptamers were then used to ligate the two connector ligands, which formed a longer oligonucleotide suitable for detection using real-time PCR. Interfering proteins, such as BSA and Factor IX plasma serine protease, had no effect on detection down to 1 fM. http://www.elsevier.com/locate/trac
685
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Table 2. Comparison of analytical techniques with respect to target, limit of detection (LOD) and linear dynamic range (LDR) Method
Technique
Target
Kd (nM)
LOD
LDR
References
Affinity-PCR
Proximity ligation Proximity extension Exonuclease protection assay CE-PCR
0.5, 25 0.1 0.5, 25 25 1
N/A 3 fM 30 pM 1 fM 30 fM
>1000 >1000 10 pM–5 nM 10 lM–1 nM N/A
[46]
Affinity-PCR Affinity-PCR Affinity-PCR
Thrombin PDGF-BB Thrombin Thrombin HIV-RT
Electrochemical Electrochemical Electrochemical Electrochemical Electrochemical Electrochemical Electrochemical
Electrochemical sandwich Differential pulse voltammetry Cyclic and differential pulse voltammetry Cyclic and differential pulse voltammetry AC-voltammetry AC-voltammetry Square-wave voltammetry using QD-bound aptamers Aptamers bound to carbon-nanotube electrodes Impedance spectroscopy Chronopotentiometric stripping Impedance spectroscopy
25, 0.5 25 25 25 25 25 25 31 25 10 31 31
10 nM 0.5 nM 11 nM 10 nM 3 nM 6.4 nM 0.5 pM N/A 10 nM 0.1 nM 7 nM N/A
40–100 nM 5–35 nM 0–50.8 nM 0–50 nM 0–80 nM 50–768 nM 0.5–12.5 pM* N/A 0–100 nM 2.5–100 nM N/A N/A
[50] [51] [52] [53] [54] [55] [56]
Electrochemical Electrochemical Electrochemical Electrochemical
Thrombin Thrombin Thrombin Thrombin Thrombin Thrombin Thrombin, lysozyme Thrombin Human IgE lysozyme lysozyme
Nanoparticles Nanoparticles
Catalytic NP enlargement Aggregation of aptamer-GNPs
25 0.1 0.1 >10
QD-aptamer beacon
20 nM 35 nM 15 nM 2.5 nM 32 nM N/A
40–140 nM N/A N/A N/A N/A N/A
[61] [62]
Nanoparticles
Thrombin PDGF-BB PDGF-AB PDGF-AA PDGFR-b Thrombin
25
[47] [48] [49]
[57] [58] [59] [60]
[63]
N/A is not available. * The units were changed by the authors to allow easier comparisons.
Table 3. Comparison of the analytical figures of the chiral stationary phases (CSPs) References
Target
Kd
Enantioselectivity (a)
Retention Factor (kd)**
[68] [69]*
D-arginine-vasopressin Tyrosine a-Methyl-tyrosine DOPA Tryptophan 1-Methyl-tryptophan 5-Hydroxyl-trypophan N-Acetyl-tryptophan 3-Benzothienyl-alanine 2-Quinoyl-alanine 2-Naphtyl-alanine 1-Naphtyl-alanine D /L -arganine
1.1 lM 25 lM
N/A 0.57 0.59 0.83 1.42 1.50 1.09 0.49 2.13 1.43 5.29 3.03 7.4
N/A 28.06 21.91 7.20 4.09 6.98 3.62 9.48 1.73 1.88 3.06 1.32 1.43
[70]
412/60 lM
*
Other analytes were also tested, but results for enantioselectivity and retention factors were not obtained. The retention factor is based on the D -enantiomer, since the L -enantiomer eluted in the void volume.
**
Our group has recently developed an ultrasensitive aptamer-based affinity-PCR technique and demonstrated its application for the determination of trace amounts of HIV-1 Reverse transcriptase (RT) [49]. Fig. 2 shows the principle of operation of the assay. The typical CE injection of 10 nL of 30 fM HIV-1 RT represents approximately 180 molecules of HIV-1 RT, which is order of magnitude better LOD than the traditional 686
http://www.elsevier.com/locate/trac
methods. This method was shown to be specific to HIV-1 RT in the presence of human IgG and RNase H. Both the specific recognition by aptamers and the CE separation helped increase the specificity of the assay. Affinity-PCR assays are fast and extremely sensitive due to the PCR-amplification step. However, PCR is also a liability, since contamination can be a problem. One limitation of ligation assays is that the target must have
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Trends
Figure 2. Affinity aptamer amplification assay for proteins [49]. The protein was first incubated with the aptamer to form an aptamer-protein complex. Separation and fractionation of the protein-aptamer complex from excess aptamers was conducted by capillary electrophoresis (CE). After the fraction containing only the aptamer-protein complex was collected, the aptamer was dissociated from the protein and amplified by PCR. The amplified aptamers were indirect measures of the protein concentration.
two distinct binding sites, and appropriate aptamers to bind to the two sites. An advantage of this requirement is the increased specificity of the method. 3.1.2. Electrochemical sensors. Several electrochemical methods for protein detection have made use of aptamers [50–60]. The aptamers are usually coated on nanoparticles (NPs) or a gold-electrode surface with streptavidin before exposing the functionalized surface to a sample containing the target. Other surfaces, such as carbon nanotubes [57], streptavidin-coated magnetic beads [59], and streptavidin/polymer-coated indium-tin oxide electrodes [60], have also been used to immobilize aptamers. Amplification of the target is by preconcentration onto the electrode surface, and the target is detected by a conformational change in the aptamer that causes an electric current. Thrombin is the most commonly studied protein in using electrochemical methods [50–57], although lysozyme [56,59,60] and human IgE [58] have also been studied. Most of these electrochemical methods have achieved LODs below 10nM and a linear dynamic range of two orders of magnitude (Table 2). Xiao [54,55] developed ‘‘signal-on’’ and ‘‘signal-off’’ formats for detecting thrombin using immobilized aptamers on an electrode for binding and using methylene blue for electron transfer (Fig. 3). It was found that the ‘‘signalon’’ version was approximately 10-fold more sensitive than the ‘‘signal-off’’ version; however, the ‘‘signal-off’’ electrode was advantageous, since it could be regenerated, whereas the ‘‘signal-on’’ electrode was stable for only 24 hours, and had to be replaced every day. Other researchers noted that their electrodes could be regenerated a limited number of times [51,52], or were unusable after the first analysis [50]. The robustness of the methods was generally good, showing little to no response from interfering proteins, such as BSA [51,54,56], IgG [56], blood serum [55], IgG, streptavidin [51], elastase [57], beta-casein and ovalbumin [52], a mixture of albumin, cytochrome C and thrombin [60], and a similar mixture that also contained myoglobin,
L -tyrosine
and L -tryptophan [59]. Others showed significant changes in signal with IgG and HSA [53]. Electrochemical assays are inexpensive and, once the electrode has been prepared, they are also fast, usually being completed within a few minutes. One of the main challenges is to maintain the stability and the reusability of the electrode in order to minimize errors resulting from differently coated electrodes. 3.1.3. Aptamer-functionalized nanoparticles with optical detection. Aptamer-functionalized NPs and quantum dots (QDs) have been developed to overcome some of the difficulties of using organic fluorophore labels, which include narrow excitation spectra, short fluorescence lifetime and low quantum yield. Pavlov et al. [61] described a method that used aptamers bound to gold NPs for the detection of thrombin. In this assay, approximately 80 aptamers were bound to each NP. Binding of the thrombin to the aptamers on adjacent NPs resulted in an aggregated precipitate of NPs, which was subsequently separated from the supernatant by centrifugation, and amplified by catalytic enlargement. This resulted in an LOD of 2 nM and a linear dynamic range of 50–120 nM. The assay was shown to be selective for thrombin in the presence of BSA and human IgG. Similarly, Huang et al. described a colorimetric assay for the detection of a platelet derived growth factor (PDGF-BB) and its isoforms (PDGF-AB and PDGF-AA), which also used aggregation of NPs for detection (Fig. 4) [62]. Another approach reported by Levy et al. described a QD-based aptamer biosensor for thrombin, which was similar to a more traditional molecular beacon approach [63]. Approximately 40 quencher/aptamer pairs were bound to the QDs, which were initially in the ‘‘OFF’’ state. Upon binding to thrombin, the conformational change of the aptamer destabilized the hybridized quencher, released it, and turned the fluorescence of the QD ‘‘ON’’. One of the main advantages of coupling NPs with aptamers is the opportunity for significant signal
http://www.elsevier.com/locate/trac
687
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Figure 3. Schematic representation of signal-on (A) and signal-off (B) electrochemical sensors [54,55]. For the signal-on format, a rigid duplex DNA structure prevents the methylene blue (MB)-tagged oligomer from reaching the surface of the electrode. Thrombin binding destabilizes the duplex DNA structure by forcing the aptamer into the characteristic G-quartet structure, resulting in electron transfer from MB on the liberated oligomer to the electrode. For the signal-off architecture, the MB-tagged aptamer is initially able to reach the electrode surface and generate electrical signal. Upon thrombin binding, a conformational change in the aptamer results in the removal of the MB from the electrode surface and inhibition of the electron transfer.
amplification that is not available using conventional fluorophores. This has the potential to lower LODs greatly, which are currently in the low-nM range. A disadvantage of the assays dependent on aggregation is that the degree of aggregation and signal intensity depend on the concentration of both the NPs and the protein (Fig. 4). As a result, optimization of the concentration of NPs may have to be done for each sample, thus complicating the analysis of real samples. 3.2. Small-molecule sensors The use of aptamers as sensors for small molecules has not been studied nearly as extensively as protein targets. Small-molecule sensors are more difficult to develop than protein sensors since there are far fewer moieties for aptamer binding. There are only a few papers that aim to develop sensors for small molecules, and that may have implications for monitoring blood samples for illegal and prescription drugs, as well as enzyme cofactors and other molecules as biomarkers for disease states. 688
http://www.elsevier.com/locate/trac
A recently developed electrochemical sensor used a gold electrode coated with aptamers to detect cocaine in diluted fetal calf serum and human saliva, with LODs of about 10 lM [64]. This represents approximately a 200-fold lower LOD than the Scott Test for cocaine currently used in law enforcement. Other colorimetric sensors for small molecules include modular fluorescence sensors for ATP, flavin mononucleotide, theophylline [65], modified aptameric sensors with a 2 0 -Ribose-linked fluorophore for AMP, tyrosinimide, and arginimide [66], and molecular beacons for ATP [67] based on fluorescence-signal transduction. These sensors have been tested in a buffer mimicking the intercellular milieu [65] and in serum and urine [66] with good results. The LODs for small molecules are severely hampered by available aptamers, which have lM Kd values that result in LODs that are at best in the low lM range. This is orders of magnitude higher than is required to detect analytes in real samples. Despite the high Kd values, the aptamers can still be very selective. For example, the
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
Trends
Figure 4. Schematic representation of the aggregation of aptamer gold nanoparticles (GNPs) in the presence of PDGFs at (A) low, (B) medium and (C) high concentrations [62]. (A) Low concentrations of PDGF result in very little aggregation of the GNPs and a very small color change. (B) A medium concentration of PDGF results in considerable aggregation of the GNPs, producing a large color change. (C) A high concentration of PDGF saturates the aptamer-binding sites on the GNPs, resulting in little to no aggregation, and only a very small color change.
theophylline aptamer is able to distinguish theophylline from caffeine, which differ by only a single methyl group [2]. 3.3. Chiral selectors in affinity chromatography Aptamers have been immobilized onto surfaces, typically as stationary phases in affinity chromatography, to provide separation of target and non-target compounds. In particular, enantiometric separations using CSPs are of increasing interest in both pharmaceutical and biological fields, because often one enantiomer of a chiral drug is pharmaceutically active, whereas the other is inactive or toxic. PeyrinÕs group used DNA aptamers for target-specific chiral separations [68]. They were able to separate the D and L -peptides of vasopressin using an HPLC column with a D -vasopressin-specific DNA aptamer immobilized on a highly porous polystyrene-divinylbenzene support via a biotin-streptavidin bridge. The D -enantiomer was retained while the L -enantiomer was eluted in the void volume. Although no quantitative data were presented, the authors reported that the enantioselectivity was better than conventional chiral selectors and comparable to using monoclonal antibodies. RNA aptamers have also been developed for CSP applications; however, the stability of the ‘‘natural’’
D -aptamers
is very low due to degradation by RNases. To address this problem, the L -RNA aptamer that was intrinsically resistant to enzyme degradation was used instead of the ‘‘natural’’ D -RNA aptamer [69,70]. This mirror-image strategy greatly increases the stability of RNA CSPs, which can otherwise be degraded over time by RNases in the sample. The L -oligonucleotides are not susceptible to degradation by the naturally-occurring enzymes, and have the same selectivity and specificity for the mirror-image of the target. Peyrin and coworkers used this strategy to develop target-specific RNA CSPs to separate enantiomers of arginine [70] and tyrosine and related compounds [69] with enantioselectivities of 7.4 and 10, respectively. The CSPs were stable for over 1500 column volumes of mobile phase (3 months of use). Table 3 presents some analytical data for these separation methods. For the RNA CSPs, the L -enantiomers were commonly eluted in the void volume, making it difficult to quantify the retention factors accurately, so the retention-factor data were obtained from the D -enantiomers. In most cases, the enantioselectivity was acceptable, except when it was very low, indicating incomplete separation. Despite the promise of aptamers for chiral separation, aptamer CSPs have lower binding capacities, lower efficiencies, and higher costs compared
http://www.elsevier.com/locate/trac
689
Trends
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
to conventional CSPs. An improvement may lie in the covalent binding of aptamers on chromatographic particles of lower diameter, which may significantly improve the chromatographic performance of the aptamer CSPs.
4. Concluding remarks Aptamers have great potential for use in bioanalytical applications. The main impetus for aptamer development is their ability to recognize virtually any target with high affinity and specificity. Antibodies are currently the reagent of choice in many assays because of their wide availability and proven success; however, new developments in SELEX techniques have shown promise for the generation of aptamers in shorter times with higher affinities for a wider array of targets. Aptamers can complement antibodies as affinity reagents in bioanalytical assays. Most studies have focused on protein targets, resulting in a lack of aptamer development for smaller inorganic and organic molecules, as well as complex targets. This is mainly because of the difficulty in obtaining aptamers with high binding affinity and specificity for these targets, which are both important factors to consider when selecting aptamers for a given application. Aptamers with high target-binding affinity do not necessarily have high target specificity; however, both of these factors will determine whether or not the aptamer could be used in a real sample with hundreds or thousands of potentially interfering components. Future studies should always assess the specificity of aptamers for other targets, as this element has been missing in many previous studies. Ideally, all assays should be tested in a medium representative of the matrix for which they were developed (e.g., biological fluids and environmental matrices). Active collaborations among different disciplines will help advance the field of aptamer technology, which will lead to new, innovative applications encompassing a number of different targets. Acknowledgement The authors acknowledge the financial support from the Natural Sciences and Engineering Research Council of Canada and Alberta Health and Wellness (to XCL and XFL), the Canada Research Chairs Program (to XCL), an Alberta Heritage Foundation for Medical Research Studentship and a Natural Sciences and Engineering Research Council of Canada Studentship (to CLAH), an Alberta Ingenuity Postdoctoral Fellowship (to JWG), and a Graduate Assistantship from the University of Alberta (to HZ).
690
http://www.elsevier.com/locate/trac
References [1] A.D. Ellington, J.W. Szostak, Nature (London) 346 (1990) 818. [2] R.D. Jenison, S.C. Gill, A. Pardi, B. Polisky, Science (Washington, DC) 263 (1994) 1425. [3] C. Tuerk, L. Gold, Science (Washington, DC) 249 (1990) 505. [4] C. Pestourie, B. Tavitian, F. Duconge, Biochimie 87 (2005) 921. [5] H. Ulrich, A.H.B. Martins, J.B. Perquero, Cytometry 59A (2004) 220. [6] S. Tombelli, M. Minunni, M. Mascini, Biosens. Bioelectron. 20 (2005) 2424. [7] V. Pavski, X.C. Le, Curr. Opin. Biotechnol. 14 (2003) 65. [8] W. Tan, K. Wang, T.J. Drake, Curr. Opin. Chem. Biol. 8 (2004) 547. [9] J.W. Guthrie, C.L.A. Hamula, H. Zhang, X.C. Le, Methods 38 (2006) 324. [10] S.M. Nimjee, C.P. Rusconi, B.A. Sullenger, Annu. Rev. Med. 56 (2005) 555. [11] S.E. Osborne, A.D. Ellington, Chem. Rev. 97 (1997) 349. [12] M.C. Golden, B.D. Collins, M.C. Willis, T.H. Koch, J. Biotechnol. 81 (2001) 167. [13] K.B. Jensen, B.L. Atkinson, M.C. Willis, T.D. Koch, L. Gold, Proc. Natl. Acad. Sci. USA 92 (1995) 12220. [14] B.S. Singer, T. Shtatland, D. Brown, L. Gold, Nucleic Acids Res. 25 (1997) 781. [15] T. Shtatland, S.C. Gill, B.E. Javornik, H.E. Johansson, B.S. Singer, O.C. Uhlenbeck, D.A. Zichi, L. Gold, Nucleic Acids Res. 28 (2000) E93. [16] T. Shimada, N. Fujita, M. Maeda, A. Ishihama, Genes Cells 10 (2005) 907. [17] J.F. Lee, J.R. Hesselberth, L.A. Meyers, A.D. Ellington, Nucleic Acids Res. 32 (2004) D95. [18] J. Kawakami, H. Imanaka, Y. Yokota, N. Sugimoto, J. Inorg. Biochem. 82 (2000) 197. [19] M. Koizumi, R.R. Breaker, Biochemistry 39 (2000) 8983. [20] P.E. Huizenga, J.W. Szostak, Biochemistry 34 (1995) 656. [21] E. Vianini, M. Palumbo, B. Gatto, Biochemistry 39 (2000) 946. [22] A. Geiger, P. Burgstaller, H. von der Eltz, A. Roeder, M. Famulok, Nucleic Acids Res. 24 (1996) 1029. [23] G.R. Zimmerman, T.R. Shields, R.D. Jenison, C.L. Wick, A. Pardi, Biochemistry 37 (1998) 918. [24] U. Brockstedt, A. Uzarowska, A. Montpetit, W. Pfau, D. Labuda, Biochem. Biophys. Res. Comm. 313 (2004) 1004. [25] K.A. Plummer, J.M. Carothers, M. Yoshimura, J.W. Szostak, G.L. Verdine, Nucleic Acids Res. 33 (2005) 5602. [26] K.N. Morris, K.B. Jensen, C.M. Julin, L. Gold, Proc. Natl. Acad. Sci. USA 95 (1998) 2902. [27] J.G. Bruno, J.L. Kiel, Biosens. Bioelectron. 14 (1999) 457. [28] M. Blank, T. Weinschenk, M. Priemer, H. Schluesener, J. Biol. Chem. 276 (2001) 16464. [29] D.A. Daniels, H. Chen, B.J. Hicke, K.M. Swiderek, L. Gold, Proc. Natl. Acad. Sci. USA 100 (2003) 15416. [30] B.J. Hicke, C. Marion, Y.-F. Chang, T. Gould, C.K. Lynott, D. Parma, P.G. Schmidt, S. Warren, J. Biol. Chem. 276 (2001) 48644. [31] C. Wang, M. Zhang, G. Yang, D. Zhang, H. Ding, H. Wang, M. Fan, B. Shen, N. Shao, J. Biotechnol. 102 (2003) 15. [32] L. Cerchia, F. Duconge, C. Pestourie, J. Boulay, Y. Aissouni, K. Gombert, B. Tavitian, V. de Franciscis, D. Libri, PLoS Biol. 3 (2005) e123. [33] K. Guo, H.P. Wendel, L. Scheideler, G. Ziemer, A.M. Scheule, J. Cell. Mol. Med. 9 (2005) 731. [34] J. Vivekananda, J.L. Kiel, Lab Invest. (2006) URL: http://www. nature.com/labinvest/journal/vaop/ncurrent/abs/3700417a.html;
Trends in Analytical Chemistry, Vol. 25, No. 7, 2006
[35] [36] [37]
[38] [39] [40]
[41] [42] [43] [44] [45] [46]
[47] [48] [49] [50]
jsessionid=B29434A4BA4D0E5D835424333F66AD77 [accessed 3 April 2006]. S.P. Ohuchi, T. Ohtsu, Y. Nakamura, Biochimie (2006) (doi:10.1016/j.biochi.2006.02.004).. K. Mosing, S.D. Mendonsa, M.T. Bowser, Anal. Chem. 77 (2005) 6017. M. Berezovski, A. Drabovich, S.M. Krylova, M. Musheev, V. Okhonin, A. Petrov, S.N. Krylov, J. Am. Chem. Soc. 127 (2005) 3165. R. White, C. Rusconi, E. Scardino, A. Wolberg, J. Lawson, M. Hoffman, B. Sulenger, Mol. Ther. 4 (2001) 567. R. Stoltenberg, C. Reinemann, B. Strehlitz, Anal. Bioanal. Chem. 383 (2005) 83. C. Bock, M. Coleman, B. Collins, J. Davis, G. Foulds, L. Gold, C. Greef, J. Heil, J.S. Heilig, B. Hicke, M.N. Hurst, G.M. Husar, D. Miller, R. Ostroff, H. Petach, D. Schneider, B. Vant-Hull, S. Waugh, A. Weiss, S.K. Wilcox, D. Zichi, Proteomics 4 (2004) 609. S.D. Mendonsa, M.T. Bowser, J. Am. Chem. Soc. 126 (2004) 20. S.D. Mendonsa, M.T. Bowser, Anal. Chem. 76 (2004) 5837. S.D. Mendonsa, M.T. Bowser, J. Am. Chem. Soc. 127 (2005) 9382. A. Drabovich, M. Berezovski, S.N. Krylov, J. Am. Chem. Soc. 127 (2005) 11224. M. Berezovski, M. Musheev, A. Drabovich, S.N. Krylov, J. Am. Chem. Soc. 128 (2006) 1410. S. Fredriksson, M. Gullberg, J. Jarvius, C. Olsson, K. Pietras, S.M. Gu´stafsdo´ttir, A. O¨stman, U. Landegren, Nat. Biotechnol. 20 (2002) 473. D.A. Di Giusto, W.A. Wlassoff, J.J. Gooding, B.A. Messerle, G.C. King, Nucleic Acids Res. 33 (2005) e64. X-L. Wang, F. Li, Y-H. Su, X. Sun, X.-B. Li, H.J. Schluesener, F. Tang, S.-Q. Xu, Anal. Chem. 76 (2004) 5605. H. Zhang, Z. Wang, X.-F. Li, X.C. Le, Angew. Chem. Int. Ed. 45 (2006) 1576. K. Ikebukuro, C. Kiyohara, K. Sode, Biosens. Bioelectron. 20 (2005) 2168.
Trends [51] A.-E. Radi, J.L.A. Sa´nchez, E. Baldrich, C.K. OÕSullivan, J. Am. Chem. Soc. 128 (2005) 117. [52] G.S. Bang, S. Cho, B.-G. Kim, Biosens. Bioelectron. 21 (2005) 863. [53] T. Hianik, V. Ostatna´, Z. Zajacova´, E. Stoikova, G. Evtugyn, Biorg. Med. Chem. Lett. 15 (2005) 291. [54] Y. Xiao, B.D. Piorek, K.W. Plaxco, A.J. Heeger, J. Am. Chem. Soc. 127 (2005) 17990. [55] Y. Xiao, A.A. Lubin, A.J. Heeger, K.W. Plaxco, Angew. Chem. Int. Ed. 44 (2005) 5456. [56] J.A. Hansen, J. Wang, A.-N. Kawde, Y. Xiang, K.V. Gothelf, G. Collins, J. Am. Chem. Soc. 128 (2006) 2228. [57] H.-M. So, K. Won, Y.H. Kim, B.-K. Kim, B.H. Ryu, P.S. Na, H. Kim, J.-O. Lee, J. Am. Chem. Soc. 127 (2005) 11906. [58] D. Xu, D. Xu, X. Yu, Z. Liu, W. He, Z. Ma, Anal. Chem. 77 (2005) 5107. [59] A.-N. Kawde, M.C. Rodriquez, T.M.H. Lee, J. Wang, Electrochem. Commun. 7 (2005) 537. [60] M.C. Rodriguez, A-N. Kawde, J. Wang, Chem. Commun. 34 (2005) 4267. [61] V. Pavlov, Y. Xiao, B. Shlyahovsky, I. Willner, J. Am. Chem. Soc. 126 (2004) 11768. [62] C.-C. Huang, Y.-F. Huang, Z. Cao, W. Tan, H.-T. Chang, Anal. Chem. 77 (2005) 5735. [63] M. Levy, S.F. Cater, A.D. Ellington, ChemBioChem 6 (2005) 2163. [64] B.R. Baker, R.Y. Lai, M.S. Wood, E.H. Doctor, A.J. Heeger, K.W. Plaxco, J. Am. Chem. Soc. 128 (2006) 3138. [65] M.N. Stojanovic, D.M. Kolpashchikov, J. Am. Chem. Soc. 126 (2004) 9266. [66] E.J. Merino, K.M. Weeks, J. Am. Chem. Soc. 127 (2005) 12766. [67] R. Nutiu, Y. Li, J. Am. Chem. Soc. 125 (2003) 4771. [68] M. Michaud, E. Jourdan, A. Villet, A. Ravel, C. Grosset, E. Peyrin, J. Am. Chem. Soc. 125 (2003) 8672. [69] C. Ravelet, R. Boulkedid, A. Ravel, C. Grosset, A. Villet, J. Fize, E. Peyrin, J. Chromatogr. A 1076 (2005) 62. [70] A. Brumbt, C. Ravelet, C. Grosset, A. Ravel, A. Vellet, E. Peyrin, Anal. Chem. 77 (2005) 1993.
http://www.elsevier.com/locate/trac
691